1 INFO-VAX	Tue, 08 Aug 2000	Volume 2000 : Issue 441       Contents: Re: Affordable: A Clarification - Re: CETS2000 Registration is open and working  FTP server problems 
 genicom 3810s  Re: genicom 3810s - Re: How to copy the current screen to a file? $ How to fake a device full situation.( Re: How to fake a device full situation.( Re: How to fake a device full situation.( Re: How to fake a device full situation.( Re: How to fake a device full situation. I NEED imagemagick5.1.1   Last order dates for VAX systems. NRGE - (Buy Alert)                         1621 Re: OpenVMS ALPHA Port Of AOL Instant Messenger ? 1 Re: OpenVMS ALPHA Port Of AOL Instant Messenger ? ) Re: Patwhorks Performance and TCP NODELAY ) Re: Patwhorks Performance and TCP NODELAY ) Re: Patwhorks Performance and TCP NODELAY ) Re: Patwhorks Performance and TCP NODELAY * PC usability (was Re: dcps documentation?)0 Re: PCSI installation problem - missing database Re: Raid array 300 on VMS  Secure LYNX client for VMS Re: Spiraling GS140 Disks # Supressing Commands from a COM File ' Re: Supressing Commands from a COM File ' Re: Supressing Commands from a COM File 1 TCPIP problem on an Alphaserver with OpenVMS v7.2 5 Re: TCPIP problem on an Alphaserver with OpenVMS v7.2 5 Re: TCPIP problem on an Alphaserver with OpenVMS v7.2 5 Re: TCPIP problem on an Alphaserver with OpenVMS v7.2 5 Re: TCPIP problem on an Alphaserver with OpenVMS v7.2 5 Re: TCPIP problem on an Alphaserver with OpenVMS v7.2  RE: TELNET vs LAT speed  Re: TELNET vs LAT speed  Re: TELNET vs LAT speed  TSM and multiple NICs  Re: VMS Vs any other OS  Re: VMS Vs any other OS  Re: VMS Vs any other OS  Re: VMS Vs any other OS  Re: VMS Vs any other OS  Re: VMS Vs any other OS 2 Re: What Can I Do With An AlphaServer 1000A 4/233?2 Re: What Can I Do With An AlphaServer 1000A 4/233?  F ----------------------------------------------------------------------   Date: 8 Aug 2000 12:10:59 +0200 * From: eplan@kapsch.net (Peter LANGSTOEGER)( Subject: Re: Affordable: A Clarification* Message-ID: <398fdcb3$1@news.kapsch.co.at>  m In article <UCva5.297944$MB.5012604@news6.giganews.com>, "William Hymen" <t18_pilot@hotmail.spam.com> writes: 8 >Whats your definition of affordable?  A Compaq proliant3 >with NT is $6,200 and a Compaq DS10 OpenVMS system ? >with tcp/ip and equivalent processor / disk / memory is $6500.  > < >So who's cheaper?  Look at the value with the DS10 or DS20.  ; A Compaq Proliant is not what I call an entry level system.   J Most servers I know of started as a $1000 system and got upgraded over theH time. And only if they serve over say 500 users, the price of a Proliant; is justified. So, we are comparing apples with lemons here.   K I would love to run OpenVMS but I no longer get money for Non-Wintel boxes. J I'm not in the highest availability business and so I'm SOL. It's a shame.   --  < Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651; Network and OpenVMS system manager  Fax.    +43 1 81111-888 < FBFV/Information Services           E-mail  eplan@kapsch.netF <<< KAPSCH AG  Wagenseilgasse 1     PSImail PSI%(0232)281001141::EPLANH A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a realist"N "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998   ------------------------------   Date: 8 Aug 2000 14:13:22 -0500 1 From: kaplow_r@eisner.decus.org.mars (Bob Kaplow) 6 Subject: Re: CETS2000 Registration is open and working+ Message-ID: <cQcnXynGpMfJ@eisner.decus.org>   m In article <25JUL00.17573585@thuria.waisman.wisc.edu>, karcher@thuria.waisman.wisc.edu (Carl Karcher) writes: K > In a previous article, kaplow_r@eisner.decus.org.mars (Bob Kaplow) wrote:  > ->  O > ->The only option to get through the registration process is via credit card. N > ->What about those of us who need to register via company Purchase Order? InM > ->the past I've filled out the hard copy form, had management sign off, and = > ->our purchasing folks issue a PO, followed u with a check.  > ->  N > ->How do you register this time if your company is paying for the symposium? > H > If you check the box indicating you're a government organization, thenH > you'll get a choice of paying by a PO. No credit card number required.    L I'm not a government organization. I work for a publicly traded company. But> NFW am I paying for DECUS\\\\\CETS on my personal credit card!  K I finally did get a call back, and they asked me all the questions, and put K in a special phony credit card number so the registration would go through.   L A day later I got an email they my credit card number was invalid, and I hadK to send a new one. And the amount was the symposium only, not the seminars. K More email requesting an invoice so I can get a PO or have the company send  a check.  J Since then they've sent me 3 emails caliming to be invoices, which had the header, but no body.  I Now they tell me they can't book my hotel room because the Holiday Inn is K booked for the night of 6-October. But I don't need a room that night. I DO I need a room the Saturday and Sunday prior to the two seminars I signed up ; for, but they don't seem to want to book those days for me.   L This whole process is FUBARed beyond anything I've ever dealt with. It makes! dealing with HP or IBM look easy.   6 If this is the future, I want the old DECUs back. NOW!   	Bob Kaplow	  E SPAM:	spamrecycle@ChooseYourmail.com	uce@ftc.gov	postmaster@127.0.0.1    ------------------------------  # Date: Tue, 08 Aug 2000 11:33:29 GMT ) From: Carl Harrison <charrison@inbis.com>  Subject: FTP server problems) Message-ID: <8mor69$q28$1@nnrp1.deja.com>   E I am having trouble with FTP. I can ftp from the VAX, no problem. But C when I FTP from a PC to the VAX, the system console gives a warning 1 unable to open loginout.exe, file already in use.    Any ideas ?    Thanks in advance     & Sent via Deja.com http://www.deja.com/ Before you buy.    ------------------------------  % Date: Tue, 08 Aug 2000 07:42:09 -0400 ) From: Bob Ricci <maxx0623@concentric.net>  Subject: genicom 3810s< Message-ID: <006e01c0012d$af5e41c0$585b5cc0@socrates.Subway>  @ have set up a genicom 3810s line printer in australia under open1 vms.....dont know what to use for the port number B init/queue auslpr1/processor=tcpware_tssym/on="192.168.111.5,xxxx". the manual doesnt have the port number for tcp does anybody know offhand? thanks Robert V. Ricci  Systems Manager  Drs. Associates (SUBWAY) 325 Bic Dr.  Milford, Ct 06460   tel  203 877 4281 ext 1144  fax to pc 203 783 7144  fax 203 876 6682  email ricci_r@subway.com  or     maxx0623@concentric.net  http://www.subway.com    ------------------------------  % Date: Tue, 08 Aug 2000 10:33:02 -0400 + From: Paul Anderson <panderson@genicom.com>  Subject: Re: genicom 3810sB Message-ID: <panderson-303235.10330208082000@news.ma.ultranet.com>  G In article <006e01c0012d$af5e41c0$585b5cc0@socrates.Subway>, Bob Ricci    <maxx0623@concentric.net> wrote:  B > have set up a genicom 3810s line printer in australia under open3 > vms.....dont know what to use for the port number D > init/queue auslpr1/processor=tcpware_tssym/on="192.168.111.5,xxxx"0 > the manual doesnt have the port number for tcp > does anybody know offhand?  , I don't know, but have you checked online at  .    http://www.genicom.com/drivers/3840-doc.htm  ) or called the support people described at   2    http://www.genicom.com/drivers/contact.htm#Tech   Paul   --  "    Paul Anderson, DCPS Engineering"    GENICOM Corporation, Gardner MA   ------------------------------   Date: 8 Aug 2000 11:45:29 GMT + From: "Gerke Grashuis" <g.grashuis@kpn.com> 6 Subject: Re: How to copy the current screen to a file?8 Message-ID: <01c0012e$267ddcc0$8d4c15ac@HKTGN9911301604>  J Better yet, from the VT400 series terminals on, cut and paste is build in.J Read the manual that came with the terminal because i'm not very sure what keys to use.  4 D.Webb <david20@alpha2.mdx.ac.uk> schreef in artikel' <8mm27n$lm2$1@aquila.news.mdx.ac.uk>... 7 > In article <398E363C.DF3E3433@videotron.ca>, JF Mezei & <jfmezei.spamnot@videotron.ca> writes: > >Robert Deininger wrote:? > >> >  Does any body know if it's posiible to copy the current < > >> > screen to a text file by means of a command procedure > >> > or an exe file or not?  > > H > >If you're talking about a pure VT terminal, then the only sure way to dump anyG > >screen contents to a file is to use the "print screen" button of the 	 terminal, K > >and send the output to a port on the computer (instead of a printer) and  haveE > >a process there waiting to capture such info an dump it to a file.  > >  > G > Or if you know in advance that you want a copy of the screen then use 	 either :-  >  > 1) set host/log=filename 0 > % > ie record your whole login session.  >  > or > / > 2)  use the PHOTO program on a vax system  or 5 >     build and use ALPHA_LOGGER.C  in  SYS$EXAMPLES.  > K >     These programs use pseudoterminals to capture screen output to a file  as >     well as to the terminal. >  > I > >If you are using a PC with terminal emulator, then the emulator itself  may J > >have a built-in function to do that. If you are using DECwindows, there is a > >function to do that as well.  >  >  > David Webb > VMS and Unix team leader > CCSS > Middlesex University >    ------------------------------  $ Date: Tue, 8 Aug 2000 10:09:20 +02001 From: "Rienk Zwaagstra" <r.zwaagstra@elsevier.nl> - Subject: How to fake a device full situation. 1 Message-ID: <965722177.943462@caiman.elsevier.nl>   H For testing the error handling after a device full I need a full device.E Only I'm not able to fill up a device. Is there an other way like for E example creating a directory with a size limit or a concealed device?    ------------------------------  % Date: Tue, 08 Aug 2000 10:45:21 +0200 5 From: Oswald Knoppers <Oswald.Knoppers@whitehouse.nl> 1 Subject: Re: How to fake a device full situation. - Message-ID: <398FC8A1.AD425B6F@whitehouse.nl>    Rienk Zwaagstra wrote: > J > For testing the error handling after a device full I need a full device.G > Only I'm not able to fill up a device. Is there an other way like for G > example creating a directory with a size limit or a concealed device?   G Look at the LD driver which will emulate a disk in a file. It is on the  freeware cd or at:  - http://www.openvms.compaq.com/freeware/LD062/    Regards,   Oswald   ------------------------------  $ Date: Tue, 8 Aug 2000 11:42:44 +02007 From: "Uri Klil-Hahoresh" <uri.klilhahoresh@compaq.com> 1 Subject: Re: How to fake a device full situation. 6 Message-ID: <8mohdv$gd1$1@mailint03.im.hou.compaq.com>  K Concealed devices and directories are not use their own quota's so it won't I help you. what you can do is create a big files with a big size that fill , up your disk. use .fdl to control file-size.    : Rienk Zwaagstra <r.zwaagstra@elsevier.nl> wrote in message+ news:965722177.943462@caiman.elsevier.nl... J > For testing the error handling after a device full I need a full device.G > Only I'm not able to fill up a device. Is there an other way like for G > example creating a directory with a size limit or a concealed device?  >  >    ------------------------------   Date: 8 Aug 2000 11:52:02 GMT + From: "Gerke Grashuis" <g.grashuis@kpn.com> 1 Subject: Re: How to fake a device full situation. 8 Message-ID: <01c0012f$10d180b0$8d4c15ac@HKTGN9911301604>  * Or maybe a $copy/allocate=<lots-of-blocks>   Gerke.  < Rienk Zwaagstra <r.zwaagstra@elsevier.nl> schreef in artikel( <965722177.943462@caiman.elsevier.nl>...J > For testing the error handling after a device full I need a full device.G > Only I'm not able to fill up a device. Is there an other way like for G > example creating a directory with a size limit or a concealed device?  >  >  >    ------------------------------  # Date: Tue, 08 Aug 2000 17:11:33 GMT / From: "John Nixon" <jorlnixon@worldnet.att.net> 1 Subject: Re: How to fake a device full situation. F Message-ID: <9bXj5.9840$gW5.645324@bgtnsc04-news.ops.worldnet.att.net>  J You don't say WHY you are not able to fill up a device.  Is it a technical problem, a timing problem, or I a managemnet edict?  Let us know the problem, and maybe we can help solve  it.     < "Rienk Zwaagstra" <r.zwaagstra@elsevier.nl> wrote in message+ news:965722177.943462@caiman.elsevier.nl... J > For testing the error handling after a device full I need a full device.G > Only I'm not able to fill up a device. Is there an other way like for G > example creating a directory with a size limit or a concealed device?  >  >    ------------------------------  $ Date: Tue, 8 Aug 2000 19:24:28 +0200# From: "H Behling" <hbehling@wxs.nl>   Subject: I NEED imagemagick5.1.1+ Message-ID: <8mpfpj$b6qa7$1@reader1.wxs.nl>    Hello newsgroup,  E I know it is a rather odd question, but is there somebody who has the # imagemagick 5.1.1 zip file for vms. F Imagemagick is a graphics utility, however the current release (5.2.2)? contains some nasty habits when obtaining a dump of the screen. E I know 5.1.1. was ok, but I don't have the original zip file anymore.  Any help is welcome.     Herman Behling   ------------------------------  % Date: Tue, 08 Aug 2000 09:51:29 +0200 0 From: Andreas Strahm <andreas.strahm@siemens.ch>) Subject: Last order dates for VAX systems * Message-ID: <398FBC01.2E158868@siemens.ch>   Just for information:     <#   Last order dates for VAX systems    1   VAX 4000, MicroVAX 3100 - 88 and 3100-98 models   ;   VAX 4000, MicroVAX 3100-88, and MicroVAX 3100-98 systems  0   and all associated options are being retired. 4   Limited supply will be supported on a first come,    first serve basis.  %   Last order date: September 30, 2000   #   Last ship date: December 31, 2000   >  < Source: http://www.compaq.com/alphaserver/vax/vax_dates.html     Regards,   Andreas Strahm,  Software Engineer    ------------------------------  % Date: Tue, 08 Aug 2000 22:00:33 -0700  From: ctr66@bigfoot.com 7 Subject: NRGE - (Buy Alert)                         162 $ Message-ID: <263730725@MVB.SAIC.COM>   DIRIGO BUY ALERT:  Nu Electric Corporation  Symbol: NRGE Traded: OTC Bulletin Board Recent Price: 1 7/8  Shares Outstanding: 4,112,855  Approximate Float: 900,000" 52 Week Trading Range: 3/8 - 2 3/4   SUMMARY:  Q Investing in a low-priced stock that is virtually unknown but is on the threshold T of a dramatic breakout can have the highest reward-risk ratio and provide the astuteR investor with outstanding gains. We believe that we've found just such a situationP in Nu Electric Corporation - a company formed to take advantage of opportunities1 in technology, whenever and wherever they appear.   c Please take the time to visit the Nu Electric web site for complete reports and additional details:   ! http://home.earthlink.net/~nrge20   S Nu Electric is an incubator for new technologies; the company identifies, acquires, P develops, and markets emerging technologies that conserve energy and protect andP enhance the world's natural resources. Last year, Nu Electric acquired worldwideO license to a patented technology which removes arsenic from drinking water, and S recently signed a Letter of Intent to acquire an exciting new technology which will + extract two dangerous parasites from water.   X Without knowing the EPA's final standards for safe levels of arsenic, it's far too earlyR to project revenues for Nu Electric with any degree of accuracy, but one study hasY indicated a multi-billion dollar market in California alone. Don't wait for the Street to B take notice of this diamond in the rough; NRGE is a buy now stock!  R For background information on the history, science, power and mystery of water, weU refer the reader to "The Story of Eau" in the July 24, 2000 issue of Forbes Magazine. U This article refers to the "alarming surge in demand" for fresh water and states that 6 "water could well become the oil of the 21st century".   RECOMMENDATION:   \ For the benefit of new readers, this is not the first time we've written about Nu Electric -S we recommended the stock a little over a year ago when NRGE was selling at 75 cents U per share. Since that time, the company has made considerable progress in meeting its U goals of acquiring emerging technologies. The stock,however, has been volatile duringCZ this period - trading between 3/8 and 2 3/4. Nu Electric is a fully-reporting company, hasV no long-term debt, is poised for dramatic growth, and in our judgement is a better buy4 now than at the time of our previous recommendation.  P Currently trading in the 1 1/2 area, we place a STRONG BUY rating on NRGE today,T and advise those who acted on our earlier recommendation to add to positions for far8 better than average gains in months and seasons to come.  k To acquire ADDTIONAL INFORMATION regarding Nu Electric please click on one of the links below and hit send.M  F mailto:nu-electric@21cn.com?subject=Additional-Information-Nu-ElectricM mailto:altzone5@consultant.com.com?subject=Additional-Information-Nu-Electricc    - or go to our web site to obtain full details:C  ! http://home.earthlink.net/~nrge20s   DISCLAIMER:)  a An affiliate of the company has paid for the writing and distribution of this report.  AffiliatesrX currently own shares of NRGE, and may buy or sell at any time after the dissemination ofX this report. Dirigo has a 30 year track record as an investment newsletter, but is not aY registered investment advisor, broker or dealer. Purchase of this stock may be consideredIN speculative, and may result in the loss of some or all of any investment made.  A =================================================================VP Please double click on the below link to be excluded from further communication.( mailto:exclude77@uole.com?subject=deleteA =================================================================                              DIRIGO BUY ALERT:r Nu Electric Corporationr Symbol: NRGE Traded: OTC Bulletin Board Recent Price: 1 1/2h                           DIRIGO BUY ALERT:  Nu Electric Corporation  Symbol: NRGE Traded: OTC Bulletin Board Recent Price: 1 7/8-   ------------------------------   Date: 8 Aug 2000 08:51:23 -0500 9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)A: Subject: Re: OpenVMS ALPHA Port Of AOL Instant Messenger ?+ Message-ID: <AoJybMe+Uc4n@eisner.decus.org>2  p In article <009EE440.9839B521@SendSpamHere.ORG>, system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) writes:b > In article <jBVLYx3GTaRq@eisner.decus.org>, malmberg@eisner.decus.org (John E. Malmberg) writes:  6 >>There are some that may find it useful to have a VMS6 >>client so that they do not have to have a Wintel box >>to perform the same function.a > 6 > Well now, there's a noble reason...  Anything which 7 > might erode the wintel cartel is good and noble and ai > worthy project.a   _Anything_ ?  G Personally, I find fewer interruptions rather than more to be the goal.   B This wintel "feature" ranks right up there with the Blue Screen ofE Death.  The primary beauty of VMS PHONE is that it is so rarely used.'   ------------------------------  # Date: Tue, 08 Aug 2000 14:06:47 GMT = From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)4: Subject: Re: OpenVMS ALPHA Port Of AOL Instant Messenger ?0 Message-ID: <009EE4A8.D3D6A7FE@SendSpamHere.ORG>  g In article <AoJybMe+Uc4n@eisner.decus.org>, Kilgallen@eisner.decus.org.nospam (Larry Kilgallen) writes:lq >In article <009EE440.9839B521@SendSpamHere.ORG>, system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) writes:-c >> In article <jBVLYx3GTaRq@eisner.decus.org>, malmberg@eisner.decus.org (John E. Malmberg) writes:R >C7 >>>There are some that may find it useful to have a VMSc7 >>>client so that they do not have to have a Wintel box5  >>>to perform the same function. >> u7 >> Well now, there's a noble reason...  Anything which i8 >> might erode the wintel cartel is good and noble and a >> worthy project. > 
 >_Anything_ ?g >tH >Personally, I find fewer interruptions rather than more to be the goal.  7 True.  Does this AOL thing have an "off hook" feature?      C >This wintel "feature" ranks right up there with the Blue Screen of F >Death.  The primary beauty of VMS PHONE is that it is so rarely used.   :)   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMf   ------------------------------  # Date: Tue, 08 Aug 2000 11:21:32 GMTo% From: Alan Greig <agreig@my-deja.com> 2 Subject: Re: Patwhorks Performance and TCP NODELAY) Message-ID: <8moqfq$pkb$1@nnrp1.deja.com>e  H In article <y4snsh5fr1.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>,C   Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>o wrote:C > Mike Price <mike.priceNOmiSPAM@littlewoods.co.uk.invalid> writes:  >tC > > As  you can see from the early posts the only performance issue A > > that the Delay and NODELAY parameter seems to address is whens; > > the pathworks systems is SENDING to VMS. I have seen noe1 > > performance changes when the PC is receiving.c > F > From a technical point of view, this makes sense, because we seem to beD > talking about a TCP parameter delaying acknowledgement of packets, i.e.,uC > at the receiving end. The real question is why does this matter -t	 windowingm< > should take care of that, even with the slow start of TCP.  > I have now received a reply from the UK CSC and from PathworksA engineering saying that Microsoft's use of SMB over TCPIP exposesd@ a weakness in the "Nagle algorithm used by VMS by default" which> causes "unnecessary delays" with a Pathworks server. InitiallyB I was told that no patch was available but have just been informedA that the problem may have been fixed in a late ECO to UCX 4.2 bute@ not in either of the 5 releases. I am told that a test fix is on its way to me via AES/DSN.  A Pathworks engineering also inform me that they have been aware ofnD the problem for 18 months. Why it has taken this long to do anything? about it or even document it in release notes is something of ab puzzle.n  B I'll report back when I receive the patch kit but may have to wait5 for a convenient reboot time (assuming it needs one).e --
 Alan Greig    & Sent via Deja.com http://www.deja.com/ Before you buy.0   ------------------------------  % Date: Tue, 08 Aug 2000 12:42:15 +0100s  From: davep <davep@hmgcc.gov.uk>2 Subject: Re: Patwhorks Performance and TCP NODELAY( Message-ID: <398FF217.5CB0@hmgcc.gov.uk>   Alan Greig wrote:  > @ > I have now received a reply from the UK CSC and from PathworksC > engineering saying that Microsoft's use of SMB over TCPIP exposesrB > a weakness in the "Nagle algorithm used by VMS by default" which@ > causes "unnecessary delays" with a Pathworks server. InitiallyD > I was told that no patch was available but have just been informedC > that the problem may have been fixed in a late ECO to UCX 4.2 but B > not in either of the 5 releases. I am told that a test fix is on > its way to me via AES/DSN. > C > Pathworks engineering also inform me that they have been aware of1F > the problem for 18 months. Why it has taken this long to do anythingA > about it or even document it in release notes is something of aR	 > puzzle.x > D > I'll report back when I receive the patch kit but may have to wait7 > for a convenient reboot time (assuming it needs one).   E Brilliant. If the fix makes as big a difference as some have claimed,2F then maybe if it had been available 18 months ago we wouldn't today be5 in the process of decommissioning our last PW server.c   Dave   ------------------------------  % Date: Tue, 08 Aug 2000 05:51:13 -0700 ? From: Mike Price <mike.priceNOmiSPAM@littlewoods.co.uk.invalid>/2 Subject: Re: Patwhorks Performance and TCP NODELAY9 Message-ID: <0758909c.568e8c3e@usw-ex0105-034.remarq.com>-  < If anyone is interested I have just tried a test with Oracle= transferring data from one system to another with and without 4 the NODELAY option - basically it made no difference  
 Mike Price7 All entries are personal opinion and do not necessarilyr represent those of my employer.s8 Any resemblence they have to reality may be coincidental  ; -----------------------------------------------------------l  7 Got questions?  Get answers over the phone at Keen.com.w Up to 100 minutes free!t http://www.keen.com3   ------------------------------    Date: 08 Aug 2000 17:45:49 +0200G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>c2 Subject: Re: Patwhorks Performance and TCP NODELAYH Message-ID: <y4bsz392oi.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  ' Alan Greig <agreig@my-deja.com> writes:   K > > From a technical point of view, this makes sense, because we seem to benL > > talking about a TCP parameter delaying acknowledgement of packets, i.e.,O > > at the receiving end. The real question is why does this matter - windowing > > > should take care of that, even with the slow start of TCP.@ > I have now received a reply from the UK CSC and from PathworksC > engineering saying that Microsoft's use of SMB over TCPIP exposessB > a weakness in the "Nagle algorithm used by VMS by default" which@ > causes "unnecessary delays" with a Pathworks server. Initially  I IIRC, the Nagle algorithm _is_ TCP slow start. It's not one of TCP's bestsD features, and it's typically Microsoft to "expose a weakness", whichF translates as "we'll get LAN performance at all cost, and standards beK damned." (The same happens with not sending the final message on connectionoN teardown, which violates the protocol definition, doesn't release resources atI the far end, but of course uses less bandwidth and makes Windowns systemsA appear faster.)+  C > Pathworks engineering also inform me that they have been aware oftF > the problem for 18 months. Why it has taken this long to do anythingI > about it or even document it in release notes is something of a puzzle.   J A model of British understatement if I ever saw one. I hope you restrainedM yourself in not shooting the messenger, which I would have understood in this  case.-   	Jan   ------------------------------   Date: 8 Aug 2000 01:28 CST' From: carl@gerg.tamu.edu (Carl Perkins)o3 Subject: PC usability (was Re: dcps documentation?)a, Message-ID: <8AUG200001283018@gerg.tamu.edu>  3 "Phil Howell" <howellp@snowyhydro.com.au> writes...hK }Acrobat & ie on my pc are not talking to each other so I have to save pdfslK }in ie and then open them with acrobat later and I can't be bothered to fixd }the problem  F Ah, the joys of PC software. Well, OK - there aren't many, and none of them show up in this post.  B I just ran into this today. It may not be quite the same flavor ofH "acrobat and IE not talking to each other" as you have, but it could be.  K We want to put a PDF file on a CD. There is a main HTML file that has linksoF to to the 3 other things on the CD (one of which is actually the V4.05H Acrobat Reader installer, one is a tiny html file with a little text and% a link, and the 3rd is the PDF file).   J In Netscape, clicking on the link to the PDF file opens the document afterH (unless you've checked/uncheked - don't remember which - the box to makeG it not ask again) asking if you want to open Acrobat inside Netscape ormI launch it standalone. But clicking on the link that points to the AcrobatnF installer .EXE pops up a "save" dialog (instead of just running it, orG asking you if you want to save or run). So the PDF works and the reader * installation is sub-optimal but not fatal.  I Now in IE, and happened using IE versions 3.0, 4.0 and 5.5 on 3 different<G PCs where two were running NT Workstation v4 Sp5 and one was running NT F Server v4 Sp(unknown), clicking on the link to the PDF file starts theF Acrobat reader and tries to open the document inside the IE window. ItE doesn't work. Blank window. No error messages or anything. But if you.@ click the reload button it does pop up a dialog box that claims:  L    "This page provides potentially unsafe information to an ActiveX control.L     Your current security settings prohibit running controls in this manner.6     As a result, this page may not display correctly."  I Except that you aren't. This is not some data coming in over the internetnH being sent to some random ActiveX control, this is data coming from a CDH in my own CD drive directly attached to my computer. I put the CD in it.G I *want* to read the damned file from it. But it won't let me. Now heremG comes the "good" part. In the help file for Internet Explorer (at least . in IE v4.0) there is a passage that says this:  9    "In addition, any files already on your local computer36     are assumed to be completely safe, and no security9     settings are applied to them. This allows you to openn7     and run files and programs on your computer withouto9     prompting or interruption. You cannot assign a foldero2     or drive on your computer to a security zone."  F But, and this is the "really good" part, it turns out that all of thisF "security" stuff is a red-herring. (Queue the fish slapping dance from Monty Python's Flying Circus.)  E The *real* problem is what is known as "DLL hell". That is, there areaE mismatched versions of a pair of DLL files. In particular, the verion G of the comcat.dll file is actually too *new* (in my case it was versionwJ 5.0something but to fix the problem I needed v4.7something). It won't workG with the version of the ole32.dll file. How to fix this? You rename thegG comcat.dll file, reboot, and install something that includes an earlier0H verion (OK, in theory you may not have to reboot - it may be possible toF just "unregister" the DLL and rename it, then install the new one - inK practice, a lot of rebooting seems to help)(I selected the Visual Basic 6.00I run-time installation since it is, by Microsoft "standards", fairly smallnJ and, of course, the Visual Basic v6.0 run-time is already installed on theJ computer I did this on so most of it will be re-installing things that areK already there). You appear to need to rename the comcat.dll file because it,G won't install an older version of a DLL file over a newer one (which islI probably why the installation of VB 6.0 run-time didn't put the right one-H there whenever it was installed before). On any other OS this would be aH good plan. On Windows NT (and other Windows flavors) this will sometimesI prevent it from actually working since they don't know the meaning of the J phrase "backwards compatability". How did the "too new" version get there?I Well, presumably at some point something was installed that installed the8F too-new version of comcat.dll without installing the corresponding newE version of ole32.dll. That's right - in theory I could have installed J something that installs a newer version of ole32.dll. But that is not whatI Microsofts' own web site said to do - it said to install an older versionrK of comcat.dll (and even suggested, in one place, that the VB6 RTL be used).,  E (See Microsoft's document Q241385, the 3rd listed amongst some others H returned via searching their web site for "activex potentially unsafe" -B you probably shouldn't add "at any speed", but if you do you get 3F hits - and document Q201364 which has links to it from the other one.)  G What a phenomenal pain in the posterior regions. They delete 3 routinesoI from the newer version of the DLL file ("forwarding" the handling of them I to ole32.dll, somehow) and we get to muck around in DLL hell to fix it byeH reverting to an older version (which, by the way, is the version shippedF in a bunch of packages that have come out *after* the newer versions -G they don't seem to be able to get everybody to use the same versions ofyK anything, even something like 3 years after the new versions have been out)uG instead of making sure that whatever updated this file also updated alltF of the others that would be affected by the change (which is aparentlyH just one, the ole32.dll file, but changing that to the new version wouldG probably have a cascading effect requiring the updating of half a dozeni1 or more others - I wasn't foolish enough to try).s  E At any rate, I now have *one* computer here that can actually do whattE it is supposed to do in this respect. Since the software installed onOC our PCs is semi-consistant, I fully expect this problem to exist on G most of them. Fortunately, it isn't really my problem - it just happens C that the person who takes care of the PCs is currently on vacation.tG More fortunately, he comes back later this week. Even more fortuantely,l$ I go on vacation starting next week.   --- Carl   ------------------------------   Date: 8 Aug 2000 08:56:01 -0500n9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)l9 Subject: Re: PCSI installation problem - missing databased+ Message-ID: <cGLj5$6oCJmU@eisner.decus.org>o  N In article <VA.0000009d.1efb5a5a@sture.ch>, Paul Sture <paul@sture.ch> writes:  S > BTW, my very first exposure to PCSI involved a parity error - the CD had a minor iN > scratch on it. That would have been far less likely to happen with good old & > fashioned VMSINSTAL backup savesets.  > I disagree if the tape was from DEC.  They created their tapes@ with /INTERCHANGE so that commodity duplicating techniques could@ be used.  That practice eliminated the "skip the bad spot on the tape" processing.t   ------------------------------  $ Date: Tue, 8 Aug 2000 08:09:02 -05001 From: "Dave Gudewicz" <david.gudewicz@abbott.com> " Subject: Re: Raid array 300 on VMS8 Message-ID: <8mp0in$5ff$1@fizban.fizban.pprd.abbott.com>  J We have a RA3000 on a VMS system and it seems to be working well.  TroubleJ is when one has 2 controllers, one needs a PC with 2, yes count em' 2, COMK ports to configure the darn thing.  We usually use laptops for this job andeJ most (if not all) don't have 2 COM ports.  In fact, many desktop PCs don'tI have 2 COM ports.  I know my DECpc doesn't.  What were they thinking when=K they designed this thing?  How to make it difficult.  Sounds like it to me.l   Anyone know a way around this?  K BTW, the price was right on this RA3000.  I found it in one of our computer=J rooms and "borrowed" it forever.  Seems like our NT group has lots of toysH (and budget nowadays) and like my kids (and I'm sure others) forget what
 they have.   Dave...   5 "Dan Talbot" <dttalbot@together.net> wrote in messageE1 news:DdCj5.12966$SM4.1226873@nntp2.onemain.com...lL > We are presently using a RA3000 with dual redundant controllers for use inL > our Windowns NT domain. Have had great luck with it, so far (over a year).E > Two Windowns NT servers, each running on a seperate controller withs failover > if needed. Cannot complain.  >n> > Uwe Zessin wrote in message <8mmpsk$9qd$1@nnrp1.deja.com>...; > >In article <86ktossbl9c61cfg6fk2713ptvvrqdpoiv@4ax.com>,< > >  news@burns.uk.net wrote:  > >> Mike Price wrote: > >>D > >> >According to the Salespeople and the documantation I have seenE > >> >there is no nice command line interface to the Raid Array 3000so > >> >like HSZTERM.c% > >> >Does anyone know any different? G > >> >I would rather not have to use a GUI on NT to configure the thingo > >>K > >> the NT server can communicate with the RA3000 either over the SCSI bust? > >> (unlikely if your VAX/ALPHA is going to be using the disk)a > >iG > >There is a SWCC agent for OpenVMS Alpha - I don't beleive the RA3000HI > >is supported on a VAX, although the version (of the agent) I have seene  > >is dumb like a pice of bread.G > >If it cannot translate the client's name it crashes with SS$_STKOVF.R > >Fine engineering work!L > >aE > >> or across an RS232 link, or across the LAN using a client/serveraE > >> protocol to another NT server which is actually connected to thea > >> RA3000a > >>H > >> I've never "looked" at the protocol on the RS232 or ethernet links,H > >> but I've certainly never seen anything which lets you talk directly > >> to it :-( > >gI > >Fire up a terminal-emulator on your PC (hey, I said terminal-emulator,s3 > >not this Hyperterm garbage!), hit <ESC> and '&'. 1 > >You should get a menu. The password is 'RAID'.t2 > >(This info _is_ in the hardware documentation.) > >aF > >The interface is quite crazy, it takes some time to get used to it,< > >but I think it is still faster than that brain-sick SWCC. > >P= > >!!!!!!     !!!                                          !!e= > > !!  !!     !!                                          !!?= > > !!  !!     !!     !!!!!    !!!!!   !!!!!!!   !!!!!     !!u= > > !!!!!      !!    !!    !       !   !!       !!    !    !!i= > > !!         !!    !!!!!!!  !!!!!!   !!!!!!!  !!!!!!!    !!t2 > > !!         !!    !!       !   !!        !!  !!= > >!!!!       !!!!    !!!!!   !!!!! !  !!!!!!!   !!!!!     !!1 > >MH > >The RA3000 was originally thought as a low-cost Winblowsup NT storage? > >system. Do yourself a favour and avoid it whenever you can!!  > >a > >--<
 > >Uwe Zessinm > >s > >r) > >Sent via Deja.com http://www.deja.com/b > >Before you buy. >e >i   ------------------------------   Date: 8 Aug 00 11:20:47 EDTO From: grant@rigel.cc.wmich.edu# Subject: Secure LYNX client for VMSr/ Message-ID: <0SEmq2Sxp5b+@mcduck.acs.wmich.edu>   C Has anyone out there successfully built and used LYNX with OPENSSL?h  I I got LYNX 2.8.3rel1 and OPENSSL (9.5?) and patches from www.moxienet.comeJ and managed to built the software with some effort.  The OPENSSL self testE succeeded, and LYNX runs for non-secure pages (not HTTPS), but I get i  I  "Unable to make secure connection to remote host" when trying to connectt  1 to any HTTPS page. I don't know what I'm missing.t   ------------------------------  # Date: Tue, 08 Aug 2000 05:29:17 GMT  From: Dirk Munk <munk@home.nl>" Subject: Re: Spiraling GS140 Disks' Message-ID: <398F9AAC.96D695F4@home.nl>d   "Doug W." wrote: > K > We have successfully tested a workaround to the GS140 disk write I/O wallaK > encountered at 25MB.  It is indeed possible to spiral multiple disks on a O > GS140.  The workaround is not obvious and perhaps not practical.  I'd like to K > thank the person who mentioned hoses and all the individuals who provideda > help.i > P > I am not a hardware person, but here is the way a GS140 looks to me.  When youL > open up a GS140 there is a big rectangular cabinet, I think it's a 19 inchP > rack.  The system bus runs thru the center of the cabinet. Slots are availableP > at the front and rear for connection to the system bus.  On our units the CPUsP > go in front 2 to a slot.  And IOPs are in the rear.  Each IOP can have up to 4Q > hoses.  Each hose is connected to a PCI chassis.  The PCI chassis itself has 12:N > slots.  Each slot can contain a disk controller which is connected to one or
 > more disks.L > N > We have been able to spiral, defined as something approaching 20MB sustainedP > write, when we dedicate a hose to each disk.  We have tested up to 3 disks andO > believe this method can be scaled up.  There may be a limit on the max numbere > of hoses in a GS140. > P > Once again, we run a hose from the IOP to a PCI chassis.  We use a single slotH > in the chasis for the disk controller and connect a single disk to theL > controller.  In this configuration a slow controller is fine.  No need forJ > ULTRA2 or 3.  If you try running with 2 disk controllers in the same PCI! > chassis, you hit the 25MB wall.e > Q > Hopefully this will help someone.  I guess the next question is what can I push Q > thru a gigbit Ethernet in a GS140 PCI chassis on writes?  Also if someone could O > explain why disk writes take longer than disk reads I'd appreciate it.  Guessa8 > it has something to do with how the bus is being used?  H More likely it has something to do with the cache on the SCSI disks. TheG write cache on the disks is usualy disabled, and only the read cache is H enabled. And all SCSI commands to the disks are done in async SCSI mode,7 and that means with 4.5 MB/sec, and not with 40 MB/sec. H In general it is a good idea to spread your IO adapters over as many PCIH busses as possible. Every PCI bus has a maximum theoretical thoughput of@ about 128 MB/sec (32 bit @ 33MHz) for all of its slots combined.F If the Gigabit adapter is a 64 bit device, it can communicate with theG PCI bus with 256 MB/sec (64 bit @ 33 MHz), or 512 MB/sec if the bus andf the adapter support 66 MHz.      > Q > We are actually contemplating using the above method.  We would like to stay oneN > the GS140s until EV68 appears.  At that time we are considering upgrading to	 > GS160s.-   ------------------------------  # Date: Tue, 08 Aug 2000 13:29:25 GMTn+ From: sfm1115@bjcmail.carenet.org (Shawn M)k, Subject: Supressing Commands from a COM File0 Message-ID: <39900a48.65379600@news.starnet.net>   Hi all,m  D We are running OpenVms 7.1 on an AlphaServer here in our shop.  I amC in the midst of writing a COM file which will freeze a Mumps DB forn backup.r  D After I issue the command to freeze the DB, I am prompted for manualD input.  It is the same question for each of the 10 DB's which i have
 to freeze.  E Instead of manually typing the information, I would like the COM filet8 to automate the response, for example a sample would be:  - Do you really want to freeze this database?  v  F I would like to have the COM file automatically answer Y and continue?    4 Can someone give me some ideas on what I need to do.     Thanks   ShawnC   ------------------------------   Date: 8 Aug 2000 14:34:13 GMTe8 From: hammond@not@peek.ppb.cpqcorp.net (Charlie Hammond)0 Subject: Re: Supressing Commands from a COM File6 Message-ID: <8mp5p5$jdo$1@mailint03.im.hou.compaq.com>  ^ In article <39900a48.65379600@news.starnet.net>, sfm1115@bjcmail.carenet.org (Shawn M) writes: >Hi all, > E >We are running OpenVms 7.1 on an AlphaServer here in our shop.  I am D >in the midst of writing a COM file which will freeze a Mumps DB for >backup. >sE >After I issue the command to freeze the DB, I am prompted for manualhE >input.  It is the same question for each of the 10 DB's which i haven >to freeze.  >tF >Instead of manually typing the information, I would like the COM file9 >to automate the response, for example a sample would be:t > . >Do you really want to freeze this database?   > G >I would like to have the COM file automatically answer Y and continue?e  H (1) Does the MUMPS command have anything similar to a /NOCONFIRM option?  G (2) Can you put the "Y" answer in the next line of the COM file?   .eg.T  8     $ <MUMPS> <FREEZE>  ! Whatever the actual command is     Yw     $ <next freeze command     ... etc.  D Sometimes it is more robust and clearer to future maintainers if you. enclose the response in a DECK/EOD pair.  e.g.  8     $ <MUMPS> <FREEZE>  ! Whatever the actual command is	     $decke     Ya     $eod     $ <next freeze command
     ... etc. o   --  K     Charlie Hammond -- Compaq Computer Corporation -- Pompano Beach  FL USAsF          (hammond@peek.ppb.cpqcorp.net -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  % Date: Tue, 08 Aug 2000 07:40:38 -0700 1 From: Dave C. <DCantorNODCSPAM@shore.net.invalid>h0 Subject: Re: Supressing Commands from a COM File9 Message-ID: <109ec386.6b22facc@usw-ex0108-062.remarq.com>t  0 In article <39900a48.65379600@news.starnet.net>,, sfm1115@bjcmail.carenet.org (Shawn M) wrote: >[...]5 > Instead of manually typing the information, I would' > like the COM filee6 > to automate the response, for example a sample would > be:b- > Do you really want to freeze this database?u8 > I would like to have the COM file automatically answer > Y and continue?"6 > Can someone give me some ideas on what I need to do.  8 IIRC, DSM reads from SYS$COMMAND, not SYS$INPUT.  I once9 had to do something similar, though my recollection isn't,% perfect, it went something like this:0  8 1.  Create a file consisting of (at least as many) lines; (records) containing just the letter Y (or the word YES) as, required, e.g., YESES.TXT.   2.  Before invoking DSM, doe)       $ DEFINE/USER SYS$COMMAND YESES.TXTt  
 Good luck.   Dave C.      * Sent from AltaVista http://www.altavista.com Where you can also find related Web Pages, Images, Audios, Videos, News, and Shopping.  Smart is Beautifuln   ------------------------------  % Date: Tue, 08 Aug 2000 09:28:32 +0100 * From: Kevin Collins <K.J.Collins@hw.ac.uk>: Subject: TCPIP problem on an Alphaserver with OpenVMS v7.2( Message-ID: <398FC4B0.784CAE1F@hw.ac.uk>   Hi -    G I'm having a problem with TCPIP running on an Alphaserver 2100a/OpenVMSi@ 7.2.  The machine occasionally crashes without warning.  RunningH ANANLYZE/CRASH shows that the the error is INCONSTATE.  The SDA reports:   --------- Time of system crash:  6-AUG-2000 00:59:59.64I     [snipped stuff]n   CPU bugcheck codes:s8         CPU 00 -- INCONSTATE, Inconsistent I/O data baseB CPU 00 reason for Bugcheck: INCONSTATE, Inconsistent I/O data base    2 Process currently executing on this CPU: TCPIP$FTP     Current image file:s7 DATA$DKA0:[SYS0.SYSCOMMON.][SYSEXE]TCPIP$FTP_SERVER.EXEo   --------  A The TCPIP$FTP.LOG for this date/time exists but is totally empty.     G I've had a look for an ECO kit for this problem but I can't see one, soe) I'm stuck.  Any info gratefully received.h   TIAc     Kevin_Ci  ; ___________________________________________________________h9 Kevin Collins, Computing Services  |   Tel  0131 451 3791 % Heriot-Watt University, EDINBURGH  | m9 EH14 4AS, Scotland, UK             |   Fax  0131 451 3261v; ___________________________________|_______________________a   ------------------------------  # Date: Tue, 08 Aug 2000 10:19:22 GMTt% From: Alan Greig <agreig@my-deja.com>o> Subject: Re: TCPIP problem on an Alphaserver with OpenVMS v7.2) Message-ID: <8momr7$n2q$1@nnrp1.deja.com>   ( In article <398FC4B0.784CAE1F@hw.ac.uk>,   wrote: > Hi - >e; > I'm having a problem with TCPIP running on an Alphaservere
 2100a/OpenVMSnB > 7.2.  The machine occasionally crashes without warning.  RunningA > ANANLYZE/CRASH shows that the the error is INCONSTATE.  The SDA, reports:   Kevin,  D I'm guessing this is TCPIP 5.0. There were a number of problems withB this release and I would strongly suggest you upgrade to TCP 5.0A.  C Might also be worth an upgrade to VMS 7.2-1 if you have the chance.    Alan (Dunfermline - Scotland)e --
 Alan Greig    & Sent via Deja.com http://www.deja.com/ Before you buy.l   ------------------------------  % Date: Tue, 08 Aug 2000 03:45:34 -0700-? From: Mike Price <mike.priceNOmiSPAM@littlewoods.co.uk.invalid>o> Subject: Re: TCPIP problem on an Alphaserver with OpenVMS v7.29 Message-ID: <12c648de.35c75623@usw-ex0105-034.remarq.com>t  @ On top of the upgrade to TCPIP 5.0a I have recently been told by@ the DIGTIAL response center that ECO1 is due out soon and it has fixes in it for system crashes.t@ There are patches available for Alpha and VAX before the offical8 ECO comes out - contact the support center to get them - assuming you have support    Mike  
 Mike Price7 All entries are personal opinion and do not necessarilyo represent those of my employer.s8 Any resemblence they have to reality may be coincidental  ; -----------------------------------------------------------e  7 Got questions?  Get answers over the phone at Keen.com.m Up to 100 minutes free!n http://www.keen.comr   ------------------------------  # Date: Tue, 08 Aug 2000 13:26:51 GMT"% From: Alan Greig <agreig@my-deja.com> > Subject: Re: TCPIP problem on an Alphaserver with OpenVMS v7.2) Message-ID: <8mp1qm$ur7$1@nnrp1.deja.com>t  9 In article <12c648de.35c75623@usw-ex0105-034.remarq.com>,nB   Mike Price <mike.priceNOmiSPAM@littlewoods.co.uk.invalid> wrote:B > On top of the upgrade to TCPIP 5.0a I have recently been told byB > the DIGTIAL response center that ECO1 is due out soon and it has! > fixes in it for system crashes.mB > There are patches available for Alpha and VAX before the offical: > ECO comes out - contact the support center to get them - > assuming you have support    Mike,o  ? As you've probably already gathered one of the fixes supposedlyiD cures the Pathworks performance problem. I should have had it by now: but our DSN link is currently failing with DSN-F-CSCNOREC.   Sigh...t --
 Alan Greig    & Sent via Deja.com http://www.deja.com/ Before you buy.u   ------------------------------  # Date: Tue, 08 Aug 2000 14:24:35 GMT)+ From: tom@automatedtech.com (Tom Hickerson)h> Subject: Re: TCPIP problem on an Alphaserver with OpenVMS v7.23 Message-ID: <3990174a.56275169@news.mindspring.com>m  ) It doesn't sound related to your problem,e* but we see a memory leak which will bring * the system down on some VMS 7.2-1 machines! when running DECnet over TCP/IP. C  
 Tom Hickerson  tom@automatedtech.coma   ------------------------------  % Date: Tue, 08 Aug 2000 11:10:38 -0400i From: quayle@pobox.com> Subject: Re: TCPIP problem on an Alphaserver with OpenVMS v7.2, Message-ID: <398FEAAE.8204.32DEFD@localhost>  - On 8 Aug 2000, at 14:24, Tom Hickerson wrote: , > but we see a memory leak which will bring , > the system down on some VMS 7.2-1 machines# > when running DECnet over TCP/IP. c  C I see this happening a lot, too.  It seems to affect only Alphas.    It's a real nuisance.t     --Stan  
 ----------G Stanley F. Quayle, P.E.   N8SQ   +1 614-868-1363   Fax: +1 614 868-1671h1 8572 North Spring Ct. NW, Pickerington, OH  43147 = Preferred address:  stan@stanq.com       http://www.stanq.coms   ------------------------------  % Date: Tue, 08 Aug 2000 10:20:13 +0100.8 From: John Macallister <J.Macallister1@physics.ox.ac.uk>  Subject: RE: TELNET vs LAT speedL Message-ID: <35666012DF4CD411BE940090279FA240111EBD@ppnt41.physics.ox.ac.uk>  E We have both LAT and TELNET ( and also DECnet ) access on all our VMSsA systems. On a practical level there's no measurable difference inAK performance between one protocol or another for terminal window access: allaD three give excellent performance. TELNET is much more robust to highL Ethernet traffic levels and network hiccups in general. LAT is indeed one ofI the best indicators of Ethernet traffic loads and consequent packet loss:CD once packets start being dropped LAT connections break with terminal6 sessions being disconnected and print queues stalling.  H Although we still have LAT and DECnet capability and also some LAT printF queues, our users now almost only use TELNET (and SSH) for interactiveK logins and most queues use LPR/LPD. Our Ethernet is now entirely a switchedaK configuration and so high traffic levels are not the plague they used to be.I when we first started using LAT with TCP/IP on our Ethernet. Although LAT9H could now work well other factors such as switches not handling LAT withJ 100% "reliability" cause LAT to be unusable from time to time: the symptomJ is that a LAT service appears to be available/reachable but any attempt toL access it results in an "unreachable" response and rebooting the interveningK switches clears the problem. This problem and user confusion with not being I able to use LAT/DECnet to access all systems on the network has helped ton reduce the use of LAT here.r  I DECnet access continues to be invaluable, of course, between VMS systems.e  5 We're using VMS 7.1, DECnet-Plus and Multinet 4.1B . R   John  B Name: John B. Macallister  E-mail: j.macallister1@physics.ox.ac.ukH Post: Nuclear and Astrophysics Laboratory, Keble Road, Oxford OX1 3RH,UKA Phone: +44-1865-273388 (direct)  273333 (reception)  273418 (Fax)v   ------------------------------  % Date: Tue, 08 Aug 2000 17:01:57 +0200 = From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com>n  Subject: Re: TELNET vs LAT speed) Message-ID: <399020E5.20D67953@gtech.com>e   Jojimbo wrote:< > The fact that LAT does local echo of characters can (does)= > make it look considerably faster than telnet.  Telnet sendse< > each character to the remote system and back for the echo. > ; > If you need fast, routable, terminal communication, RTPADw8 > (aka Set Host) is very good.  Also does local echo and > intelligent buffering.   ????  ( Neither LAT nor DECnet uses local echo !  9 In fact you can not work on a VMS system with local echo.e1 Local echo and a full screen editor is not a goodm0 combination for VT terminals since the sequences2 send from the keyboard and the necesarry sequences+ to be send to the screen are not identical.    Arne   ------------------------------  % Date: Tue, 08 Aug 2000 17:13:09 +0200e= From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com>I  Subject: Re: TELNET vs LAT speed) Message-ID: <39902385.9A906645@gtech.com>    JF Mezei wrote: > > Something I have been wishing to conform/inform for a while: > N > Is it true that LAT is a ethernet-only protocol and deals only with ethernetT > adresses to deliver packets and this is the fundamental reason it is not routable?   LAT address = ethernet address  . but that is not the reason it is not routable.  0 The point is that it does only have one address.0 If the LAT protocol packet format contained both8 the immediate receiver address and the final destination6 address, then it could actually be routed (by a router than understood the format).   Arne   ------------------------------  " Date: Tue, 8 Aug 2000 17:19:13 GMT/ From: "Richard L. Dyson" <rick-dyson@uiowa.edu>  Subject: TSM and multiple NICs) Message-ID: <398FFAC1.22BD1868@uiowa.edu>r  E I have TSM (v2.1-07) running on my AlphaServer 2100 with two ethernet E cards: EWA and EWB.  I just recently switched all ethernet traffic to E just the EWB card.  The EWA card (on the original I/O module card) is.  no longer connected to anything.  G I have DECnet, TCP/IP (UCX), and LAT all running smoothly on the secondrE visible ethernet interface.  I recently had to teach AMDS what NIC tooC use (overriding the default).  However, I believe the TSM (Terminal.D Server Manager) software may be still stuck on using the first foundF NIC (EWA).  I can't find any TSM$DEVICE-like logicals or configurationE settings in the (Dec 1999) on-line documentation CDs (BTW, is it justlC my copy or is the TSM document in Bookreader all messed up?), TSM$*d files, or release notes.  @ Does anyone know how (or if!) I can configure TSM to use the EWB controller?o   Regards, Rick -- kH Richard L. Dyson                                    rick-dyson@uiowa.eduH  _   _      _____                http://www-pi.physics.uiowa.edu/~dyson/H | | | |    |_   _|   Systems Analyst                     O: 319/335-1879H | | | | of   | |     The University of Iowa            FAX: 319/335-17536 | \_/ |     _| |_    Department of Physics & Astronomy-  \___/     |_____|   Iowa City, IA 52242-1479    ------------------------------  % Date: Mon, 07 Aug 2000 21:46:55 -0400P* From: David A Froble <davef@tsoft-inc.com>  Subject: Re: VMS Vs any other OS- Message-ID: <398F668F.AF929587@tsoft-inc.com>C   Harald Droste wrote: >  > ----<snip>----" > > I would add that VMS has yearsM > > of engineering behind it, where NT, for instance, only has a few years oflN > > questionable (I have my doubts about microsoft's engineering) equivelance. >  > ----<snip>---- > + > even if I get slaughtered for this one...vK > the N/T architect is nobody else than Dave Cuttler and his (ex-DEC) team.f >  > Harald  . Slaughtered, probably too old and stringy. :-)  M What you need to do is look at the date Cutler left DEC, and if possible, how P much he had to do with VMS beyond the initial few versions.  VMS 7.2 is very farO from VMS 1.5, 2.0, etc.  If NT's target is VMS Version 1.5, a product availabler in 1980, that's rather pitiful.,  N Also, the mention of "(ex-DEC) team" seems to imply that a sizable part of theP VMS development team moved to Microsoft.  I don't know how many of them may have: done so, but I seriously doubt it is a significant number.  K Finally, some people seem to be under the impression that one person, DavidoP Cutler, was responsible for VMS.  Not even close, and I do not think that he wasM in charge, just one of the many people involved in the initial design effort.l  N One thing I'm rather sure of (which isn't saying much) is that one of the mostL well known features of VMS, the VMS CLuster, is something Cutler didn't have anything to do with.  O Maybe reality isn't as good a 'story' as some of the fiction making the rounds,  including possibly mine.   Dave   --  4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com6 T-Soft, Inc.  170 Grimplin Road  Vanderbilt, PA  15486   ------------------------------   Date: 8 Aug 2000 11:17:34 GMTh9 From: Roar =?iso-8859-1?Q?Thron=E6s?= <roart@nvg.ntnu.no>e  Subject: Re: VMS Vs any other OS3 Message-ID: <965730024.859233@tornado.itea.ntnu.no>-  : Larry Kilgallen <Kilgallen@eisner.decus.org.nospam> wrote:\ : In article <398D0B4B.183776F6@tsoft-inc.com>, David A Froble <davef@tsoft-inc.com> writes: :> "Hipenbecker, Doug" wrote:e :>> L :>> I must mention that Oracle dropped production support for RDB/NT shortlyM :>> after it was released, but distributes it free as "Oracle RDB Workbench".pO :>> Oracle blames the problem on Compaq dropping support for the BLISS compiler.# :>> on WNT which RDB is written in.  :> 0R :> Which is rather bogus, since BLISS is available free.  It will do tomorrow whatS :> it does today and did yesterday.  If they're so worried about it, they could gety  :> the compiler and maintain it.  : : The Bliss on the VMS Freeware is only for VAX and Alpha.  5 : To my knowledge the source has never been released.o  = The closest is the source for Bliss-11 dating from 1972/1973.-   ------------------------------   Date: 8 Aug 2000 11:48:26 GMTk9 From: Roar =?iso-8859-1?Q?Thron=E6s?= <roart@nvg.ntnu.no>   Subject: Re: VMS Vs any other OS2 Message-ID: <965731877.91562@tornado.itea.ntnu.no>  & Alan Greig <agreig@my-deja.com> wrote:: : In article <009EE398.D6759364@SSRL04.SLAC.STANFORD.EDU>,) :   winston@SSRL.SLAC.STANFORD.EDU wrote:h? :> In article <398D0B4B.183776F6@tsoft-inc.com>, David A Froblef : <davef@tsoft-inc.com> writes:f :> >"Hipenbecker, Doug" wrote: :> >>mF :> >> I must mention that Oracle dropped production support for RDB/NT	 : shortlyfC :> >> after it was released, but distributes it free as "Oracle RDBa
 : Workbench".hH :> >> Oracle blames the problem on Compaq dropping support for the BLISS
 : compiler% :> >> on WNT which RDB is written in.  :> >E :> >Which is rather bogus, since BLISS is available free.  It will dol : tomorrow whato  B But is it the same version of Bliss the developers use themselves?  E :> >it does today and did yesterday.  If they're so worried about it,C : they could get  F : Yes it will do tomorrow what it will do today and I believe that wasD : the problem. It generated code that ran fine on EV4/5 machines but= : barfed on EV6. Also the Intel version presumably would have F : required serious work to ever produce IA64 code. Perhaps at one timeB : Compaq intended to to a Bliss IA64 port then decided against it.  4 How well does the freeware Bliss work on AXP anyway?A I do not know Bliss very well, and I have not tested it too much.k  Bliss-64 is not even documented.D When creating the Bliss-32 libraries, I got -I- or -W- messages when the 64-bit values overflow.-" Bliss-64 libraries seems to be ok.= But there are few *.b64, *.l64, *.r64 compared to the 32-bit,sD and I am not sure how many of the *.b32, *.l32, *.r32 can be used on( Bliss-64. (I know tutio can not be used)   ------------------------------  * Date: Tue, 8 Aug 2000 13:13:51 +0000 (UTC)' From: Osmo Kujala <kujala@tukki.jyu.fi>]  Subject: Re: VMS Vs any other OS, Message-ID: <8mp12f$el0$1@mordred.cc.jyu.fi>  7 ChuckT <chuck.taylorNOchSPAM@vishay.com.invalid> wrote:eI > The problem with an open-ended question like "which operating system istE > the best one" or "how can you justify this operating system or thateG > one" is that the discussion quickly turns into a religous arguement.  I > Without going that way let me give you my two cents worth of opinion...  OkayI > Each operating system that has been created was created for a purpose.  D > Therefore,  most operating systems fit into some sort of niche and5 > there have been a lot of operating systems already.o  J PC wasn't invented yet, so VMS was meant to be system for time sharing andE batch processing. It's very near all purpose and not for small niche.y  D > I like VMS or OpenVMS as it is now called.  It is almost a naturalD > language environment where you use commands that are very close toH > standard English to communicate with the computer.  It isn't quit thatH > easy to use but it is easy enough that with the built in help function/ > most people can begin to use it very quickly.   H > I believe that there is a place for each of the operating systems thatE > are extent today (with maybe just a couple of exceptions).  I don'tAF > believe that there is a "one-size-fits-all" answer yet.  Maybe there5 > will be someday but it won't be what we have today.s  H If we had to choose one, and that is what I'd like to do, then which oneF has the needed features. All are usable on desktop, small servers etc.A but only VMS shines on clusters and reliable big systems. This isPE situation now, but very soon NT and unix (esp. Linux) will be used onPG every niche. Yes, even NT will be used (and works) in bet-your-businesstA big systems. It needs more work, more money etc. but because it'si; industry standard and used everywhere else it makes sense. e  E > I remember one operating system that was really interesting and had_E > some of the most innovative ideas in it that I have seen.  It had alI > command line interface or CLI that was called DWIM or "Do What I Mean" dG > it was used on Xerox artificial intelligence workstations back in the_G > early '80s.  If you typed in a command and mis-spelled something thenuI > the CLI would automatically correct the spelling for you and ask you ifcI > what you really meant was the corrected version or not.  Of course, yourI > could do the same sort of thing with aliases in some O/Ss or symbols ort > logicals in VMS.  E Maybe CLI or GUI aren't so important any more, when more and more are  done in www?  H > Today most O/Ss can be grouped into a few catagories.  The ProprietaryH > O/Ss (like Windows, or VMS) and the Unix flavors (some are proprietaryH > and some are not).  The biggest problem with Unix is that the commandsH > are so obscure that they are hard to remember.  Basically, you have toI > memorize them.  And there are so many command line modifiers that it is G > just simply beyond anyones ability to remember them all.  Then on top H > off that the help system isn't called help its called man pages.  EvenI > the explainations of how the commands are used and what they do that isiH > provided in the man pages is obscure.  It has always been that way andH > I suppose that it will remain that way too.  It is what is referred to" > as part of the "Unix mentality".  C > Then there is the question of what is Unix?  There are many, many G > flavors of Unix.  All of them differ by various degrees.  There is an I > organization called the Open Group that has the authority to "brand" aneH > operating system as a Unix operating system.  The branding is based onH > whether or not the operating system can pass a series of tests. I knowD > that Compaq Tru64 Unix has receive Unix95 branding and Solaris hasB > received Unix98 branding.  Here is a URL to the Open Group site.   > http://www.opengroup.org  D > One interesting fact that might be of interest is that most of theE > operating systems that are in use today pretty much all came from a D > common progenator.  Most of them were developed on the Digital PDPF > series of computers.  For instance, the original Unix was created atI > Bell labs on a PDP (I think it was a PDP-8 but I am not positive).  VMS I > stands for Virtual Memory System and the first type of hardware that itoF > was designed to work with was the Digital VAX computers.  VAX standsA > for Virtual Address Extension with respect to the PDP family of A > computers.  The PDP comptuers were 16 bit computers and the VAXAF > computers are 32 bit machines.  The newest machines that VMS runs on) > are the Compaq Alpha, 64 bit computers.i  * And now VMS runs on PC:s too (Charon-VAX).  H > Another interesting bit of information is that most software developedH > on the very first VAX computers will still run (probably would need toG > be recompiled) on the newest VAX and Alpha computers.  I believe that.G > VMS runs on the widest range of computers of any O/S excepting Unix.  = > From small desktop machines to departmental machines to bigs/ > refrigerator size servers in the data center.r  H > VMS would probably be much more popular and would be in much wider use > had a few things not occured.p  I > First, Digital and then Compaq had the attitude of "if we build it they B > will come" and as for advertising, the best discription of theirA > advertising technique that I have heard is that they use a veryaI > effective form of "stealth marketing" designed to keep their productionT8 > lines from being taxed too heavily by too many orders.  G > The second thing that happened was that they didn't believe that UnixsH > was ever going to amount to anything.  Ken Olsen, the former presidentH > of Digital once said the "Unix, is snake oil".  They missed the windowH > of opertunity then and have had a difficult time getting back on track > every since.  H > The third thing that happened was the move to open systems that caughtH > the computer industry by surprise.  Digital was VERY proprietary aboutF > their products and they were very closed.  There were many law suits= > over patent infringments.  It was ugly and no one liked it.0   But now NT is even more closed.a  F > The final thing was that Digital and now Compaq didn't (and doesn't)H > know how to price anything correctly.  I believe that there was a timeH > when you could not order anything from Digital for less than $500.00. G > While other companies were giving away their O/Ss to get market shareeF > Digital was boasting about how much market share they had.  Well notD > for long and not anymore.  I think the hobbyist licenses should beD > completely free.  Schools were dumping everything in favor of Unix5 > because it was free...  Not because it was better. q  F One main reason was opennes of BSD unix. Source code was available for learning and patching.   > Budget not qualityH > was the bottom line and Digital would not giveaway VMS.  They did haveI > programs where if a school purchased hardware from them they would give I > them a break on the software but it was still very expensive.  PCs wereo( > cheap and Unix was free.  Go figure...  H > Currently, there is still a large group of people who love VMS.  ThereH > is another very large group of Unix priests and gurus.  And then thereG > is WindowsNT.  We don't really need to talk about any form of WindowslG > before NT as they were not really operating systems but were actuallyAI > just exteremely simple, very stupid job schedulers.  I have to say thateG > there were actually some great operating systems before MS-DOS becames< > the de facto standard.  But MS-DOS was what I would call a% > lowest-common-denominator solution.s  A There were very good reasons why DOS-PC boom happened at schools.eC Centralized computer systems were very expensive, strictly rationedSD (meaning rules and quotas etc.) and didn't offer graphics terminals E at reasonable cost. Not many understood that decentralization in forma> of DOS/Windows PC:s would lead to even more expensive systems.   >  I guess it satisfied the-G > requirements of the contract that MS had with IBM but it was (and is)- > pretty pittiful.  F > WindowsNT on the other hand is a true operating system and is thus aF > lot more complex.  It was largely written by some of the same peopleF > who wrote VMS long ago.  In a lot of ways it is a very stripped downI > version of VMS (I just heard a giant agghhh in the background) but thenaG > again - when I consider what Unix is like, the best definition that IfH > have seen is that Unix is a subset of VMS, or even more precise VMS isG > like a very large feature rich superset of Unix.  Unix is very simplegG > compared to VMS.  It has very few (if any, depending on the flavor offE > Unix) of the things that make VMS the "bet your business on it" O/Sa
 > that it is.i  I > With all of that said: I want to make a point, and that is, most peopleoH > like to use what they are familiar with and comfortable with.  This isG > such a powerful force in most peoples lives that any discussion about E > another O/S being better immediately causes a "Holly Jehad" type ofjF > reaction.  If you want to start a big arguement then just go into anH > AIX (IBM Unix) shop and start telling people that some other flavor ofI > Unix is better or better yet try the same experiment in s Solaris shop.n  F > Once again, I think that an O/S is a tool.  My recommendation is useB > the right tool for the job.  VMS is great for a server but it is > terrible on the desktop.  F I disagree. I don't see any lack in VMS for desktop. Of course there'sB programs missing now because DEQ has almost succeeded to kill VMS.> Someone might like Windows-like GUI better that CDE/Motif etc.B These will be fixed when VMS gets popularity back. Or some missingG applications and features (Yes, it includes unix compatibility featuresg3 too) must be added before VMS gets popularity back.s  0 >  But than again don't even think about running6 > a corporations most critical databases on an NT box.  5 Will be done and will succeed as irrational as it is.    >  It just doesn'tG > have the reliability or scalability that is needed in a server but it A > does pretty well as a desktop system.  Unix and NT have so many I > security problems that it is hard to go into it.  Unix does okay on thenG > desktop but doesn't have the richness of application software that NThI > does.  The current boast for Unix is that it is really fast and it is. i  C Fast vs. reliable is choice of user in VMS and in unix too I guess.u= Default is reliable in VMS and fast in unix. NT I don't know.n  F > It is hard to use and it is hard to memorize all of the commands butF > the GUI helps a lot.  However, most of the things that make a serverH > secure, reliable and scalable just aren't there.  It is getting betterB > but I personally don't believe that there is a Unix based systemH > available that is as robust as VMS is on the server side.  The problemH > with VMS is that a lot of software that used to be available on VMS is6 > no longer available on VMS.  A point is case is SAP.  E > A few years ago Digital started telling everyone that the only O/SsXG > that would be used in a few years would be NT and Unix.  They sort of\G > shot themselves in the foot to help their dying Unix business.  There-H > were many VMS shops that said okay fine we will move off of VMS and goI > with NT on our Alpha boxes instead (then they announced that they wouldFB > no longer support NT on Alpha).  It was like DEC and then CompaqI > couldn't push people off of VMS fast enough.  I think that they finallyTE > realized how stupid what they were doing was, but by then there waso5 > almost fatal damage done to the installed VMS base.   G > So this is how I see it.  WindowsNT or Unix on the desktop as needed. B > OpenVMS or Unix in the middle tier - for application servers andH > compute engines as needed.  Then OpenVMS, Unix or some other mainframe@ > type OS on the servers where the corporate jewels are located.  I No, it's much more reasonable to use same OS everywhere if possible. None J of those are now quite ready for that, but need a little developement. I'd: like to see NT and whole MS disappear (it's caused so muchA regression in software developement) and see VMS rise from ashes.t  G > Unix could be on all of the tiers but it may not be the best solution D > in all cases.  For example, MS-Office doesn't run on Unix (not yetB > anyway) and Mentor Graphics Schematic Capture and Design system,I > electronic engineering design software, runs best on a Unix workstationt1 > and I don't believe it available for WindowsNT.e  I > I must say in the end that I too am a bit biased.  I like VMS.   VMS is(I > like a good old-friend.  It is easy to use and runs pretty much withoutSD > any problems for long periods of times.  I worked on some VAX 9440I > systems years ago that had a clock in them that had a date field of twohI > digits.  We had to restart the system every 99 days or it would restart I > itself (nice little built in feature that always came around to exhibit-G > itself at quarter closing time for some reason).  See if you can find @ > an NT or even a Unix box that has been running for that long. I > Muchless, longer.  the VAX 9000 was the only VMS system that I am aware8F > of that had that problem.  I have seen VMS systems that have been upF > and running for much, much longer than that.  Some have been running! > non-stop for years - literally.p  I > This was a long reply to your questions but I also wanted to comment onaG > the "religous wars" that are raging in this news group.  I would likeuG > for everyone to please stop it.  Just be reasonable.  After all, they F > do make Fords, Chevys, Toyotas, and Porsches don't they.  They don'tE > make Rambler cars anymore because of market changes or pressures orIH > whatever.  Unix is older than VMS.  It is on its downhill side of life0 > too.  As are most other O/Ss - excepting NT.     > Something else will beD > coming along pretty soon that will make a lot of people happy I am > sure.   F Lets hope so. That might suit better in net environment than these old? OS:s. I think there is need for "Distributed OS" that would useeG distributed hardware more optimally. I think vmscluster now has some of- the needed features.  B > Wouldn't it be great to have a new O/S that behaved as nicely as> > VMS, with all its features and richness.  With the speed andI > flexability of Unix and the richness of applications that are availablegH > for NT?  I think it will happen as market pressure pushes us towards aI > common O/S some time in the future.  And I'll bet it won't come from MS0I > either.  The again, I wouldn't be surpirsed to find out that their is asA > WindowsNT display environment that runs on top of Linux either.a  H Market pressure pushes very much towards Windows, but rather I'd like toH consider Linux to be the "Esperanto of operating systems" before we get 0 new Zamenhof of OS's to create the real thing.    I > Don't let yourself get sucked into one of these established religions.  ; > Finding your own way is the only path to enlightenment...    > Good Lucko   > Chuck Taylor% > Sr. Global Infrastructure Developere > Vishay Siliconix   regardst           Osmo Kujalat   ------------------------------   Date: 8 Aug 2000 15:26:19 GMT 2 From: mathog@seqaxp.bio.caltech.edu (David Mathog)  Subject: Re: VMS Vs any other OS, Message-ID: <8mp8qr$k40@gap.cco.caltech.edu>  g In article <398F5AFD.41091A74@earthlink.net>, "David J. Dachtera" <djesys.nospam@earthlink.net> writes:  >4C >So, I could wish for a utility to scan HELP libraries and create a H >database of topics appropriate to various keywords and topics extractedI >from the help libraries. An "APRO[POS]" command for OVMS, if you will...  >s  J We index them and serve the results through a web interface.  The indexer D was Bruce Tanner's and I don't recall where we got the HELP gateway G software.  Hmm, come to think of it, it's time to redo this since we're2F at 7.2 and I last indexed at 6.2 (and in the interim POSIX went away.)  F The scripts and programs I wrote to do this (heavily UNdocumented and K containing some hardwired directory and file names!) can be picked up from:l  =   http://seqaxp.bio.caltech.edu/pub/SOFTWARE/VMSHELPINDEX.ZIP8  G Unpack with /VMS since it contains an Alpha executable and object file.nF That doesn't include Bruce Tanner's tools which you can get elsewhere.J There are two build_all versions, one uses "preprocess" and runs directly K on .HLP files.  The other uses LYNX to download the data from a web server.r9 (The former is much faster.)  To see what it looks like,    +    see  http://seqaxp.bio.caltech.edu/HELP i   and follow the SEARCH link.    Regards,   David Mathog mathog@seqaxp.bio.caltech.edus? Manager, sequence analysis facility, biology division, Caltech f   ------------------------------  % Date: Tue, 08 Aug 2000 18:07:51 +0200 * From: "J. McLean" <mcleanj@pop.dplanet.ch>  Subject: Re: VMS Vs any other OS. Message-ID: <39903057.D42B2205@pop.dplanet.ch>   "Mike E. Fackler" wrote: >  > Hi All ....J >   My reading has recently taken me out of the Intel world and into otherN > type systems. (OS/400 and OpenVMS) I don't mean for this question to start aK > war, but being so indoctrinated in a Win type OS (Win/2000) I am having a E > hard time seeing the advantage to the VMS OS. What exactly are it's M > strengths over other OS? Why would an industry choose it over anything elseo# > in general and W2K in particular?i >  .... >  > Thank you....i > Mike  	 ---------M    E To put it simply, VMS epitomizes just what a serious operating system > should provide, and the manner in which it should be provided.  ? VMS is very serious about security, functionality, flexibility,r@ reliability and provision of service, so serious that it goes toG extremes which are rarely found individually in other operating systemssE let alone in an integrated form.  It is an old-fashioned word to use,b2 but VMS has Quality - and that's with a capital Q.  B The software engineers who develop and maintain VMS work in a veryD disciplined environment.  They are fanatical about code inspections,> about the integrity of the product and about ensuring that theG functionality that is described is what the customer will receive.  AndlD they have the tools and discipline to ensure that this fanaticism isF channeled into the final product.  Discipline like having a designatedD function for every directory which holds operating system files, and% strictly adhering to that definition.i  ? Those people who use VMS on a day-to-day basis are accessing anoE environment where the rich array of features and the ease-of-use hide B this continuing fundamental principle of discipline.  DCL providesE ease-of-use for the general user, there is ease-of-use for developersOH and there is ease-of-use for system management.  At all levels the rangeF of features and capabilities are very substantial, but always there is5 the underlying disciplines of security and integrity.m  G And the end product of all this discipline can be summed up in very few H word - a very, very reliable operating system.  Reliability which is notF merely enough to make it the best general purpose operating system, orF reliability which on which you can "bet-the-business", but reliability' which can be literally "life-or-death".c  H Finally, I would like to say that the creation of a product this good isH only possible because it is proprietary.  There is no need for agreementC and compliance with some common standard; the VMS engineers can setaF their own standards as high as they like (and they are high !).  BeingH proprietary is not a weakness for VMS; it is the reason for its ultimate strength    i John McLean  Zurich Switzerlanda   ------------------------------  % Date: Tue, 08 Aug 2000 12:30:48 +0100eB From: Andrew Harrison SUNUK Consultancy <andrew.nospam@uk.sun.com>; Subject: Re: What Can I Do With An AlphaServer 1000A 4/233?h* Message-ID: <398FEF68.91636599@uk.sun.com>   "Ramon L. Tate" wrote:  D > In article <ZY6SLHAFfoj5Iw$j@planet-tharg.demon.co.uk>, "T Bluck."& > <tb@planet-tharg.demon.co.uk> wrote: >n? > > I did some work for a Ministry of Defence supplier based ino< > > Surrey U.K.  the computer room was the size of an indoorA > > tennis court, and housed 10+ VAX 11/780 systems and all theiro= > > attached disc and tape drives. All was working fine untilP? > > someone decided to test the fire alarm system's gas extractsA > > fans,  Then the false/lowered ceiling came down, plaster dust  > > everywhere..    (ouch!)a > >  > > --< > > Tim Bluck.   TB565   http://www.planet-tharg.demon.co.uk >rH > This is beginning to sound like "Can You Top This?" (an old U.S. radioB > program from more years ago that I care to remember). Vis  vis:   Steam is nothing.n  D I was involved in deploying and testing a cluster of servers for oneD of the national newspapers in the UK, it held all their pictures for< example. We completed the system and applications testing onI the thursday with the initial go live being monday. As I left I commentedXB to the datacenter manager that their machine room felt pretty hot.  D On the monday we got a call telling us not to come in and letting us' know that the system wasn't going live.e  G Worried we phoned back and after a few abortive calls got the low down.hB Seems that the aircon unit was faulty and so the customer had shutL the systems down on friday and two aircon engineers had arrived on saturday,C dissassembled the whole system including a high pressure water feed I above the systems in the roof void and then re-assembled the whole thing.a  L Or so they thought, they forgot to tighten up one of the joints in the waterD feed and when they turned the water back on at the valve which was aI few minutes walk from the machine room a few thousand gallons of slightlyiE gritty (calcium) water was dumped over the 6 machines in the cluster.l  I The engineers turned the water feed back off and one of them did a runner ' leaving his colleague to carry the can.g  G The machines were soaked, water in the disk trays, water behind the LEDo? covers on the diag units and fine calcium grit over everything.s  F We dried the machines out and cleaned them which took about a week andB all 6 systems re-started without a single problem. Sadly they were subsequentlyH junked because they could not be warrantied having been drowned and they were replaced.  H 3 months later the people in the offices below started to complain aboutE damp and an inspection of the sub-floor revealed that there was stillnF 3 inches of water under all the systems. Still that must have kept the relative humidity up.g   >rK > Early one Monday morning, the early shift operator for the aforementionedlG > NIH DECsystem-10 arrived to find the entire computer room filled witheL > steam! The Gods of Computing were smiling that day, however wryly, becauseK > (1) it was relatively low temperature steam, and (2) the system power hadeF > been turned off on Sunday by a DEC maintenance engineer after he wasH > finished doing some maintenance. It took the better part of 2 weeks toM > clean and dry out all the electronics - including the disk packs and backup I > tapes, which were [you guessed it] stored in the same room. Not one bitn9 > lost, though, at least as far as anyone could tell. :-}  > F > There was a steam line from the central power plant that entered theJ > building through a closet off the basement computer room, and one of theJ > valves sprang a small leak sometime during the weekend. The DEC engineerM > actually noted a little bit of steam coming from the closet and decided (astL > I remember) to leave the power off, "just in case". By Sunday morning, the< > small leak had turned into something bigger - much bigger! >h > Over.o >a > -- > Ramon L. Tatew > Casa Maaa? > taterskins@patriot.net   "Skin" that "tater" before replying!o   -- Andrew Harrisono Enterprise IT Architect    ------------------------------  # Date: Tue, 08 Aug 2000 14:20:29 GMTa% From: Alan Greig <agreig@my-deja.com>o; Subject: Re: What Can I Do With An AlphaServer 1000A 4/233?t( Message-ID: <8mp4va$sa$1@nnrp1.deja.com>  * In article <398FEF68.91636599@uk.sun.com>,E   Andrew Harrison SUNUK Consultancy <andrew.nospam@uk.sun.com> wrote:a >tD > 3 months later the people in the offices below started to complain aboutnG > damp and an inspection of the sub-floor revealed that there was stilllH > 3 inches of water under all the systems. Still that must have kept the > relative humidity up.   E Once upon a time when I worked as a junior student systems programmer  fore@ Dundee College of Technology I remember dialing in to the DEC-20E remotely (300 baud) and continually getting %DECSYSTEM-20 NOT RUNNING C followed by [DECSystem-20 Continued] a few seconds later. The firstpB error indicated some kind of serious problem and usually meant theD system had crashed. If it was followed by the continued message then< TOPS-20 had managed to take corrective action to work aroundB the problem. I had a quick look at the error logs and found parity> errors, front end reloads, powerfail restarts, Massbus errors,= I/O subsystem restarts etc. There was the occasional completeeF reboot but on most occasions the system was able to recover. Obviously something was far wrong.  H It was a Sunday night and on the Monday the DEC engineer (Ed Merle (sp?)B later to become a senior figure in DEC FS, turned up looked at theF temp/humidity charts and lifted a floor tile . Almost up to tile levelF was about 6 inches of water. Every single cable was entirely submergedE and water was dripping down the CPU and memory blown by fans. SeveraltC single phase outlets had blown but amazingly the 3-phase supply waskB intact and a very soggy DEC-20 was finally rapidly brought down by@ one very distressed engineer (realising he was sitting on a lake> with live power all around) hitting the nearest emergency stop button.o --
 Alan Greig    & Sent via Deja.com http://www.deja.com/ Before you buy.n   ------------------------------   End of INFO-VAX 2000.441 ************************