1 INFO-VAX	Fri, 24 Aug 2001	Volume 2001 : Issue 470       Contents: Alpha - Complete for USD599 " Re: cluster system disk shadowness Re: Compaq Mark Twain mailing % Conference: CETS-2001 Detailed Update 
 DCL challenge  Re: DCL challenge  Freeware Dskmon Utility  Re: Freeware Dskmon Utility  Re: Freeware Dskmon Utility # Hex dump-and-lookatit-in-themorning - Re: IA-64 Emulation (was: Re: Nits in Slides) - Re: IA-64 Emulation (was: Re: Nits in Slides) - Re: IA-64 Emulation (was: Re: Nits in Slides) - Re: IA-64 Emulation (was: Re: Nits in Slides)  Intersystems and Alpha+ Re: Intersystems and Alpha and VAX and IA64 + Re: Introduction and BACKUP recommendations / Re: Looking for a good alert management utility ( Re: LPR from VMS 7.2-1 to Win2000 Server( Re: LPR from VMS 7.2-1 to Win2000 Server( Re: LPR from VMS 7.2-1 to Win2000 Server Re: More Alpha rubbish in print  Re: Nits in Slides Re: Nits in Slides Re: Nits in Slides Re: Nits in Slides Re: Nits in Slides Re: Nits in Slides Re: Nits in Slides Re: Nits in Slides Re: Nits in Slides9 Re: Nits in Slides (was: Re: The Final Knell Has Sounded) 9 Re: Nits in Slides (was: Re: The Final Knell Has Sounded) 9 Re: Nits in Slides (was: Re: The Final Knell Has Sounded) 9 Re: Nits in Slides (was: Re: The Final Knell Has Sounded) 9 Re: Nits in Slides (was: Re: The Final Knell Has Sounded) 9 Re: Nits in Slides (was: Re: The Final Knell Has Sounded) 5 Re: Nuts-n-bolts in News (was: Re: Nits in Slides...) 5 Re: Nuts-n-bolts in News (was: Re: Nits in Slides...) 5 Re: Nuts-n-bolts in News (was: Re: Nits in Slides...) 5 Re: Nuts-n-bolts in News (was: Re: Nits in Slides...) 5 Re: Nuts-n-bolts in News (was: Re: Nits in Slides...) " Re: ODS-5 and parse_style question" Re: ODS-5 and parse_style question" Re: ODS-5 and parse_style question Old print program wanted Re: Old print program wanted! On board SCSI supported on DS20E? # Re: OT: TOPS-20 and TOPS-10 live on  Queue/Entry management Re: Queue/Entry management7 Re: RWMBX - What to do, What to do! Help me if you can. 7 Re: RWMBX - What to do, What to do! Help me if you can. 7 Re: RWMBX - What to do, What to do! Help me if you can. * Re: Single disk root versus multiple disks Re: Some postive points I hope.  Re: Unexpected DIAGPACK error , Re: VMS high reliability needed by Air Force, Re: VMS high reliability needed by Air Force4 re: Wailing at Eunuchs (was: Wailing and Moaning...)4 Re: Wailing at Eunuchs (was: Wailing and Moaning...)4 Re: Wailing at Eunuchs (was: Wailing and Moaning...)4 Re: Wailing at Eunuchs (was: Wailing and Moaning...)4 Re: Wailing at Eunuchs (was: Wailing and Moaning...)4 Re: Wailing at Eunuchs (was: Wailing and Moaning...)4 Re: Wailing at Eunuchs (was: Wailing and Moaning...): Re: [Q]: How to get BACKUP to relabel tapes automatically?  F ----------------------------------------------------------------------  % Date: Fri, 24 Aug 2001 07:49:40 -0400 - From: "www.islandco.com" <sales@islandco.com> $ Subject: Alpha - Complete for USD599/ Message-ID: <tocfbto4pn9t50@news.supernews.com>   A Yep - September is here and we are shipping these little puppies.    Configured as follows:   433Mhz CPU 21164 EV56 at 433Mhz & 128MB Memory 10/100 Ethernet On Board, S3Trio64 2MB 1.44Mb Floppy + 
 4x SCSI CDROM 3 Unix/Linux SCSI Ctr PCI 2GB  7200rpm Disk SCSI Wide  Keyboard Generic 3 Button Mouse   D VMS compatible system requires different (QLogic) controller add $51     23 in stock and ready to go !      -- David B Turner	 Sales Dpt  Island Computers US Corporation  2700 Gregory Street 	 Suite 180  Savannah GA 31404  Tel: 912 447 6622  Fax: 912 201 0096  sales@islandco.com www.islandco.com' http://www.islandco.com/legal-email.htm    We sell Alpha's !    ------------------------------  # Date: Fri, 24 Aug 2001 14:13:44 GMT 2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)+ Subject: Re: cluster system disk shadowness 2 Message-ID: <sith7.720$bB1.31904@news.cpqcorp.net>  h In article <B1hh7.179$3t3.21487@news-west.eli.net>, "frank brown" <frank.brown@ci.seattle.wa.us> writes: ..G :What determines if a VAX tries to create a shadowset when mounting the 0 :system disk?  How can I change this behavior?     $ mc sysgen  SYSGEN>  SHO SH   .   Specifically, SHADOW_SYS_DISK and SHADOWING.  ? :Alas, I cannot find the Volume Shadowing manual in our docset.   =   http://www.openvms.compaq.com/doc/73final/5423/5423pro.html   N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Fri, 24 Aug 2001 08:03:52 -0400  From: William_Bochnik@acml.com& Subject: Re: Compaq Mark Twain mailing> Message-ID: <OFCEEAC266.ED2E4D3E-ON85256AB2.004240D7@acml.com>   hence the ROTFL       a                                                                                                   a                     "David J.                                                                     a                     Dachtera"                    To:  Info-VAX@Mvb.Saic.Com                       a                     <djesys.nospa                cc:                                              a                     m@fsi.net>           Subject:     Re: Compaq Mark Twain mailing               a                                                                                                   a                     08/23/2001                                                                    a                     11:00 PM                                                                      a                                                                                                   a                                                                                                          William_Bochnik@acml.com wrote:  > 0 > when you get it in the mail, you'll understand > A > page 55 - Put all your eggs in one basket and WATCH THAT BASKET  >  > ROTFL   > Well, some financial advisors and success counselors also hold	 that same  conviction.   A Personally, I think it's a dumb move for the industry as whole to  be' putting all its eggs in Intel's basket.   # ..but then what the hell do I know? > I come from Waunakee (Wisconsin) (not really, its just a song) I wear animal pelts, I eat raw meat,  I sleep in a hollow tree...    -- David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/           F ______________________________________________________________________;  The information contained in this transmission may contain @ privileged and confidential information and is intended only forA the use of the person(s) name above.  If you are not the intended = recipient, or an employee or agent responsible for delivering 3 this message to the intended recipient, any review, @ dissemination, distribution or duplication of this communication? is strictly prohibited.  If you are not the intended recipient, A please contact the sender immediately by reply e-mail and destroy # all copies of the original message.    ------------------------------  # Date: Fri, 24 Aug 2001 17:37:00 GMT & From: "Jeff Killeen" <Jeff@IDM-IO.com>. Subject: Conference: CETS-2001 Detailed Update8 Message-ID: <0hwh7.1151$YN4.805979@typhoon2.gnilink.net>  E This message is a detailed update for the Compaq Enterprise Technical & Symposium 2001. This message includes:  %     1) Registration Discount Extended      2) Symposium Program Update #     3) Meet the Experts Night (New)      4) Itanium Sessions (New) <     5) Compaq Listens Panel Update (Open for Questions Now!)1     6) Compaq Engineering Technical Center Update ?     7) Encompass Exposition Update (Complementary - No Reg Fee) !     8) Encompass Campground (New) =     9) Technical Seminars Update (Symposium Reg Not Required)     ! 1) Registration Discount Extended ! ---------------------------------   I You can still register for the Compaq Enterprise Technical Symposium 2001  at:   6     http://www.cets2001.com/cets/reg/registration.jsp.  F We currently have more Encompass members registered for 2001 than wereD registered for all of last year. Due to summer vacations and currentF economic conditions we have decided to extend the Early Bird discountsA through August 27 at midnight central USA time. Depending on what L registration options you choose those discounts can combine up to a total of $300.      2) Symposium Program Update  ---------------------------   E We have achieved our goal for the to deliver solid in-depth technical L content for the Tru64, OpenVMS, and/or ProLiant IT professional. The detailsJ are available at http://www.cets2001.com - please select "Session Catalog" on the left.     3) Meet the Experts Night (New) " ----------------------------------   Meet the Experts Night* Tuesday September 11th, 5:30 PM to 7:30 PM Pacific Ballroom, Hilton Hotel  J Calling all techies! This is your chance to meet all the technical expertsI of Compaq Enterprise Technical Symposium assembled in one place! Use this J opportunity to discuss your own projects and challenges with the engineersE and consultants with the expertise in those areas. Get your questions G answered, and meet new people who can be your virtual resources for the  future.      4) Itanium Sessions (New)  ---------------------------   ? Currently the following Itanium sessions are being scheduled...   3     You Can Bet Your Business on Compaq and Itanium      by Rich Marcello  <     The AlphaServer to Itanium Processor Family (IPF) Server#     Transition and IPF Road Forward      by David Fenwick  5     Alpha to Itanium Processor Family (IPF) Q&A Panel      Moderator: Rich Marcello  0     Itanium Processor Family (IPF) on Tru64 UNIX      by Bob Beckett and Ray Lanza  1     Compilers and Tools for Tru64 UNIX on Itanium      by Ken Reilly   7     OpenVMS Internals on Itanium Processor Family (IPF)      by Clair Grant  +     Porting OpenVMS Applications to Itanium      by Gaitan D'Antoni       OpenVMS Technical Update     by Stephen "Hoff" Hoffman   6     Building Your Business' Future with Compaq OpenVMS     by Mark Gorham  J There are a few more sessions under development. In early September please check for updates at:   3     http://www.cets2001.com/privatedocs/Itanium.pdf      or3     http://www.cets2001.com/privatedocs/Itanium.txt      5) Compaq Listens Panel Update  -----------------------------  B The Compaq Listens panel is the place to seek the answers to thoseK thoughtful questions you are pondering about Compaq's strategies and future I business directions. This year, the Compaq Listens panel is scheduled for E Monday evening, from 5:45 PM to 7:00 PM. It is open to all registered C attendees of Compaq Enterprise Technical Symposium and ITUG Summit.   J Come to the panel and ask your question in person, or submit your questionI in advance. We will attempt to address all your concerns in a forthright, 
 frank manner.   J Submit your questions to Information@cets2001.com by September 9 and label/ the subject line "question for Compaq Listens."     - 6) Compaq Engineering Technical Center Update - ---------------------------------------------   F This is your opportunity to talk directly with the Compaq Engineer whoK developed the product you are deploying. Compaq engineers will be scheduled F in this center throughout the week to meet with you. Bring the currentE technical issues you are working on and confer with the engineers who  actually build the products.     7) Encompass Exposition Update  -----------------------------  L If you are going to be in the Anaheim area, but can't attended the Symposium? this year, you still can register for the Encompass Exposition. I For 2001 only the Compaq Engineering Technical Center is being located in G the Exhibition Center. This is a special opportunity for Tradeshow Only H attendees. For more information about our complementary Tradeshow please visit:  7     http://www.cets2001.com/cets/content/exposition.jsp   , Encompass Exposition 2001 Hours of Operation  0     Monday, September 10 12:00 p.m. to 5:45 p.m.0     Tuesday, September 11 9:30 a.m. to 6:30 p.m.2     Wednesday, September 12 9:00 a.m. to 4:00 p.m.     8) Encompass Campground (New)  -----------------------------   K Need a place to take a break a chat with your fellow IT professionals? Come C visit the Encompass Campground located at the rear of the Encompass I Exposition hall. Compaq engineers have been known to take a break also in  the campground.     % 9) In-depth Technical Seminars Update % -------------------------------------   I There are still seats available if you are attending Symposium and if you J are going to be in the Anaheim California area you do not need to registerL for the Symposium to register for a Seminar! You can still gain the benefitsH of a Pre-Conference Seminar offered on Saturday September 8th and SundayJ September 9th even if you can't take advantage of the Symposium during the week.   K If you have not registered for Symposium - Log in and select "Register Now" L on the menu at the left side of the screen. Next, click "Edit" (near bottom)C to bring up the screen with tabs showing sessions for which you can L register. Click on a tab, and select "add to cart" for the sessions you wish
 to attend.  F If you have already register for Symposium - Log in and select "UpdateL Registration" on the menu at the left side of the screen. Next, click "Edit"  I (near bottom) to bring up the screen with tabs showing sessions for which K you can register. Click on a tab, and select "add to cart" for the sessions  you wish to attend.   ) Our Techworks Bootcamps Seminars include:   =     1459 Tru64/TruClusters 101 Bootcamp (Saturday and Sunday) -         An Intro for the OpenVMS Professional   4     1640 Windows 2000 Bootcamp (Saturday and Sunday)8          Windows 2000 for the Windows NT 4 Administrator  5     1639 Exchange 2000 Bootcamp (Saturday and Sunday) 9          Exchange 2000 for the Exchange 5.x Administrator   )     1455 SAN 101 Bootcamp (Saturday only) '          Intro to Storage Area Networks   '     1577 SAN 101 Bootcamp (Sunday only) '          Intro to Storage Area Networks   E For detailed information on 30 plus Seminars, including the Techworks  Bootcamps, please download:   4     http://www.cets2001.com/privatedocs/Seminars.pdf     or4     http://www.cets2001.com/privatedocs/Seminars.txt  E PLEASE NOTE - Our Seminar on console management has been sigificantly  revised and extended:   5 Seminar 1078: Enterprise Web-based Console Management   	 Abstract:   L ConsoleWorks is a Web-based console management system that runs on Tru64(tm)F UNIX, OpenVM(tm), Linux, Windows NT and SUN Solaris. ConsoleWorks is aH perfect upgrade from PCM, Robo Central or other older Console ManagementG Systems. ConsoleWorks has Scanfiles for Tru64 UNIX, OpenVMS 7.3, Compaq L Storage Controllers, SAN Switches and Compaq Alpha(tm) SRM Processors. It isE distributed with the GS AlphaServer and manages ALL Compaq equipment.   L Come see how console management should be done, what benefits can be gained,J setup, configuration, management and reporting. Also learn what you shouldF consider when setting up a console management system, hints and kinks,J problems and solutions. This seminar covers Console Out-Of-Band managementJ in general using ConsoleWorks. It is a prime course for anyone consideringC upgrading or implementing a console management solution. We discuss L how-to's, what to do, what not to do as well as alternative ways of managing+ a data center with an out-of-band solution.   E This seminar is a workshop. The instructor demonstrates how to build, H configure and manage a data center in Dallas, Texas from Anaheim via theK WEB. Also, additional configuration parts are built right in the classroom.     ; Compaq Enterprise Technical Symposium 2001 Registration Kit ; -----------------------------------------------------------   6 A PDF version of the registration kit is available at:  :     http://www.cets2001.com/privatedocs/CETS2001RegKit.pdf     Hope to see you in Anaheim!2     Jeff Killeen> CETS-2001 Business Manager for Encompass - A Compaq User Group http://www.EncompassUS.org   ------------------------------  % Date: Fri, 24 Aug 2001 17:44:56 +0200 > From: "Jean-Francois Marchal" <jean-francois.marchal@x9000.fr> Subject: DCL challenge. Message-ID: <9m5sg2$n6m$1@reader1.imaginet.fr>  
 Hi all ...   here is a small DCL code   $ SYSNUM = 1 $ REFNUM = 1$ $ S_'SYSNUM'_'REFNUM'_REFINI = "AAA" $6$ $ S_'SYSNUM'_'REFNUM'_REF    = "BBB"> $ pipe a = f$fao("S_!UL_!UL_REF!0UL!1%CINI!%F",SYSNUM,REFNUM,-2  'f$fao("S_!UL_!UL_REF",SYSNUM,REFNUM).eqs."") ; --  z = f$fao("!9AS",&A) ; define/job KEYPART &z  $ z = f$trnlnm("KEYPART")e $ sh symbol z ! should be BBB  $m! $ S_'SYSNUM'_'REFNUM'_REF    = ""P> $ pipe a = f$fao("S_!UL_!UL_REF!0UL!1%CINI!%F",SYSNUM,REFNUM,-2  'f$fao("S_!UL_!UL_REF",SYSNUM,REFNUM).eqs."") ; --  z = f$fao("!9AS",&A) ; define/job KEYPART &z  $ z = f$trnlnm("KEYPART")  $ sh symbol z ! should be AAA  $d  I could any DCL guru rewrite the code to get z in one line without a pipe ?l  " Yes of course, I could just use  :  $ Z = S_'SYSNUM'_'REFNUM'_REFINID $ if S_'SYSNUM'_'REFNUM'_REF.nes."" then Z = S_'SYSNUM'_'REFNUM'_REF   Cordialement Jean-Franois Marchald   ------------------------------  % Date: Fri, 24 Aug 2001 19:49:10 +0200e, From: Didier Morandi <Didier.Morandi@gmx.ch> Subject: Re: DCL challenge& Message-ID: <3B869396.FD905B31@gmx.ch>   Jean-Francois Marchal wrote:   ../..bK > could any DCL guru rewrite the code to get z in one line without a pipe ?:  " You should make a choice here, JF:  J Either you please yourself with the most ununderstandable line(s) of code, orE you consider that one day someone else than you will have to maintaint
 this code.  G I have definitely choosen the second option and each time I arrive on anC Customer's site, I split complicated DCL code to different lines to ) improve readability and ease maintenance.t  7 Think about the poor consultants who come behind... :-)t   D.   ------------------------------  # Date: Fri, 24 Aug 2001 14:30:20 GMTc From: sfm1115@bjc.org (Shawn)h  Subject: Freeware Dskmon Utility0 Message-ID: <3b86648f.81637238@news.starnet.net>  ? I am trying to get the freeware version of dskmon running on myi/ AlphaServer 1000/A Server running Openvms 7.3-1r  2 When I do a run dskmon, I get the following error:   y2kalp>run dskmona% X Toolkit Error: Can't open display: W% %DWT-F-NOMSG, Message number 03AB8204E/ %TRACE-F-TRACEBACK, symbolic stack dump follows 9   image    module    routine             line      rel PCn abs PC      >                                             0 FFFFFFFF807EA4F0 FFFFFFFF807EA4F0>                                             0 FFFFFFFF807EA084 FFFFFFFF807EA084>                                             0 FFFFFFFF807EA2C0 FFFFFFFF807EA2C0>                                             0 FFFFFFFF807E8A3C FFFFFFFF807E8A3C>                                             0 FFFFFFFF8081DE84 FFFFFFFF8081DE84>                                             0 FFFFFFFF8081E0E4 FFFFFFFF8081E0E4>  DSKMON  DSKMON  main                   21606 000000000000014C 000000000003014C>  DSKMON  DSKMON  __main                     0 000000000000009C 000000000003009C>                                             0 FFFFFFFF8024B3D4 FFFFFFFF8024B3D4    C Do I need to have DECWindows running on this server in order to user the freeware utilities.    Thanks Shawn    ------------------------------  % Date: Fri, 24 Aug 2001 16:40:43 +01000( From: Nic Clews <sendspamhere@127.0.0.1>$ Subject: Re: Freeware Dskmon Utility) Message-ID: <3B86757B.2807D46D@127.0.0.1>c   Shawn wrote: > A > I am trying to get the freeware version of dskmon running on my-1 > AlphaServer 1000/A Server running Openvms 7.3-1  > 4 > When I do a run dskmon, I get the following error: >  > y2kalp>run dskmoni& > X Toolkit Error: Can't open display:' > %DWT-F-NOMSG, Message number 03AB8204<E > Do I need to have DECWindows running on this server in order to usee > the freeware utilities.W >   F You guess right. ish. You need the decwindows base files on the systemC you're trying to run it on, plus a system with an X server, PC withS1 Xceed for example, or a workstation. A simple SETW@ DISPLAY/CREATE/NODE=node[/TRANSPORT=transport] will fix it, also: assuming the security allows the creation of said display.  G I've not used, I've thought about playing with it/trying it out, so I'd-- be interested in your opinions on running it.e   -- e( Regards, Nic Clews CSC Computer Sciences nclews at csc dot com-   ------------------------------  % Date: Fri, 24 Aug 2001 19:43:26 +0200 , From: Didier Morandi <Didier.Morandi@gmx.ch>$ Subject: Re: Freeware Dskmon Utility& Message-ID: <3B86923E.FD230DAA@gmx.ch>   Nic Clews wrote: ../..o > A simple SETB > DISPLAY/CREATE/NODE=node[/TRANSPORT=transport] will fix it, also< > assuming the security allows the creation of said display.   A few details:  2 the real command in your login.com (or elsewhere):  7 $ set display/create/node='target_node'/transport=tcpip   G where 'target-node' is either the name or fully qualified name or TCPIP F address of the system where you wish to have the display ... displayed  A For the security param, if you use DEcwindows, click the SECURITYcG function and add *:*, whish means all IP addresses ad all users alloweds( to display something on your X terminal.  G Of course, you should then change this to avoid an erroneous command to A display something on your terminal, then an eventual resetof  the F originator X server, will causes your terminal to poofff and force youD to reopen the 12 windows you had opened and carefully sized for yourH development session. The reset command sent to a server is actually also@ sent to the X terminal where it runs and resets the thing... :-(  E I'm not an X specialist, but I learned a few months ago that a server E should not be rest when a window is opened on someone else's terminal E (the client was a SUN station and I was running a Macintosh X server)  D.   ------------------------------  % Date: Fri, 24 Aug 2001 10:51:36 +0200o' From: "Thys de Wet" <thys@it.sun.ac.za>o, Subject: Hex dump-and-lookatit-in-themorning/ Message-ID: <9m54ij$p0a$1@news.adamastor.ac.za>    Reminds me of my operator days:g  I Program chrashes at 03:00 right in the middle of the night production runaF (No OS, no terminals, no disks in those days: Old ICT Orion transistorG 'puter, Papertape, punch cards, flexowriter, I-line paper, drum memory)   I Phone up the standby programmer and lo and behold: "do a hex dump and Ille look at it in the morning"...mI He didn't get far.  He had to go and "look at it" in the manager's office_ "in the morning.   T de Wet in South Africa  B "David Beatty" <David.Beatty@qwertysasasdfgh.com> wrote in message, news:FvuEOzrbjhPNnu6j=KuWnenM6m6p@4ax.com...9 > On Wed, 22 Aug 2001 22:37:24 -0500, "David J. Dachtera",  > <djesys.nospam@fsi.net> wrote: >f > >Randy Burlew wrote: > >>1 > >> In article <3B82C1F4.CE30489F@videotron.ca>,o jfmezei.spamnot@videotron.ca > >> says... > >> >J > >> >But in light of the Alpha murder, consider that this deal was in the works aeJ > >> >long long time ago, having a *NEW* deal signed in light of the alpha murder  > >> >would have far more value. > >>> > >> I'm a big supporter of free speech and all, but this talk@ > >> of the Alpha "murder" is getting a wee little bit tiresome. > >>< > >> Besides, computer language is already too violent (kill% > >> the process, abort, crash, etc.)t > >s+ > >Story told by a co-worker some time ago:c > >:J > >Overheard at a party: a tech. support person on the phone to a computerJ > >operator: "It did, huh? Well, o.k. Just take a dump and I'll look at it > >in the morning. Bye." >eD > ... that one's right up there with "Take an aspirin and call me in> > the morning", although opposite ends of the intestinal tract$ > are taking the appropriate action! >N   ------------------------------  # Date: Fri, 24 Aug 2001 14:28:18 GMT 2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)6 Subject: Re: IA-64 Emulation (was: Re: Nits in Slides)2 Message-ID: <6wth7.724$bB1.31900@news.cpqcorp.net>  [ In article <cgabotcqvfvhlilmajf4vh7ebukagsgrvb@4ax.com>, jlsue <jlsuexxxz@home.com> writes:sG :On 23 Aug 2001 09:01:57 -0500, Kilgallen@SpamCop.net (Larry Kilgallen)u :wrote:t :sX :>In article <3B8504DE.7020704@compaq.com>, John Reagan <john.reagan@compaq.com> writes: :>> Larry Kilgallen wrote: :>>> S> :>>> Were those added for the sake of performance (after EV4),H :>>> or were they there from the beginning for the sake of correctness ? :>>>   :>> K :>> There were there in EV4 and my recollection was for correctness.  I do dJ :>> remember that the need for them came up late in the process after the L :>> VEST folks realized that they couldn't do something without having them. :>H :>Ok.  Andrew's original claim was that changes were introduced to AlphaF :>to make VESTed images perform better.  I am sure he was referring toE :>changes _after_ Alpha and Alpha VMS 1.0 were released to customers.  : C :True.  But then, none of us expect Andy to actually know what he'smC :talking about.  He'd much rather make claims on topics on which he G :could have no possible idea what the truth is.  It helps him spin FUD.s    I   The RC and RS instructions were incorporated in the Alpha Architecture rH   circa 1990, and reflect the nature of instruction stream notificationsJ   on Alpha -- highly asynchronous.  Based solely on the date, the changes E   would have been incorporated into the original EV4 microprocessors.   J   As for the applicability of notifications or the need for these or otherJ   special assists within IA-64, that remains to be determined.  Given the H   IA-64 features that are already available in support of the HP Ares PAK   emulator, it certainly appears quite feasible to have emulation on IA-64.y  L   Though certain folks here are likely already completely familiar with it, K   for those that are not I will mention that the HP Ares mechanisms within IK   HP IA-64 platforms permit HP Precision Architecture (PA RISC) applicationl(   code to execute on HP IA-64 platforms.  J   That said, I would expect most folks will be recompiling their code and H   will thus be operating entirely native.  Emulation is certainly usefulI   for certain applications and certain situations, but native code tends c<   to have better performance -- as well as other benefits...    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Fri, 24 Aug 2001 15:46:00 +0100t0 From: andrew harrison <andrew.nospam@uk.sun.com>6 Subject: Re: IA-64 Emulation (was: Re: Nits in Slides)* Message-ID: <3B8668A8.3459E5E2@uk.sun.com>   Hoff Hoffman wrote:t > ] > In article <cgabotcqvfvhlilmajf4vh7ebukagsgrvb@4ax.com>, jlsue <jlsuexxxz@home.com> writes:aI > :On 23 Aug 2001 09:01:57 -0500, Kilgallen@SpamCop.net (Larry Kilgallen)a	 > :wrote:h > :jZ > :>In article <3B8504DE.7020704@compaq.com>, John Reagan <john.reagan@compaq.com> writes: > :>> Larry Kilgallen wrote: > :>>>@ > :>>> Were those added for the sake of performance (after EV4),J > :>>> or were they there from the beginning for the sake of correctness ? > :>>> > :>>uL > :>> There were there in EV4 and my recollection was for correctness.  I doK > :>> remember that the need for them came up late in the process after thelN > :>> VEST folks realized that they couldn't do something without having them. > :>J > :>Ok.  Andrew's original claim was that changes were introduced to AlphaH > :>to make VESTed images perform better.  I am sure he was referring toG > :>changes _after_ Alpha and Alpha VMS 1.0 were released to customers.i > :hE > :True.  But then, none of us expect Andy to actually know what he's E > :talking about.  He'd much rather make claims on topics on which heeI > :could have no possible idea what the truth is.  It helps him spin FUD.o > J >   The RC and RS instructions were incorporated in the Alpha ArchitectureJ >   circa 1990, and reflect the nature of instruction stream notificationsK >   on Alpha -- highly asynchronous.  Based solely on the date, the changes-G >   would have been incorporated into the original EV4 microprocessors.y > L >   As for the applicability of notifications or the need for these or otherK >   special assists within IA-64, that remains to be determined.  Given thetJ >   IA-64 features that are already available in support of the HP Ares PAM >   emulator, it certainly appears quite feasible to have emulation on IA-64.l >   F Of course it is but HP developed the IA64 ISA and they made damn sure @ that they provided the instructions in IA64 to map onto from PA.  : This is a totally different situation to the one you find 6 youselves in and a totally different situation to the % one that you were in with VAX->Alpha.a  : It also depends on how you define feasible, x86 emulation : is known to be slow and there are benchmarks to prove it, 7 I don't think that HP have published any benchmarks to  : IA64 running PA binaries so the jury is still out on that.   Regardsi Andrew Harrisonh Enterprise IT Architectl   ------------------------------  % Date: Fri, 24 Aug 2001 12:35:45 -0400.- From: JF Mezei <jfmezei.spamnot@videotron.ca>-6 Subject: Re: IA-64 Emulation (was: Re: Nits in Slides), Message-ID: <3B868260.C54AF880@videotron.ca>   Hoff Hoffman wrote:bK >   That said, I would expect most folks will be recompiling their code andlJ >   will thus be operating entirely native.  Emulation is certainly usefulJ >   for certain applications and certain situations, but native code tends> >   to have better performance -- as well as other benefits...  N When APPLE migrated friom 68000 to PowerPC, it didn't have to wait to have theI whole OS ported before it was released since it had designed an automatic I on-the-fly emulator that worked for each individual component (eg: nativeg9 application using old emulated extension, or vice versa).   M If the VMS folks did the same thing, which would mean that they wouldn't havenL to be bothered initially with all the "fluff" and utilities, how much soonerH could VMS be released ? (with the followup version being fully native) ?   ------------------------------  # Date: Fri, 24 Aug 2001 16:39:27 GMT . From: "Duane Sand" <duane.sand@mindspring.com>6 Subject: Re: IA-64 Emulation (was: Re: Nits in Slides)< Message-ID: <3rvh7.8901$sa.3907766@news1.rdc1.sfba.home.com>   andrew harrison wrotee >n > Hoff Hoffman wrote: 	 > >   ...rH > >   As for the applicability of notifications or the need for these or othertI > >   special assists within IA-64, that remains to be determined.  Givenr theML > >   IA-64 features that are already available in support of the HP Ares PAH > >   emulator, it certainly appears quite feasible to have emulation on IA-64. >oG > Of course it is but HP developed the IA64 ISA and they made damn sure B > that they provided the instructions in IA64 to map onto from PA. >h; > This is a totally different situation to the one you findi7 > youselves in and a totally different situation to the ' > one that you were in with VAX->Alpha.m > ; > It also depends on how you define feasible, x86 emulationr; > is known to be slow and there are benchmarks to prove it,o8 > I don't think that HP have published any benchmarks to< > IA64 running PA binaries so the jury is still out on that.   See-C http://www.hp.com/products1/itanium/infolibrary/whitepapers/dct.pdfiI for an initial HP report on Aries' performance.  It is working quite wellhH enough to support HP's customer migration goals.  It doesn't have to runG fast enough to win benchmarketing wars. It's there to make whole-systemfE migrations practical while doing native recompiles of major apps at as convenient pace.   For more info on Aries, see-C http://www.computer.org/computer/articles/March/coverfeature300.htms, and http://devresource.hp.com/STK/index.html  A Emulation of Alpha code on IA64 will be simpler than emulation ofeB PA-RISC code, which has trickier decodes, condition codes, and regH side effects. And both are much much simpler than emulation of x86 code.  E The slow x86 emulation Andrew apparently refers to is the x86 decodermI on the initial Itanium chip.  This is a minimal-state hardware-only thingaK that has all of 1 or 2 machine cycles to analyze x86 code words.  I suspectiJ it doesn't leave any of its past decisions in the code cache.  This is notF anything like what can be accomplished by static or dynamic translatorF software or even a simple code-caching interpreter, or even Pentium 4.  = A more relevent example of x86 emulation is TransMeta's code-AD morphing firmware, converting hot traces of x86 code into VLIW code.A The performance of TransMeta is limited by the chip-vendor choiceE= to never save morphing decisions back onto disk files.  AlphaM1 translators would not have that silly constraint.e  B As to HP tailoring IA64 architecture to support PA-RISC emulation,E there was surprisingly very very little of that.   It looks like HP'sw: PlayDoh architects made minimal concessions to PA-RISC and< none to dynamic translation.  I spot two PA-RISC wartlettes:C the pointer-swizzle operator for a 4-segment 32-bit virtual addressa; space, and a 4-level ring model for proc calls.  And that's  apparently all.0   ------------------------------  + Date: Fri, 24 Aug 2001 05:45:59 -0700 (PDT)s. From: Fabio Cardoso <fabiopenvms@yahoo.com.br> Subject: Intersystems and Alphao@ Message-ID: <20010824124559.15048.qmail@web20210.mail.yahoo.com>   Click   , http://www.intersystems.com/cache/alpha.html   It is dated from Aug/8   Regardsr  
 F=E1bio C.  2 __________________________________________________ Do You Yahoo!?H Make international calls for as low as $.04/minute with Yahoo! Messenger http://phonecard.yahoo.com/    ------------------------------    Date: 24 Aug 2001 08:06:03 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen) 4 Subject: Re: Intersystems and Alpha and VAX and IA643 Message-ID: <tf0Oe8iH0Q3y@eisner.encompasserve.org>   q In article <20010824124559.15048.qmail@web20210.mail.yahoo.com>, Fabio Cardoso <fabiopenvms@yahoo.com.br> writes:8 > Click8 > . > http://www.intersystems.com/cache/alpha.html >  > It is dated from Aug/8 > 	 > Regardse >  > F=E1bio C.  B It would be much better to summarize such articles, as the ability@ to read comp.os.vms does not imply the ability to read the world	 wide web.M   =======   A The article says they will continue to support Alpha as they have ? continued to support VAX.  It says they have tentative plans to @ support Itanium-based VMS but must wait for more from Compaq for! final commitment (VMS and Tru64).    ------------------------------  % Date: Fri, 24 Aug 2001 08:29:57 -0400 - From: "Richard D. Piccard" <piccard@ohio.edu>04 Subject: Re: Introduction and BACKUP recommendations( Message-ID: <3B8648C6.7433FEDB@ohio.edu>  G Remember that the system disk backup can be directly restored to create E a bootable system disk ONLY if both the save and the restore are doneFH /IMAGE.  Learned that the hard way the first week-end my predecessor was4 showing me how to be a VMS system manager (in 1983)!  #                                 RDP0     Robin wrote:  @ > I've been lurking the past couple of weeks and decided to step
 > forward. >3E > I'm not completely new to OpenVMS.  I have been a programmer/DB Mgr B > on an OpenVMS system that we shared with other State users.  The> > system management/facility management had been outsourced to: > Northrup Grumman via a state-wide contract.  My agency'sD > powers-that-be decided to bring our stuff in-house and now I'm theB > proud possessor of an Alpha OpenVMS 7.3 server.  Suddenly I'm noD > longer a glorified user I'm now a system manager.  I've spent thatB > last 3 weeks moving applications and migrating data and I'm VERY > tired. >lD > Anyway, I was wondering if anyone could recommend that best way toC > do backups.  The consultant that did the initial install/setup of D > the server set up a little command procedure that works fine.  ButE > when I was digging into it further today, I noticed he is using thelC > /image switch and got to wondering if that's really what we want.t= > From what I gather, I guess this means that if a restore isp@ > necessary the entire image must be restored.  If I only need aC > single file, this would be a pain to "back rev" the entire system/2 > for one file.  Anyone have any opinions on this? >o: > Also, we are running volume shadowing that was installedA > last-minute, so if that backup procedure needs to be changed toaD > reflect this I don't think the consultant did it.  Any wisdom here > would be appreciated also. >r > RobinU > Austin, TX   --B ==================================================================B Dick Piccard                           Academic Technology ManagerB piccard@ohio.edu                                 Computer ServicesB http://oak.cats.ohiou.edu/~piccard/                Ohio University   ------------------------------    Date: 24 Aug 2001 12:51:45 -0500+ From: kuhrt@encompasserve.org (Marty Kuhrt)l8 Subject: Re: Looking for a good alert management utility3 Message-ID: <zzgQxkbZmvbI@eisner.encompasserve.org>y  i In article <80125384.0108231137.425acec4@posting.google.com>, linuxmtl@yahoo.com (Moi Je Le Sais) writes:  > Good day,s > H > I'm currently looking for a third party product that would simply sendH > all the AUDIT$SERVER and OPCOM alerts to and e-mail (or alias) address > ORG > send all those alerts to our main log server (using syslog) and mergen) > the messages OR any good idea to do so.2 >   B A company I used to work for, Symark, has a product called WatcherA which can intercept and act on audit$server messages.  Check out  ! http://www.symark.com/watcher.htmh  5 No endorsement implied, I just know of its existence.a   ------------------------------  % Date: Fri, 24 Aug 2001 08:21:03 +0200o< From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <noone@home.com>1 Subject: Re: LPR from VMS 7.2-1 to Win2000 Server-( Message-ID: <3B85F24F.3193C062@home.com>   OK.16 Using the TELNET print symbiont is something completly! different from using the LPR/LPD.-1 You have first to descide what print protocoll tor4 use, then set it up at both ends in the correct way.7 Personlay, I'd prefer to use LPR/LPD instead of TELNET.u   Jan-Erik Sderholm.    "Michael D. Ober" wrote: > 	 > Thanks,r > K > The ports aren't relevent, except that TELNETSYM requires port numbers toe5 > print to IP printers and I had never used LPRSETUP.h > -- > Mike Ober.   ------------------------------  % Date: Fri, 24 Aug 2001 07:05:53 -0600a4 From: "Michael D. Ober" <mdo.@.wakeassoc.com.nospam>1 Subject: Re: LPR from VMS 7.2-1 to Win2000 Server<1 Message-ID: <Rish7.117$JO2.54657@news.uswest.net>   J LPR is definitely easier, but Compaq/DEC recommends the TelnetSym as being more robust. --
 Mike Ober.  6 "Jan-Erik Sderholm" <noone@home.com> wrote in message" news:3B85F24F.3193C062@home.com... > OK.v8 > Using the TELNET print symbiont is something completly# > different from using the LPR/LPD.l3 > You have first to descide what print protocoll to 6 > use, then set it up at both ends in the correct way.9 > Personlay, I'd prefer to use LPR/LPD instead of TELNET.a >. > Jan-Erik Sderholm.s >o > "Michael D. Ober" wrote: > >t > > Thanks,> > >hJ > > The ports aren't relevent, except that TELNETSYM requires port numbers to7 > > print to IP printers and I had never used LPRSETUP.e > > -- > > Mike Ober.   ------------------------------  % Date: Fri, 24 Aug 2001 19:01:29 +0200 < From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <noone@home.com>1 Subject: Re: LPR from VMS 7.2-1 to Win2000 Serverp( Message-ID: <3B868869.71C37A72@home.com>  9 Hm, I'v not seen that recomendation. My point was that byi8 using "standard" LPR/LPD, you can, just by adjusting the5 princap file, select just about anything to print at.t8 A printer-server, some NT box, a Unix box or another VMS? system. Im not sure thet telnet printing have this flexibility.f  D And the question was if an NT box can support telnet printing, not ?2 I'm not sure, I have never seen one doing that....   Jan-Erik Sderholm.i   "Michael D. Ober" wrote: > L > LPR is definitely easier, but Compaq/DEC recommends the TelnetSym as being > more robust. > -- > Mike Ober. >t   ------------------------------  % Date: Fri, 24 Aug 2001 18:05:55 +0100c0 From: andrew harrison <andrew.nospam@uk.sun.com>( Subject: Re: More Alpha rubbish in print* Message-ID: <3B868973.F68E8286@uk.sun.com>   Arne Vajh=F8j wrote: > =r   > Richard Tomkins wrote:J > > The point I was stumbling towards was that 2 good word processors and=  58.J > > lame word processors for a system in comparison to 2 good word proces= sorsJ > > and 3 lame word processors for another system, does not make the syst= em with 6 > > 60 word processors better than the one that has 5. > >eJ > > There is no need to have thousands of applications and programs for a=  J > > system. There is just a need to have good quality applications andsol= utionsJ > > available, and in that department, across all those different Operati= ngJ > > Systems, there were some good quality applications and solutions avai= lable. > =a  = > Noone is questioning the quality of the available software.t > =o  A > We are questioning, whether the software people want to use aree > available. > =   J > You mentioned word-processors yourself. The 3 big one is MS Word, Lotus=  A > Word and Corel Wordperfect. Which are available for Alpha (in ah > recent version !) ?c > =s    9 According to the latest windows survey StarOffice now haso7 15% of the Office market ahead of Corel and Lotus but =m   behind MS Office..   Regards  Andrew Harrisons Enterprise IT Architectl   ------------------------------  % Date: Fri, 24 Aug 2001 12:22:56 +0100i0 From: andrew harrison <andrew.nospam@uk.sun.com> Subject: Re: Nits in Slidesa* Message-ID: <3B863910.DE584C1E@uk.sun.com>   jlsue wrote: > H > On 23 Aug 2001 09:01:57 -0500, Kilgallen@SpamCop.net (Larry Kilgallen) > wrote: > Y > >In article <3B8504DE.7020704@compaq.com>, John Reagan <john.reagan@compaq.com> writes:i > >> Larry Kilgallen wrote:w > >>>r? > >>> Were those added for the sake of performance (after EV4),iI > >>> or were they there from the beginning for the sake of correctness ?s > >>>n > >>K > >> There were there in EV4 and my recollection was for correctness.  I dotJ > >> remember that the need for them came up late in the process after theM > >> VEST folks realized that they couldn't do something without having them.o > >cI > >Ok.  Andrew's original claim was that changes were introduced to Alpha-G > >to make VESTed images perform better.  I am sure he was referring tolF > >changes _after_ Alpha and Alpha VMS 1.0 were released to customers. > D > True.  But then, none of us expect Andy to actually know what he'sD > talking about.  He'd much rather make claims on topics on which heH > could have no possible idea what the truth is.  It helps him spin FUD.    6 Really, given your posting record you really should be3 more carefull about accusing people of not knowing g. anything about the topics they are posting on.  0 But back to your point. Byte alligned access was. added to Alpha after the first implimentation + because of performance issues when running W8 VAX and other apps on Alpha. While this probably wasn't 7 added to specifically improve VESTED performance it was 2 also added to improve the performance of FX!32 as 7 well as if my memory is correct Cobol and it would haveo$ had an impact on VESTED performance.   Back to you buddy.  c Regards  Andrew Harrisonf Enterprise IT Architecto   ------------------------------    Date: 24 Aug 2001 08:01:24 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)A Subject: Re: Nits in Slidese3 Message-ID: <pYNMkfaCKmiu@eisner.encompasserve.org>>  ] In article <3B863910.DE584C1E@uk.sun.com>, andrew harrison <andrew.nospam@uk.sun.com> writes:o  2 > But back to your point. Byte alligned access was0 > added to Alpha after the first implimentation - > because of performance issues when running a: > VAX and other apps on Alpha. While this probably wasn't 9 > added to specifically improve VESTED performance it wasc   Even though you claimed it was.h   ------------------------------  % Date: Fri, 24 Aug 2001 15:03:59 +0100r0 From: andrew harrison <andrew.nospam@uk.sun.com> Subject: Re: Nits in Slidesd* Message-ID: <3B865ECF.AD73D69D@uk.sun.com>   Larry Kilgallen wrote: > _ > In article <3B863910.DE584C1E@uk.sun.com>, andrew harrison <andrew.nospam@uk.sun.com> writes:t > 4 > > But back to your point. Byte alligned access was1 > > added to Alpha after the first implimentation . > > because of performance issues when running; > > VAX and other apps on Alpha. While this probably wasn'tl; > > added to specifically improve VESTED performance it wasd > ! > Even though you claimed it was.r  2 Did I, I don't think you will find I did you seem 2 to be suffering from an attack of the Jlsue's, he 4 claimed I said that support had been added to Alpha 4 to improve the performance of VESTED apps after the 5 first release. If you read my response to Mr Hoffmansr+ posts you will see that I did not say that.p   Here is what I postedoD "It does however as you know perfectly well have support for VESTED C applications specifically the RC and RS instructions, strange that -D you forgot to mention this. It also has support for byte allignment 0 because certain VAX apps ran poorly without it."  4 "Apart from helping out the OS, wow so you entirely 0 discount providing support for the OS as being a3 usefull feature set that it might just be importantr4 to have in the new processor architecture. You also 1 forgot the things that were left out of the first / implimentations of Alpha that were added later ,0 specifically to improve the performance of apps . coming from VAX which also effected any VESTED apps."  9 Does either of these paragraphs say that byte allignemente5 was added specifically to improve VESTED performance.e  3 Yet again jlsues lets himself down, I am only sorryo! that he has let you down as well.t  ? You would be better off being a bit more critical of his claims 6 it is after all something of a speciality on his part.  6 But who really cares anyway. The fact is that support 3 for VESTED apps was specifically added to Alpha and 6 that byte aligned access was also added later because 2 it was discovered that some VAX apps ran badly on  Alpha.  2 If you compare this with Alpha->IA64 you find the  following differences.  4 1. Intel is not going to add any support to Mckinley5 assuming Mckinley is the target CPU for OpenVMS FCS,  2 to enable emulated Alpha apps to run well on IA64.  . 2. I would not take bets on them adding Alpha 4 friendly instructions to any post Mckinley processor either.t   Regards- Andrew Harrisont Enterprise IT Architect    ------------------------------  % Date: Fri, 24 Aug 2001 15:12:14 +0100A0 From: andrew harrison <andrew.nospam@uk.sun.com> Subject: Re: Nits in Slidesn* Message-ID: <3B8660BE.2F172C16@uk.sun.com>   Larry Kilgallen wrote: > X > In article <3B8504DE.7020704@compaq.com>, John Reagan <john.reagan@compaq.com> writes: > > Larry Kilgallen wrote: > >>> > >> Were those added for the sake of performance (after EV4),H > >> or were they there from the beginning for the sake of correctness ? > >> > >hJ > > There were there in EV4 and my recollection was for correctness.  I doI > > remember that the need for them came up late in the process after theaL > > VEST folks realized that they couldn't do something without having them. > H > Ok.  Andrew's original claim was that changes were introduced to AlphaF > to make VESTed images perform better.  I am sure he was referring toE > changes _after_ Alpha and Alpha VMS 1.0 were released to customers.a    & Read my posts that is not what I said.  7 I said RC and RS were added to Alpha to improve VESTED g4 performance and byte alligned access was added later8 which would also help VESTED performance allthough there& were other reasons for its additions.      -- n Andrew Harrisonr Enterprise IT Architect    ------------------------------  % Date: Fri, 24 Aug 2001 15:16:27 +0100S0 From: andrew harrison <andrew.nospam@uk.sun.com> Subject: Re: Nits in Slides * Message-ID: <3B8661BB.4E8686A0@uk.sun.com>   jlsue wrote: > H > On 23 Aug 2001 09:01:57 -0500, Kilgallen@SpamCop.net (Larry Kilgallen) > wrote: > Y > >In article <3B8504DE.7020704@compaq.com>, John Reagan <john.reagan@compaq.com> writes:  > >> Larry Kilgallen wrote:  > >>>i? > >>> Were those added for the sake of performance (after EV4),TI > >>> or were they there from the beginning for the sake of correctness ?n > >>>  > >>K > >> There were there in EV4 and my recollection was for correctness.  I dolJ > >> remember that the need for them came up late in the process after theM > >> VEST folks realized that they couldn't do something without having them.  > >eI > >Ok.  Andrew's original claim was that changes were introduced to AlphasG > >to make VESTed images perform better.  I am sure he was referring tosF > >changes _after_ Alpha and Alpha VMS 1.0 were released to customers. > D > True.  But then, none of us expect Andy to actually know what he'sD > talking about.  He'd much rather make claims on topics on which heH > could have no possible idea what the truth is.  It helps him spin FUD.  ? Since when were you a reliable source of facts and truth ??????o  = Oh I forgot, pre the 25th is now pre-history which we should - all expunge from our memories.  = If you had an exemplary posting record then I might read any v= specific criticism you might be prepared to make and I might   take your criticism on-board.2  C As it is with your track record you don't really even deserve this h	 response.t   Regards  Andrew Harrisond Enterprise IT Architect_   ------------------------------  # Date: Fri, 24 Aug 2001 14:40:20 GMTa2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: Nits in Slidesh2 Message-ID: <oHth7.727$bB1.31883@news.cpqcorp.net>  ] In article <3B863910.DE584C1E@uk.sun.com>, andrew harrison <andrew.nospam@uk.sun.com> writes:t1 :But back to your point. Byte alligned access wasi/ :added to Alpha after the first implimentation 0, :because of performance issues when running 9 :VAX and other apps on Alpha. While this probably wasn't l8 :added to specifically improve VESTED performance it was3 :also added to improve the performance of FX!32 as =8 :well as if my memory is correct Cobol and it would have% :had an impact on VESTED performance.m  D   Most code out there that I am aware of does not use the byte-word K   instructions.  OpenVMS, for instance, is built without byte-word enabled.eF   (The latest C compilers do use amask-protected sequences containing F   byte-word instructions within the tight loops, but you have to have &   rebuilt with very recent compilers.)  G   Folks that do have performance-critical code -- and particularly coderD   with byte- or word-aligned data access -- are using the byte-word H   instructions.  But I would be quite surprised to learn these are same 2   applications that are also using FX!32 nor VEST.  E   Byte-word is only available (in hardware) on EV56 and later.  As anIE   interestingly tangental discussion, the byte-word instructions are  ,   emulated on earlier Alpha implementations.  G   Having just been looking at the VEST code for reasons other than thisbJ   discussion, it is sufficiently old that it has no clue of the existence :   of the byte-word instructions in the Alpha Architecture.    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------    Date: 24 Aug 2001 09:43:07 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)t Subject: Re: Nits in Slides'3 Message-ID: <dF40+WPcnkzf@eisner.encompasserve.org>y  ] In article <3B8660BE.2F172C16@uk.sun.com>, andrew harrison <andrew.nospam@uk.sun.com> writes:s >  > Larry Kilgallen wrote: >> tY >> In article <3B8504DE.7020704@compaq.com>, John Reagan <john.reagan@compaq.com> writes:o >> > Larry Kilgallen wrote:a >> >>c? >> >> Were those added for the sake of performance (after EV4),aI >> >> or were they there from the beginning for the sake of correctness ?e >> >>m >> >K >> > There were there in EV4 and my recollection was for correctness.  I dovJ >> > remember that the need for them came up late in the process after theM >> > VEST folks realized that they couldn't do something without having them.  >> aI >> Ok.  Andrew's original claim was that changes were introduced to AlphaeG >> to make VESTed images perform better.  I am sure he was referring to F >> changes _after_ Alpha and Alpha VMS 1.0 were released to customers. >  > ( > Read my posts that is not what I said. > 9 > I said RC and RS were added to Alpha to improve VESTED  
 > performance   ? And John Reagan has indicated they were put in for correctness, ? not for performance.  Steve Hoffman says in 1990, so aside from D DEC employees who had access to a Manniquin Alpha Emulator (pre-EV3)* "added" does not even seem the right word.   ------------------------------  % Date: Fri, 24 Aug 2001 16:00:51 +0100y0 From: andrew harrison <andrew.nospam@uk.sun.com> Subject: Re: Nits in Slideso* Message-ID: <3B866C23.55A83E94@uk.sun.com>   Larry Kilgallen wrote: > _ > In article <3B8660BE.2F172C16@uk.sun.com>, andrew harrison <andrew.nospam@uk.sun.com> writes:e > >  > > Larry Kilgallen wrote: > >>[ > >> In article <3B8504DE.7020704@compaq.com>, John Reagan <john.reagan@compaq.com> writes:i > >> > Larry Kilgallen wrote:  > >> >> A > >> >> Were those added for the sake of performance (after EV4), K > >> >> or were they there from the beginning for the sake of correctness ?. > >> >>o > >> >M > >> > There were there in EV4 and my recollection was for correctness.  I do L > >> > remember that the need for them came up late in the process after theO > >> > VEST folks realized that they couldn't do something without having them.s > >>K > >> Ok.  Andrew's original claim was that changes were introduced to AlphamI > >> to make VESTed images perform better.  I am sure he was referring tocH > >> changes _after_ Alpha and Alpha VMS 1.0 were released to customers. > >a > >e* > > Read my posts that is not what I said. > >N: > > I said RC and RS were added to Alpha to improve VESTED > > performance  > A > And John Reagan has indicated they were put in for correctness,sA > not for performance.  Steve Hoffman says in 1990, so aside from F > DEC employees who had access to a Manniquin Alpha Emulator (pre-EV3), > "added" does not even seem the right word.   Added is the right word.   Read Johns post.H "There were there in EV4 and my recollection was for correctness.  I do F remember that the need for them came up late in the process after the B VEST folks realized that they couldn't do something without having them."  @ Nor does "they couldn't do something without having them" imply ; just correctness, not being able to do a specific function   could be infinitely slow.R     Regards  Andrew Harrisonc Enterprise IT Architectt   ------------------------------  % Date: Fri, 24 Aug 2001 13:36:29 -0400d5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>r Subject: Re: Nits in Slidesl2 Message-ID: <7gwh7.749$bB1.32396@news.cpqcorp.net>  B andrew harrison wrote in message <3B863910.DE584C1E@uk.sun.com>... >  >r1 >But back to your point. Byte alligned access was . >added to Alpha after the first implimentation+ >because of performance issues when running08 >VAX and other apps on Alpha. While this probably wasn't8 >added to specifically improve VESTED performance it was2 >also added to improve the performance of FX!32 as8 >well as if my memory is correct Cobol and it would have% >had an impact on VESTED performance.y >j    % As my Uncle used to say, horse pucky.   K Partial word access was added explicitly because of NT, and the use of byteS( access both to memory *and* to IO space.   ------------------------------  % Date: Fri, 24 Aug 2001 04:12:17 -0400e' From: "Bill Todd" <billtodd@foo.mv.com>eB Subject: Re: Nits in Slides (was: Re: The Final Knell Has Sounded)( Message-ID: <9m528a$e6r$1@pyrite.mv.net>  - "jlsue" <jlsuexxxz@home.com> wrote in messageb2 news:h99bot8vcq0lv2godv2etr6qtmf1h55ocs@4ax.com...   ...-  D > The initial IA64 may be slower than EV7 on translated code, but it > won't stay that way forever.  K Er, the initial IA64 (Merced) is already out, and is slower than *EV6* when6H both are running *native* code (with the possible exception of unusuallyK regular floating-point-intensive code).  That Merced's successor (McKinley,.J due in the EV7 time frame) will be similarly slower than EV7 when both areL running native code seems likely (if only due to EV7's on-chip glue for bothJ memory and SMP), so there's no chance in hell that McKinley will *emulate*H EV7 at anything like native speed.  And the same situation (or worse forG IA64, with the addition of SMT to the mix) would likely have applied ina< EV8's era, had EV8 not been traded for a bag of magic beans.  I In other words, Alpha's native performance would have remained noticeablypL superior to IA64's native performance for the foreseeable future, had CompaqL chosen to continue developing Alpha.  And IA64 performance *emulating* AlphaL code likely won't catch up to EV7's until well after Compaq stops giving EV7 process upgrades.h   - bill   ------------------------------  % Date: Fri, 24 Aug 2001 18:53:36 +0010o% From: paddy.o'brien@zzz.tg.nsw.gov.au B Subject: Re: Nits in Slides (was: Re: The Final Knell Has Sounded)5 Message-ID: <01K7ITM3WHB6004A6W@tgmail.tg.nsw.gov.au>    Bill Todd wrote:  . >"jlsue" <jlsuexxxz@home.com> wrote in message3 >news:h99bot8vcq0lv2godv2etr6qtmf1h55ocs@4ax.com...  >, >... >0E >> The initial IA64 may be slower than EV7 on translated code, but itu >> won't stay that way forever.g  ? In 10/20 years time, this has to be true if EV7 is end-of-line.u  L >Er, the initial IA64 (Merced) is already out, and is slower than *EV6* whenI >both are running *native* code (with the possible exception of unusuallyc( >regular floating-point-intensive code).  K This possible exception I would be very interested in.  My production ES40 >L with 4 CPUs runs at 400% almost 24x7, possibly weekends are generally 300%, J but still often 400%.  This is good old floating point number crunching.  J Our engineers on the production machine will **NOT** be interested in any E downgrade.  Will I be able to test my real applications (forget SPEC o' whatever) before I buy an IA64 machine?o   [Rest snipped]   Regards, Paddy   Paddy O'Brien, Transmission Development,s
 TransGrid, PO Box A1000, Sydney South,  NSW 2000, Australiaf& (Street address, 201 Elizabeth Street)     Tel:   +61 2 9284-3063 Fax:   +61 2 9284-3050& Email: paddy.o'brien@zzz.tg.nsw.gov.au  F Either "\'" or "\s" (to escape the apostrophe) seems to work for most  people,n; but that little whizz-bang apostrophe gives me little spam.    ------------------------------    Date: 24 Aug 2001 05:59:57 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)iB Subject: Re: Nits in Slides (was: Re: The Final Knell Has Sounded)3 Message-ID: <EcMZOnzBhM17@eisner.encompasserve.org>r  ] In article <01K7ITM3WHB6004A6W@tgmail.tg.nsw.gov.au>, paddy.o'brien@zzz.tg.nsw.gov.au writes:i > Bill Todd wrote: > / >>"jlsue" <jlsuexxxz@home.com> wrote in messagee4 >>news:h99bot8vcq0lv2godv2etr6qtmf1h55ocs@4ax.com... >> >>...l >>F >>> The initial IA64 may be slower than EV7 on translated code, but it  >>> won't stay that way forever. > A > In 10/20 years time, this has to be true if EV7 is end-of-line.- > M >>Er, the initial IA64 (Merced) is already out, and is slower than *EV6* when.J >>both are running *native* code (with the possible exception of unusually) >>regular floating-point-intensive code).e > M > This possible exception I would be very interested in.  My production ES40 oN > with 4 CPUs runs at 400% almost 24x7, possibly weekends are generally 300%, L > but still often 400%.  This is good old floating point number crunching.  L > Our engineers on the production machine will **NOT** be interested in any G > downgrade.  Will I be able to test my real applications (forget SPEC  ) > whatever) before I buy an IA64 machine?r  A Presumably by then you should have upgraded to an ES45 or to someu8 EV7-based machine (given your present lack of headroom).   ------------------------------  % Date: Fri, 24 Aug 2001 13:05:06 +010060 From: andrew harrison <andrew.nospam@uk.sun.com>B Subject: Re: Nits in Slides (was: Re: The Final Knell Has Sounded)* Message-ID: <3B8642F2.BC94E2D5@uk.sun.com>   jlsue wrote: > 5 > On Wed, 22 Aug 2001 15:45:24 +0100, andrew harrison # > <andrew.nospam@uk.sun.com> wrote:e >  > >Emulation/translation.e > >s6 > >Firstly some emulation/translation is faster on the6 > >new platform than it was on the native system. This1 > >for example was true for some x86 apps runninga4 > >translated on Alpha using FX!32. This was because1 > >the Alpha was a faster processor than the x86.m > >o: > >However running translated or emulated code on the IA646 > >is particularly likely to show big performance hits6 > >because of the the huge dependency that IA64 has on > >the compiler to support ILP.p > E > Oh geeze.  Here we go again, Andrew's making things up (conjecture)aF > and spouting them as facts.  It's not that your opinions aren't well< > thought out, it's just that they're stated as known facts. >   5 Oh geeze here we go again another ill considered and t0 badly thought out posting from one of the choir.  E > Andy, you'll just have to wait for the end product like the rest ofr > us.e > E > P.S. -- A translater can work somewhat like a compiler.  It doesn'tm* > have to be interpretive, like FX!32 was. >   8 Correct me if I am wrong but I was under the impression & FX!32 could produce native Alpha code.   > >s7 > >The effects of not having natively compiled code area6 > >already known with some x86 apps running at Pentium5 > >100 Mhz speeds on an 800 Mhz IA64 CPU. Most peopleh* > >would consider this to be unacceptable. > >b8 > >It is also not possible to draw parallels between the6 > >experience when moving from VAX to Alpha since care5 > >was taken in the Alpha architecture to allow this.  > >D6 > >This has not happened for Alpha->IA64 because Intel: > >never made any provision for this in their architecture7 > >as is obvious from the some of the posts refering toa > >levels of protection etc. > >h5 > >Now with IA64 you are exposing your customers to aa5 > >triple whammy which you didn't expose them to withr > >VAX->Alpha. > C > Conjecture.  You have *no* idea what we're exposing customers to.u; > Unless you secretly work in the OpenVMS engineering team.  >   3 If you had bothered reading the posts from your ownw7 engineering team you would realise that they also don't-4 yet have a clear view on what you are exposing your 1 customers to and they had not idea whatsoever on  4 the day of the announcement. The interveaning period! has been a mad scramble for them.e  . Because of that any statements you make about 4 EV7->IA64 any timeframes, any claims of performance,. and suggestions about ISV porting issues, any - suggestions concerning emulation/translation ,, of existing code is pure conjecture on your  part.s    D > The initial IA64 may be slower than EV7 on translated code, but it > won't stay that way forever. >   B The inital IA64 will be slower than EV7 on translated code unless > you somehow manage to make the translator more efficient than A native compilation, this is because except for SPECfp the initialn' IA64 is slower than the current 6/833. E   regards  Andrew HarrisonI Enterprise IT Architect    ------------------------------  # Date: Fri, 24 Aug 2001 15:10:25 GMTa  From: jlsue <jlsuexxxz@home.com>B Subject: Re: Nits in Slides (was: Re: The Final Knell Has Sounded)8 Message-ID: <ndrcot0pp9ag5a5p1fv0imq62hl9h74041@4ax.com>  B Er, Bill, I didn't make that statement.  Something happened in the cut-n-paste I think.    E On Fri, 24 Aug 2001 04:12:17 -0400, "Bill Todd" <billtodd@foo.mv.com>- wrote:   >-. >"jlsue" <jlsuexxxz@home.com> wrote in message3 >news:h99bot8vcq0lv2godv2etr6qtmf1h55ocs@4ax.com...c >R >... >nE >> The initial IA64 may be slower than EV7 on translated code, but it  >> won't stay that way forever., >lL >Er, the initial IA64 (Merced) is already out, and is slower than *EV6* whenI >both are running *native* code (with the possible exception of unusuallyyL >regular floating-point-intensive code).  That Merced's successor (McKinley,K >due in the EV7 time frame) will be similarly slower than EV7 when both areoM >running native code seems likely (if only due to EV7's on-chip glue for bothsK >memory and SMP), so there's no chance in hell that McKinley will *emulate*nI >EV7 at anything like native speed.  And the same situation (or worse fortH >IA64, with the addition of SMT to the mix) would likely have applied in= >EV8's era, had EV8 not been traded for a bag of magic beans.r >mJ >In other words, Alpha's native performance would have remained noticeablyM >superior to IA64's native performance for the foreseeable future, had CompaqAM >chosen to continue developing Alpha.  And IA64 performance *emulating* AlphaoM >code likely won't catch up to EV7's until well after Compaq stops giving EV7s >process upgrades. >  >- bill  >  >e   ------------------------------  # Date: Fri, 24 Aug 2001 15:09:08 GMTn  From: jlsue <jlsuexxxz@home.com>B Subject: Re: Nits in Slides (was: Re: The Final Knell Has Sounded)8 Message-ID: <4arcotkvi7lu9d6l4otak9m8mng13k00bu@4ax.com>  , On Thu, 23 Aug 2001 21:32:45 -0400, JF Mezei% <jfmezei.spamnot@videotron.ca> wrote:r  
 >jlsue wrote: D >> Conjecture.  You have *no* idea what we're exposing customers to. >dM >That is the biggest problem with  Compaq. It publicly engaged itself to IA64uK >before really knowing what it was getting into. That is like jumping off ajO >very very high cliff, thinking you will have plenty of time on the way down ton! >think of a way to land smoothly.c  A I don't know how you can say they don't know what they're gettingoD into.  They may have made assumptions that the OSs could cross over,B but from a hardware perspective I haven't seen anything to support that statement.    ------------------------------  % Date: Fri, 24 Aug 2001 12:22:33 +0100E( From: Nic Clews <sendspamhere@127.0.0.1>> Subject: Re: Nuts-n-bolts in News (was: Re: Nits in Slides...)) Message-ID: <3B8638F9.E7B1BAAA@127.0.0.1>o   Bill Todd wrote: > 7 > "Nic Clews" <sendspamhere@127.0.0.1> wrote in messageh% > news:3B84FBE5.B6253678@127.0.0.1...  > > andrew harrison wrote: >  > ...d > : > > >It does but any otimism you might want to inject into8 > > >this discussion needs to be countered with the fact9 > > >that Intel have tried as much as they can to improvet9 > > >the performance of the x86 emulation but that it haso< > > >proved to be very hard to emulate x86 and other non ILP' > > >binaries into an ILP architecture.n > >t( > > Thats like saying that a performance7 > > car manufacturer is busy altering valve timings andF@ > > tweaking inlet pressures, when you're fuelling it with dregsA > > of French wine and measuring actual performance against that.y > >rA > > The VMS guys are software engineers, I'd hardly describe what 9 > > Micro$oft cobble together as 'engineered', would you?t > F > Your point escapes me.  Andrew appears to have been referring to theE > difficulty Intel has obtaining performance from IA32 code under itsrN > emulation on IA64, and suggesting that similar difficulties would hamper theI > performance of IA64's emulation of Alpha code.  The quality of the codepE > being emulated is not of apparent relevance to the emulation speed.-  H Sorry I left the above in as a big quote but its in context. When 'user'G code runs on a processor, any processor, under control of the operating0F system, the operating system still has ultimate control. The operatingF system, could, if 'it' wanted, change what the user code was doing, orG trying to do. Its running in a virtual machine, having a self modifyingw8 virtual machine is not beyond the realms of possibility.  ? VMS works this way already to a degree. I recall some years agooD installing a floating point processor into an VAX 750. Did I have toE recompile? Nope. Did it 'go' any faster? Well no it didn't but that's H not the point. The operating system took care of what the hardware couldE or could not do. VAX chips actually varied quite a lot, probably moree< than Alphas do, but the operating system just did its thing.  E I do however accept your [and Andrews] argument that there could be a B problem, but if such a problem did manifest itself, then it is notD necessarily the end of the line for VMS. Perhaps I am also inferringF that things can be done, when they can't. However, if anyone could, myH money would be on OpenVMS engineering achieving it, rather than the guysC working for a certain other company we've mentioned more than once.p   > ...rA > > Quite a number of implementations we have don't even need thewC > > EV6 performance, in the majority of cases it won't be an issue.sB > > EV7 may be too fast, and it won't be the first time I've had a+ > > complaint that a systems is *too* fast.e > L > Perhaps you're confused about Alpha's bread-and-butter market segment thatK > brings in the bucks.  It's the server segment, where you'll find that the:J > better the performance of each individual processor (and assuming decentM > memory latency/bandwidth, which EV7 has in spades but which I've yet to seenN > any comparable mention of in IA64's future - ever), the fewer processors youD > need (and the smaller the server you need) to satisfy a given load
 > acceptably.t  D HPTC or just general transaction shifting? A programmer can optimise4 their code before it even hits the compiler (as if).  C I don't practice martial arts, but I believe, in Judo, you use your>H oponents 'attack' as their weakness. Even if Alpha had proven advantagesF over IA64, it is not unfeasible that one could change the way you workH with your chip to take advantage of its features. Alpha was not designedG as VMS's chip, VMS just happens to work quite well on it. As I've said, B some kernel tweaks have been made for NUMA, I expect when the dustH settles, it'll be tweaked for IA64 (though your and my interpretation of 'tweak' may differ).   > ...e > E > > VMS is currently undergoing a transformation that could level the  > > playing field. > M > More likely, VMS has just undergone a market transformation that will levela* > VMS (and possibly Compaq along with it).  D Well we hope not. Marketing, good marketing would be helpful at this point. Nuff said.   B > > the chip features, they may as well just be serial processors.B > > When do application writers ever listen to software engineers? > K > You talk as if Alpha's market was desktop systems, where at least much ofdN > the time only a single instance of a single application is actively running.G > In Alpha's main market, 'the majority of commercial applications' are H > completely irrelevant (unless something like MTS is running a bunch ofI > instances in parallel, in which case SMT would have worked just great).l  A Nope, I'm talking big servers. Take the Apache code base. Its not-H optimised for anything. ISVs typically will try to have a 'similar' codeB base with the necessary '.H' files. Optimised for what? OK to someF degree its up to the compiler writers, but they can't work miracles. I? wonder how many copies of the "Programming Concepts" manual lie D undisturbed because some "I am" programmer decides (s)he knows best?  D I've had many performance battles with programmers, all the way fromH distributed locking issues, through IO, system services through to thoseF that think "Mr Jackson" was a really cool bloke for saying code should be neat, not efficient.3  F (Writing a bit of JSP certainly gives the page fault handler something to do).E   > ...o > : > > >Both HP-UX and Tru64 will have much larger catalogues5 > > >of natively compiled apps than OpenVMS so if you : > > >have to rely on emulation it had better be real fast. > >-: > > What is there to stop VMS creating a native 'shell' in$ > > which to run these applications? > A > Project funding, for a start.  Even if VMS winds up running COE K > applications, that's only a subset (possibly a very small subset, but I'm.N > not sufficiently conversant with the area to know) of the application set onM > HP/UX and Tru64 (the former of which will of course also have a 3-year heado' > start on the IA64 hardware platform).t  D I think I mentioned somewhere in my previous post that someone wouldD have to authorize/fund whatever time was spent making this possible.  F I get the feeling you've not really seem what emulators and shells areF capable of. What would you like your PC to do today? Be a playstation,E gameboy, Sega megdrive, BBC, VAX, deep space telescope? I may be sad,u7 but even I'm impressed with Sonic the hedgehog on a PC.   G I'm hinting of course that it might be a nice idea that VMS has some ofa> these hooks in place, even if they are not used right now. VMS@ engineering are more or less at the right stage for this. If theF abstraction layer is positioned just right, I could have "Sonic" under VMS.   >  All of a sudden the > > balance shifts...4 > K > Oh, really?  Just why do you believe that people will prefer to run theirdK > Unix applications on VMS unless they already feel they need VMS for otheraE > reasons?  Especially if Compaq's 'commitment' to VMS doesn't changedC > dramatically (which at the moment there's absolutely no sign of)?  > N > Saying that 'the balance shifts' seems a ridiculously strong statement if itM > only means that VMS becomes marginally less unacceptable to the market as ag > whole.  B OK, what I meant to say is that the balance has the opportunity to shift.  H What I can say is that we've had quite a bit of interest from one of our: clients that this is happening [the port of VMS to Intel].  i > ...w > G > > I for one really appreciate the time they are taking out discussingV; > > what they are doing here, a personal thank-you to them.d > K > The VMS engineers have both my respect (at least as long as they stick tooN > engineering and don't try to spread Compaq's misinformation, however sincereN > their motivation) and my sympathy.  While they can scarcely be enjoying thisM > discussion, I hope they can understand that the criticisms are not aimed atl > them.0  G Similarly I respect your views because I can't necessarily see what yout can, from where I'm stood.  
 Regards, Nic.t   -- n( Regards, Nic Clews CSC Computer Sciences nclews at csc dot come   ------------------------------  % Date: Fri, 24 Aug 2001 12:59:16 +0100a0 From: andrew harrison <andrew.nospam@uk.sun.com>> Subject: Re: Nuts-n-bolts in News (was: Re: Nits in Slides...)* Message-ID: <3B864194.91B107A4@uk.sun.com>   Bill Todd wrote: > 7 > "Nic Clews" <sendspamhere@127.0.0.1> wrote in messaget% > news:3B84FBE5.B6253678@127.0.0.1...  > > andrew harrison wrote: >  > ...r > : > > >It does but any otimism you might want to inject into8 > > >this discussion needs to be countered with the fact9 > > >that Intel have tried as much as they can to improver9 > > >the performance of the x86 emulation but that it hasi< > > >proved to be very hard to emulate x86 and other non ILP' > > >binaries into an ILP architecture.6 > >W( > > Thats like saying that a performance7 > > car manufacturer is busy altering valve timings and @ > > tweaking inlet pressures, when you're fuelling it with dregsA > > of French wine and measuring actual performance against that.y > >lA > > The VMS guys are software engineers, I'd hardly describe whato9 > > Micro$oft cobble together as 'engineered', would you?V > F > Your point escapes me.  Andrew appears to have been referring to theE > difficulty Intel has obtaining performance from IA32 code under itspN > emulation on IA64, and suggesting that similar difficulties would hamper theI > performance of IA64's emulation of Alpha code.  The quality of the code0E > being emulated is not of apparent relevance to the emulation speed.a >   C That is exactly what I was refering to. Except I am not suggesting  C that there will definitely be similar difficulties emulating Alpha bD code on IA64 but I am questioning what gives Mr Hoffman and various F other posters any confidence that they can succeed where Intel failed.  C ILP appears to be very tricky to optimise for and without bundling -D performance is poor. What is it about the Alpha and VAX instruction A set that would lead anyone to think that producing optimised IA64n> binaries from Alpha/VAX ones will be any easier than from say  HP-PA or x86 ?  @ The x86 emulated apps tests I have seen have compared the actual; throughput of x86 apps running emulated on IA64 with their 0= performance running on Pentium III and other IA32 processors  4 this should eliminate poor coding from the equation.   Regardst Andrew Harrisont Enterprise IT Architecti   ------------------------------  % Date: Fri, 24 Aug 2001 14:56:58 +0100 0 From: andrew harrison <andrew.nospam@uk.sun.com>> Subject: Re: Nuts-n-bolts in News (was: Re: Nits in Slides...)* Message-ID: <3B865D2A.8DD0CC65@uk.sun.com>   Hoff Hoffman wrote:  >i  8 > :It is also not possible to draw parallels between the6 > :experience when moving from VAX to Alpha since care5 > :was taken in the Alpha architecture to allow this.g > ? >   Alpha was not particularly designed with emulation in mind.f >   ? This is as you should know now incorrect, RC and RS were added tA to the architecture specifically to support VESTED applications. i  ; Byte alligned access was added later to improve among othero things FX!32 performance.   : As you should also know microprocessor designers are very 8 carefull about adding instructions to an implimentation ; so the decision to add this support was not something that X would have been made lightly.      RegardsI Andrew Harrisonp Enterprise IT Architectd   ------------------------------  # Date: Fri, 24 Aug 2001 14:50:04 GMTp2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)> Subject: Re: Nuts-n-bolts in News (was: Re: Nits in Slides...)2 Message-ID: <wQth7.732$bB1.32132@news.cpqcorp.net>  ] In article <3B864194.91B107A4@uk.sun.com>, andrew harrison <andrew.nospam@uk.sun.com> writes:m  D :That is exactly what I was refering to. Except I am not suggesting D :that there will definitely be similar difficulties emulating Alpha E :code on IA64 but I am questioning what gives Mr Hoffman and various oG :other posters any confidence that they can succeed where Intel failed.t     My assumptions are 4  G     a: you can't predict the performance of one emulator or translator xF        based on an emulation of a different platform -- other than the7        general statements such as "slower than native".l  B     b: folks that have performance-critical code will not be using         translation or emulation.  D :ILP appears to be very tricky to optimise for and without bundling E :performance is poor. What is it about the Alpha and VAX instruction aB :set that would lead anyone to think that producing optimised IA64? :binaries from Alpha/VAX ones will be any easier than from say t :HP-PA or x86 ?b  B   Alpha has a fixed-length instruction architecture and relativelyC   simple instruction set, which means that decoding the instructionn=   stream is easier than decoding variable-length instructionsaE   such as VAX.  VAX instructions are more complex -- condition codes, H   instruction formats, etc -- meaning emulation will be more interesting
   and slower.t  >   I'd not want to have to emulate IA-64 on any other platform.    A :The x86 emulated apps tests I have seen have compared the actualE< :throughput of x86 apps running emulated on IA64 with their > :performance running on Pentium III and other IA32 processors 5 :this should eliminate poor coding from the equation.2  D   I am not familiar with the innards of the Intel emulator, nor withA   what technologies it uses.  Ares has some interesting features,1C   some akin to what VEST provides during its translation/emulation.   H   Application code that has been through VEST code is normally compiled J   code, and not emulated -- as new code paths are found and are emulated, I   information is made available to subsequent VEST compilation passes to sH   permit more of the application code to be translated into native code.  N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Fri, 24 Aug 2001 12:16:24 -0400s' From: "Bill Todd" <billtodd@foo.mv.com> > Subject: Re: Nuts-n-bolts in News (was: Re: Nits in Slides...)( Message-ID: <9m5uka$5lt$1@pyrite.mv.net>  5 "Nic Clews" <sendspamhere@127.0.0.1> wrote in message # news:3B8638F9.E7B1BAAA@127.0.0.1...  > Bill Todd wrote:   ...v  I > > Perhaps you're confused about Alpha's bread-and-butter market segment  thatI > > brings in the bucks.  It's the server segment, where you'll find thats theoL > > better the performance of each individual processor (and assuming decentK > > memory latency/bandwidth, which EV7 has in spades but which I've yet tod seeeL > > any comparable mention of in IA64's future - ever), the fewer processors youaF > > need (and the smaller the server you need) to satisfy a given load > > acceptably.: >0, > HPTC or just general transaction shifting?  E Jan Vorgrueggen mentioned recently a March 2001 presentation in whichDJ VMS-related revenue was stated as being $4 billion annually and Tru64's asI $3 billion annually.  My strong impression is that virtually all of VMS's J revenue is non-HPTC-related, plus at least a large chunk of Tru64's.  ThatI leaves Alpha Linux (plus a smidgeon of Alpha *BSD), but I doubt that they6K constitute too much of a market compared with the above $7 billion (if they / did, Compaq's revenue numbers wouldn't add up).     A programmer can optimise6 > their code before it even hits the compiler (as if). >wE > I don't practice martial arts, but I believe, in Judo, you use youreJ > oponents 'attack' as their weakness. Even if Alpha had proven advantagesH > over IA64, it is not unfeasible that one could change the way you workJ > with your chip to take advantage of its features. Alpha was not designedI > as VMS's chip, VMS just happens to work quite well on it. As I've said,sD > some kernel tweaks have been made for NUMA, I expect when the dustJ > settles, it'll be tweaked for IA64 (though your and my interpretation of > 'tweak' may differ).  I I'd suggest changing the word 'unfeasible' above to 'inconceivable'.  Buto' even then I'd still take issue with it.   I Alpha already does have proven advantages over IA64, in their two currentiG instantiations - save in the area of highly-regular floating-point codetA where EPIC's approach to ILP does manage to keep up with Alpha's.d  E Alpha has strongly-arguable (though one can't literally say 'proven') L advantages over IA64 for the foreseeable future as well.  EV7's on-chip glueG for RAMBUS and SMP seem likely to eclipse any improvements to IA64 thateJ McKinley brings.  EV8's SMT seems likely to have eclipsed any improvementsI in later IA64 versions for multi-thread operation (at least until 2006 oriH later, the earliest that IA64 could likely have been massively redone toI support some of Alpha's architectural features - assuming that this couldhI have been done at all without the major influx of talent that Compaq justdG gave Intel), while EV8's 8-wide single-thread issue should have kept itIJ highly competitive even for the single-thread, regular floating-point code+ which seems to be IA64's sole strong point.s  I So Alpha's advantages aren't spotty - they're nearly pervasive.  As such,uK one can't 'tweak' one's code to get around IA64's disadvantages (unless you0L can convert all your applications to regular floating-point-intensive ones -J something that seems likely 'unfeasible' in the example of Apache that you present below).    - bill   ...i  D > > > the chip features, they may as well just be serial processors.D > > > When do application writers ever listen to software engineers? > >rJ > > You talk as if Alpha's market was desktop systems, where at least much ofG > > the time only a single instance of a single application is actively  running.I > > In Alpha's main market, 'the majority of commercial applications' are.J > > completely irrelevant (unless something like MTS is running a bunch ofK > > instances in parallel, in which case SMT would have worked just great).r >]C > Nope, I'm talking big servers. Take the Apache code base. Its notmJ > optimised for anything. ISVs typically will try to have a 'similar' codeD > base with the necessary '.H' files. Optimised for what? OK to someF > degree its up to the compiler writers, but they can't work miracles.   ------------------------------    Date: 24 Aug 2001 01:56:13 -0700* From: polato@igi.pd.cnr.it (Sandro Polato)+ Subject: Re: ODS-5 and parse_style question'= Message-ID: <2af2b3d8.0108240056.599d04b4@posting.google.com>   o polato@igi.pd.cnr.it (Sandro Polato) wrote in message news:<2af2b3d8.0108220648.749312e1@posting.google.com>...s > Ciao a tutti.t > G > Do you know whether there is a way to force the uppercase in the filen< > name when I create it, for example, by a fortran program ? > G > I know that I can force DCL to use only uppercase after the command : % > SET PROCESS/PARSE_STYLE=TRADITIONALb0 > but the scope seems to be limited to only DCL. >  > Thanks in advanceb >  > Sandro Polato'  # Only to keep alive the question ...y Does anyone know the answer ?  Thanks Sandro   ------------------------------    Date: 24 Aug 2001 06:02:57 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)i+ Subject: Re: ODS-5 and parse_style questionf3 Message-ID: <bQVrwbqoSZ+E@eisner.encompasserve.org>n  j In article <2af2b3d8.0108240056.599d04b4@posting.google.com>, polato@igi.pd.cnr.it (Sandro Polato) writes:q > polato@igi.pd.cnr.it (Sandro Polato) wrote in message news:<2af2b3d8.0108220648.749312e1@posting.google.com>...a >> Ciao a tutti. >> sH >> Do you know whether there is a way to force the uppercase in the file= >> name when I create it, for example, by a fortran program ?t >> iH >> I know that I can force DCL to use only uppercase after the command :& >> SET PROCESS/PARSE_STYLE=TRADITIONAL1 >> but the scope seems to be limited to only DCL.  >>   >> Thanks in advance >> o >> Sandro Polato > % > Only to keep alive the question ...t > Does anyone know the answer ?   D What is wrong with upcasing the string in place within the program ?C The user has said they want extended names, the programmer has saidtD they don't want extended names.  This seems no different that if the> programmer decided all filetypes had to be ".From_My_Program".   ------------------------------  + Date: Fri, 24 Aug 2001 11:04:20 +0000 (UTC)g' From: david20@alpha2.mdx.ac.uk (D.Webb)t+ Subject: Re: ODS-5 and parse_style questione+ Message-ID: <9m5cbj$7q8$1@aquila.mdx.ac.uk>   j In article <2af2b3d8.0108240056.599d04b4@posting.google.com>, polato@igi.pd.cnr.it (Sandro Polato) writes:p >polato@igi.pd.cnr.it (Sandro Polato) wrote in message news:<2af2b3d8.0108220648.749312e1@posting.google.com>... >> Ciao a tutti. >> tH >> Do you know whether there is a way to force the uppercase in the file= >> name when I create it, for example, by a fortran program ?i >> rH >> I know that I can force DCL to use only uppercase after the command :& >> SET PROCESS/PARSE_STYLE=TRADITIONAL1 >> but the scope seems to be limited to only DCL.o >> s >> Thanks in advance >> a >> Sandro Polato >r$ >Only to keep alive the question ... >Does anyone know the answer ? >Thankse >Sandro     F Maybe I'm missing something but can't you just specify the filename inM uppercase when you create it ? If the name is supplied to the running programe6 by the user then you need to uppercase the user input.  K So long as a lowercase/mixedcase version of the file does not already exist  then this should surely work.   F If a lowercase/mixedcase version already exists then this case will beN preserved for all later versions. Hence you will need to get rid of such filesB if you want to make sure it is created with an uppercase filename.I (Note. In this context 'later versions' means versions created later not sK necessarily with higher version numbers. If a mixed case filename exists aslN version number 2 without a version 1 of the file existing. Then it's case willA be used if you try to explicitly create a version 1 of the file.)      HELP     EXT_FILE_SPECS     Using        File_Specifications          Extended_(ODS-5)_Syntaxs           Case_Preservation   I              On an ODS-5 volume, the case for all versions of a file namelK              is identical; the case is preserved as the file name was firstgK              created. When you create more than one file with the same nameeK              differing only in case, DCL treats the subsequent files as new I              versions, and converts them to the same case as the originale              file.  =              For example, the following sequence of commands:f                $ CREATE CaPri.;1              $ CREATE CAPRI"              $ CREATE caprio  *              produces the resulting files:  )              CaPri.;1  CaPri.;2  CaPri.;3p  I              In prior versions of OpenVMS, DCL and RMS converted all filefK              specifications to uppercase. On ODS-5 volumes, the case of all <              file names is preserved as created by the user.       Note.e  F using  SET PROCESS/PARSE_STYLE=TRADITIONAL on an ODS-5 volume does not; guarantee that all files will be displayed as in uppercase.g   eg   Alpha1:set process/parse=tra  
 Alpha1:dir   Directory TESTODS5:[DAVID20]   Daves.txt;1         MAIL.MAI;1   Total of 2 files.t Alpha1:create test.txt  Exiti
 Alpha1:dir   Directory TESTODS5:[DAVID20]  2 Daves.txt;1         MAIL.MAI;1          TEST.TXT;1   Total of 3 files.  Alpha1:create DAVES.TXT   Exite
 Alpha1:dir   Directory TESTODS5:[DAVID20]  F Daves.txt;2         Daves.txt;1         MAIL.MAI;1          TEST.TXT;1   Total of 4 files.t  J Since Daves.txt;1 already existed Daves.txt;2 was created preserving it's H mixedcase even though the command set process/parse=tra  had been given.    
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------  % Date: Fri, 24 Aug 2001 16:56:34 +0100l* From: Andrew Robinson <arobinson@hspg.com>! Subject: Old print program wantedrM Message-ID: <CDA4BAD1E10ED41181AC00508B6051D3C3E369@grumpy.internal.hspg.com>s  I Many moons ago, whilst working on a MicroVax II, I found it had a programuK for multiple page banner printing to dot matrix printers - does anyone havejJ a copy of that program & will it work on an Alpha ? I know its not cuttingE edge, but I don't want to sit there with a stapler putting together as lasered banner.    Many thanks in advance 3   Andrew Robinson  0   ------------------------------  % Date: Fri, 24 Aug 2001 13:13:18 -0400o  From: jamese@beast.dtsw.army.mil% Subject: Re: Old print program wantedu0 Message-ID: <01082413131803@beast.dtsw.army.mil>  M Andrew Robinson <arobinson@hspg.com> wrote on Fri, 24 Aug 2001 16:56:34 +0100tE in <CDA4BAD1E10ED41181AC00508B6051D3C3E369@grumpy.internal.hspg.com>:   K > Many moons ago, whilst working on a MicroVax II, I found it had a programrM > for multiple page banner printing to dot matrix printers - does anyone havegL > a copy of that program & will it work on an Alpha ? I know its not cuttingG > edge, but I don't want to sit there with a stapler putting together af > lasered banner.   > Attached after my .sig is the Perl Power Tools banner program.  G I also have a zip file From Brian Tillman that contains a Gothic letter G (Olde English) banner program I can send to anyone that wants it. It isc written in FORTRAN.l  : Ed James                           ed.james@telecomsys.com5 TeleCommunications Systems, Inc.   voice 410-295-1919c; 2024 West Street, Suite 300              800-810-0827 x1919p5 Annapolis, MD 21401-3556           fax   410-280-1094    #!/usr/bin/perl -w # * banner - prints large signsi& # * banner [-w#] [-d] [-t] message ...  J # This is a translation from C of the BSD banner.c source, which bears the # following copyright notice:e #o # /*# #  * Copyright (c) 1980, 1993, 1994wG #  *	The Regents of the University of California.  All rights reserved.i #  *G #  * Redistribution and use in source and binary forms, with or withoutuG #  * modification, are permitted provided that the following conditionsn
 #  * are met: F #  * 1. Redistributions of source code must retain the above copyrightE #  *    notice, this list of conditions and the following disclaimer.tI #  * 2. Redistributions in binary form must reproduce the above copyrightPK #  *    notice, this list of conditions and the following disclaimer in themL #  *    documentation and/or other materials provided with the distribution.M #  * 3. All advertising materials mentioning features or use of this softwarer3 #  *    must display the following acknowledgement:NB #  *	This product includes software developed by the University of/ #  *	California, Berkeley and its contributors. L #  * 4. Neither the name of the University nor the names of its contributorsM #  *    may be used to endorse or promote products derived from this softwaree2 #  *    without specific prior written permission. #  *L #  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' ANDJ #  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THEO #  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSEhM #  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLEeO #  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL L #  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODSJ #  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)O #  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICTdN #  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAYK #  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OFB #  * SUCH DAMAGE.s #  */,   use strict;n require 5.004;   use constant MAXMSG => 1024; use constant DWIDTH => 132;t   END { 7     close STDOUT || die "$0: can't close stdout: $!\n";l$     $? = 1 if $? == 255;  # from die }   * sub usage { die "usage: $0 [-w width]\n" }   # Character data my @data_table;c  / # Pointers into @data_table for each ASCII chart my %ascii_to_table;a    	 # optionsy my $width = DWIDTH;p% while (@ARGV && $ARGV[0] =~ s/^-//) {r   local $_ = shift;n   /^[h?]$/ && usage();   if (s/^w//) {e"     if    (length) { $width = $_ }%     elsif (@ARGV)  { $width = shift } K     else           { die "$0: option requires an integer argument -- w\n" }a(     if ($width =~ /\D/ || $width == 0) {9       # the original source silently defaults to 80 cols.n       # I think that's rude.I       warn "Don't understand -w argument `$width'.  Using 80 columns.\n";r       $width = 80;     }n	     next;A   }V$   warn "$0: illegal option -- $_\n";
   usage(); }    # scale characters to widthe my @printmask;$ for (my $i = 0; $i < $width; $i++) {$   $printmask[$i * 132 / $width] = 1; }n  
 # get messagen my $message;) if (@ARGV) { $message = join ' ', @ARGV }c else {   print STDERR "Message: ";o   chomp($message = <STDIN>); }c   # check messageo my $ok = 1;c while ($message =~ /(.)/g) {    unless ($ascii_to_table{$1}) {     $ok = 0;@     warn "$0: The character `$1' is not in my character set.\n";   }e }p exit 1 unless $ok;   # print banner while ($message =~ /(.)/g) {   my @line = (' ') x DWIDTH;   my $pc = $ascii_to_table{$1};o%   my $term = my $max = my $linen = 0;    while (!$term) {A     if ($pc < 0 || $pc > $#data_table) { die "$0: bad pc $pc\n" }e$     my $x = $data_table[$pc] & 0377;     if ($x >= 128) {       ++$term if $x > 192;       $x &= 63;s       while ($x--) { 	if ($printmask[$linen++]) {& 	  for (my $i = 0; $i <= $max; $i++) {' 	    print $line[$i] if $printmask[$i];m 	  } 	  print "\n"; 	}       }l       @line = (' ') x DWIDTH;o       $pc++;     }      else {  # $x < 128!       my $y = $data_table[$pc+1];a       $max = $x + $y;m-       while ($x < $max) { $line[$x++] = '#' }d       $pc += 2;.     }-   }- }-   exit 0;-     BEGIN {<'   @ascii_to_table{ map {chr} 0..127 } =      ( I         0,      0,      0,      0,      0,      0,      0,      0,  # ^@ -I         0,      0,      0,      0,      0,      0,      0,      0,  # ^H  I         0,      0,      0,      0,      0,      0,      0,      0,  # ^P  I         0,      0,      0,      0,      0,      0,      0,      0,  # ^X  I         1,      3,     50,     81,    104,    281,    483,    590,  #     I       621,    685,    749,    851,    862,    893,    898,    921,  #  ( NI      1019,   1150,   1200,   1419,   1599,   1744,   1934,   2111,  #  0 AI      2235,   2445,   2622,   2659,      0,   2708,      0,   2715,  #  8 bI      2857,   3072,   3273,   3403,   3560,   3662,   3730,   3785,  #  @ .I      3965,   4000,   4015,   4115,   4281,   4314,   4432,   4548,  #  H rI      4709,   4790,   4999,   5188,   5397,   5448,   5576,   5710,  #  P nI      5892,   6106,   6257,      0,      0,      0,      0,      0,  #  X  I        50,   6503,   6642,   6733,   6837,   6930,   7073,   7157,  #  `  I      7380,   7452,   7499,   7584,   7689,   7702,   7797,   7869,  #  h  I      7978,   8069,   8160,   8222,   8381,   8442,   8508,   8605,  #  p uI      8732,   8888,   9016,      0,      0,      0,      0,      0   #  x e     );   $   # Table of stuff to print. Format:(   # 128+n -> print current line n times.'   # 64+n  -> this is last byte of char.'F   # else, put m chars at position n (where m is the next elt in array)*   # and goto second next element in array.   @data_table =      (b=      #0     1     2     3     4     5     6     7     8     9 H      129,  227,  130,   34,    6,   90,   19,  129,   32,   10,  #    0 H       74,   40,  129,   31,   12,   64,   53,  129,   30,   14,  #   10 H       54,   65,  129,   30,   14,   53,   67,  129,   30,   14,  #   20 H       54,   65,  129,   31,   12,   64,   53,  129,   32,   10,  #   30 H       74,   40,  129,   34,    6,   90,   19,  129,  194,  130,  #   40 H       99,    9,  129,   97,   14,  129,   96,   18,  129,   95,  #   50 H       22,  129,   95,   16,  117,    2,  129,   95,   14,  129,  #   60 H       96,   11,  129,   97,    9,  129,   99,    6,  129,  194,  #   70 H      129,   87,    4,  101,    4,  131,   82,   28,  131,   87,  #   80 H        4,  101,    4,  133,   82,   28,  131,   87,    4,  101,  #   90 H        4,  131,  193,  129,   39,    1,   84,   27,  129,   38,  #  100 H        3,   81,   32,  129,   37,    5,   79,   35,  129,   36,  #  110 H        5,   77,   38,  129,   35,    5,   76,   40,  129,   34,  #  120 H        5,   75,   21,  103,   14,  129,   33,    5,   74,   19,  #  130 H      107,   11,  129,   32,    5,   73,   17,  110,    9,  129,  #  140 H       32,    4,   73,   16,  112,    7,  129,   31,    4,   72,  #  150 H       15,  114,    6,  129,   31,    4,   72,   14,  115,    5,  #  160 H      129,   30,    4,   71,   15,  116,    5,  129,   27,   97,  #  170 H      131,   30,    4,   69,   14,  117,    4,  129,   30,    4,  #  180 H       68,   15,  117,    4,  132,   30,    4,   68,   14,  117,  #  190 H        4,  129,   27,   97,  131,   30,    5,   65,   15,  116,  #  200 H        5,  129,   31,    4,   65,   14,  116,    4,  129,   31,  #  210 H        6,   64,   15,  116,    4,  129,   32,    7,   62,   16,  #  220 H      115,    4,  129,   32,    9,   61,   17,  114,    5,  129,  #  230 H       33,   11,   58,   19,  113,    5,  129,   34,   14,   55,  #  240 H       21,  112,    5,  129,   35,   40,  111,    5,  129,   36,  #  250 H       38,  110,    5,  129,   37,   35,  109,    5,  129,   38,  #  260 H       32,  110,    3,  129,   40,   27,  111,    1,  129,  193,  #  270 H      129,   30,    4,  103,    9,  129,   30,    7,  100,   15,  #  280 H      129,   30,   10,   99,   17,  129,   33,   10,   97,    6,  #  290 H      112,    6,  129,   36,   10,   96,    5,  114,    5,  129,  #  300 H       39,   10,   96,    4,  115,    4,  129,   42,   10,   95,  #  310 H        4,  116,    4,  129,   45,   10,   95,    3,  117,    3,  #  320 H      129,   48,   10,   95,    3,  117,    3,  129,   51,   10,  #  330 H       95,    4,  116,    4,  129,   54,   10,   96,    4,  115,  #  340 H        4,  129,   57,   10,   96,    5,  114,    5,  129,   60,  #  350 H       10,   97,    6,  112,    6,  129,   63,   10,   99,   17,  #  360 H      129,   66,   10,  100,   15,  129,   69,   10,  103,    9,  #  370 H      129,   39,    9,   72,   10,  129,   36,   15,   75,   10,  #  380 H      129,   35,   17,   78,   10,  129,   33,    6,   48,    6,  #  390 H       81,   10,  129,   32,    5,   50,    5,   84,   10,  129,  #  400 H       32,    4,   51,    4,   87,   10,  129,   31,    4,   52,  #  410 H        4,   90,   10,  129,   31,    3,   53,    3,   93,   10,  #  420 H      129,   31,    3,   53,    3,   96,   10,  129,   31,    4,  #  430 H       52,    4,   99,   10,  129,   32,    4,   51,    4,  102,  #  440 H       10,  129,   32,    5,   50,    5,  105,   10,  129,   33,  #  450 H        6,   48,    6,  108,   10,  129,   35,   17,  111,   10,  #  460 H      129,   36,   15,  114,    7,  129,   40,    9,  118,    4,  #  470 H      129,  193,  129,   48,   18,  129,   43,   28,  129,   41,  #  480 H       32,  129,   39,   36,  129,   37,   40,  129,   35,   44,  #  490 H      129,   34,   46,  129,   33,   13,   68,   13,  129,   32,  #  500 H        9,   73,    9,  129,   32,    7,   75,    7,  129,   31,  #  510 H        6,   77,    6,  129,   31,    5,   78,    5,  129,   30,  #  520 H        5,   79,    5,  129,   20,   74,  132,   30,    4,   80,  #  530 H        4,  129,   31,    3,   79,    4,  129,   31,    4,   79,  #  540 H        4,  129,   32,    3,   78,    4,  129,   32,    4,   76,  #  550 H        6,  129,   33,    4,   74,    7,  129,   34,    4,   72,  #  560 H        8,  129,   35,    5,   72,    7,  129,   37,    5,   73,  #  570 H        4,  129,   39,    4,   74,    1,  129,  129,  193,  130,  #  580 H      111,    6,  129,  109,   10,  129,  108,   12,  129,  107,  #  590 H       14,  129,   97,    2,  105,   16,  129,   99,   22,  129,  #  600 H      102,   18,  129,  105,   14,  129,  108,    9,  129,  194,  #  610 H      130,   63,   25,  129,   57,   37,  129,   52,   47,  129,  #  620 H       48,   55,  129,   44,   63,  129,   41,   69,  129,   38,  #  630 H       75,  129,   36,   79,  129,   34,   83,  129,   33,   28,  #  640 H       90,   28,  129,   32,   23,   96,   23,  129,   32,   17,  #  650 H      102,   17,  129,   31,   13,  107,   13,  129,   30,    9,  #  660 H      112,    9,  129,   30,    5,  116,    5,  129,   30,    1,  #  670 H      120,    1,  129,  194,  130,   30,    1,  120,    1,  129,  #  680 H       30,    5,  116,    5,  129,   30,    9,  112,    9,  129,  #  690 H       31,   13,  107,   13,  129,   32,   17,  102,   17,  129,  #  700 H       32,   23,   96,   23,  129,   33,   28,   90,   28,  129,  #  710 H       34,   83,  129,   36,   79,  129,   38,   75,  129,   41,  #  720 H       69,  129,   44,   63,  129,   48,   55,  129,   52,   47,  #  730 H      129,   57,   37,  129,   63,   25,  129,  194,  129,   80,  #  740 H        4,  130,   80,    4,  129,   68,    2,   80,    4,   94,  #  750 H        2,  129,   66,    6,   80,    4,   92,    6,  129,   67,  #  760 H        7,   80,    4,   90,    7,  129,   69,    7,   80,    4,  #  770 H       88,    7,  129,   71,    6,   80,    4,   87,    6,  129,  #  780 H       72,   20,  129,   74,   16,  129,   76,   12,  129,   62,  #  790 H       40,  131,   76,   12,  129,   74,   16,  129,   72,   20,  #  800 H      129,   71,    6,   80,    4,   87,    6,  129,   69,    7,  #  810 H       80,    4,   88,    7,  129,   67,    7,   80,    4,   90,  #  820 H        7,  129,   66,    6,   80,    4,   92,    6,  129,   68,  #  830 H        2,   80,    4,   94,    2,  129,   80,    4,  130,  193,  #  840 H      129,   60,    4,  139,   41,   42,  131,   60,    4,  139,  #  850 H      193,  130,   34,    6,  129,   32,   10,  129,   31,   12,  #  860 H      129,   30,   14,  129,   20,    2,   28,   16,  129,   22,  #  870 H       22,  129,   24,   19,  129,   27,   15,  129,   31,    9,  #  880 H      129,  194,  129,   60,    4,  152,  193,  130,   34,    6,  #  890 H      129,   32,   10,  129,   31,   12,  129,   30,   14,  131,  #  900 H       31,   12,  129,   32,   10,  129,   34,    6,  129,  194,  #  910 H      129,   30,    4,  129,   30,    7,  129,   30,   10,  129,  #  920 H       33,   10,  129,   36,   10,  129,   39,   10,  129,   42,  #  930 H       10,  129,   45,   10,  129,   48,   10,  129,   51,   10,  #  940 H      129,   54,   10,  129,   57,   10,  129,   60,   10,  129,  #  950 H       63,   10,  129,   66,   10,  129,   69,   10,  129,   72,  #  960 H       10,  129,   75,   10,  129,   78,   10,  129,   81,   10,  #  970 H      129,   84,   10,  129,   87,   10,  129,   90,   10,  129,  #  980 H       93,   10,  129,   96,   10,  129,   99,   10,  129,  102,  #  990 H       10,  129,  105,   10,  129,  108,   10,  129,  111,   10,  # 1000 H      129,  114,    7,  129,  117,    4,  129,  193,  129,   60,  # 1010 H       31,  129,   53,   45,  129,   49,   53,  129,   46,   59,  # 1020 H      129,   43,   65,  129,   41,   69,  129,   39,   73,  129,  # 1030 H       37,   77,  129,   36,   79,  129,   35,   15,  101,   15,  # 1040 H      129,   34,   11,  106,   11,  129,   33,    9,  109,    9,  # 1050 H      129,   32,    7,  112,    7,  129,   31,    6,  114,    6,  # 1060 H      129,   31,    5,  115,    5,  129,   30,    5,  116,    5,  # 1070 H      129,   30,    4,  117,    4,  132,   30,    5,  116,    5,  # 1080 H      129,   31,    5,  115,    5,  129,   31,    6,  114,    6,  # 1090 H      129,   32,    7,  112,    7,  129,   33,    9,  109,    9,  # 1100 H      129,   34,   11,  106,   11,  129,   35,   15,  101,   15,  # 1110 H      129,   36,   79,  129,   37,   77,  129,   39,   73,  129,  # 1120 H       41,   69,  129,   43,   65,  129,   46,   59,  129,   49,  # 1130 H       53,  129,   53,   45,  129,   60,   31,  129,  193,  129,  # 1140 H       30,    4,  129,   30,    4,  100,    1,  129,   30,    4,  # 1150 H      100,    3,  129,   30,    4,  100,    5,  129,   30,   76,  # 1160 H      129,   30,   78,  129,   30,   80,  129,   30,   82,  129,  # 1170 H       30,   83,  129,   30,   85,  129,   30,   87,  129,   30,  # 1180 H       89,  129,   30,   91,  129,   30,    4,  132,  193,  129,  # 1190 H       30,    3,  129,   30,    7,  129,   30,   10,  112,    1,  # 1200 H      129,   30,   13,  112,    2,  129,   30,   16,  112,    3,  # 1210 H      129,   30,   18,  111,    5,  129,   30,   21,  111,    6,  # 1220 H      129,   30,   23,  112,    6,  129,   30,   14,   47,    8,  # 1230 H      113,    6,  129,   30,   14,   49,    8,  114,    5,  129,  # 1240 H       30,   14,   51,    8,  115,    5,  129,   30,   14,   53,  # 1250 H        8,  116,    4,  129,   30,   14,   55,    8,  116,    5,  # 1260 H      129,   30,   14,   56,    9,  117,    4,  129,   30,   14,  # 1270 H       57,    9,  117,    4,  129,   30,   14,   58,   10,  117,  # 1280 H        4,  129,   30,   14,   59,   10,  117,    4,  129,   30,  # 1290 H       14,   60,   11,  117,    4,  129,   30,   14,   61,   11,  # 1300 H      116,    5,  129,   30,   14,   62,   11,  116,    5,  129,  # 1310 H       30,   14,   63,   12,  115,    6,  129,   30,   14,   64,  # 1320 H       13,  114,    7,  129,   30,   14,   65,   13,  113,    8,  # 1330 H      129,   30,   14,   65,   15,  111,    9,  129,   30,   14,  # 1340 H       66,   16,  109,   11,  129,   30,   14,   67,   17,  107,  # 1350 H       12,  129,   30,   14,   68,   20,  103,   16,  129,   30,  # 1360 H       14,   69,   49,  129,   30,   14,   70,   47,  129,   30,  # 1370 H       14,   71,   45,  129,   30,   14,   73,   42,  129,   30,  # 1380 H       15,   75,   38,  129,   33,   12,   77,   34,  129,   36,  # 1390 H       10,   79,   30,  129,   40,    6,   82,   23,  129,   44,  # 1400 H        3,   86,   15,  129,   47,    1,  129,  193,  129,  129,  # 1410 H       38,    3,  129,   37,    5,  111,    1,  129,   36,    7,  # 1420 H      111,    2,  129,   35,    9,  110,    5,  129,   34,    8,  # 1430 H      110,    6,  129,   33,    7,  109,    8,  129,   32,    7,  # 1440 H      110,    8,  129,   32,    6,  112,    7,  129,   31,    6,  # 1450 H      113,    6,  129,   31,    5,  114,    6,  129,   30,    5,  # 1460 H      115,    5,  129,   30,    5,  116,    4,  129,   30,    4,  # 1470 H      117,    4,  131,   30,    4,  117,    4,  129,   30,    4,  # 1480 H       79,    2,  117,    4,  129,   30,    5,   78,    4,  117,  # 1490 H        4,  129,   30,    5,   77,    6,  116,    5,  129,   30,  # 1500 H        6,   76,    8,  115,    6,  129,   30,    7,   75,   11,  # 1510 H      114,    6,  129,   30,    8,   73,   15,  112,    8,  129,  # 1520 H       31,    9,   71,   19,  110,    9,  129,   31,   11,   68,  # 1530 H       26,  107,   12,  129,   32,   13,   65,   14,   82,   36,  # 1540 H      129,   32,   16,   61,   17,   83,   34,  129,   33,   44,  # 1550 H       84,   32,  129,   34,   42,   85,   30,  129,   35,   40,  # 1560 H       87,   27,  129,   36,   38,   89,   23,  129,   38,   34,  # 1570 H       92,   17,  129,   40,   30,   95,   11,  129,   42,   26,  # 1580 H      129,   45,   20,  129,   49,   11,  129,  193,  129,   49,  # 1590 H        1,  129,   49,    4,  129,   49,    6,  129,   49,    8,  # 1600 H      129,   49,   10,  129,   49,   12,  129,   49,   14,  129,  # 1610 H       49,   17,  129,   49,   19,  129,   49,   21,  129,   49,  # 1620 H       23,  129,   49,   14,   65,    9,  129,   49,   14,   67,  # 1630 H        9,  129,   49,   14,   69,    9,  129,   49,   14,   71,  # 1640 H       10,  129,   49,   14,   74,    9,  129,   49,   14,   76,  # 1650 H        9,  129,   49,   14,   78,    9,  129,   49,   14,   80,  # 1660 H        9,  129,   49,   14,   82,    9,  129,   49,   14,   84,  # 1670 H        9,  129,   30,    4,   49,   14,   86,   10,  129,   30,  # 1680 H        4,   49,   14,   89,    9,  129,   30,    4,   49,   14,  # 1690 H       91,    9,  129,   30,    4,   49,   14,   93,    9,  129,  # 1700 H       30,   74,  129,   30,   76,  129,   30,   78,  129,   30,  # 1710 H       81,  129,   30,   83,  129,   30,   85,  129,   30,   87,  # 1720 H      129,   30,   89,  129,   30,   91,  129,   30,    4,   49,  # 1730 H       14,  132,  193,  129,   37,    1,  129,   36,    3,   77,  # 1740 H        3,  129,   35,    5,   78,   11,  129,   34,    7,   78,  # 1750 H       21,  129,   33,    7,   79,   29,  129,   32,    7,   79,  # 1760 H       38,  129,   32,    6,   80,    4,   92,   29,  129,   31,  # 1770 H        6,   80,    5,  102,   19,  129,   31,    5,   80,    6,  # 1780 H      107,   14,  129,   31,    4,   81,    5,  107,   14,  129,  # 1790 H       30,    5,   81,    6,  107,   14,  129,   30,    4,   81,  # 1800 H        6,  107,   14,  130,   30,    4,   81,    7,  107,   14,  # 1810 H      129,   30,    4,   80,    8,  107,   14,  130,   30,    5,  # 1820 H       80,    8,  107,   14,  129,   30,    5,   79,    9,  107,  # 1830 H       14,  129,   31,    5,   79,    9,  107,   14,  129,   31,  # 1840 H        6,   78,   10,  107,   14,  129,   32,    6,   76,   11,  # 1850 H      107,   14,  129,   32,    8,   74,   13,  107,   14,  129,  # 1860 H       33,   10,   71,   16,  107,   14,  129,   33,   15,   67,  # 1870 H       19,  107,   14,  129,   34,   51,  107,   14,  129,   35,  # 1880 H       49,  107,   14,  129,   36,   47,  107,   14,  129,   37,  # 1890 H       45,  107,   14,  129,   39,   41,  107,   14,  129,   41,  # 1900 H       37,  107,   14,  129,   44,   32,  107,   14,  129,   47,  # 1910 H       25,  111,   10,  129,   51,   16,  115,    6,  129,  119,  # 1920 H        2,  129,  193,  129,   56,   39,  129,   51,   49,  129,  # 1930 H       47,   57,  129,   44,   63,  129,   42,   67,  129,   40,  # 1940 H       71,  129,   38,   75,  129,   37,   77,  129,   35,   81,  # 1950 H      129,   34,   16,   74,    5,  101,   16,  129,   33,   11,  # 1960 H       76,    5,  107,   11,  129,   32,    9,   77,    5,  110,  # 1970 H        9,  129,   32,    7,   79,    4,  112,    7,  129,   31,  # 1980 H        6,   80,    4,  114,    6,  129,   31,    5,   81,    4,  # 1990 H      115,    5,  129,   30,    5,   82,    4,  116,    5,  129,  # 2000 H       30,    4,   82,    4,  116,    5,  129,   30,    4,   82,  # 2010 H        5,  117,    4,  131,   30,    5,   82,    5,  117,    4,  # 2020 H      129,   31,    5,   81,    6,  117,    4,  129,   31,    6,  # 2030 H       80,    7,  117,    4,  129,   32,    7,   79,    8,  117,  # 2040 H        4,  129,   32,    9,   77,    9,  116,    5,  129,   33,  # 2050 H       11,   75,   11,  116,    4,  129,   34,   16,   69,   16,  # 2060 H      115,    5,  129,   35,   49,  114,    5,  129,   37,   46,  # 2070 H      113,    5,  129,   38,   44,  112,    6,  129,   40,   41,  # 2080 H      112,    5,  129,   42,   37,  113,    3,  129,   44,   33,  # 2090 H      114,    1,  129,   47,   27,  129,   51,   17,  129,  193,  # 2100 H      129,  103,    2,  129,  103,    6,  129,  104,    9,  129,  # 2110 H      105,   12,  129,  106,   15,  129,  107,   14,  135,   30,  # 2120 H       10,  107,   14,  129,   30,   17,  107,   14,  129,   30,  # 2130 H       25,  107,   14,  129,   30,   31,  107,   14,  129,   30,  # 2140 H       37,  107,   14,  129,   30,   42,  107,   14,  129,   30,  # 2150 H       46,  107,   14,  129,   30,   50,  107,   14,  129,   30,  # 2160 H       54,  107,   14,  129,   30,   58,  107,   14,  129,   59,  # 2170 H       32,  107,   14,  129,   64,   30,  107,   14,  129,   74,  # 2180 H       23,  107,   14,  129,   81,   18,  107,   14,  129,   86,  # 2190 H       16,  107,   14,  129,   91,   14,  107,   14,  129,   96,  # 2200 H       25,  129,  100,   21,  129,  104,   17,  129,  107,   14,  # 2210 H      129,  111,   10,  129,  114,    7,  129,  117,    4,  129,  # 2220 H      120,    1,  129,  193,  129,   48,   13,  129,   44,   21,  # 2230 H      129,   42,   26,  129,   40,   30,   92,   12,  129,   38,  # 2240 H       34,   88,   20,  129,   36,   37,   86,   25,  129,   35,  # 2250 H       39,   84,   29,  129,   34,   13,   63,   12,   82,   33,  # 2260 H      129,   33,   11,   67,    9,   80,   36,  129,   32,    9,  # 2270 H       70,    7,   79,   38,  129,   31,    8,   72,   46,  129,  # 2280 H       30,    7,   74,   22,  108,   11,  129,   30,    6,   75,  # 2290 H       19,  111,    9,  129,   30,    5,   75,   17,  113,    7,  # 2300 H      129,   30,    5,   74,   16,  114,    6,  129,   30,    4,  # 2310 H       73,   16,  115,    6,  129,   30,    4,   72,   16,  116,  # 2320 H        5,  129,   30,    4,   72,   15,  117,    4,  129,   30,  # 2330 H        4,   71,   16,  117,    4,  129,   30,    5,   70,   16,  # 2340 H      117,    4,  129,   30,    5,   70,   15,  117,    4,  129,  # 2350 H       30,    6,   69,   15,  116,    5,  129,   30,    7,   68,  # 2360 H       17,  115,    5,  129,   30,    9,   67,   19,  114,    6,  # 2370 H      129,   30,   10,   65,   22,  113,    6,  129,   31,   12,  # 2380 H       63,   27,  110,    9,  129,   32,   14,   60,   21,   84,  # 2390 H        9,  106,   12,  129,   33,   47,   85,   32,  129,   34,  # 2400 H       45,   86,   30,  129,   35,   43,   88,   26,  129,   36,  # 2410 H       40,   90,   22,  129,   38,   36,   93,   17,  129,   40,  # 2420 H       32,   96,   10,  129,   42,   28,  129,   44,   23,  129,  # 2430 H       48,   15,  129,  193,  129,   83,   17,  129,   77,   27,  # 2440 H      129,   36,    1,   74,   33,  129,   35,    3,   72,   37,  # 2450 H      129,   34,    5,   70,   41,  129,   33,    6,   69,   44,  # 2460 H      129,   33,    5,   68,   46,  129,   32,    5,   67,   49,  # 2470 H      129,   31,    5,   66,   17,  101,   16,  129,   31,    5,  # 2480 H       66,   11,  108,   10,  129,   30,    4,   65,    9,  110,  # 2490 H        9,  129,   30,    4,   64,    8,  112,    7,  129,   30,  # 2500 H        4,   64,    7,  114,    6,  129,   30,    4,   64,    6,  # 2510 H      115,    5,  129,   30,    4,   64,    5,  116,    5,  129,  # 2520 H       30,    4,   64,    5,  117,    4,  131,   30,    4,   65,  # 2530 H        4,  117,    4,  129,   30,    5,   65,    4,  116,    5,  # 2540 H      129,   31,    5,   66,    4,  115,    5,  129,   31,    6,  # 2550 H       67,    4,  114,    6,  129,   32,    7,   68,    4,  112,  # 2560 H        7,  129,   32,    9,   69,    5,  110,    9,  129,   33,  # 2570 G       11,   70,    5,  107,   11,  129,   34,   16,   72,    5,  # 2580 H      101,   16,  129,   35,   81,  129,   37,   77,  129,   38,  # 2590 H       75,  129,   40,   71,  129,   42,   67,  129,   44,   63,  # 2600 H      129,   47,   57,  129,   51,   49,  129,   56,   39,  129,  # 2610 H      193,  130,   34,    6,   74,    6,  129,   32,   10,   72,  # 2620 H       10,  129,   31,   12,   71,   12,  129,   30,   14,   70,  # 2630 H       14,  131,   31,   12,   71,   12,  129,   32,   10,   72,  # 2640 H       10,  129,   34,    6,   74,    6,  129,  194,  130,   34,  # 2650 H        6,   74,    6,  129,   32,   10,   72,   10,  129,   31,  # 2660 H       12,   71,   12,  129,   30,   14,   70,   14,  129,   20,  # 2670 H        2,   28,   16,   70,   14,  129,   22,   22,   70,   14,  # 2680 H      129,   24,   19,   71,   12,  129,   27,   15,   72,   10,  # 2690 H      129,   31,    9,   74,    6,  129,  194,  129,   53,    4,  # 2700 H       63,    4,  152,  193,  130,   99,    7,  129,   97,   13,  # 2710 H      129,   96,   16,  129,   96,   18,  129,   96,   19,  129,  # 2720 H       97,   19,  129,   99,    6,  110,    7,  129,  112,    6,  # 2730 H      129,  114,    5,  129,   34,    6,   57,    5,  115,    4,  # 2740 H      129,   32,   10,   54,   12,  116,    4,  129,   31,   12,  # 2750 H       53,   16,  117,    3,  129,   30,   14,   52,   20,  117,  # 2760 H        4,  129,   30,   14,   52,   23,  117,    4,  129,   30,  # 2770 H       14,   52,   25,  117,    4,  129,   31,   12,   52,   27,  # 2780 H      117,    4,  129,   32,   10,   53,   10,   70,   11,  116,  # 2790 H        5,  129,   34,    6,   55,    5,   73,   10,  115,    6,  # 2800 H      129,   74,   11,  114,    7,  129,   75,   12,  112,    9,  # 2810 H      129,   76,   13,  110,   10,  129,   77,   16,  106,   14,  # 2820 H      129,   78,   41,  129,   80,   38,  129,   81,   36,  129,  # 2830 H       82,   34,  129,   84,   30,  129,   86,   26,  129,   88,  # 2840 H       22,  129,   92,   14,  129,  194,  129,   55,   15,  129,  # 2850 H       50,   25,  129,   47,   32,  129,   45,   13,   70,   12,  # 2860 H      129,   43,    9,   76,   10,  129,   42,    6,   79,    8,  # 2870 H      129,   41,    5,   81,    7,  129,   40,    4,   84,    6,  # 2880 H      129,   39,    4,   59,   12,   85,    6,  129,   38,    4,  # 2890 H       55,   19,   87,    5,  129,   37,    4,   53,   23,   88,  # 2900 H        4,  129,   36,    4,   51,    8,   71,    6,   89,    4,  # 2910 H      129,   36,    4,   51,    6,   73,    4,   89,    4,  129,  # 2920 H       36,    4,   50,    6,   74,    4,   90,    3,  129,   35,  # 2930 H        4,   50,    5,   75,    3,   90,    4,  129,   35,    4,  # 2940 H       50,    4,   75,    4,   90,    4,  131,   35,    4,   50,  # 2950 H        5,   75,    4,   90,    4,  129,   36,    4,   51,    5,  # 2960 H       75,    4,   90,    4,  129,   36,    4,   51,    6,   75,  # 2970 H        4,   90,    4,  129,   36,    4,   53,   26,   90,    4,  # 2980 H      129,   37,    4,   54,   25,   90,    4,  129,   37,    4,  # 2990 H       52,   27,   90,    3,  129,   38,    4,   52,    4,   89,  # 3000 H        4,  129,   39,    4,   51,    4,   88,    4,  129,   40,  # 3010 H        4,   50,    4,   87,    5,  129,   41,    4,   50,    4,  # 3020 H       86,    5,  129,   42,    4,   50,    4,   85,    5,  129,  # 3030 H       43,    3,   50,    4,   83,    6,  129,   44,    2,   51,  # 3040 H        5,   80,    7,  129,   46,    1,   52,    6,   76,    9,  # 3050 H      129,   54,   28,  129,   56,   23,  129,   60,   16,  129,  # 3060 H      193,  129,   30,    4,  132,   30,    5,  129,   30,    8,  # 3070 H      129,   30,   12,  129,   30,   16,  129,   30,    4,   37,  # 3080 H       12,  129,   30,    4,   41,   12,  129,   30,    4,   44,  # 3090 H       13,  129,   30,    4,   48,   13,  129,   52,   13,  129,  # 3100 H       56,   12,  129,   58,   14,  129,   58,    4,   64,   12,  # 3110 H      129,   58,    4,   68,   12,  129,   58,    4,   72,   12,  # 3120 H      129,   58,    4,   75,   13,  129,   58,    4,   79,   13,  # 3130 H      129,   58,    4,   83,   13,  129,   58,    4,   87,   13,  # 3140 H      129,   58,    4,   91,   12,  129,   58,    4,   95,   12,  # 3150 H      129,   58,    4,   96,   15,  129,   58,    4,   93,   22,  # 3160 H      129,   58,    4,   89,   30,  129,   58,    4,   85,   36,  # 3170 H      129,   58,    4,   81,   38,  129,   58,    4,   77,   38,  # 3180 H      129,   58,    4,   73,   38,  129,   58,    4,   70,   37,  # 3190 H      129,   58,    4,   66,   37,  129,   58,   41,  129,   58,  # 3200 H       37,  129,   54,   38,  129,   30,    4,   50,   38,  129,  # 3210 H       30,    4,   46,   38,  129,   30,    4,   42,   38,  129,  # 3220 H       30,    4,   38,   39,  129,   30,   43,  129,   30,   39,  # 3230 H      129,   30,   35,  129,   30,   31,  129,   30,   27,  129,  # 3240 H       30,   24,  129,   30,   20,  129,   30,   16,  129,   30,  # 3250 H       12,  129,   30,    8,  129,   30,    5,  129,   30,    4,  # 3260 H      132,  193,  129,   30,    4,  117,    4,  132,   30,   91,  # 3270 H      137,   30,    4,   80,    4,  117,    4,  138,   30,    4,  # 3280 H       80,    5,  116,    5,  129,   30,    5,   79,    6,  116,  # 3290 H        5,  130,   30,    6,   78,    8,  115,    6,  129,   31,  # 3300 H        6,   77,    9,  115,    6,  129,   31,    7,   76,   11,  # 3310 H      114,    6,  129,   31,    8,   75,   14,  112,    8,  129,  # 3320 H       32,    8,   74,   16,  111,    9,  129,   32,    9,   73,  # 3330 H       19,  109,   10,  129,   33,   10,   71,   24,  106,   13,  # 3340 H      129,   33,   13,   68,   12,   83,   35,  129,   34,   16,  # 3350 H       64,   15,   84,   33,  129,   35,   43,   85,   31,  129,  # 3360 H       36,   41,   86,   29,  129,   37,   39,   88,   25,  129,  # 3370 H       38,   37,   90,   21,  129,   40,   33,   93,   15,  129,  # 3380 H       42,   29,   96,    9,  129,   45,   24,  129,   49,   16,  # 3390 H      129,  193,  129,   63,   25,  129,   57,   37,  129,   53,  # 3400 H       45,  129,   50,   51,  129,   47,   57,  129,   45,   61,  # 3410 H      129,   43,   65,  129,   41,   69,  129,   39,   73,  129,  # 3420 H       38,   25,   92,   21,  129,   36,   21,   97,   18,  129,  # 3430 H       35,   18,  102,   14,  129,   34,   16,  106,   11,  129,  # 3440 H       33,   14,  108,   10,  129,   32,   12,  111,    8,  129,  # 3450 H       32,   10,  113,    6,  129,   31,   10,  114,    6,  129,  # 3460 H       31,    8,  115,    5,  129,   30,    8,  116,    5,  129,  # 3470 H       30,    7,  116,    5,  129,   30,    6,  117,    4,  130,  # 3480 H       30,    5,  117,    4,  131,   31,    4,  116,    5,  129,  # 3490 H       32,    4,  116,    4,  129,   32,    5,  115,    5,  129,  # 3500 H       33,    4,  114,    5,  129,   34,    4,  112,    6,  129,  # 3510 H       35,    4,  110,    7,  129,   37,    4,  107,    9,  129,  # 3520 H       39,    4,  103,   12,  129,   41,    4,  103,   18,  129,  # 3530 H       43,    4,  103,   18,  129,   45,    5,  103,   18,  129,  # 3540 H       48,    5,  103,   18,  129,   51,    1,  129,  193,  129,  # 3550 H       30,    4,  117,    4,  132,   30,   91,  137,   30,    4,  # 3560 H      117,    4,  135,   30,    5,  116,    5,  130,   30,    6,  # 3570 H      115,    6,  130,   31,    6,  114,    6,  129,   31,    7,  # 3580 H      113,    7,  129,   32,    7,  112,    7,  129,   32,    8,  # 3590 H      111,    8,  129,   33,    9,  109,    9,  129,   33,   12,  # 3600 H      106,   12,  129,   34,   13,  104,   13,  129,   35,   15,  # 3610 H      101,   15,  129,   36,   19,   96,   19,  129,   37,   24,  # 3620 H       90,   24,  129,   39,   73,  129,   40,   71,  129,   42,  # 3630 H       67,  129,   44,   63,  129,   46,   59,  129,   49,   53,  # 3640 H      129,   52,   47,  129,   56,   39,  129,   61,   29,  129,  # 3650 H      193,  129,   30,    4,  117,    4,  132,   30,   91,  137,  # 3660 H       30,    4,   80,    4,  117,    4,  140,   30,    4,   79,  # 3670 H        6,  117,    4,  129,   30,    4,   77,   10,  117,    4,  # 3680 H      129,   30,    4,   73,   18,  117,    4,  132,   30,    4,  # 3690 H      117,    4,  130,   30,    5,  116,    5,  130,   30,    7,  # 3700 H      114,    7,  129,   30,    8,  113,    8,  129,   30,   11,  # 3710 H      110,   11,  129,   30,   18,  103,   18,  132,  193,  129,  # 3720 H       30,    4,  117,    4,  132,   30,   91,  137,   30,    4,  # 3730 H       80,    4,  117,    4,  132,   80,    4,  117,    4,  136,  # 3740 H       79,    6,  117,    4,  129,   77,   10,  117,    4,  129,  # 3750 H       73,   18,  117,    4,  132,  117,    4,  130,  116,    5,  # 3760 H      130,  114,    7,  129,  113,    8,  129,  110,   11,  129,  # 3770 H      103,   18,  132,  193,  129,   63,   25,  129,   57,   37,  # 3780 H      129,   53,   45,  129,   50,   51,  129,   47,   57,  129,  # 3790 H       45,   61,  129,   43,   65,  129,   41,   69,  129,   39,  # 3800 H       73,  129,   38,   25,   92,   21,  129,   36,   21,   97,  # 3810 H       18,  129,   35,   18,  102,   14,  129,   34,   16,  106,  # 3820 H       11,  129,   33,   14,  108,   10,  129,   32,   12,  111,  # 3830 H        8,  129,   32,   10,  113,    6,  129,   31,   10,  114,  # 3840 H        6,  129,   31,    8,  115,    5,  129,   30,    8,  116,  # 3850 H        5,  129,   30,    7,  116,    5,  129,   30,    6,  117,  # 3860 H        4,  130,   30,    5,  117,    4,  131,   30,    5,   75,  # 3870 H        4,  116,    5,  129,   31,    5,   75,    4,  116,    4,  # 3880 H      129,   31,    6,   75,    4,  115,    5,  129,   32,    7,  # 3890 H       75,    4,  114,    5,  129,   32,    9,   75,    4,  112,  # 3900 H        6,  129,   33,   11,   75,    4,  110,    7,  129,   34,  # 3910 H       15,   75,    4,  107,    9,  129,   35,   44,  103,   12,  # 3920 H      129,   36,   43,  103,   18,  129,   38,   41,  103,   18,  # 3930 H      129,   39,   40,  103,   18,  129,   41,   38,  103,   18,  # 3940 H      129,   44,   35,  129,   48,   31,  129,   52,   27,  129,  # 3950 H       61,   18,  129,  193,  129,   30,    4,  117,    4,  132,  # 3960 H       30,   91,  137,   30,    4,   80,    4,  117,    4,  132,  # 3970 H       80,    4,  140,   30,    4,   80,    4,  117,    4,  132,  # 3980 H       30,   91,  137,   30,    4,  117,    4,  132,  193,  129,  # 3990 H       30,    4,  117,    4,  132,   30,   91,  137,   30,    4,  # 4000 H      117,    4,  132,  193,  129,   44,    7,  129,   40,   13,  # 4010 H      129,   37,   17,  129,   35,   20,  129,   34,   22,  129,  # 4020 H       33,   23,  129,   32,   24,  129,   32,   23,  129,   31,  # 4030 H        6,   41,   13,  129,   31,    5,   42,   11,  129,   30,  # 4040 H        5,   44,    7,  129,   30,    4,  132,   30,    5,  130,  # 4050 H       31,    5,  129,   31,    6,  117,    4,  129,   31,    8,  # 4060 H      117,    4,  129,   32,    9,  117,    4,  129,   33,   11,  # 4070 H      117,    4,  129,   34,   87,  129,   35,   86,  129,   36,  # 4080 H       85,  129,   37,   84,  129,   38,   83,  129,   40,   81,  # 4090 H      129,   42,   79,  129,   45,   76,  129,   50,   71,  129,  # 4100 H      117,    4,  132,  193,  129,   30,    4,  117,    4,  132,  # 4110 H       30,   91,  137,   30,    4,   76,    8,  117,    4,  129,  # 4120 H       30,    4,   73,   13,  117,    4,  129,   30,    4,   70,  # 4130 H       18,  117,    4,  129,   30,    4,   67,   23,  117,    4,  # 4140 H      129,   65,   26,  129,   62,   31,  129,   59,   35,  129,  # 4150 H       56,   29,   89,    7,  129,   53,   29,   91,    7,  129,  # 4160 H       50,   29,   93,    7,  129,   47,   29,   95,    6,  129,  # 4170 H       30,    4,   45,   29,   96,    7,  129,   30,    4,   42,  # 4180 H       29,   98,    7,  129,   30,    4,   39,   30,  100,    6,  # 4190 H      129,   30,    4,   36,   30,  101,    7,  129,   30,   33,  # 4200 H      103,    7,  117,    4,  129,   30,   30,  105,    6,  117,  # 4210 H        4,  129,   30,   27,  106,    7,  117,    4,  129,   30,  # 4220 H       25,  108,    7,  117,    4,  129,   30,   22,  110,   11,  # 4230 H      129,   30,   19,  111,   10,  129,   30,   16,  113,    8,  # 4240 H      129,   30,   13,  115,    6,  129,   30,   11,  116,    5,  # 4250 H      129,   30,    8,  117,    4,  129,   30,    5,  117,    4,  # 4260 H      129,   30,    4,  117,    4,  130,   30,    4,  130,  193,  # 4270 H      129,   30,    4,  117,    4,  132,   30,   91,  137,   30,  # 4280 H        4,  117,    4,  132,   30,    4,  144,   30,    5,  130,  # 4290 H       30,    7,  129,   30,    8,  129,   30,   11,  129,   30,  # 4300 H       18,  132,  193,  129,   30,    4,  117,    4,  132,   30,  # 4310 H       91,  132,   30,    4,  103,   18,  129,   30,    4,   97,  # 4320 H       24,  129,   30,    4,   92,   29,  129,   30,    4,   87,  # 4330 H       34,  129,   81,   40,  129,   76,   45,  129,   70,   49,  # 4340 H      129,   65,   49,  129,   60,   49,  129,   55,   49,  129,  # 4350 H       50,   48,  129,   44,   49,  129,   39,   48,  129,   33,  # 4360 H       49,  129,   30,   47,  129,   34,   37,  129,   40,   26,  # 4370 H      129,   46,   19,  129,   52,   19,  129,   58,   19,  129,  # 4380 H       64,   19,  129,   70,   19,  129,   76,   19,  129,   82,  # 4390 H       19,  129,   30,    4,   88,   18,  129,   30,    4,   94,  # 4400 H       18,  129,   30,    4,  100,   18,  129,   30,    4,  106,  # 4410 H       15,  129,   30,   91,  137,   30,    4,  117,    4,  132,  # 4420 H      193,  129,   30,    4,  117,    4,  132,   30,   91,  132,  # 4430 H       30,    4,  107,   14,  129,   30,    4,  104,   17,  129,  # 4440 H       30,    4,  101,   20,  129,   30,    4,   99,   22,  129,  # 4450 H       96,   25,  129,   93,   28,  129,   91,   28,  129,   88,  # 4460 H       29,  129,   85,   29,  129,   82,   29,  129,   79,   29,  # 4470 H      129,   76,   29,  129,   74,   29,  129,   71,   29,  129,  # 4480 H       68,   29,  129,   65,   29,  129,   62,   29,  129,   60,  # 4490 H       29,  129,   57,   29,  129,   54,   29,  129,   51,   29,  # 4500 H      129,   49,   28,  129,   46,   29,  129,   43,   29,  129,  # 4510 H       40,   29,  117,    4,  129,   37,   29,  117,    4,  129,  # 4520 H       35,   29,  117,    4,  129,   32,   29,  117,    4,  129,  # 4530 H       30,   91,  132,  117,    4,  132,  193,  129,   63,   25,  # 4540 H      129,   57,   37,  129,   53,   45,  129,   50,   51,  129,  # 4550 H       47,   57,  129,   45,   61,  129,   43,   65,  129,   41,  # 4560 H       69,  129,   39,   73,  129,   38,   21,   92,   21,  129,  # 4570 H       36,   18,   97,   18,  129,   35,   14,  102,   14,  129,  # 4580 H       34,   11,  106,   11,  129,   33,   10,  108,   10,  129,  # 4590 H       32,    8,  111,    8,  129,   32,    6,  113,    6,  129,  # 4600 H       31,    6,  114,    6,  129,   31,    5,  115,    5,  129,  # 4610 H       30,    5,  116,    5,  130,   30,    4,  117,    4,  132,  # 4620 H       30,    5,  116,    5,  130,   31,    5,  115,    5,  129,  # 4630 H       31,    6,  114,    6,  129,   32,    6,  113,    6,  129,  # 4640 H       32,    8,  111,    8,  129,   33,   10,  108,   10,  129,  # 4650 H       34,   11,  106,   11,  129,   35,   14,  102,   14,  129,  # 4660 H       36,   18,   97,   18,  129,   38,   21,   92,   21,  129,  # 4670 H       39,   73,  129,   41,   69,  129,   43,   65,  129,   45,  # 4680 H       61,  129,   47,   57,  129,   50,   51,  129,   53,   45,  # 4690 H      129,   57,   37,  129,   63,   25,  129,  193,  129,   30,  # 4700 H        4,  117,    4,  132,   30,   91,  137,   30,    4,   80,  # 4710 H        4,  117,    4,  132,   80,    4,  117,    4,  134,   80,  # 4720 H        5,  116,    5,  131,   80,    6,  115,    6,  130,   81,  # 4730 H        6,  114,    6,  129,   81,    8,  112,    8,  129,   81,  # 4740 H        9,  111,    9,  129,   82,   10,  109,   10,  129,   82,  # 4750 H       13,  106,   13,  129,   83,   35,  129,   84,   33,  129,  # 4760 H       85,   31,  129,   86,   29,  129,   88,   25,  129,   90,  # 4770 H       21,  129,   93,   15,  129,   96,    9,  129,  193,  129,  # 4780 H       63,   25,  129,   57,   37,  129,   53,   45,  129,   50,  # 4790 H       51,  129,   47,   57,  129,   45,   61,  129,   43,   65,  # 4800 H      129,   41,   69,  129,   39,   73,  129,   38,   21,   92,  # 4810 H       21,  129,   36,   18,   97,   18,  129,   35,   14,  102,  # 4820 H       14,  129,   34,   11,  106,   11,  129,   33,   10,  108,  # 4830 H       10,  129,   32,    8,  111,    8,  129,   32,    6,  113,  # 4840 H        6,  129,   31,    6,  114,    6,  129,   31,    5,  115,  # 4850 H        5,  129,   30,    5,  116,    5,  130,   30,    4,   39,  # 4860 H        2,  117,    4,  129,   30,    4,   40,    4,  117,    4,  # 4870 H      129,   30,    4,   41,    5,  117,    4,  129,   30,    4,  # 4880 H       41,    6,  117,    4,  129,   30,    5,   40,    8,  116,  # 4890 H        5,  129,   30,    5,   39,   10,  116,    5,  129,   31,  # 4900 H        5,   38,   11,  115,    5,  129,   31,   18,  114,    6,  # 4910 H      129,   32,   17,  113,    6,  129,   32,   16,  111,    8,  # 4920 H      129,   33,   15,  108,   10,  129,   33,   14,  106,   11,  # 4930 H      129,   32,   17,  102,   14,  129,   31,   23,   97,   18,  # 4940 H      129,   31,   28,   92,   21,  129,   30,   82,  129,   30,  # 4950 H       80,  129,   30,   11,   43,   65,  129,   30,   10,   45,  # 4960 H       61,  129,   31,    8,   47,   57,  129,   32,    6,   50,  # 4970 H       51,  129,   33,    5,   53,   45,  129,   35,    4,   57,  # 4980 H       37,  129,   38,    2,   63,   25,  129,  193,  129,   30,  # 4990 H        4,  117,    4,  132,   30,   91,  137,   30,    4,   76,  # 5000 H        8,  117,    4,  129,   30,    4,   73,   11,  117,    4,  # 5010 H      129,   30,    4,   70,   14,  117,    4,  129,   30,    4,  # 5020 H       67,   17,  117,    4,  129,   65,   19,  117,    4,  129,  # 5030 H       62,   22,  117,    4,  129,   59,   25,  117,    4,  129,  # 5040 H       56,   28,  117,    4,  129,   53,   31,  117,    4,  129,  # 5050 H       50,   34,  117,    4,  129,   47,   29,   80,    5,  116,  # 5060 H        5,  129,   30,    4,   45,   29,   80,    5,  116,    5,  # 5070 H      129,   30,    4,   42,   29,   80,    5,  116,    5,  129,  # 5080 H       30,    4,   39,   30,   80,    6,  115,    6,  129,   30,  # 5090 H        4,   36,   30,   80,    6,  115,    6,  129,   30,   33,  # 5100 H       81,    6,  114,    6,  129,   30,   30,   81,    8,  112,  # 5110 H        8,  129,   30,   27,   81,    9,  111,    9,  129,   30,  # 5120 H       25,   82,   10,  109,   10,  129,   30,   22,   82,   13,  # 5130 H      106,   13,  129,   30,   19,   83,   35,  129,   30,   16,  # 5140 H       84,   33,  129,   30,   13,   85,   31,  129,   30,   11,  # 5150 H       86,   29,  129,   30,    8,   88,   25,  129,   30,    5,  # 5160 H       90,   21,  129,   30,    4,   93,   15,  129,   30,    4,  # 5170 H       96,    9,  129,   30,    4,  130,  193,  129,   30,   18,  # 5180 H      130,   30,   18,   89,   15,  129,   30,   18,   85,   23,  # 5190 H      129,   34,   11,   83,   27,  129,   34,    9,   81,   31,  # 5200 H      129,   33,    8,   79,   35,  129,   33,    6,   78,   16,  # 5210 H      106,    9,  129,   32,    6,   77,   15,  109,    7,  129,  # 5220 H       32,    5,   76,   14,  111,    6,  129,   31,    5,   75,  # 5230 H       14,  113,    5,  129,   31,    4,   74,   15,  114,    5,  # 5240 H      129,   31,    4,   74,   14,  115,    4,  129,   30,    4,  # 5250 H       73,   15,  116,    4,  129,   30,    4,   73,   14,  116,  # 5260 H        4,  129,   30,    4,   73,   14,  117,    4,  129,   30,  # 5270 H        4,   72,   15,  117,    4,  130,   30,    4,   71,   15,  # 5280 H      117,    4,  130,   30,    4,   70,   15,  117,    4,  129,  # 5290 H       30,    5,   70,   15,  117,    4,  129,   30,    5,   69,  # 5300 H       15,  116,    5,  129,   30,    6,   68,   16,  115,    5,  # 5310 H      129,   31,    6,   67,   16,  114,    6,  129,   31,    7,  # 5320 H       66,   17,  113,    6,  129,   32,    7,   64,   18,  111,  # 5330 H        8,  129,   32,    8,   62,   19,  109,    9,  129,   33,  # 5340 H        9,   60,   20,  107,   10,  129,   34,   11,   57,   22,  # 5350 H      103,   13,  129,   35,   43,  103,   18,  129,   36,   41,  # 5360 H      103,   18,  129,   38,   38,  103,   18,  129,   39,   35,  # 5370 H      103,   18,  129,   41,   31,  129,   43,   27,  129,   46,  # 5380 H       22,  129,   49,   14,  129,  193,  129,  103,   18,  132,  # 5390 H      110,   11,  129,  113,    8,  129,  114,    7,  129,  116,  # 5400 H        5,  130,  117,    4,  132,   30,    4,  117,    4,  132,  # 5410 H       30,   91,  137,   30,    4,  117,    4,  132,  117,    4,  # 5420 H      132,  116,    5,  130,  114,    7,  129,  113,    8,  129,  # 5430 H      110,   11,  129,  103,   18,  132,  193,  129,  117,    4,  # 5440 H      132,   56,   65,  129,   50,   71,  129,   46,   75,  129,  # 5450 H       44,   77,  129,   42,   79,  129,   40,   81,  129,   38,  # 5460 H       83,  129,   36,   85,  129,   35,   86,  129,   34,   20,  # 5470 H      117,    4,  129,   33,   17,  117,    4,  129,   32,   15,  # 5480 H      117,    4,  129,   32,   13,  117,    4,  129,   31,   12,  # 5490 H      129,   31,   10,  129,   31,    9,  129,   30,    9,  129,  # 5500 H       30,    8,  130,   30,    7,  132,   31,    6,  130,   31,  # 5510 H        7,  129,   32,    6,  129,   32,    7,  129,   33,    7,  # 5520 H      129,   34,    7,  129,   35,    8,  129,   36,    9,  117,  # 5530 H        4,  129,   38,    9,  117,    4,  129,   40,   10,  117,  # 5540 H        4,  129,   42,   12,  117,    4,  129,   44,   77,  129,  # 5550 H       46,   75,  129,   50,   71,  129,   56,   43,  100,   21,  # 5560 H      129,  117,    4,  132,  193,  129,  117,    4,  132,  115,  # 5570 H        6,  129,  110,   11,  129,  105,   16,  129,  101,   20,  # 5580 H      129,   96,   25,  129,   92,   29,  129,   87,   34,  129,  # 5590 H       83,   38,  129,   78,   43,  129,   74,   47,  129,   70,  # 5600 H       42,  117,    4,  129,   65,   42,  117,    4,  129,   60,  # 5610 H       43,  117,    4,  129,   56,   42,  129,   51,   42,  129,  # 5620 H       46,   43,  129,   42,   43,  129,   37,   44,  129,   33,  # 5630 H       43,  129,   30,   42,  129,   33,   34,  129,   38,   25,  # 5640 H      129,   42,   16,  129,   47,   15,  129,   52,   15,  129,  # 5650 H       57,   15,  129,   61,   16,  129,   66,   16,  129,   71,  # 5660 H       16,  129,   76,   16,  129,   80,   16,  129,   85,   16,  # 5670 H      117,    4,  129,   90,   16,  117,    4,  129,   95,   16,  # 5680 H      117,    4,  129,  100,   21,  129,  105,   16,  129,  110,  # 5690 H       11,  129,  114,    7,  129,  117,    4,  132,  193,  129,  # 5700 H      117,    4,  132,  115,    6,  129,  110,   11,  129,  105,  # 5710 H       16,  129,  101,   20,  129,   96,   25,  129,   92,   29,  # 5720 H      129,   87,   34,  129,   83,   38,  129,   78,   43,  129,  # 5730 H       74,   47,  129,   70,   42,  117,    4,  129,   65,   42,  # 5740 H      117,    4,  129,   60,   43,  117,    4,  129,   56,   42,  # 5750 H      129,   51,   42,  129,   46,   43,  129,   42,   43,  129,  # 5760 H       37,   44,  129,   33,   43,  129,   30,   42,  129,   33,  # 5770 H       34,  129,   38,   25,  129,   42,   16,  129,   47,   15,  # 5780 H      129,   52,   15,  129,   57,   15,  129,   61,   16,  129,  # 5790 H       65,   17,  129,   60,   27,  129,   56,   36,  129,   51,  # 5800 H       42,  129,   46,   43,  129,   42,   43,  129,   37,   44,  # 5810 H      129,   33,   43,  129,   30,   42,  129,   33,   34,  129,  # 5820 H       38,   25,  129,   42,   16,  129,   47,   15,  129,   52,  # 5830 H       15,  129,   57,   15,  129,   61,   16,  129,   66,   16,  # 5840 H      129,   71,   16,  129,   76,   16,  129,   80,   16,  129,  # 5850 H       85,   16,  117,    4,  129,   90,   16,  117,    4,  129,  # 5860 H       95,   16,  117,    4,  129,  100,   21,  129,  105,   16,  # 5870 H      129,  110,   11,  129,  114,    7,  129,  117,    4,  132,  # 5880 H      193,  129,   30,    4,  117,    4,  132,   30,    4,  115,  # 5890 H        6,  129,   30,    4,  112,    9,  129,   30,    6,  109,  # 5900 H       12,  129,   30,    9,  106,   15,  129,   30,   11,  103,  # 5910 H       18,  129,   30,   14,  100,   21,  129,   30,    4,   38,  # 5920 H        9,   98,   23,  129,   30,    4,   40,   10,   95,   26,  # 5930 H      129,   30,    4,   43,    9,   92,   29,  129,   46,    9,  # 5940 H       89,   32,  129,   49,    8,   86,   28,  117,    4,  129,  # 5950 H       51,    9,   83,   28,  117,    4,  129,   54,    9,   80,  # 5960 H       28,  117,    4,  129,   57,    8,   77,   28,  117,    4,  # 5970 H      129,   59,    9,   74,   28,  129,   62,   37,  129,   64,  # 5980 H       33,  129,   66,   28,  129,   63,   28,  129,   60,   28,  # 5990 H      129,   57,   28,  129,   54,   33,  129,   51,   39,  129,  # 6000 H       48,   29,   83,    9,  129,   30,    4,   45,   29,   86,  # 6010 H        9,  129,   30,    4,   42,   29,   89,    9,  129,   30,  # 6020 H        4,   39,   29,   92,    8,  129,   30,    4,   36,   29,  # 6030 H       94,    9,  129,   30,   32,   97,    9,  129,   30,   29,  # 6040 H      100,    8,  117,    4,  129,   30,   26,  103,    8,  117,  # 6050 H        4,  129,   30,   23,  105,    9,  117,    4,  129,   30,  # 6060 H       20,  108,   13,  129,   30,   18,  111,   10,  129,   30,  # 6070 H       15,  113,    8,  129,   30,   12,  116,    5,  129,   30,  # 6080 H        9,  117,    4,  129,   30,    6,  117,    4,  129,   30,  # 6090 H        4,  117,    4,  132,  193,  129,  117,    4,  132,  114,  # 6100 H        7,  129,  111,   10,  129,  108,   13,  129,  105,   16,  # 6110 H      129,  102,   19,  129,  100,   21,  129,   96,   25,  129,  # 6120 H       93,   28,  129,   90,   31,  129,   87,   34,  129,   84,  # 6130 H       30,  117,    4,  129,   30,    4,   81,   30,  117,    4,  # 6140 H      129,   30,    4,   78,   30,  117,    4,  129,   30,    4,  # 6150 H       75,   30,  117,    4,  129,   30,    4,   72,   30,  129,  # 6160 H       30,   69,  129,   30,   66,  129,   30,   63,  129,   30,  # 6170 H       60,  129,   30,   57,  129,   30,   54,  129,   30,   51,  # 6180 H      129,   30,   48,  129,   30,   51,  129,   30,    4,   73,  # 6190 H       12,  129,   30,    4,   76,   12,  129,   30,    4,   80,  # 6200 H       12,  129,   30,    4,   83,   12,  129,   87,   12,  129,  # 6210 H       90,   12,  117,    4,  129,   94,   11,  117,    4,  129,  # 6220 H       97,   12,  117,    4,  129,  101,   12,  117,    4,  129,  # 6230 H      104,   17,  129,  108,   13,  129,  111,   10,  129,  115,  # 6240 H        6,  129,  117,    4,  134,  193,  129,   30,    1,  103,  # 6250 H       18,  129,   30,    4,  103,   18,  129,   30,    7,  103,  # 6260 H       18,  129,   30,    9,  103,   18,  129,   30,   12,  110,  # 6270 H       11,  129,   30,   15,  113,    8,  129,   30,   18,  114,  # 6280 H        7,  129,   30,   21,  116,    5,  129,   30,   24,  116,  # 6290 H        5,  129,   30,   27,  117,    4,  129,   30,   30,  117,  # 6300 H        4,  129,   30,   33,  117,    4,  129,   30,    4,   37,  # 6310 H       28,  117,    4,  129,   30,    4,   40,   28,  117,    4,  # 6320 H      129,   30,    4,   42,   29,  117,    4,  129,   30,    4,  # 6330 H       45,   29,  117,    4,  129,   30,    4,   48,   29,  117,  # 6340 H        4,  129,   30,    4,   51,   29,  117,    4,  129,   30,  # 6350 H        4,   54,   29,  117,    4,  129,   30,    4,   57,   29,  # 6360 H      117,    4,  129,   30,    4,   59,   30,  117,    4,  129,  # 6370 H       30,    4,   62,   30,  117,    4,  129,   30,    4,   65,  # 6380 H       30,  117,    4,  129,   30,    4,   68,   30,  117,    4,  # 6390 H      129,   30,    4,   71,   30,  117,    4,  129,   30,    4,  # 6400 H       74,   30,  117,    4,  129,   30,    4,   77,   30,  117,  # 6410 H        4,  129,   30,    4,   80,   30,  117,    4,  129,   30,  # 6420 H        4,   83,   30,  117,    4,  129,   30,    4,   86,   35,  # 6430 H      129,   30,    4,   89,   32,  129,   30,    4,   91,   30,  # 6440 H      129,   30,    4,   94,   27,  129,   30,    5,   97,   24,  # 6450 H      129,   30,    5,  100,   21,  129,   30,    7,  103,   18,  # 6460 H      129,   30,    8,  106,   15,  129,   30,   11,  109,   12,  # 6470 H      129,   30,   18,  112,    9,  129,   30,   18,  115,    6,  # 6480 H      129,   30,   18,  117,    4,  129,   30,   18,  120,    1,  # 6490 H      129,  193,  129,   42,    8,  129,   38,   16,  129,   36,  # 6500 H       20,  129,   34,   24,   71,    5,  129,   33,   26,   69,  # 6510 H       10,  129,   32,   28,   68,   13,  129,   31,   30,   68,  # 6520 H       14,  129,   31,    9,   52,    9,   68,   15,  129,   30,  # 6530 H        8,   54,    8,   69,   14,  129,   30,    7,   55,    7,  # 6540 H       71,    4,   78,    6,  129,   30,    6,   56,    6,   79,  # 6550 H        5,  129,   30,    6,   56,    6,   80,    4,  130,   31,  # 6560 H        5,   56,    5,   80,    4,  129,   31,    5,   56,    5,  # 6570 H       79,    5,  129,   32,    5,   55,    5,   78,    6,  129,  # 6580 H       33,    5,   54,    5,   77,    7,  129,   34,    6,   52,  # 6590 H        6,   74,    9,  129,   35,   48,  129,   33,   49,  129,  # 6600 H       32,   49,  129,   31,   49,  129,   30,   49,  129,   30,  # 6610 H       47,  129,   30,   45,  129,   30,   41,  129,   30,    6,  # 6620 H      129,   30,    4,  129,   30,    3,  129,   30,    2,  129,  # 6630 H      193,  129,   30,    4,  117,    4,  130,   31,   90,  136,  # 6640 H       37,    5,   72,    5,  129,   35,    5,   74,    5,  129,  # 6650 H       33,    5,   76,    5,  129,   32,    5,   77,    5,  129,  # 6660 H       31,    5,   78,    5,  129,   31,    4,   79,    4,  129,  # 6670 H       30,    5,   79,    5,  131,   30,    6,   78,    6,  129,  # 6680 H       30,    7,   77,    7,  129,   31,    8,   75,    8,  129,  # 6690 H       31,   11,   72,   11,  129,   32,   15,   67,   15,  129,  # 6700 H       33,   48,  129,   34,   46,  129,   35,   44,  129,   37,  # 6710 H       40,  129,   39,   36,  129,   42,   30,  129,   46,   22,  # 6720 H      129,  193,  129,   48,   18,  129,   43,   28,  129,   41,  # 6730 H       32,  129,   39,   36,  129,   37,   40,  129,   35,   44,  # 6740 H      129,   34,   46,  129,   33,   13,   68,   13,  129,   32,  # 6750 H        9,   73,    9,  129,   32,    7,   75,    7,  129,   31,  # 6760 H        6,   77,    6,  129,   31,    5,   78,    5,  129,   30,  # 6770 H        5,   79,    5,  129,   30,    4,   80,    4,  133,   31,  # 6780 H        3,   79,    4,  129,   31,    4,   79,    4,  129,   32,  # 6790 H        3,   78,    4,  129,   32,    4,   76,    6,  129,   33,  # 6800 H        4,   74,    7,  129,   34,    4,   72,    8,  129,   35,  # 6810 H        5,   72,    7,  129,   37,    5,   73,    4,  129,   39,  # 6820 H        4,   74,    1,  129,  129,  193,  129,   46,   22,  129,  # 6830 H       42,   30,  129,   39,   36,  129,   37,   40,  129,   35,  # 6840 H       44,  129,   34,   46,  129,   33,   48,  129,   32,   15,  # 6850 H       67,   15,  129,   31,   11,   72,   11,  129,   31,    8,  # 6860 H       75,    8,  129,   30,    7,   77,    7,  129,   30,    6,  # 6870 H       78,    6,  129,   30,    5,   79,    5,  131,   31,    4,  # 6880 H       79,    4,  129,   31,    5,   78,    5,  129,   32,    5,  # 6890 H       77,    5,  129,   33,    5,   76,    5,  129,   35,    5,  # 6900 H       74,    5,  117,    4,  129,   37,    5,   72,    5,  117,  # 6910 H        4,  129,   30,   91,  136,   30,    4,  130,  193,  129,  # 6920 H       48,   18,  129,   43,   28,  129,   41,   32,  129,   39,  # 6930 H       36,  129,   37,   40,  129,   35,   44,  129,   34,   46,  # 6940 H      129,   33,   13,   55,    4,   68,   13,  129,   32,    9,  # 6950 H       55,    4,   73,    9,  129,   32,    7,   55,    4,   75,  # 6960 H        7,  129,   31,    6,   55,    4,   77,    6,  129,   31,  # 6970 H        5,   55,    4,   78,    5,  129,   30,    5,   55,    4,  # 6980 H       79,    5,  129,   30,    4,   55,    4,   80,    4,  132,  # 6990 H       30,    4,   55,    4,   79,    5,  129,   31,    3,   55,  # 7000 H        4,   78,    5,  129,   31,    4,   55,    4,   77,    6,  # 7010 H      129,   32,    3,   55,    4,   75,    7,  129,   32,    4,  # 7020 H       55,    4,   73,    9,  129,   33,    4,   55,    4,   68,  # 7030 H       13,  129,   34,    4,   55,   25,  129,   35,    5,   55,  # 7040 H       24,  129,   37,    5,   55,   22,  129,   39,    4,   55,  # 7050 H       20,  129,   55,   18,  129,   55,   16,  129,   55,   11,  # 7060 H      129,  193,  129,   80,    4,  129,   30,    4,   80,    4,  # 7070 H      130,   30,   78,  129,   30,   82,  129,   30,   85,  129,  # 7080 H       30,   87,  129,   30,   88,  129,   30,   89,  129,   30,  # 7090 H       90,  130,   30,    4,   80,    4,  115,    6,  129,   30,  # 7100 H        4,   80,    4,  117,    4,  129,   80,    4,  105,    6,  # 7110 H      117,    4,  129,   80,    4,  103,   10,  116,    5,  129,  # 7120 H       80,    4,  102,   19,  129,   80,    4,  101,   19,  129,  # 7130 H      101,   19,  129,  101,   18,  129,  102,   16,  129,  103,  # 7140 H       12,  129,  105,    6,  129,  193,  129,   12,   10,   59,  # 7150 H       11,  129,    9,   16,   55,   19,  129,    7,   20,   53,  # 7160 H       23,  129,    6,    7,   23,    5,   32,    6,   51,   27,  # 7170 H      129,    4,    7,   25,   16,   50,   29,  129,    3,    6,  # 7180 H       27,   16,   49,   31,  129,    2,    6,   28,   16,   48,  # 7190 H       33,  129,    1,    6,   27,   18,   47,   35,  129,    1,  # 7200 H        6,   27,   31,   71,   12,  129,    1,    5,   26,   15,  # 7210 H       44,   10,   75,    8,  129,    1,    5,   25,   14,   45,  # 7220 H        7,   77,    7,  129,    1,    5,   25,   13,   45,    5,  # 7230 H       79,    5,  129,    1,    5,   24,   14,   45,    4,   80,  # 7240 H        4,  129,    1,    5,   24,   13,   45,    4,   80,    4,  # 7250 H      129,    1,    5,   23,   14,   45,    4,   80,    4,  129,  # 7260 H        1,    5,   23,   13,   45,    4,   80,    4,  129,    1,  # 7270 H        6,   22,   13,   45,    5,   79,    5,  129,    1,    6,  # 7280 H       21,   14,   45,    7,   77,    7,  129,    1,    7,   21,  # 7290 H       13,   46,    8,   75,    8,  129,    1,    8,   20,   13,  # 7300 H       46,   12,   71,   12,  129,    1,   10,   18,   15,   47,  # 7310 H       35,  129,    2,   30,   48,   33,  129,    3,   29,   49,  # 7320 H       32,  129,    4,   27,   50,   31,  129,    5,   25,   51,  # 7330 H       27,   80,    2,   86,    4,  129,    7,   21,   53,   23,  # 7340 H       80,    3,   85,    6,  129,    9,   17,   55,   19,   80,  # 7350 H       12,  129,   12,   12,   59,   11,   81,   11,  129,   82,  # 7360 H       10,  129,   84,    7,  129,   86,    4,  129,  193,  129,  # 7370 H       30,    4,  117,    4,  130,   30,   91,  136,   30,    4,  # 7380 H       72,    5,  129,   30,    4,   74,    5,  129,   75,    5,  # 7390 H      129,   76,    5,  129,   76,    6,  129,   77,    6,  130,  # 7400 H       77,    7,  130,   76,    8,  129,   30,    4,   75,    9,  # 7410 H      129,   30,    4,   72,   12,  129,   30,   54,  129,   30,  # 7420 H       53,  130,   30,   52,  129,   30,   51,  129,   30,   49,  # 7430 H      129,   30,   46,  129,   30,   42,  129,   30,    4,  130,  # 7440 H      193,  129,   30,    4,   80,    4,  129,   30,    4,   80,  # 7450 H        4,  100,    6,  129,   30,   54,   98,   10,  129,   30,  # 7460 H       54,   97,   12,  129,   30,   54,   96,   14,  131,   30,  # 7470 H       54,   97,   12,  129,   30,   54,   98,   10,  129,   30,  # 7480 H       54,  100,    6,  129,   30,    4,  130,  193,  129,    7,  # 7490 H        6,  129,    4,   11,  129,    3,   13,  129,    2,   14,  # 7500 H      129,    1,   15,  130,    1,    3,    6,    9,  129,    1,  # 7510 H        3,    7,    6,  129,    1,    3,  130,    1,    4,  129,  # 7520 H        1,    5,   80,    4,  129,    1,    7,   80,    4,  100,  # 7530 H        6,  129,    2,   82,   98,   10,  129,    3,   81,   97,  # 7540 H       12,  129,    4,   80,   96,   14,  129,    5,   79,   96,  # 7550 H       14,  129,    7,   77,   96,   14,  129,   10,   74,   97,  # 7560 H       12,  129,   14,   70,   98,   10,  129,   19,   65,  100,  # 7570 H        6,  129,  193,  129,   30,    4,  117,    4,  130,   30,  # 7580 H       91,  136,   30,    4,   57,    9,  129,   30,    4,   55,  # 7590 H       12,  129,   52,   17,  129,   50,   20,  129,   48,   24,  # 7600 H      129,   46,   27,  129,   44,   21,   69,    6,  129,   41,  # 7610 H       22,   70,    6,   80,    4,  129,   30,    4,   39,   21,  # 7620 H       72,    6,   80,    4,  129,   30,    4,   36,   22,   73,  # 7630 H       11,  129,   30,   26,   75,    9,  129,   30,   23,   76,  # 7640 H        8,  129,   30,   21,   78,    6,  129,   30,   19,   79,  # 7650 H        5,  129,   30,   16,   80,    4,  129,   30,   14,   80,  # 7660 H        4,  129,   30,   12,  129,   30,   10,  129,   30,    7,  # 7670 H      129,   30,    5,  129,   30,    4,  130,  193,  129,   30,  # 7680 H        4,  117,    4,  130,   30,   91,  136,   30,    4,  130,  # 7690 H      193,  129,   30,    4,   80,    4,  130,   30,   54,  136,  # 7700 H       30,    4,   72,    5,  129,   30,    4,   74,    5,  129,  # 7710 H       75,    5,  129,   76,    5,  129,   30,    4,   75,    7,  # 7720 H      129,   30,    4,   74,    9,  129,   30,   54,  132,   30,  # 7730 H       53,  129,   30,   52,  129,   30,   51,  129,   30,   48,  # 7740 H      129,   30,    4,   72,    5,  129,   30,    4,   74,    5,  # 7750 H      129,   75,    5,  129,   76,    5,  129,   30,    4,   75,  # 7760 H        7,  129,   30,    4,   74,    9,  129,   30,   54,  132,  # 7770 H       30,   53,  129,   30,   52,  129,   30,   51,  129,   30,  # 7780 H       48,  129,   30,    4,  130,  193,  129,   30,    4,   80,  # 7790 H        4,  130,   30,   54,  136,   30,    4,   72,    5,  129,  # 7800 H       30,    4,   74,    5,  129,   75,    5,  129,   76,    5,  # 7810 H      129,   76,    6,  129,   77,    6,  130,   77,    7,  130,  # 7820 H       76,    8,  129,   30,    4,   75,    9,  129,   30,    4,  # 7830 H       72,   12,  129,   30,   54,  129,   30,   53,  130,   30,  # 7840 H       52,  129,   30,   51,  129,   30,   49,  129,   30,   46,  # 7850 H      129,   30,   42,  129,   30,    4,  130,  193,  129,   48,  # 7860 H       18,  129,   43,   28,  129,   41,   32,  129,   39,   36,  # 7870 H      129,   37,   40,  129,   35,   44,  129,   34,   46,  129,  # 7880 H       33,   13,   68,   13,  129,   32,    9,   73,    9,  129,  # 7890 H       32,    7,   75,    7,  129,   31,    6,   77,    6,  129,  # 7900 H       31,    5,   78,    5,  129,   30,    5,   79,    5,  129,  # 7910 H       30,    4,   80,    4,  132,   30,    5,   79,    5,  130,  # 7920 H       31,    5,   78,    5,  129,   31,    6,   77,    6,  129,  # 7930 H       32,    7,   75,    7,  129,   32,    9,   73,    9,  129,  # 7940 H       33,   13,   68,   13,  129,   34,   46,  129,   35,   44,  # 7950 H      129,   37,   40,  129,   39,   36,  129,   41,   32,  129,  # 7960 H       43,   28,  129,   48,   18,  129,  193,  129,    1,    3,  # 7970 H       80,    4,  130,    1,   83,  137,   37,    5,   72,    5,  # 7980 H      129,   35,    5,   74,    5,  129,   33,    5,   76,    5,  # 7990 H      129,   32,    5,   77,    5,  129,   31,    5,   78,    5,  # 8000 H      129,   31,    4,   79,    4,  129,   30,    5,   79,    5,  # 8010 H      131,   30,    6,   78,    6,  129,   30,    7,   77,    7,  # 8020 H      129,   31,    8,   75,    8,  129,   31,   11,   72,   11,  # 8030 H      129,   32,   15,   67,   15,  129,   33,   48,  129,   34,  # 8040 H       46,  129,   35,   44,  129,   37,   40,  129,   39,   36,  # 8050 H      129,   42,   30,  129,   46,   22,  129,  193,  129,   46,  # 8060 H       22,  129,   42,   30,  129,   39,   36,  129,   37,   40,  # 8070 H      129,   35,   44,  129,   34,   46,  129,   33,   48,  129,  # 8080 H       32,   15,   67,   15,  129,   31,   11,   72,   11,  129,  # 8090 H       31,    8,   75,    8,  129,   30,    7,   77,    7,  129,  # 8100 H       30,    6,   78,    6,  129,   30,    5,   79,    5,  131,  # 8110 H       31,    4,   79,    4,  129,   31,    5,   78,    5,  129,  # 8120 H       32,    5,   77,    5,  129,   33,    5,   76,    5,  129,  # 8130 H       35,    5,   74,    5,  129,   37,    5,   72,    5,  129,  # 8140 H        1,   83,  136,    1,    3,   80,    4,  130,  193,  129,  # 8150 H       30,    4,   80,    4,  130,   30,   54,  136,   30,    4,  # 8160 H       68,    6,  129,   30,    4,   70,    6,  129,   71,    7,  # 8170 H      129,   72,    7,  129,   73,    7,  129,   74,    7,  129,  # 8180 H       74,    8,  129,   75,    8,  130,   69,   15,  129,   67,  # 8190 H       17,  129,   66,   18,  129,   65,   19,  130,   65,   18,  # 8200 H      130,   66,   16,  129,   67,   13,  129,   69,    8,  129,  # 8210 H      193,  129,   30,   13,   64,    8,  129,   30,   13,   61,  # 8220 H       14,  129,   30,   13,   59,   18,  129,   30,   13,   57,  # 8230 H       22,  129,   33,    8,   56,   24,  129,   32,    7,   55,  # 8240 H       26,  129,   32,    6,   54,   28,  129,   31,    6,   53,  # 8250 H       16,   77,    6,  129,   31,    5,   53,   14,   79,    4,  # 8260 H      129,   30,    5,   52,   14,   80,    4,  129,   30,    5,  # 8270 H       52,   13,   80,    4,  129,   30,    4,   52,   13,   80,  # 8280 H        4,  129,   30,    4,   52,   12,   80,    4,  129,   30,  # 8290 H        4,   51,   13,   80,    4,  130,   30,    4,   50,   13,  # 8300 H       79,    5,  129,   30,    4,   50,   13,   78,    5,  129,  # 8310 H       30,    5,   49,   14,   77,    6,  129,   31,    4,   49,  # 8320 H       13,   76,    6,  129,   31,    5,   48,   14,   75,    7,  # 8330 H      129,   32,    5,   47,   14,   73,    8,  129,   32,    6,  # 8340 H       45,   16,   71,   13,  129,   33,   27,   71,   13,  129,  # 8350 H       34,   26,   71,   13,  129,   35,   24,   71,   13,  129,  # 8360 H       37,   20,  129,   39,   16,  129,   43,    9,  129,  193,  # 8370 H      129,   80,    4,  131,   41,   56,  129,   37,   60,  129,  # 8380 H       35,   62,  129,   33,   64,  129,   32,   65,  129,   31,  # 8390 H       66,  129,   30,   67,  130,   30,   11,   80,    4,  129,  # 8400 H       30,    9,   80,    4,  129,   30,    8,   80,    4,  129,  # 8410 H       31,    7,   80,    4,  129,   31,    6,  129,   32,    5,  # 8420 H      129,   33,    5,  129,   35,    4,  129,   38,    3,  129,  # 8430 H      193,  129,   80,    4,  130,   42,   42,  129,   38,   46,  # 8440 H      129,   35,   49,  129,   33,   51,  129,   32,   52,  129,  # 8450 H       31,   53,  130,   30,   54,  129,   30,   12,  129,   30,  # 8460 H        9,  129,   30,    8,  129,   30,    7,  130,   31,    6,  # 8470 H      130,   32,    6,  129,   33,    5,  129,   34,    5,  129,  # 8480 H       35,    5,   80,    4,  129,   37,    5,   80,    4,  129,  # 8490 H       30,   54,  136,   30,    4,  130,  193,  129,   80,    4,  # 8500 H      130,   77,    7,  129,   74,   10,  129,   70,   14,  129,  # 8510 H       66,   18,  129,   62,   22,  129,   59,   25,  129,   55,  # 8520 H       29,  129,   51,   33,  129,   47,   37,  129,   44,   32,  # 8530 H       80,    4,  129,   40,   32,   80,    4,  129,   36,   32,  # 8540 H      129,   32,   33,  129,   30,   31,  129,   33,   24,  129,  # 8550 H       36,   17,  129,   40,   12,  129,   44,   12,  129,   48,  # 8560 H       12,  129,   51,   13,  129,   55,   13,  129,   59,   13,  # 8570 H       80,    4,  129,   63,   13,   80,    4,  129,   67,   17,  # 8580 H      129,   71,   13,  129,   74,   10,  129,   78,    6,  129,  # 8590 H       80,    4,  131,  193,  129,   80,    4,  130,   77,    7,  # 8600 H      129,   74,   10,  129,   70,   14,  129,   66,   18,  129,  # 8610 H       62,   22,  129,   59,   25,  129,   55,   29,  129,   51,  # 8620 H       33,  129,   47,   37,  129,   44,   32,   80,    4,  129,  # 8630 H       40,   32,   80,    4,  129,   36,   32,  129,   32,   33,  # 8640 H      129,   30,   31,  129,   33,   24,  129,   36,   17,  129,  # 8650 H       40,   12,  129,   44,   12,  129,   47,   13,  129,   44,  # 8660 H       20,  129,   40,   28,  129,   36,   31,  129,   32,   32,  # 8670 H      129,   30,   30,  129,   33,   24,  129,   36,   17,  129,  # 8680 H       40,   12,  129,   44,   12,  129,   48,   12,  129,   51,  # 8690 H       13,  129,   55,   13,  129,   59,   13,   80,    4,  129,  # 8700 H       63,   13,   80,    4,  129,   67,   17,  129,   71,   13,  # 8710 H      129,   74,   10,  129,   78,    6,  129,   80,    4,  131,  # 8720 H      193,  129,   30,    4,   80,    4,  130,   30,    4,   79,  # 8730 H        5,  129,   30,    5,   77,    7,  129,   30,    6,   74,  # 8740 H       10,  129,   30,    8,   72,   12,  129,   30,   11,   69,  # 8750 H       15,  129,   30,   13,   67,   17,  129,   30,    4,   37,  # 8760 H        8,   64,   20,  129,   30,    4,   39,    8,   62,   22,  # 8770 H      129,   41,    8,   59,   25,  129,   43,    8,   57,   27,  # 8780 H      129,   45,    8,   55,   22,   80,    4,  129,   47,   27,  # 8790 H       80,    4,  129,   49,   23,  129,   47,   22,  129,   44,  # 8800 H       23,  129,   42,   22,  129,   30,    4,   39,   27,  129,  # 8810 H       30,    4,   37,   31,  129,   30,   27,   62,    8,  129,  # 8820 H       30,   25,   64,    8,  129,   30,   22,   66,    8,   80,  # 8830 H        4,  129,   30,   20,   68,    8,   80,    4,  129,   30,  # 8840 H       17,   70,    8,   80,    4,  129,   30,   15,   73,   11,  # 8850 H      129,   30,   12,   75,    9,  129,   30,   10,   77,    7,  # 8860 H      129,   30,    7,   79,    5,  129,   30,    5,   80,    4,  # 8870 H      129,   30,    4,   80,    4,  130,  193,  129,    4,    5,  # 8880 H       80,    4,  129,    2,    9,   80,    4,  129,    1,   11,  # 8890 H       77,    7,  129,    1,   12,   74,   10,  129,    1,   12,  # 8900 H       70,   14,  129,    1,   12,   66,   18,  129,    1,   11,  # 8910 H       62,   22,  129,    2,    9,   59,   25,  129,    4,   11,  # 8920 H       55,   29,  129,    7,   12,   51,   33,  129,   10,   12,  # 8930 H       47,   37,  129,   14,   12,   44,   32,   80,    4,  129,  # 8940 H       17,   13,   40,   32,   80,    4,  129,   21,   13,   36,  # 8950 H       32,  129,   25,   40,  129,   29,   32,  129,   33,   24,  # 8960 H      129,   36,   17,  129,   40,   12,  129,   44,   12,  129,  # 8970 H       48,   12,  129,   51,   13,  129,   55,   13,  129,   59,  # 8980 H       13,   80,    4,  129,   63,   13,   80,    4,  129,   67,  # 8990 H       17,  129,   71,   13,  129,   74,   10,  129,   78,    6,  # 9000 H      129,   80,    4,  131,  193,  129,   30,    1,   71,   13,  # 9010 H      129,   30,    3,   71,   13,  129,   30,    6,   71,   13,  # 9020 H      129,   30,    9,   75,    9,  129,   30,   11,   77,    7,  # 9030 H      129,   30,   14,   79,    5,  129,   30,   17,   79,    5,  # 9040 H      129,   30,   19,   80,    4,  129,   30,   22,   80,    4,  # 9050 H      129,   30,   25,   80,    4,  129,   30,   27,   80,    4,  # 9060 H      129,   30,    4,   36,   24,   80,    4,  129,   30,    4,  # 9070 H       38,   25,   80,    4,  129,   30,    4,   41,   24,   80,  # 9080 H        4,  129,   30,    4,   44,   24,   80,    4,  129,   30,  # 9090 H        4,   46,   25,   80,    4,  129,   30,    4,   49,   25,  # 9100 H       80,    4,  129,   30,    4,   52,   24,   80,    4,  129,  # 9110 H       30,    4,   54,   30,  129,   30,    4,   57,   27,  129,  # 9120 H       30,    4,   59,   25,  129,   30,    4,   62,   22,  129,  # 9130 H       30,    4,   65,   19,  129,   30,    5,   67,   17,  129,  # 9140 H       30,    5,   70,   14,  129,   30,    7,   73,   11,  129,  # 9150 H       30,    9,   76,    8,  129,   30,   13,   78,    6,  129,  # 9160 H       30,   13,   81,    3,  129,   30,   13,  129,  193,    2,  # 9170 H        9,   59,   25,  129,    4,   11,   55,   29,  129,    7,  # 9180 H       12,   51,   33,  129,   10,   12,   47,   37,  129,   14,  # 9190 H       12,   44,   32,   80,    4,  129,   17,   13,   40,   32,  # 9200 H       80,    4,  129,   21,   13,   36,   32,  129,   25,   40,  # 9210 H      129,   29,   32,  129,   33,   24,  129,   36,   17,  129,  # 9220 H       40,   12,  129,   44,   12,  129,   48,   12,  129,   51,  # 9230 H       13,  129,   55,   13,  129,   59,   13,   80,    4,  129,  # 9240 H       63,   13,   80,    4,  129,   67,   17,  129,   71,   13,  # 9250 H      129,   74,   10,  129,   78,    6,  129,   80,    4,  131,  # 9260 H      193                                                         # 9270      ); } # BEGIN block7   __END__1   =head1 NAME   & banner - print large banner on printer   =head1 SYNOPSIS    banner [B<-w>I<n>] message ...   =head1 DESCRIPTION  : Banner prints a large, high quality banner on the standard: output.  If the message is omitted,  it  prompts  for  and= reads one line of its standard input.  If B<-w> is given, the = output is scrunched down from a width of 132 to I<n> ,  suit-9 able  for a narrow terminal.    : The output should be printed on a hard-copy device, up  to: 132  columns  wide,  with no breaks between the pages. The: volume is great enough that you may want a  printer  or  a: fast   hardcopy  terminal,  but  if  you  are  patient,  a- decwriter or other 300 baud terminal will do.    =head1 BUGS3  A The following bugs are preserved from the implementation on which  this one is based:   =over4   =item *5: Several ASCII characters are not defined, notably <, >, [,: ],  \,  ^,  _, {, }, |, and ~.  Also, the characters ", ',. and & are funny looking (but in a useful way.)   =item * = The B<-w> option is implemented by  skipping  some  rows  and : columns.   The  smaller  it gets, the grainier the output.# Sometimes it runs letters together.    =back7    Any new bugs are as yet unknown.  
 =head1 AUTHOR5  F The Perl implementation of I<banner> was written by Jonathan Feinberg, I<jdf@pobox.com>.4  A The C source on which this implementation is based was written by  Mark Horton.   =head1 COPYRIGHT and LICENSE  G Because I consider this program to be, in essence, a B<modification> of @ the BSD version of /usr/games/banner, I am bound to preserve the" following copyright notice intact:    Copyright (c) 1980, 1993, 19941D  	The Regents of the University of California.  All rights reserved.   C  Redistribution and use in source and binary forms, with or without1C  modification, are permitted provided that the following conditions 	  are met: B  1. Redistributions of source code must retain the above copyrightA     notice, this list of conditions and the following disclaimer. E  2. Redistributions in binary form must reproduce the above copyright9G     notice, this list of conditions and the following disclaimer in the9H     documentation and/or other materials provided with the distribution.I  3. All advertising materials mentioning features or use of this software /     must display the following acknowledgement: ?  	This product includes software developed by the University of ,  	California, Berkeley and its contributors.H  4. Neither the name of the University nor the names of its contributorsI     may be used to endorse or promote products derived from this software4.     without specific prior written permission.   H  THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' ANDF  ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THEK  IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE I  ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE0K  FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL H  DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODSF  OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)K  HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT3J  LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAYG  OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 
  SUCH DAMAGE.4  1   ------------------------------    Date: 24 Aug 2001 05:24:28 -07005 From: graham@the-shades.demon.co.uk (Graham Harrison)4* Subject: On board SCSI supported on DS20E?= Message-ID: <ed1713eb.0108240424.628f2613@posting.google.com>    HI,   @ Does anyone know if the on board SCSI is supported on the DS20E?9 I know the Adaptec's on a DS20 aren't and they appear as  < "Device PK<x>0:, device type unknown", but the 7895's on our* DS20E's actually appear as Adaptec 7895's.  ? I've run out of PCI slots any still need 1 more, so if I could  @ move the internal DDS3 onto one of the 7895's that would save me@ the headache of having to reconfigure the hardware with a couple of multi-function cards.   Cheers,  Graham.4   ------------------------------  % Date: Fri, 24 Aug 2001 12:58:55 -0400  From: Tym_Stegner@cca-int.com , Subject: Re: OT: TOPS-20 and TOPS-10 live onA Message-ID: <OF0207E2FE.C302BC2C-ON85256AB2.005CD546@cca-int.com>9   Vic Mendham wrote: |6 |Hey,  |4G |Someone out there is still using Tops-10 and Tops-20? I'll have to let9F |my boss, know we're running about 8 boxes, with Tops-20, but on newerC |platforms called SC25 and SC40. These babies just keep running....  |1  B I happen to know that there must still be a few sites running thisE hardware, or the emulators, because around the Y2K checkpoint, we had F inquiries from sites around the world for Y2K versions of System 1022,  which runs on the 10's & 20's...  C Unfortunately, I can't tell you specifically who/where they are, as1D somebody else handled that project...  I think there was one site in> Cambridge, MA/USA, and another is the Danish Railway system...   Tym Stegner6
 S1032 Support    ------------------------------  + Date: Fri, 24 Aug 2001 10:30:00 -0700 (PDT)4. From: Fabio Cardoso <fabiopenvms@yahoo.com.br> Subject: Queue/Entry management @ Message-ID: <20010824173000.15839.qmail@web20203.mail.yahoo.com>   People    6 Do you know a way to avoid deleting entries in a queue/ using a "way" to protect the entries from being " deleted  by a priviliged account ? The problem is:9  / The development people have privilges to delete11 entries and sometimes, a "sleepy" cobol developer5, deletes an important entry in the queues.=20  5 We would like to create a way to protect entries from7 being deleted, just this.    Regards        =3D=3D=3D=3D=3D L =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D7 F=E1bio dos Santos Cardoso OpenVMS System Manager Rio de Janeiro - Brazil  fabiopenvms@yahoo.com.brL =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D   2 __________________________________________________ Do You Yahoo!?H Make international calls for as low as $.04/minute with Yahoo! Messenger http://phonecard.yahoo.com/1   ------------------------------  % Date: Fri, 24 Aug 2001 13:38:14 -04000 From: William_Bochnik@acml.com# Subject: Re: Queue/Entry management > Message-ID: <OFF456B08A.483CA20E-ON85256AB2.0060C38A@acml.com>  ? read up on settin priv's on queues and only give them priv's to # delete jobs from their own queue's.1   help init/que/prot   give them each their own queue        H                                                                        =                             =20 H                     Fabio Cardoso                                      =                             =20 H                     <fabiopenvms@yah                To:  Info-VAX@Mvb.S= aic.Com                     =20 H                     oo.com.br>                      cc:                =                             =20 H                                             Subject:     Queue/Entry ma= nagement                    =20 H                     08/24/2001 01:30                                   =                             =20 H                     PM                                                 =                             =201H                                                                        =                             =20 H                                                                        =                             =20        People    6 Do you know a way to avoid deleting entries in a queue/ using a "way" to protect the entries from being5" deleted  by a priviliged account ? The problem is:   / The development people have privilges to delete51 entries and sometimes, a "sleepy" cobol developer#) deletes an important entry in the queues.   5 We would like to create a way to protect entries from6 being deleted, just this.1   Regards1       =3D=3D=3D=3D=3D I =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=  =3D=3D F=E1bio dos Santos Cardoso OpenVMS System Manager Rio de Janeiro - Brazil1 fabiopenvms@yahoo.com.brI =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=  =3D=3D  2 __________________________________________________ Do You Yahoo!?> Make international calls for as low as $.04/minute with Yahoo!	 Messenger  http://phonecard.yahoo.com/9          F ______________________________________________________________________;  The information contained in this transmission may contain @ privileged and confidential information and is intended only forA the use of the person(s) name above.  If you are not the intended5= recipient, or an employee or agent responsible for delivering 3 this message to the intended recipient, any review, @ dissemination, distribution or duplication of this communication? is strictly prohibited.  If you are not the intended recipient, A please contact the sender immediately by reply e-mail and destroy $ all copies of the original message.=   ------------------------------    Date: 24 Aug 2001 05:44:10 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)4@ Subject: Re: RWMBX - What to do, What to do! Help me if you can.3 Message-ID: <RI0P6FwhFIUI@eisner.encompasserve.org>   Z In article <tobfcd65gmhcec@corp.supernews.com>, "Bill Ames" <billames@accunet.net> writes:N > I checked the source code and when the create the mailbox they are not usingJ > bufquo. So it must be using DEFMBXBUFQUO.  With that said this is what I > found: > ! > SOUTHEAST>RUN SYS$SYSTEM:SYSGEN  > SYSGEN>  SHOW DEFMBXBUFQUOI > Parameter Name     Current    Default    Min.      Max.   Unit  Dynamic I > --------------     -------    -------    -------   ----   ----  ------- F > DEFMBXBUFQUO        1056       1056       256      64000  Bytes    D >0D > Are these really low numbers and can I increase them to get betterJ > performance.  The business system has way to many programs for me to getK > into the source and fix it all. I seem to only have these problems on one B > AXP server. Probably more business is being done on that region.  D The only number under discussion is "Current", and since it is equalF to the default it is by definition not low for VMS systems in general.  > If your particular application was not designed to specify its? requirement in the call to create the mailbox, then perhaps the5< author had in mind controlling it with the system parameter.@ This can lead to a slippery slope of system parameter adjustment= if you ever get two applications with different requirements.   ? You imply that the parameter value is the same on the other AXP  servers.  > But waiting for the mailbox is just a symptom.  What keeps the< mailbox emptying process from working faster ?  As Brian has: indicated, storing more of the overflow in mailbox buffers: will not really get the work done any faster, it will just< make SHOW SYSTEM look more uniform :-).   Unless, of course,< the process that is waiting for the mailbox has some _other_; activity to do, whose end result is _not_ to put stuff into  the mailbox.   ------------------------------    Date: 24 Aug 2001 08:12:50 -0500 From: briggs@encompasserve.org@ Subject: Re: RWMBX - What to do, What to do! Help me if you can.3 Message-ID: <vSe6gJfOhkGc@eisner.encompasserve.org>#  \ In article <3B8545A8.B18F3009@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes:L > If, however, the process hangs and stays in RWMBX forever, the trick is toL > ANA/SYSTEM, SET PROCESS <name of process in RWMBX), show proc/CHANNELS and- > that will tell you the mailbox name/number.5 > 6 > Then, you can write a simple DCL procedure tsuch as: >  > $OPEN/READ temp MBAxxx:9 > $loop  > $READ/end=finish temp test > $goto loop
 > $finish: > $exit    Or   $ COPY /LOG MBAxxx NL:  D Saves you a dangling open file named "temp" that will screw you over3 the next time you try to use it for something else.3   	John Briggs   ------------------------------  % Date: Fri, 24 Aug 2001 12:29:05 -04007- From: JF Mezei <jfmezei.spamnot@videotron.ca> @ Subject: Re: RWMBX - What to do, What to do! Help me if you can., Message-ID: <3B8680CF.6D7AC0DE@videotron.ca>   Larry Kilgallen wrote:@ > But waiting for the mailbox is just a symptom.  What keeps the0 > mailbox emptying process from working faster ?    H Perhaps transactions come in a bunch, followed by a quiet period, so the4 server process gets overwhelmed and then catches up.  K It also depends on whether the server process reads the mailbox in AST mode N and queues the transactions internally, or whether the server uses the mailboxL as a queue, where a transaction gets fully processed before the server reads/ the next one, hence a long delay between reads.2  J If the problem is simply a question of throughput where the client submitsN transactions faster than the server can read them, hence putting the client inM RWMBX *temporarily* while the server catches up, is there anything inherently  wrong with a process in RWMBX ?   N (Obviously, if the client is frozen during that time and the above would applyK only to unidirectional messages - if the server must send a message back to0L the client, and the client is stuck in RWMBX, then the client can't read theD message, and this may put the server in RWMBX, a deadlock situation.  	 QUESTION:   K RWMBX implies that it can be in read or write. Correct ? I understand how a K write request to a mailbox that is "full" will cause you to be RWMBXed. But6O under what circumstances would a process be put in RWMBX for a read operation ?    ------------------------------    Date: 24 Aug 2001 08:25:08 -0500+ From: young_r@encompasserve.org (Rob Young)73 Subject: Re: Single disk root versus multiple disks83 Message-ID: <WAYB4YvvXj+H@eisner.encompasserve.org>   [ In article <sb853644.021@AAASMTA.aaas.org>, "John Eisenschmidt" <jeisensc@aaas.org> writes:6   > N > And you may think you don't want to mirror your /tmp file system, but what =L > happens if the disk it's on dies? Think VMS - yeah performance smells if = > you shadow the page file,  >    	Small World.  #  E 	How about 10+ page files scattered around high-end controller based #G 	storage (i.e. RAIDed and write back cache enabled).  Been there, done # 	that!  ? 	I/O performance "smells" in the workstation world.  (Unless of9> 	course you are using Linux and boy is it so much better!  And@ 	VMS performance is so slow if you have a little bitty disk and ) 	you want to go *fast*!  Bla Bla blech! )9   				Rob4   ------------------------------  % Date: Fri, 24 Aug 2001 16:30:46 +0200 & From: Michael Joosten <joost@c-lab.de>( Subject: Re: Some postive points I hope.$ Message-ID: <3B866516.6EEA@c-lab.de>   Bill Todd wrote: >    > J > Anecdotal evidence is not necessarily persuasive.  E.g., the only *hard*L > hard-disk failure I've ever experienced was when the 5.25" 20 MB hard diskM > on my 1987 PC-XT failed to spin up about 3 years ago (a 2.1 GB Seagate 3.5" M > drive started getting read errors after about 3 years of service, but I was I > able to get everything off it and FAIK it may still more or less work).2 >   F Well, but... the early ST225 and similar RLL/MFM disks from Seagate atG THAT time were known to be cheap consumer stuff. (there was/is a saying1G or pun in Germany: 'Seagate: Sie geht oder sie geht nicht', alluding to1E the similar pronounciation of seagate and 'sie geht' (she goes)). The H M2266 from '91 was a heavy, solidly built server/workstation disks at at? least $1000. Back when the hard disk cases were bullet-proof...7  A Though... The 8" SMD drives popular just a few years earlier from > Fujitsu (m2372, M2382 with 800MB or 1.6GB formatted, IIRC) hadH unfortunately also their interesting problems: 'bit rot'. After about 18G months, you got more and more soft errors, because the magnetic density1D was apparently a little too high for the coating of the platters. AtG least after 2 years, a complete re-format was necessary. And no, just a9D 'refresh' run (read/write each sector) did not help, as this did notG refresh the sector headers. Considering the price of these beasts then,  well...   K > AFAIK the rated service life of commodity drives is still at most 5 yearsaK > (as contrasted with the MBTF, which should never be confused with servicenL > life, as MBTF may significantly exceed it), and I've seen little reason toM > doubt that it is usually attained.  But as I said in another response, I'vepK > heard from a couple of sources that the IBM 75 GB drives (one source said M > specifically those manufactured in Hungary) had defects, rather than simplyr > being marginal.h >   D Except the SMD experiences, I still have the feeling that quality ofE SCSI disks declined noticeably between, say, '92 and '96. As this wasoE the time where SCSI was also used in PCs, I tend to attribute that to F the usual quality decline due to increased cost efficiency. Sure, they! became cheaper, but - at a price!p  B I can vividly remember almost a dozen of ST11200 Hawks and ST32550F Barracudas with head crashes or other mechanical defects after just 12F or 13 months (OEM drives, just after the warranty period. Nice.)  ManyH of the old IBM 663 also went belly-up after about 4 years. Three QuantumH Vikings (9GB) also dead, after about 18 months (OEM of course, so out of
 warranty).  
 'Nuf said.   -- i* Michael Joosten, SBS C-LAB, joost@c-lab.de* Fuerstenallee 11, 33094 Paderborn, Germany, Phone: +49 5251 606127, Fax: +49 5251 6060658 C-LAB is a cooperation of University Paderborn & SIEMENS   ------------------------------  " Date: Fri, 24 Aug 01 15:30:03 +100 From: rok@nuk.uni-lj.siG& Subject: Re: Unexpected DIAGPACK error& Message-ID: <3b8672fe.0@NUK.Uni-Lj.Si>     Thanks. The actual command was:  ? $ initialize/cluster_size=1/highwater/nodata_check/index=BEGIN-e8    /directories=800/headers=200000/maximum_files=200000-8    /verified/owner_uic=[system]/system  $4$DKB300 NUKDS1  =  The error persists with /verified even after /noverified wasi used.   ;  The same command was used on two other volumes of the samei type with no error.   =  I was under impression that OpenVMS is consistent and BACKUP > would have the same objection - the test proved I was wrong :)  <  Nevertheless, I find difficult to retry the operation using1 another volume when I use the *only* volume left.h   Regards, Rok    1 In Article <3B850FA0.859781BD@clarityconnect.com>s3 "Mark D. Jilson" <jilly@clarityconnect.com> writes:   H >Redo the INIT command with /NOVERIFIED qualifier.  It appears that INITF >is getting confused by what it is attempting to read as the bad blockE >data.  If this INIT works then retry it again without the qualifier.4 >e >rok@nuk.uni-lj.si wrote:r >> n1 >> AlphaServer 4100 5/400 4MB, OpenVMS AXP V7.2-1o >> r	 >>  I gett >> d. >> %INIT-F-DIAGPACK, disk is a diagnostic pack >> c7 >> when I try to initialize an device type DEC RZ1DF-CBD8 >> (DS-RZ1DF-WW 9.1GB) disk which used to be volume 0 of1 >> multidisk ODS-2 converted to ODS-5 volume set.D >> T5 >>  As I do not like the suggested user action at allF7 >> (Retry the operation using another volume) I will beH >> gratefull for any hints.R >> L >> Regards,A >> RG >> Rok Vidmar                       Internet:  rok.vidmar@nuk.uni-lj.siSP >> National and University Library  Phone:     +386 1 421 5461 <-- changed againP >> Turjaska 1, SI-1000 Ljubljana    Fax:       +386 1 425 7293 <-- changed again >> SloveniaA >I >-- E >Jilly	- Working from Home in the Chemung River Valley - Lockwood, NYR1 >	- jilly@clarityconnect.com			- Brett Bodine fanT/ >	- Mark.Jilson@Compaq.com			- since 1975 or soO- >	- http://www.jilly.baka.com               -- Regards,  D Rok Vidmar                       Internet:  rok.vidmar@nuk.uni-lj.siM National and University Library  Phone:     +386 1 421 5461 <-- changed again1M Turjaska 1, SI-1000 Ljubljana    Fax:       +386 1 425 7293 <-- changed againh Slovenia   ------------------------------  % Date: Fri, 24 Aug 2001 09:13:16 +0100 % From: Alan Greig <a.greig@virgin.net>c5 Subject: Re: VMS high reliability needed by Air Forcey8 Message-ID: <dv2cotk1jcmjd1n5n61q0lcinkf5n57pjd@4ax.com>  D On 24 Aug 2001 01:08:50 GMT, Joe Heimann <heimann@nog.ecs.umass.edu> wrote:  4 >Michael D. Ober <mdo.@.wakeassoc.com.nospam> wrote:I >> My favorite was when our Honeywell systems crashed, the first four hexh6 >> characters in the crashdump were frequently "DEAD". >> --1
 >> Mike Ober.  >rH >A department here used to have a Wang VS45 system that crashed during aF >thunderstorm.  When it came back up, its LED status display also read >  >        DE: >        AD   A I recall a VAX with an add in processor card running some librarytD software. The second processor/OS crashed with the error "Full power to the shields Mr. Sulu"     -- Alan   ------------------------------  % Date: Fri, 24 Aug 2001 12:32:16 +0100u0 From: andrew harrison <andrew.nospam@uk.sun.com>5 Subject: Re: VMS high reliability needed by Air Forcea* Message-ID: <3B863B40.89B2480A@uk.sun.com>   Alan Greig wrote:t > F > On 24 Aug 2001 01:08:50 GMT, Joe Heimann <heimann@nog.ecs.umass.edu> > wrote: > 6 > >Michael D. Ober <mdo.@.wakeassoc.com.nospam> wrote:K > >> My favorite was when our Honeywell systems crashed, the first four hex 8 > >> characters in the crashdump were frequently "DEAD". > >> --- > >> Mike Ober.- > > J > >A department here used to have a Wang VS45 system that crashed during aH > >thunderstorm.  When it came back up, its LED status display also read > >0
 > >        DEe
 > >        ADo > C > I recall a VAX with an add in processor card running some libraryeF > software. The second processor/OS crashed with the error "Full power > to the shields Mr. Sulu" >   ? How about "Error in local gravitational constant" which was an  ? error message that one C compiler "slightly modified" produced y for one of our programmers.d     Regardst Andrew Harrisont Enterprise IT Architecte   ------------------------------  % Date: Fri, 24 Aug 2001 11:37:06 +0100=/ From: Nigel Arnot <sysmgr@maxwell.ph.kcl.ac.uk>D= Subject: re: Wailing at Eunuchs (was: Wailing and Moaning...) 7 Message-ID: <00A01019.9122AE6F.20@maxwell.ph.kcl.ac.uk>D   > ^ > In article <3B854295.B71E9583@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes: > > Nigel Arnot wrote:K > >> Engineers know well that powerful buttons and levers must be protected J > >> by Molly-guards, interlocks, padlocks, and suchlike. The designers ofN > >> unix failed to appreciate this, and the above is a typical example of the > >> consequences. > > R > > Perhaps the VMS command "INIT" might be given some molly-guards ? Even on DOS,G > > FORMAT would prompt you if you are really sure you want to do this.r > >  >  > 	It does!  > ? > 	You can't INIT a disk drive that is mounted.  We could think 8 > 	about that and kick it around a bit, but it is enough > 	protection, IMHO. > 	 > 				Rob   I Furthermore, even if it isn't mounted, you need either to be its owner or=G to have VOLPRO privilege to succeed. Shared disks are normally INITted  E /SYSTEM, meaning that you have to be logged into a UIC=SYSTEM account D to damage one -- and logging in to such an account is (or should be)2 equivalent to raising the Molly-guard on a button.  A As for the "do you really want to do this?" approach, it's almost N useless. You woudn't have typed INIT if you didn't want to. The disaster is ifH you enter the wrong device as target - but you think at that moment thatE it's the right one, so almost certainly you answer "YES", even if you H haven't habituated yourself into doing so mechanically without thinking.  F (Not understanding the fact that the human subconscious is designed toD identify repeated stimulus-response patterns and to remove them fromG conscious oversight is perhaps the commonest failing of amateur safety-n@ system design. You cannot fight this. If it wasn't that way, you? couldn't walk or chew gum, let alone learn to ride a bicycle! )   G One thing I *would* like to see is INIT making a backup copy of the oldeC volume structure to unused space on the disk, and linking this to a A field in the new volume header. This is all that it would take to1@ make an UNINITIALIZE command easy (though of course, no promisesI are possible after you mount and start writing to the newly initted disk)C   	Yours,n
 		Nigel Arnotg- 		NRA@MAXWELL.PH.KCL.AC.UK                   .  7 		"In the beginning there was nothing, which exploded."=   ------------------------------    Date: 24 Aug 2001 05:55:16 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)n= Subject: Re: Wailing at Eunuchs (was: Wailing and Moaning...)p3 Message-ID: <whtIPOYqSTex@eisner.encompasserve.org>_  V In article <3B85CEBE.B19F5E6D@bigfoot.com>, Hamlyn Mootoo <univms@bigfoot.com> writes:   > Larry Kilgallen wrote:  F >> Is your suggestion that all programmers produce 64 versions of each! >> program ?  Or was that 2**64 ?a > E > What the heck are you talking about here? Please explain. Maybe youlI > think I'm suggesting to install normal VMS executables with higher thano! > normal privs? What do you mean?h  D You appear to be saying that programmers should "be prepared" for anD end user to install a program with privilege when the programmer has) not given specific directions to do that.e  C Somebody (perhaps not you) earlier in this thread mentioned BYPASS.0? The protection-related privileges are particularly difficult toM? defend against correctly for a complex program (that uses files0 for multiple purposes.  U http://ljk.com/LJK_SECURITY_DOCUMENTATION/LJK_SECURITY_D_012.HTML#tests_disk_instprivu   ------------------------------    Date: 24 Aug 2001 08:15:57 -0500+ From: young_r@encompasserve.org (Rob Young)u= Subject: Re: Wailing at Eunuchs (was: Wailing and Moaning...)n3 Message-ID: <1qkcoegFr2Hd@eisner.encompasserve.org>   V In article <3B856D15.155D5F60@bigfoot.com>, Hamlyn Mootoo <univms@bigfoot.com> writes: >  >  > Fred Kleinsorge wrote: >> -D >> Hamlyn Mootoo wrote in message <3B8534FE.CC397217@bigfoot.com>... >> > >> >B >> >Any VMS user holding just the BYPASS privilege can do the sameL >> >destruction to the file system as can "root" under Unix.  And since mostJ >> >programmers on VMS don't take the time to make sure their applicationsJ >> >can be installed with privs so that non-priv'ed users can just do whatK >> >they're supposed to, an incredibly LARGE percentage of user accounts on,I >> >VMS systems are running around with near full privs, or at least onessE >> >that will allow their accounts to get the rest with minimal work.o >> > >> s2 >> Remind me not to work on any of your systems.  C > After I get done inheriting the crap that other people leave, thea > systems are secure.  a  $ 	But by stating that, you basically:  5 		1)  Point out that users with BYPASS are a very bade 			thing in the paragraph abovee4 		2)  Point out you (like others that know what they4 			are doing) go about cleaning up train wrecks like 			BYPASS being doled out.  > 	So why even raise the issue?  Point out that some folks don't@ 	know what they are doing and implying that is a LARGE number of0 	folks?   And you ride to the rescue once again!   			Give us a break, man!   				Robf   ------------------------------    Date: 24 Aug 2001 08:57:37 -0500+ From: young_r@encompasserve.org (Rob Young)o= Subject: Re: Wailing at Eunuchs (was: Wailing and Moaning...)-3 Message-ID: <yqKm3XAIIC8r@eisner.encompasserve.org>c  V In article <3B85D13A.D4148E74@bigfoot.com>, Hamlyn Mootoo <univms@bigfoot.com> writes: >  >  > Rob Young wrote: >> sP >> In article <3B850B30.63DE@c-lab.de>, Michael Joosten <joost@c-lab.de> writes: >> > Hamlyn Mootoo wrote:  >> >>e >> >> Rob Young wrote: >> >> [snip]F >> >>   Two skilled Unix "persons" making a mistake in a shell script.U >> >> > >>         Two high profile clients dead in the water for a day or two.  Why? 
 >> >> > >>2 >> >> > >>         # rm -r ${variable_goes_here}/*	 >> >> > >mA >> >> > > Couldn't have been too "skilled" to make that mistake ! 	 >> >> > >  >> >> > S >> >> >         Absolutely skilled.  One of the people is a 15 year veteran, admin, O >> >> >         system configurator... one of the brighter guys you would know.g >> >> >-' >> >> >         Trust me... bad design.2 >> >>0H >> >> All this example serves to illustrate is the power of the the UnixM >> >> command syntax.  And whoever coded the above example, regardless of hisCM >> >> years of systems management experience, is obviously not experienced atoM >> >> writing shell scripts.  Actually, it's doubtful that whoever wrote thataJ >> >> had many months, let alone years of scripting experience. Any decentH >> >> shell scripter would NEVER, I repeat NEVER use that combination ofN >> >> syntax for just the exact reason of the variable being undefined.  Don'tN >> >> take my word for it. Post this example in any of the Unix newsgroups andK >> >> when they get done laughing maybe they'll respond if anyone can stillo) >> >> manage to type without cracking up.f >> >H >> > Right. This wasn't scripted, it was scribbled. Especially if it wasM >> > written to be executed as root. A more skilled shell scripter would have F >> > used at least a check if the variable was set before hand or used! >> > conditional evaluation, like ; >> >    rm -rf ${variable_goes_here-"UNDEFINED_DIRECTORY!"}eE >> > assuming of course that is no dir called /UNDEFINED_DIRECTORY! .t >> > >> sG >>         Notice I was careful enough to NOT say it was something thatnJ >>         could NOT be avoided.  But when you look at what took place, itN >>         was the transposition of a single letter in one case and all drivesM >>         were wiped out.  Good shell programming technique would have savedyM >>         the day, as you point out.  But what you are also doing is raisingXM >>         the bar to a certain level (a level that may not always be reached Q >>         via an oversight or a rush job) that may not always be there.  TestingiL >>         for the variable may have occured, can't recall.  You can test onL >>         one line and make a typo on the next line... happens all the time >>         you know. >>  P >>         Transposition of a single letter and all your drives are wiped clean.D > Transposition of a single number on defining the logical used by aJ > similar procedure on VMS wipes that disk clean too.  If it happens to beJ > the system disk or any disk on which production data resides, your going7 > to have to break out the backup tapes - same as Unix.# >   C 	"Wipes that disk clean too."  Yep, you just wiped a *single* disk.9F 	I'm talking about an entire system wiped clean such that the *system*@ 	restore takes days.  Major embarassement for all involved.  AndG 	much worse than what you are proposing for a similar disaster.  Apples- 	and Oranges as they say . . . 1   				Rob0   >> r# >>         Bad design ... trust me.  >> c. >>                                         Rob   ------------------------------  % Date: Fri, 24 Aug 2001 11:45:40 -0400e' From: "Bill Todd" <billtodd@foo.mv.com>e= Subject: Re: Wailing at Eunuchs (was: Wailing and Moaning...)n( Message-ID: <9m5sqb$433$1@pyrite.mv.net>  < "Nigel Arnot" <sysmgr@maxwell.ph.kcl.ac.uk> wrote in message1 news:00A01019.9122AE6F.20@maxwell.ph.kcl.ac.uk...    ...f  C > As for the "do you really want to do this?" approach, it's almosttJ > useless. You woudn't have typed INIT if you didn't want to. The disaster is ifsJ > you enter the wrong device as target - but you think at that moment thatG > it's the right one, so almost certainly you answer "YES", even if youhJ > haven't habituated yourself into doing so mechanically without thinking.  J It at least provides the opportunity to display prominently once again theK device you're actually INITting before you irrevocably commit to the action E (though as you said, making the action immediately revocable also hasi merit).2   - bill   ------------------------------  % Date: Fri, 24 Aug 2001 12:00:05 -0400g( From: Hamlyn Mootoo <univms@bigfoot.com>= Subject: Re: Wailing at Eunuchs (was: Wailing and Moaning...)s+ Message-ID: <3B867A05.3B711B31@bigfoot.com>e   Larry Kilgallen wrote: > X > In article <3B85CEBE.B19F5E6D@bigfoot.com>, Hamlyn Mootoo <univms@bigfoot.com> writes: >  > > Larry Kilgallen wrote: > H > >> Is your suggestion that all programmers produce 64 versions of each# > >> program ?  Or was that 2**64 ?  > >aG > > What the heck are you talking about here? Please explain. Maybe youtK > > think I'm suggesting to install normal VMS executables with higher thans# > > normal privs? What do you mean?n > F > You appear to be saying that programmers should "be prepared" for anF > end user to install a program with privilege when the programmer has+ > not given specific directions to do that.c  4 Not at all.  End users do not install images, systemF managers/programmers would only install specifically written images toB perform specific tasks.  The problem is that a lot of applicationsE written by commercial software houses are written by some programmerslG who ARE ignorant of VMS or do not take the time to install their imagesuG with privilege, instead require the users to have elevated privilges toiE run the software.  I would name names but that would be irresponsible5F dangerous in a public forum.  I can email you privately if you wish to* know of the ones I remember in my travels. > E > Somebody (perhaps not you) earlier in this thread mentioned BYPASS. E Yes, I did.  A lot of software package writers don't take the time tooF create the correct group/ACL structures and require users (in order toF run the software) to have READALL, CMKRNL, and in some cases yes, even BYPASS.   A > The protection-related privileges are particularly difficult tooA > defend against correctly for a complex program (that uses files  > for multiple purposes. > W > http://ljk.com/LJK_SECURITY_DOCUMENTATION/LJK_SECURITY_D_012.HTML#tests_disk_instpriv8   ------------------------------  % Date: Fri, 24 Aug 2001 12:23:43 -0400c( From: Hamlyn Mootoo <univms@bigfoot.com>= Subject: Re: Wailing at Eunuchs (was: Wailing and Moaning...)0+ Message-ID: <3B867F8F.20967038@bigfoot.com>4   Rob Young wrote: > X > In article <3B856D15.155D5F60@bigfoot.com>, Hamlyn Mootoo <univms@bigfoot.com> writes: > >o > >  > > Fred Kleinsorge wrote: > >>F > >> Hamlyn Mootoo wrote in message <3B8534FE.CC397217@bigfoot.com>... > >> > > >> >D > >> >Any VMS user holding just the BYPASS privilege can do the sameN > >> >destruction to the file system as can "root" under Unix.  And since mostL > >> >programmers on VMS don't take the time to make sure their applicationsL > >> >can be installed with privs so that non-priv'ed users can just do whatM > >> >they're supposed to, an incredibly LARGE percentage of user accounts onnK > >> >VMS systems are running around with near full privs, or at least ones G > >> >that will allow their accounts to get the rest with minimal work.l > >> > > >>2 > >> Remind me not to work on any of your systems.E > > After I get done inheriting the crap that other people leave, the  > > systems are secure.  > - >         But by stating that, you basically:R > E >                 1)  Point out that users with BYPASS are a very badt6 >                         thing in the paragraph aboveD >                 2)  Point out you (like others that know what theyK >                         are doing) go about cleaning up train wrecks like>1 >                         BYPASS being doled out.d > G >         So why even raise the issue?  Point out that some folks don't I >         know what they are doing and implying that is a LARGE number of 9 >         folks?   And you ride to the rescue once again!y > / >                         Give us a break, man!c > % >                                 Rob-G Remedial lessons in english, I'm sure, are offered in other newsgroups, C as are reading comprehension pointers.  I'll do my best here to use A small words for you though, and I'll try to avoid those confusingdG abbreviations.  Mister Kleinsorge made a comment about the systems thatsF I run, implying (I'm sorry. Definition: making it seem) that I installC images left and right or that I allow users to run around with full1H privs.  I was simply assuring him that this is quite the opposite.  As aE consultant, and having administered more different companies' systems1C than generally would a person who is permanent (unless his lifetime<E spanned 150 years), I have done a LOT of cleanup (security-wise) of a.H lot of VMS systems. If you took the time to read the whole thread (maybeD there is someone who can help you with this), you would see that theH whole discussion started as a comparison between Unix's root (superuser)H and the VMS privileging scheme.  My point is that in both systems, abuseD is possible.  I would even further argue here, that at least in UnixB systems, you know who has root (only one password) whereas in VMS,F regular users may be holding incredibly destructive privilges that youC might not catch because in some cases third party app installs will ( create accounts with elevated privilege.     HM   ------------------------------  % Date: Fri, 24 Aug 2001 02:47:16 -0400   From: John Santos <JOHN@egh.com>C Subject: Re: [Q]: How to get BACKUP to relabel tapes automatically? 5 Message-ID: <1010824023233.6553A-100000@Ives.egh.com>d  0 On Thu, 23 Aug 2001, Kenneth H. Fairfield wrote:  F >     I've been studying BACKUP's /LABEL=(<label-list>), /EXACT_ORDER,F > and /IGNORE=(LABEL_PROCESSING) qualifiers (yes, there is some mutual> > exclusion present), in order to solve the following problem.  6 [snip a lot of stuff I might regret snipping later...]   >     So.... > H >     Does anyone have a suggestion of how to either (a) force BACKUP toD > use the labels I want it to without pre-labeling the tapes, or (b)	 > provide I > a (robust, reliable?) method to determine how many tapes a given BACKUPpE > job as used (without scanning the log file, which will be held opent > during > this processing)?  >  >     Thanks, Kene  1 A couple of things you might want to investigate.f  E 1) The MRU (Media Robot Utility, but Compaq may have changed the namecE recently) allows you direct control of the loader.  You could write a E command procedure to load each tape in turn and label it (with INIT),WG then unload it.  When done, reload the 1st tape and run backup with the1 label=() [*] and exact_order.D  B (I once wrote a little program that calls the MRU library routinesE to check which tape from which slot is currently loaded in the drive,s@ and set a bunch of DCL symbols to that and similar related info.B This made writing DCL to manipulate the loader much easier.  I canC post details if anyone is interested.  Also, there is something fun A about giving "$ ROBOT do such-and-so" commands, and actually haven something mechanical happen :-)y    A 2) TAPESYS (I think that's the name, the tape management & backupuD product from Software Partners) may do all this.  If you have enoughC systems and tapes, this may prove very worthwhile, just for keepingi track of things.  G [*] Will the label=() qualifier do the right thing when you are alreadyeC on the 2nd tape when you start a new backup?  Or will it expect theOB current tape to have the 1st label in the list?  Or maybe you have> already solved this problem by checking the current label withD f$getdvi before each backup command, to determine which tape you are now on.a   --   John Santoso Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------   End of INFO-VAX 2001.470 ************************