0 INFO-VAX	Sun, 11 Feb 2001	Volume 2001 : Issue 83      Contents: create User  Re: create User  Re: create User  Efficiency in processing bits ! Re: Efficiency in processing bits  EV8 description 3 Re: Firmware upgrade for PWS 600au to use a ZLXp-L1  Re: FORCING 8BIT COLOR IN DCE? Re: Free space ? Re: Free space ?F Multinet installation in cluster problem - was Re: How do you make ... Re: Open VMS Cloning POSIX 1003.1 & 1003.2 9 Re: Problem Logging Into DECWindows - LMF License Check??  Re: Status of EV7  Re: Status of EV7  Re: Status of EV7  Re: Status of EV7  Re: VMS Margin-alized  VMS Umbrella Re: VMS Umbrella Re: VMS Umbrella Re: Wizard page re-organization   F ----------------------------------------------------------------------  % Date: Sun, 11 Feb 2001 00:54:38 +0500 + From: jamil ur rehman <jameel@super.net.pk>  Subject: create User, Message-ID: <3A859C7E.6066DA2B@super.net.pk>   Sir,@ hope that you will be ok , i can not create the user and can notF login.How can i create a user and login with it sucessfully.I am using
 VMS 5.5.2. Regards, Jamil    ------------------------------  % Date: Sat, 10 Feb 2001 22:05:37 +0100 > From: "Jean-Franois Marchal" <jean-francois.marchal@x9000.fr> Subject: Re: create User. Message-ID: <964a8c$ck6$1@reader1.imaginet.fr>  2 You must use AUTHORIZE to create your new user ...C If you already did so, I guess you forgot to clear the DISUSER flag $ UAF> MODIFY username /FLAG=NODISUSER   Cordialement Jean-Franois Marchal  X9000 - LYON (FR)     8 "jamil ur rehman" <jameel@super.net.pk> wrote in message& news:3A859C7E.6066DA2B@super.net.pk... > Sir,B > hope that you will be ok , i can not create the user and can notH > login.How can i create a user and login with it sucessfully.I am using > VMS 5.5.2.
 > Regards, > Jamil  >    ------------------------------  % Date: Sat, 10 Feb 2001 16:57:13 -0800 2 From: "Randy Park" <rjpark@mindspring.com.nospaam> Subject: Re: create User2 Message-ID: <964o2a$pe7$1@slb5.atl.mindspring.net>  A Don't forget to SET DEFAULT SYS$SYSTEM: before running AUTHORIZE.   G Jean-Franois Marchal <jean-francois.marchal@x9000.fr> wrote in message ( news:964a8c$ck6$1@reader1.imaginet.fr...4 > You must use AUTHORIZE to create your new user ...E > If you already did so, I guess you forgot to clear the DISUSER flag & > UAF> MODIFY username /FLAG=NODISUSER >  > Cordialement > Jean-Franois Marchal  > X9000 - LYON (FR)  >  > : > "jamil ur rehman" <jameel@super.net.pk> wrote in message( > news:3A859C7E.6066DA2B@super.net.pk... > > Sir,D > > hope that you will be ok , i can not create the user and can notJ > > login.How can i create a user and login with it sucessfully.I am using > > VMS 5.5.2. > > Regards,	 > > Jamil  > >  >  >    ------------------------------  % Date: Sat, 10 Feb 2001 20:11:33 -0500 - From: JF Mezei <jfmezei.spamnot@videotron.ca> & Subject: Efficiency in processing bits, Message-ID: <3A85E6BD.6A63CF68@videotron.ca>  @ With DECC, one can access individual bits in byte/word/longword.  & How efficient is this type of access ?  I If I have to treat a "string" of 1728 bits, would it be more efficient to N first convert it to 1728 bytes (which is easy because for each 8 bit byte, you> can have a table of 8 bytes with which you build the "string".  F Is there a huge difference in how VAX and ALPHA treat bits in terms of efficiency ?   ------------------------------  % Date: Sat, 10 Feb 2001 22:42:11 -0500 2 From: rdeininger@mindspring.com (Robert Deininger)* Subject: Re: Efficiency in processing bitsL Message-ID: <rdeininger-1002012242110001@user-2ivecjp.dialup.mindspring.com>  5 In article <3A85E6BD.6A63CF68@videotron.ca>, JF Mezei % <jfmezei.spamnot@videotron.ca> wrote:   B > With DECC, one can access individual bits in byte/word/longword. > ( > How efficient is this type of access ? > K > If I have to treat a "string" of 1728 bits, would it be more efficient to P > first convert it to 1728 bytes (which is easy because for each 8 bit byte, you@ > can have a table of 8 bytes with which you build the "string". > H > Is there a huge difference in how VAX and ALPHA treat bits in terms of > efficiency ?   Depends on what you want to do.   * Are your basic data items only 1 bit wide?  J Alphas only operate on registers, never memory.  So to set a single bit inI a memory location, it must be read into a register, modified, and written D back out to memory.  If you do that a lot, your program's speed will0 suffer unless everything tends to stay in cache.  D For maximum speed, use an array of 1728 longwords.  You can modify aJ longword faster than you can modify a bit.  Bytes are just as bad as bits,H except maybe on the newest alphas.  (I haven't looked at 21264 and newer in any detail.)    --   Robert Deininger rdeininger@mindspring.com    ------------------------------   Date: 10 Feb 2001 21:40:23 GMT' From: dashw459@aol.comeatspam (Doug W.)  Subject: EV8 description: Message-ID: <20010210164023.05438.00000573@ng-ch1.aol.com>  A There is an interesting series of articles on EV8 architecture at @ www.realworldtech.com.  SMT, hardware assisted locking and more.   ------------------------------    Date: 10 Feb 2001 18:44:08 -05009 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen) < Subject: Re: Firmware upgrade for PWS 600au to use a ZLXp-L13 Message-ID: <SqRPhwk$DBKt@eisner.encompasserve.org>   X In article <3A8524B8.33CAC0DF@infopuls.com>, Christof Brass <brass@infopuls.com> writes: > Larry Kilgallen wrote: >>  [ >> In article <3A848A63.49892BFF@infopuls.com>, Christof Brass <brass@infopuls.com> writes:  >>  D >> > Okay, the Mac offers a real convenient way of using multiple (? >> > - at least two) screens.  >>  9 >> The Macintosh software will handle at least 6 screens.  >>  I >> How many PCI slots are in any particular Macintosh model is a separate 	 >> issue.  >>  Q >> ============================================================================== Q >> Great Inventors of our time: Al Gore -> Internet; Sun Microsystems -> Clusters Q >> ==============================================================================  > B > This is a strange thing. I always regarded one aspect of writing@ > software as a decision between the one and the many case (1 vs= > n). I can't imagine how a system can work for more than one  > (logically) and stop at 6.  C What is a strange thing is that you interpreted my answer as saying > that the Macintosh software cannot handle more than 6 screens.  @ Whether it can or it cannot I do not know, and I did not say :-)   ------------------------------  % Date: Sat, 10 Feb 2001 13:10:04 -0700 1 From: Vance Haemmerle <vance@toyvax.Tucson.AZ.US> ' Subject: Re: FORCING 8BIT COLOR IN DCE? 3 Message-ID: <3A853DAC.602C6775@toyvax.Tucson.AZ.US>    Martin Vorlaender wrote: > ' > Mark London (mrl@psfc.mit.edu) wrote: O > > I have a 24 bit color monitor, but I don't want to use 24 bit color because I > > of a problem with an application (ghostscript, see my previous post). I > > Is there anyway to force DCE to start up and use 8 bit color instead?  > M > I'm not near a DECwindows system right now, but I remember seeing something M > like DEFAULT_VISUAL in one of the SYS$MANAGER:DECW$PRIVATE* files (probably M > DECW$PRIVATE_SERVER_SETUP). Copy from .TEMPLATE to .COM if the .COM doesn't ? > yet exist, edit appropriately, and then restart DECwindows by  > ' >   $ @SYS$STARTUP:DECW$STARTUP RESTART  > ) > Note that this kills all open sessions!  > I > It's also possible for some graphics cards to set the color depth via a I > special symbol; dig in the DECW$DEVICE_*.COM for your graphic device to  > find out.   >  Put the following in your DECW$PRIVATE_SERVER_SETUP.COM file:  / $       DECW$SERVER_DEFAULT_VISUAL_CLASS == "3"   B 3 means 8-bit Pseudocolor.  This only changes the default visual.  ApplicationsE which check the capabilities of the server and specifically request a  24-bit# TrueColor visual will still get it.   G Note: This does not work on VAX, though the documentation says it will. G Someone at OpenVMS engineering:  Please fix or check that this is fixed  for VMS 7.3.     --B Vance Haemmerle               Internet   vance@toyvax.Tucson.AZ.US( Tucson, AZ                    Web       " http://toyvax.Tucson.AZ.US/~vance/   ------------------------------  # Date: Sat, 10 Feb 2001 21:00:56 GMT - From: goathunter@goatley.com (Hunter Goatley)  Subject: Re: Free space ? / Message-ID: <3a85aab7.5797786@swen.process.com>   P On Sat, 10 Feb 2001 12:01:41 GMT, system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) wrote:   _ >In article <wv_g6.601$uY2.12066@news2-hme0>, "John Griffiths" <johngriffiths@ic24.net> writes: I >>Free space is shown in blocks, how do I work out the % of the full disc  >>space of 9.1gb ? >>Thanks in anticipation.  >> >> > & > ((MAXBLOCK-FREEBLOCKS)/MAXBLOCK)*100 > C You might want to try FREE, my utility that displays disk usage for C all mounted disks (there are lots of others out there---this is the  one I use):    $ freeP Device Name      Volume Label   Type      Used Blocks     Free Blocks      TotalP ---------------- -------------- ------ --------------- --------------- ---------P CAESAR$DUA0:     SYS_CAESAR     RA90      770946 (33%)   1605207 (67%)   2376153P CAESAR$DUA1:     VAXVMS061      RA90     1725354 (73%)    650799 (27%)   2376153P CAESAR$VDA0:     WORK           VD01        5045 ( 8%)     64587 (92%)     69632P CAESAR$VDA1:     SCRATCH        VD01        1279 ( 9%)     13721 (91%)     15000  P Totals:                4 mounted disks    1279MB (52%)    1194MB (48%)    2473MB   You can find it here:   ' ftp://ftp.wku.edu/vms/fileserv/free.zip 3 ftp://ftp.tmk.com/ftp-wku-edu/vms/fileserv/free.zip / ftp://ftp.process.com/wku/vms/fileserv/free.zip   ! and on all the other WKU mirrors.      Hunter ------9 Hunter Goatley, Process Software, http://www.process.com/ 9 goathunter@goatley.com     http://www.goatley.com/hunter/    ------------------------------  # Date: Sun, 11 Feb 2001 00:46:58 GMT * From: Alan E. Feldman <alan48@my-deja.com> Subject: Re: Free space ? ) Message-ID: <964ne1$va3$1@nnrp1.deja.com>   0 In article <009F76B7.C8A63123@SendSpamHere.ORG>,    system@SendSpamHere.ORG wrote:? > In article <wv_g6.601$uY2.12066@news2-hme0>, "John Griffiths"   <johngriffiths@ic24.net> writes:E > >Free space is shown in blocks, how do I work out the % of the full  disc > >space of 9.1gb ?  > >Thanks in anticipation. > >  > >  > ' >  ((MAXBLOCK-FREEBLOCKS)/MAXBLOCK)*100  >   C Have you tried this? I believe it will give zero every time because  this is integer arithmetic.   E I recommend instead FREEBLOCKS/(MAXBLOCK/100) for percentage free and > (MAXBLOCK-FREEBLOCKS)/(MAXBLOCK/100) for percentage full. ThisE algorithm will avoid overflow problems for disks larger than about 11 F GB and will work fine for all but the tiniest disks (tiny meaning less& than, maybe, 10000 blocks total size).   > --C > VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)  TMESIS(dot)COM > E > city, n., 1. a place where trees are cut down and streets are named  after them.  >    --F NOTE: If you wish to e-mail me, please do NOT use the deja address. ItD is not a valid address. Instead, use the address below, removing the long wrong part first. Thanks.   Disclaimer: JMHO Alan E. Feldman  &-)( afeldman@gfigroup.ButItSaidItPrinted.com     Sent via Deja.com  http://www.deja.com/   ------------------------------    Date: 10 Feb 2001 14:25:52 -0500: From: malmberg@eisner.encompasserve.org (John E. Malmberg)O Subject: Multinet installation in cluster problem - was Re: How do you make ... 3 Message-ID: <FQ5EJCUC7WFl@eisner.encompasserve.org>   7 > In article <3A849A10.50CE@nospam.chemistry.ucsc.edu>, 4 Todd Wipke <wipke@nospam.chemistry.ucsc.edu> writes:  > Please pick a title that better describes the real problem you@ are trying to solve, thanks.  There are folks from Multinet that= do occasinally look at this newsgroup / mailing list that may " have missed your posting.  Thanks.  @ >Installing Multinet with the new file structure requires makingD >a directory pointer to the vax_common directory in each homogeneousA >cluster satellite node, BUT they don't describe how to do it.  I D >tried copying the syscommon link in the first node's directory, but >just got an empty directory.  > D >Could someone email me a com file that does this or describe how to@ >create a link?  Remove the nospam from my email address for the+ >correct email address.  Thanks in advance.    Christoph Gartman wrote:@ | Normally Multinet does this during installation. Otherwise see |   HELP SET FILE/ENTER   : If Multinet does normally do this during the installation:  F I would advise you to review or reinstall the Multinet product to make> sure that it was installed correctly, and if it still does not: to either contact Multinet folks at Process or post on theD vmsnet.networks.tcp-ip.multinet newsgroup to ask for specific adviceF on why the installation failed before attempting to repair the problem
 your self.  E I have at a past job spent quite a bit of time troubleshooting a sick G cluster that someone got quite "creative" with using SET FILE/ENTER and H in the process removed the entire vms$common directory tree, and had two( non-identical syscommon directory trees.  F In short, the SET FILE/ENTER command should only be used by people whoC know exactly what they are doing with it and WHY they are doing it.    -John  wb8tyw@qsl.network Personal Opinion Only    ------------------------------    Date: 10 Feb 2001 14:05:58 -0500: From: malmberg@eisner.encompasserve.org (John E. Malmberg) Subject: Re: Open VMS Cloning 3 Message-ID: <vgxaq$czYtEG@eisner.encompasserve.org>   , In article <3A84B4D1.71278C0A@videotron.ca>,/ JF Mezei <jfmezei.spamnot@videotron.ca> writes: E > OK, I know that to be pedantic, you should really backup by booting B > from a CD so that the source system disk is more or less unused.1 > (does booting from CD use any page/swap file ?)   @ Only if you map a page and swap file after you get it to DCL :-)  H > However, in practice, if you boot on the source hard drive but have no8 > applications running and no users logging in,  doesn't3 > BALCUP/IMAGE/IGNORE=INTERLOCK work just as well ?   D It is a case of "That Depends".  It depends on how many applications that you have shut down.  D When you boot from Standalone Backup, or the CD-ROM or an alternate,E possibly minimum system disk, you can be sure that there are no files ' open at all on the primary system disk.   H When you boot minimum, it is probably close to the same state, but there. are also many features that are not available.    D There are some OpenVMS utilities, layered products that you may haveE forgotten about when doing an Image backup from a active system disk.   G These include disk defragmenters, The batch and print queues, even when ) idle, and The Management Station process.   E I have had to reinitialize the QUEUE database with new files after an  image restore.  H The Management Station product locks the SYSUAF file from possibly beingK backed up.  I have seen advice on comp.os.vms to use the Convert Utility to  back it up.   H I recommend that relying on your backup tapes not be your only method ofG rebuilding your system after a disaster happens.  Keep documentation on L how to rebuild your system from known good distributions, and keep a copy of4 your application configuration in also a safe place.   -John  wb8tyw@qsl.network Personal Opinion Only    ------------------------------  # Date: Sat, 10 Feb 2001 20:31:50 GMT ' From: moi_is_me <moi_is_me@my-deja.com>  Subject: POSIX 1003.1 & 1003.2) Message-ID: <9648fm$jg2$1@nnrp1.deja.com>    Hi,     According to the FAQ at  G Http://www.openvms.compaq.com/solutions/government/coe/dii_COE_Faq.html   * OpenVMS will support POSIX 1003.1 & 1003.2  B However, trying to determine what POSIX 1003.1 & 1003.2 encompass, is a different matter.  A One site suggests that IEEE 1003.1-1996 includes .1b, .1c and .1i = whereas another suggest that 1003.1b (formerly 1003.4) is NOT  part of 1003.1    A Without upsetting IEEE (dont want the whole doc), anybody care to @ enlighten me as to what is included in POSIX 1003.1 & 1003.2 ???     TIA    -PierreV           Sent via Deja.com0 http://www.deja.com/   ------------------------------  % Date: Sat, 10 Feb 2001 20:57:25 -0000c, From: "Mark Lamond" <markl@markl.force9.net>B Subject: Re: Problem Logging Into DECWindows - LMF License Check??( Message-ID: <MRhh6.1881$I5.50869@stones>   Hi,s    Got it sorted out now - thanks!!   Mark Lamondc markl@markl.force9.net    = Hoff Hoffman <hoffman@xdelta.zko.dec.nospam> wrote in message 0 news:93i25d$ol4$2@mailint03.im.hou.compaq.com... >c: > In article <DwN66.16635$I5.309491@stones>, "Mark Lamond"  <markl@markl.force9.net> writes:D > :...OpenVMS (7.2), which i have just installed on my DEC 3000-600. > :dH > :When i boot the machine, it starts up with the graphical login, and i thenB > :login as system and enter my password as set during the install
 procedure. > :mL > :However, i then get a message which says "LMF License chack has failed" - do. > :you have any idea what could be wrong here? >cL >   You probably don't have the full set of licenses installed, you have oneJ >   or more licenses that are not loaded, or you have one or more licensesI >   that are not valid, or (if this is a cluster) you have not loaded allr? >   of your licenses into all of the license databases present.  >s' >   The usual set of core license PAKs:> >  >     OpenVMS:, >       OPENVMS-ALPHA and OPENVMS-ALPHA-USER >     DECwindows:s  >       DW-MOTIF or NET-APP-SUP* >     TCP/IP Services: >       UCX or NET-APP-SUP*e
 >     DECnet:9 >       DVNETEND or DVNETRTG or DVNETEXT or NET-APP-SUP*.m >@A >   Use @SYS$UPDATE:VMSLICENSE to register and load license PAKs..1 >   Use SHOW LICENSE to list the active licenses.cB >   Use LICENSE LIST to list the contents of the license database./ >   Use LICENSE LOAD to load licenses manually. 3 >   Use LICENSE UNLOAD to unload licenses manually.- > D >   Since you cannot log in, you will require a login on the consoleD >   serial line (if you have set the console up for serial access inC >   the firmware) or you will have to shut down and boot minimally,TA >   and log in on the non-DECwindows console -- set STARTUP_P1 tooE >   "MIN" during a conversational bootstrap.  (Please see the OpenVMSnB >   FAQ for some details on conversational bootstraps, and see theE >   OpenVMS manuals for full details of conversational bootstraps andA* >   registering and maintaining licenses.) >a, >  --------------------------- pure personal# opinion --------------------------- 1 >    Hoff (Stephen) Hoffman   OpenVMS Engineering9 hoffman#xdelta.zko.dec.com >    ------------------------------  % Date: Sun, 11 Feb 2001 00:25:31 +0000t, From: Peter Boyle <pboyle@holyrood.ed.ac.uk> Subject: Re: Status of EV7H Message-ID: <Pine.GSO.4.10.10102102357070.3018-100000@holyrood.ed.ac.uk>   Hi,-  # On 9 Feb 2001, Nick Maclaren wrote::  3J > In article <Pine.GSO.4.10.10102091610270.3018-100000@holyrood.ed.ac.uk>,0 > Peter Boyle <pboyle@holyrood.ed.ac.uk> writes: > >dD > > Blah blah about 128 regs covering cache latency on wide issue fp > >r > D > Even if it did fly, there is nothing to stop competitors doing the# > same, only using more registers. f  < Yes. However they either have to architect more registers to; cover L2 latency on really wide issue or use Hitachi style w' windowing. Even Hitachi style windowingaA with a window size of 32 would make it a bit tricky for 8 way fp.i/ with, say a 4 cycle latency and no fused madds.   % > The SR2201 uses PA-RISC as a basis,dE > the SR8000 uses POWER, and the techniques are equally applicable topC > any other clean RISC architecture.  In fact, the way that Hitachi.B > have done it would allow them to use thousands of registers, but > just don't switch context :-)d  B I assume they can extend of the number of windows without changing the user visible architecture?.i  nD > A more plausible approach to this would be to use cache preloadingE > in a big way.  This could give the effect of thousands of registerssA > without touching the basic CPU architecture.  You wouldn't eveneB > need a very large number of outstanding cache prefetches, if the? > compiler got it right, but just don't mispredict branches :-)i  A My analysis was for just this sort of code, however my main pointm was that a rule of thumb:s  $ (2-4)*(FPU latency )*Superscalarity   I architected registers are needed to keep pipelines full from L2, assumingtD prescient prefetching puts everything in L2 for you. Windowing gainsC you in terms of *architected* regs as load latency doesn't have to  7 be covered by *architectural* regs. If we try to becomedH much more aggressively superscalar with architectures like Power, alpha,B sparc etc... we won't be able to express intrinsically independentF operations as independent due to the limited architectural registers.   9 So IMHO there is a space in which the IA-64 architecture -C *in principle* could fly where the *standard* risc chips couldn't. u  C > So, even if the IA-64 line does fly for these codes, don't expect0@ > it to be left to clean up that market on its own.  Anyway, the  = Sure, but I would bet there will be extra registers appearings@ on the my-own-risc-processor front sometime in the next 6 years.  ? > market isn't all that big, and will certainly not justify theaA > IA-64 line.  Intel simply HAS to make the IA-64 line perform ono. > the spaghetti C that is typical of GUIs etc.  % Heh, branching is a different matter!4  
 All the best,u   Peter  > 
 > Regards, > Nick Maclaren,, > University of Cambridge Computing Service,@ > New Museums Site, Pembroke Street, Cambridge CB2 3QG, England. > Email:  nmm1@cam.ac.uk1 > Tel.:  +44 1223 334761    Fax:  +44 1223 334679    ------------------------------  % Date: Sat, 10 Feb 2001 23:35:27 -0000h& From: "Thomas Womack" <tom@womack.net> Subject: Re: Status of EV7. Message-ID: <964j5d$h4g$1@news7.svr.pol.co.uk>  % "Chris Morgan" <cm@mihalis.net> wrote=* > "Thomas Womack" <tom@womack.net> writes: >=? > > By your measure, the only modern microprocessor not majorlyIC > > bottlenecked is the EV7, and that's significantly more vaporous= > > than the Itanium.= >=E > It's true that EV7 is still vapour, but it doesn't strike me as thed. > Thing That Should Not Be as much as Itanium.  K Yes, EV7 seems a much more plausible extrapolation from EV6 than Itanium iscJ from anything Intel made before. But I have talked to more than one personI who has working Itanium hardware at I'm-sorry-I-can't-tell-you MHz in hisfE office -- OK, sometimes used more as a space-heater than as a compute K engine -- and there are a couple of Itanium boxes at various sites that youeK can get accounts on if you sign the right NDAs. I'd suspect Itanium was nownI delayed for verification or marketing reasons, whilst I'm not sure anyone0G outside Compaq's labs has an EV7 at all. That was the sense of vapour Io meant.   > If it really is the 21264eG > with cache memory, memory controller, I/O controller, comms links etc = > on-chip, I doubt it will be _quite_ so late as the originals > 21264.  G Well, the announcement of the 21264 was October 1996, "samples Q1 1997, J availability H2 1997"; benchmark numbers had appeared by July 1998, and byJ September 1999 there was a 500MHz XP1000 on my supervisor's desk; DecemberK 2000 we got a couple of 667MHz ES40 boxes in the central computing facility. to> complement and eventually replace their quad-EV56 main server.  K The 21364 was announced October 1998, "tapeout 4Q99". So it's already laterkE since the announcement than the 21264 was; how long should tapeout ->rJ samples take? I certainly don't expect to see one on my supervisor's desk,I or even in the server room in the applied-maths building, this September.r   TomR   ------------------------------  # Date: Sun, 11 Feb 2001 03:25:38 GMT=# From: Chris Morgan <cm@mihalis.net>  Subject: Re: Status of EV71 Message-ID: <87k86xc3cn.fsf@dumpster.mihalis.net>i  ( "Thomas Womack" <tom@womack.net> writes:  M > The 21364 was announced October 1998, "tapeout 4Q99". So it's already later,G > since the announcement than the 21264 was; how long should tapeout -> L > samples take? I certainly don't expect to see one on my supervisor's desk,K > or even in the server room in the applied-maths building, this September.   F Oh I beg your pardon then. Must be showing my age, I thought the 21364F was announced "not so long ago" when I posted but come to think of it,9 it was quite a while ago that Peter Bannon's 21364 slideseD _disappeared_ off the face of the Compaq website (is it me or is theA Compaq website amazingly bad?) - last seend on their European OEMn site.   F I wonder how much excitement the chip will cause when it comes out. ItE may be a cool piece of hardware, but since it probably wont run Quake  III just what is the point? :)  B I'm just fascinated to find out what they can really do with those> on-chip "mesh" links to other chips. Much more _interestingly_D different to other chips than many recent chip launches. At least to me.    Chris  -- hH Chris Morgan <cm at mihalis.net>                  http://www.mihalis.net        Temp sig. - Enquire within   ------------------------------  # Date: Sun, 11 Feb 2001 06:48:51 GMT-; From: Mark Garrett <Mark.Garrett@wedontwantyourspam.com.au>0 Subject: Re: Status of EV7C Message-ID: <B6AC7D03.11CA0%Mark.Garrett@wedontwantyourspam.com.au>3  J in article rdeininger-0902011111550001@user-2ive7gb.dialup.mindspring.com,H Robert Deininger at rdeininger@mindspring.com wrote on 10/02/2001 03:11:  5 > In article <ducVmkQJDblD@eisner.encompasserve.org>, 5 > young_r@eisner.encompasserve.org (Rob Young) wrote:) > O >> In article <rdeininger-0902011053300001@user-2ive7gb.dialup.mindspring.com>,:7 >> rdeininger@mindspring.com (Robert Deininger) writes:t@ >>> In article <87pugr515d.fsf@prep.synonet.com>, Paul Repacholi" >>> <prep@prep.synonet.com> wrote: >>>  >>> F >>>> I fully expect the IA-64 to outperform the EV3 on selected codes. >>>  >>> Ok, I'll take the bait.  >>>  >>> What's an EV3? >>>  >> e >> The precursor to EV4  :-) >> r> >> EV3 was the lab version, EV4 was the first production Alpha> >> aka 21064 circa 1993 (announced Nov 1992, program announced0 >> in Feb 1992 IIRC) at a then stunning 200 MHz. >  > O > Hmm.  My 21064 reference manual from June 1996 mentions 150, 166, and 200 MHzj8 > chips, and 21064A chips at 200, 233, 275, and 300 MHz. > L > But there were 21064 _systems_ with at least these clock speeds: 100, 125,P > 133, 150, 175, and 200 MHz.  And 21064A systems with at least 225 and 275 MHz. > N > Did the 21064 really start at the top speed of 200 MHz, and then they made aN > series of slower ones?  That doesn't correspond with the order of release of > the machines.L > J > Is there enough tweekability in an alpha chip to run a 150 MHz part (for# > example) at 100, 125, or 133 MHz?n > # > What's the EV number of a 21064A?      EV45    I I can't remember what the console said on early 3000 alphas (Flamingo and I then Hot Flamingo) its mentions a Pass number like 21064 and the P2 or P3SL I don't think the EV was even in usage at that stage and its likely though II don't know this, that EV was just new terminology from marketing for P3++O
 becoming EV4.a    
     Cheers         Mark ;)s   ------------------------------   Date: 10 Feb 2001 13:44:44 CDT= From: wayne@tachysoft.xxx.042685.killspam.00bb (Wayne Sewell)g Subject: Re: VMS Margin-alized. Message-ID: <RKJ+dyz2ylRA@tachxxsoftxxconsult>  ` In article <JwPlYwnkYhts@malvm1.mala.bc.ca>, nothome@spammers.are.scum (Malcolm Dunnett) writes:     > H >    None of our Wintel servers carry the Compaq brand, and I doubt that > any ever will. k  N Of course.  Compaq never seems to get that message.  You can get VMS and alphaL *only* from compaq.  You can get a billybox from anybody.  Yet they push theE latter so that billy can get most of the profits.  If they could weanoL themselves off dependence on billyboxes and concentrate on vms and nsk, they could keep *all* the profits.r       -- nO =============================================================================== M Wayne Sewell, Tachyon Software Consulting  (281)812-0738  wayne@tachysoft.xxxd: http://www.tachysoft.xxx/www/tachyon.html and wayne.html  K change .xxx to .com in addresses above, assuming you are not a spambot  :-)eO ===============================================================================eB Cute Girl, to Curly: "Oh, what a beautiful head of bone you have!"   ------------------------------  % Date: Sat, 10 Feb 2001 14:59:13 -0700 % From: Dan O'Reilly <dano@process.com>a Subject: VMS UmbrellalA Message-ID: <5.0.2.1.2.20010210145840.00a77640@ntbsod.psccos.com>L  G Just got my official OpenVMS umbrella.  Thanks, Sue & Co., I needed one  anyway!i     ------I +-------------------------------+---------------------------------------+cI | Dan O'Reilly                  |                                       |cI | Principal Engineer            |  "Why should I care about posterity?  |eI | Process Software              |   What's posterity ever done for me?" |eI | http://www.process.com        |                    -- Groucho Marx    |bI +-------------------------------+---------------------------------------+T   ------------------------------  # Date: Sat, 10 Feb 2001 23:07:25 GMTt3 From: Eric Dittman <dittman@narnia.int.dittman.net>r Subject: Re: VMS UmbrellaI> Message-ID: <NQjh6.1077$wu5.27630@e420r-sjo2.usenetserver.com>  & Dan O'Reilly <dano@process.com> wrote:I : Just got my official OpenVMS umbrella.  Thanks, Sue & Co., I needed one-	 : anyway!-  K I received mine today as well. I picked up the tube and thought, "wow, theypL must have really packed in a lot of posters!"  I was pleasantly surprised to# find out I had an OpenVMS Umbrella.  -- n Eric Dittman dittman@dittman.neto   ------------------------------  % Date: Sat, 10 Feb 2001 18:31:40 -0500 9 From: "Steven Shamlian" <not dot an at earthling dot net>m Subject: Re: VMS Umbrellas2 Message-ID: <964itj$omo$1@slb6.atl.mindspring.net>  < Hey, where do I get on this Umbrella-recieving mailing list? =+=Steven Shamlian=+=i  2 "Dan O'Reilly" <dano@process.com> wrote in message; news:5.0.2.1.2.20010210145840.00a77640@ntbsod.psccos.com...dI > Just got my official OpenVMS umbrella.  Thanks, Sue & Co., I needed ones	 > anyway!  >i >  > ------K > +-------------------------------+---------------------------------------+ K > | Dan O'Reilly                  |                                       |bK > | Principal Engineer            |  "Why should I care about posterity?  | K > | Process Software              |   What's posterity ever done for me?" |oK > | http://www.process.com        |                    -- Groucho Marx    | K > +-------------------------------+---------------------------------------+t >    ------------------------------  % Date: Sun, 11 Feb 2001 00:27:00 +0000d; From: Malcolm MacArthur <malcolmm@rustic-place.demon.co.uk> ( Subject: Re: Wizard page re-organization8 Message-ID: <3A85DC54.61804AB3@rustic-place.demon.co.uk>   Warren Sander wrote: > B > I've taken the January through July 2000 atw's and put them intoO > a second archive. Now there is July 1998 to December 1999 and January 2000 toi8 > July 2000 archives (both with and without javascript). > ( > The page is now down to 140K from 250K > - > The search functions will find all of them.d > L > We have had a 'keyword' project going on to put every question/answer intoI > different catagories (20 different ones were decided on) but before the F > guy doing this over the summer got done he had to go back to school.I > At some point this will get finished and you will be able to do keywordm8 > searchs and have ATW listings based on those keywords. > 0 > IF you have any other suggestions let me know. >   J How about a 'feedback' system, to allow ordinary users to make comments onL Wizard topics? With a disclaimer, of course, stating that the information is7 not provided by Compaq and may not be fully accurate...e  K Many minds work better than several (judging by posting styles, there don't / seem to be more than two or three 'wizards'...)1  	 -Malcolm.    ------------------------------   End of INFO-VAX 2001.083 ************************enlighten me as to what is included in POSIX 1003.1 & 1003.2 ???     TIA    