1 INFO-VAX	Mon, 10 Apr 2000	Volume 2000 : Issue 201       Contents: RE: 1.6 GHz Alpha ! Alpha 1000a to 1200 SCSI Question % Re: Alpha 1000a to 1200 SCSI Question 
 Alpha Prices?  Re: CC patches ?% Re: Critical/important bug on VMS ...  Re: Freeware MIME 2 Re: Galaxy (or perhaps Wildfire) and shared memory2 Re: Galaxy (or perhaps Wildfire) and shared memory2 Re: Galaxy (or perhaps Wildfire) and shared memory2 Re: Galaxy (or perhaps Wildfire) and shared memory2 Re: Galaxy (or perhaps Wildfire) and shared memory2 Re: Galaxy (or perhaps Wildfire) and shared memory2 Re: Galaxy (or perhaps Wildfire) and shared memory GBLSECTIONS change problem Re: GBLSECTIONS change problem* Re: I can't run/detach in my own X session Is sys$dclexh ever not called?" Re: Is sys$dclexh ever not called? Re: locations of C libs in VMS Re: locations of C libs in VMS" Re: Pascal Compile-TIme stack dump Request for Wildfire/Galaxy URL # Re: Request for Wildfire/Galaxy URL  Re: So who will buy VMS ?  Re: So who will buy VMS ? & Re: Sun considers VMS a "mainframe" OS Re: Sun question for Andrew  UCX port assignments Re: UCX port assignments< Unique Strategy Trading Secrets                         8667 Re: VAX BASIC Specifications?  Re: VAX BASIC Specifications?  Re: VAXStation 3100 boot disks Re: VMS indexed files to Unix 2 Re: VMS Information Is Online - VMS Isn't Dead Yet( Re: Warning: circuitsurgeon is a ripoff!( Re: Warning: circuitsurgeon is a ripoff!( Re: Warning: circuitsurgeon is a ripoff!( Re: Warning: circuitsurgeon is a ripoff!( Re: Warning: circuitsurgeon is a ripoff!( Re: Warning: circuitsurgeon is a ripoff!( Re: Warning: circuitsurgeon is a ripoff!( Re: Warning: circuitsurgeon is a ripoff!( Re: Warning: circuitsurgeon is a ripoff!( Re: Warning: circuitsurgeon is a ripoff!O Warning: Circumcision is a ripoff!  (Was: Warning: circuitsurgeon is a ripoff!) % Re: Wierd Elsa Gloria Synergy Problem ' Will COMPAQ Australia please contact me   F ----------------------------------------------------------------------   Date: 10 Apr 2000 14:59:21 GMT2 From: mathog@seqaxp.bio.caltech.edu (David Mathog) Subject: RE: 1.6 GHz Alpha+ Message-ID: <8csq89$9q@gap.cco.caltech.edu>   x In article <910612C07BCAD1119AF40000F86AF0D8052841D0@kaoexc4.kao.dec.com>, "Main, Kerry" <Kerry.Main@Compaq.com> writes: >David,  > M >You are correct that there is a place for single processor boxes and a place  >for SMP config's. > B >Kind of like the auto industry - no one solution will address all >requirements. > J >Btw - in case anyone thinks the 1Ghz x86's are now faster than single cpu" >Alpha's, check out the following:I >http://www.spec.org/osg/cpu2000/results/cpu2000.html (single cpu EV6/667 M >still rated higher (444 vs 410) in Spec CPU2000 and still MUCH higher in the " >Specfp 2000 (577 vs 284) results)   That's not a complete analysis.   J The benchmark you cite is for a 667 Mhz DS20E - which due to its high costK is unlikely to be used in a massively parallel system (unless money were no J object as at Celera or the NSA.)  The DS10 and DS10L have a 466 Mhz clock,J and all else being equal would score about a 310 Cint2000 and 403 Cfp2000.K (All else is not equal though, the memory bandwidth on the DS20E is 4x that H of the DS10.)  Those estimated DS10 numbers are slower in absolute termsH for integer operations and only about 33% faster for floating point thanG the 1Ghz PIII in that chart.   One can only imagine that Compaq did not K submit the DS10 or DS10L numbers for that very reason.  There are as yet no I Athlon numbers in that chart, and the Athlon does a bit better on FP than H the PIII (unclear if any of the current crop of Athlon motherboards are I that fast though, due again, to memory bandwidth limitations).   Finally, G divide the performance numbers by the price of the single cpu computing H blocks and you're not looking at a clear win for Alpha in terms of totalI CPU power/$.   This should come as no surprise to anybody here since from E day 1 the Alpha mantra has always been twice the power at 3 times the G price.  That's only a winning formula for those few niches that need as & much power as possible per single CPU.  	 Regards,     David Mathog mathog@seqaxp.bio.caltech.edu ? Manager, sequence analysis facility, biology division, Caltech     ------------------------------  % Date: Mon, 10 Apr 2000 11:09:06 -0700 ( From: "Russell Leathe" <russ@gordon.edu>* Subject: Alpha 1000a to 1200 SCSI Question% Message-ID: <M4mI4.35$6i7.728@client>   > I'm migrating from an Alpha 1000a to an Alpha 1200.  The 1000a" disk array is set up as follows...   4.3gb disks....   . DKA0            SYSTEM              Type RZ29B+ DKA100        STU1997            TYPE RZ29B + DKA200        STU1998            TYPE RZ29B + DKA300        STU1999            TYPE RZ29B + DKA400        STU2000            TYPE RZ29B   	 CD-ROM...   * DKA600        CD-ROM            TYPE RRD46   All are on the same SCSI chain.   I On the Alpha 1200 the DISKs and CD-ROM are on different SCSI controllers.  Thus....  ! NCR 53c810 Slot 1, BUS 0 PCI ....    DKA500    CD-ROM        RRD47     # QLOGIC ISP1020 Slot 2, BUS 0 PCI...   & DKB0                        TYPE RZ1CF DKB100                    "" DKB200                    "" DKB300                    "" DKB400                    ""  L When I initially set out to perform the migration I thought...Well I'll justE perform an image backup and restore, run autogen and be done with it. A Unfortunately, this wouldn't work the way the disks are arranged.   K My question is, is there a way to change the hardware that will re-allocate I the drives as DKA### and the CD-ROM to DKB### on the Alpha 1200.  Or am I . facing a re-install of 7.2 and a re-install of& the layered products (PMDF, Multinet)?   Thanks in advance,   Russ   ------------------------------  % Date: Mon, 10 Apr 2000 09:22:59 -0700 ? From: Mike Price <mike.priceNOmiSPAM@littlewoods.co.uk.invalid> . Subject: Re: Alpha 1000a to 1200 SCSI Question9 Message-ID: <024205ea.6e377ff8@usw-ex0110-075.remarq.com>    2 things8 a) I have found from several experiences that you should9 always refer to disks as DISK$diskname rather than DKAxxx 7 b) I have got round the problem by defining logicals so 9 that the original disk names still appear to be there but 9 the logical changes them to the real new disk names. This ; may not help if you  want DKA100 to become DKB100 but there 9 is also a real DKA100 but can you switch the disks around % if they are just storage works disks?   4 There is no way I know of that you can have DKA (for: example) and DKB on the same bus as the A or B denotes the9 actual bus. Sometimes you can move the SCSI busses around : to change the names - I am not sure how VMS determines the9 Bus names but I have seen systeme where the console names   are not the same as VMS chooses.   Hope that helps    sorry if it doesn't        * Sent from AltaVista http://www.altavista.com Where you can also find related Web Pages, Images, Audios, Videos, News, and Shopping.  Smart is Beautiful    ------------------------------  # Date: Mon, 10 Apr 2000 11:48:09 GMT ) From: "Neil Rieck" <n.rieck@sympatico.ca>  Subject: Alpha Prices?: Message-ID: <ZbjI4.14414$2D6.394039@news20.bellglobal.com>   Folks,  K Download a copy of the Windows based "Alpha Systems Configuration Tools" at + http://www.digital.com/alphaserver/acu.html    Select "Server"  Select "OpenVMS"' Dialup 1000 MB of memory (yes; one Gig)  Dialup 50 GB of disk storageL Select all machines where you can see the CPU (e.g.. 5/600 is an EV5 running 600 MHz) Click "Configure"   L result: almost all prices are between US$60k and US$140k (except GS140 which is US$404K).  4 p.s. Does anyone know why the GS140 is so expensive?  
 Neil Rieck* Kitchener(New Berlin?)/Waterloo/Cambridge, Ontario, Canada.! http://www3.sympatico.ca/n.rieck/ 6 http://www3.sympatico.ca/n.rieck/docs/alpha_diary.html   ------------------------------  % Date: Mon, 10 Apr 2000 09:00:31 -0400 ( From: Ed Vogel <edward.vogel@compaq.com> Subject: Re: CC patches ? ) Message-ID: <38F1D06F.DFDFF64@compaq.com>    >Does this fix anything?  3     The CCAE03062 kit fixes the following problems:   J         1.      Using the __CMP_STORE_QUAD built-in function with a 64-bitL value as the third parameter of the built-in causes the compiler to generate3 code that truncates the third parameter to 32-bits.   H         2.      An optimization regression causes certain assignments toE struct members to be improperly optimized away as "redundant stores". J Unfortunately, it is not possible to characterize at the source code levelI the circumstances in which this is likely to occur, so you can't look for  problem areas by inspection.  L         3.      The compiler may sometimes perform an incorrect optimizationF when computing certain loop counters.  This bug can manifest as eitherI incorrect code or an internal error.  This bug has been in the C compiler / for some time.  We know it was present in V5.7.     P >Compaq C V6.2 has bugs, bugs, bugs, bugs, bugs! No wait! They're ANSI features!  E     Care to expand on this?  If you'd rather not post specifics, just      e-mail me directly.   0                                         Ed Vogel<                                         Compaq C Engineering   ------------------------------  # Date: Mon, 10 Apr 2000 12:14:25 GMT , From: koehler@eisner.decus.org (Bob Koehler). Subject: Re: Critical/important bug on VMS ...' Message-ID: <2000Apr10.081425.1@eisner>   R In article <8cla69$156$1@pyrite.mv.net>, "Bill Todd" <billtodd@foo.mv.com> writes: > 9 > Bob Koehler <koehler@eisner.decus.org> wrote in message   H >> Actually you'ld only be able to measure your velocity relative to the >> solar wind. > J > I may be being exceptionally dense today, but that kind of sounds like a. > pretty good analogue to 'air speed' to me... >   F Aerodynamically, yes.  Somehow that didn't seem to match the intent of the original post.  F ----------------------------------------------------------------------? Bob Koehler                     | Computer Sciences Corporation = Hubble Space Telescope Payload  | Federal Sector, Civil Group E  Flight Software Team           | please remove ".aspm" when replying    ------------------------------  % Date: Mon, 10 Apr 2000 07:55:22 -0600 & From: "Lee Gillie" <Lee@NOSPAModp.com> Subject: Re: Freeware MIME= Message-ID: <c3lI4.110851$AT6.124726@dfw-read.news.verio.net>   F Depending on use, mime++ may be free: http://www.hunnysoft.com/mimepp/ or at least very reasonable.  A We found it simple to port to VMS.  We wrote our own front end to A VMS mail years ago, and were able to add mime capability via this & package.  This is still under VMS 6.2. --F ______________________________________________________________________F Lee Gillie, CCP                                Remove NOSPAM to E-MailF Online Data Processing, Inc. - 3501 N. Haven -  Spokane, WA 99207-8500  D Frits A.M. Storms wrote in message <8ckjuo$rij$1@cyan.nl.gxn.net>...I >I am looking for a simple freeware tool to encode & decode MIME files on 	 >OpenVMS. A >Could someone point me to the appropriate download site please ?  >  >much obliged, >  >Frits A.M. Storms >    ------------------------------  # Date: Mon, 10 Apr 2000 10:58:56 GMT  From: d.webb@mdx.ac.uk; Subject: Re: Galaxy (or perhaps Wildfire) and shared memory ) Message-ID: <8csc5b$isq$1@nnrp1.deja.com>   / In article <38F13224.4C86088B@vl.videotron.ca>, 3   JF Mezei <jfmezei.spamnot@vl.videotron.ca> wrote:  > Larry Kilgallen wrote:? > > So the very fact that DEQ leaves this as a decision for the < > > customer indicates sharing is not the same as isolation. > F > I remember being told that in a Galaxy, it is *expected* that the OS thatE > coexist with each other are required to behave properly and that if  you haveG > one instance of an OS that doesn't really behave, then you'rte asking  for trouble. >  > So the question should be:D > If you have 2 instances of VMS in a galaxy, does that provide much moreH > robustness compared to an instance of VMS with an instance of Linux of True64 ? >     F I wasn't aware that you could have instances running different OS's inB a wildfire system sharing memory. I thought that the only way with1 multiple OSs was in a shared-nothing environment.   H > The other storey I had heard was that Galaxy was a way to run VMS in a machine ? > which has more processors than VMS can support as a single OS  instance. So youD > run multiple instances of VMS, each with the maximum number of SMP CPUs it can & > handle. Is there any truth to this ?  D I don't know what the maximum number of SMP cpu's VMS can handle is.? However with any SMP solution you get to a point of diminishing B returns where adding extra cpus has little if any positive effect.= The exact point this happens depends on the type of workload. @ Generally Compaq/Digital reckon on reasonably linear performance growth up to 8 cpus.E The rational behind Wildfire and galaxy is that you can overcome this E SMP bottleneck by having multiple instances "clustered" each of which C has upto 8 cpus. Hence, for many application workloads, you achieve 8 near linear scaling to much much larger numbers of cpus.      
 David Webb VMS and Unix team leader CCSS Middlesex University    & Sent via Deja.com http://www.deja.com/ Before you buy.    ------------------------------  # Date: Mon, 10 Apr 2000 12:37:47 GMT 2 From: kilgallen@eisner.decus.org (Larry Kilgallen); Subject: Re: Galaxy (or perhaps Wildfire) and shared memory ' Message-ID: <2000Apr10.083747.1@eisner>   b In article <38F13224.4C86088B@vl.videotron.ca>, JF Mezei <jfmezei.spamnot@vl.videotron.ca> writes:  I > If you have 2 instances of VMS in a galaxy, does that provide much more H > robustness compared to an instance of VMS with an instance of Linux of
 > True64 ?  H That depends on the relative correctness of the three operating systems,H and many in this group would say "of course VMS has fewer defects!". You4 can get a different answer in a different newsgroup.  C > The other storey I had heard was that Galaxy was a way to run VMS B > in a machine which has more processors than VMS can support as a@ > single OS instance. So you run multiple instances of VMS, eachA > with the maximum number of SMP CPUs it can handle. Is there any  > truth to this ?   ? I don't thing the problem of the number VMS "can support" is an ? issue so much as the number with which VMS can "scale well". As ; the rumors (or is it now marketing statements) say Wildfire < will support 32 processors it seems the hardware has not yetB surpassed the ability of VMS to support non-Galaxy configurations.  : When that changes, your statement hold, but it is not very? interesting if everybody will be using Galaxy before that point  to address the scaling issues.   ------------------------------  % Date: Mon, 10 Apr 2000 09:39:25 -0400 * From: Clair Grant <grant@evms.zko.dec.com>; Subject: Re: Galaxy (or perhaps Wildfire) and shared memory 0 Message-ID: <38F1A14D.31CD352A@evms.zko.dec.com>   JF Mezei wrote:  >  > Larry Kilgallen wrote:? > > So the very fact that DEQ leaves this as a decision for the < > > customer indicates sharing is not the same as isolation. > K > I remember being told that in a Galaxy, it is *expected* that the OS that N > coexist with each other are required to behave properly and that if you haveT > one instance of an OS that doesn't really behave, then you'rte asking for trouble. >   F That is true. Sharing is definitely not the same as isolation. Sharing. implies cooperation. You have to get it right.    > So the question should be:I > If you have 2 instances of VMS in a galaxy, does that provide much more Q > robustness compared to an instance of VMS with an instance of Linux of True64 ?  >   A Not if the code were to be written properly. Note that no one has D attempted to do this. (OK, you've seen a few DECUS demos but we wereD just fooling around, trying to verify some concepts. We never reallyH tackled the entire LINUX OS; we just did enough to make CPU reassignment" and a little shared memory work.)    P > The other storey I had heard was that Galaxy was a way to run VMS in a machineP > which has more processors than VMS can support as a single OS instance. So youP > run multiple instances of VMS, each with the maximum number of SMP CPUs it can& > handle. Is there any truth to this ?  H As of now we don't ship any hardware with more CPUs than VMS can supportG as a single instance so your particular example has not come about yet.aG In the future we will have to make changes to support some platforms ifyG we want to run them as a single instance (which may not be practical or)A a good use of the platform's resources anyway). TBD. Plus, it alloF depends on where your environment hits its SMP scaling point as to howE you configure your system. We are trying to provide enough options toS" satisfy as many people are we can.    D I have not been reading the news group for a few weeks (sort of busyD right now) but someone suggested that I ought to take a look at thisA topic. There seems to be a bit of confusion about WildFire's hard > partitioning capability and Galaxy. I'll make a brief attempt.  C There is no sharing between hard partitions on WildFire - no sharedME memory, no common I/O, no nothing, zip, nada. Within a hard partition H you may choose to run a multiple instance Galaxy (soft partitions) where. you have shared memory, CPU reassignment, etc.     -- I Clair Granto VMS Exec Group Project Leaderd COMPAQ Computer Corporationo 110 Spit Brook Rd. Nashua, NH  03062    ------------------------------  % Date: Mon, 10 Apr 2000 10:55:24 -0400c5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> ; Subject: Re: Galaxy (or perhaps Wildfire) and shared memorya+ Message-ID: <8csptc$gup$1@lead.zk3.dec.com>n  : Bill Todd wrote in message <8cljtd$c28$1@pyrite.mv.net>... >r >fG >The difference between normal SMP operation and Galaxy (or perhaps Sun L >'domain' or S/390 'hardware-partitioned') operation is that when a hardware< >fault occurs in a 'normal' SMP, the entire system stops and3 >cache/main-memory coherence ceases to be an issue.t >iL >In a Galaxy, if a fault disables a cache that is holding dirty data that noK >other processor has yet expressed an interest in, then that data is lost -aB >but subsequent use of the related but unupdated memory can occur. >8    : The hardware platforms currently using Galaxy, and the oneI just-to-be-released, provide only the same level of hardware integrity asoF anSMP system.  If a failure occurs on a processor, which may have beenL accessing shared memory *and* that processor failure occurs in such a way toI make its cache inaccessable - then the entire Galaxy must reboot, mappingi" that CPU out of the available set.  G Note that this is a somewhat unusual failure.  It is a catestrophic CPU/ failure during runtime.r  F It is entirely possible (but I don't think it is in the code) that theJ failure could cause a signal to all processes accessing shared memory, andL force them to abandon the shared memory as being suspect (possibly corrupt).L But that may well require all shared memory to be abandoned, and "scrubbed".  L Note that most failures will result in bugchecks, as opposed to catestrophicH CPU failures, in which case well coded applications can take appropriate> steps to recover from the failure (as noted in other replies).   ------------------------------  % Date: Mon, 10 Apr 2000 11:09:51 -0400e5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>o; Subject: Re: Galaxy (or perhaps Wildfire) and shared memoryo+ Message-ID: <8csqog$nmv$1@lead.zk3.dec.com>c  H MBs are a integral part of any SMP operation... including talking to DMAJ "devices".  The Alpha Architecture Reference Manual goes into great detailL about this.  To correctly coordinate access between two processors, MBs muchH be used by both the writer and reader of the memory to ensure coherence.  L Its a bit ugly, but the good news is that from a user programmer perspectiveG it's mostly transparent, as it is buried in the access methods used forw= locks, queues, and other "interlocked" communication methods.   K Galaxy doesn't introduce anything new here.  Access to a shared memory areaeK must be treated just like it was any ordinary SMP memory section.  Failures K from a program point of view can be handled in the same way (i.e. a failureaG of a galaxy "instance" is no different than the failure of a process or  thread on a SMP system.y  L And yes, there are some catestrophic failures in the hardware that can cause the entire Galaxy to crash.e        : Bill Todd wrote in message <8clj0c$9o1$1@pyrite.mv.net>... >n> >Hoff Hoffman <hoffman@xdelta.zko.dec.nospam> wrote in message1 >news:8ckv5r$8d3$1@mailint03.im.hou.compaq.com...r >>7 >> In article <8cjdd4$18m$1@pyrite.mv.net>, "Bill Todd"  <billtodd@foo.mv.com>  >writes: >>! >>   re: Galaxy and Shared Memory  >>J >>   Alpha provides specific instruction sequence requirements to maintainJ >>   and to update the CPU caches appropriately -- the memory barriers...>G >>   Appropriate use of these barriers is required on any OpenVMS AlphaC >system. >>I >>   OpenVMS Galaxy partitions and instances (or even hard partitions) do  not.J >>   specifically enter into the discssion of processor cache coherency -- >theH >>   box has to work right as an SMP box.  Of interest here, however, isK >>   coherency of cluster resources, and the same coherency mechanisms used L >>   elsewhere within an OpenVMS Cluster are used across any cluster members( >>   located within an OpenVMS Galaxy... >>G >>   The OpenVMS Ask The Wizard area has a general discussion of memory J >>   barriers and hardware memory synchronization in topic 2681, with moreH >>   details in the OpenVMS Programming Concepts Manual and in the Alpha% >>   Architecture Reference Manual...B > C >Thanks for the pointer(s):  6 or so years ago I actually thought ItI >understood memory barrier use, but I've never had reason to use an Alphas and-G >I'd forgotten whatever understanding I may have had - or even that MBsr mighte% >be relevant to the question at hand.c >kL >Ah, yes.  Use w.r.t. instruction reordering - likely unrelated.  Use w.r.t.L >flushing write-back-cached data - quite possibly relevant, depending on theL >details of the memory architecture.  The critical fact that I had forgottenK >was that Alpha appears to eliminate a good deal of SMP hardware complexitytK >by pushing responsibility for explicit management of shared memory updateshK >back onto the programer - by no means necessarily a bad trade-off, but oneoJ >which I think is less common in other architectures (though may be fairlyJ >standard for RISC architectures - the little hardware knowledge I have is) >for the most part horribly out of date).o >hJ >In particular, if an MB write-flush causes data to be written all the wayA >back to main memory, thus completely eliminating inter-partitione connections H >at any cache level (of the kind that exist in some SMP architectures toF >allow inter-processor data-sharing without the latency of main-memoryK >accesses), then my concern that inter-cache connections between processorsWF >executing in different partitions might compromise fault-isolation isL >groundless:  the only inter-partition connection will be through the sharedD >memory itself, which will presumably pass along any necessary cacheK >invalidations as write activity occurs (though my understanding of exactlydG >how this occurs is regrettably fuzzy), and the hardware connecting thepH >caches to main memory is presumably already sufficiently robust to keepC >local faults from affecting access by others to any shared memory.s >aL >Furthermore, my fear that a cache might hold dirty data for lengthy periodsE >before writing it back to shared memory, thus drastically increasingeL >exposure to any failure which could keep that write-back from occurring, isG >groundless if in fact MB flushes will *always* be explicitly issued ont *all*aK >processors in a partition that have participated in any interlocked updateyJ >activity to an inter-partition-shared piece of memory prior to release ofI >whatever synchronization mechanism is used - which it sounds as if is noaB >stronger a requirement than exists today for memory-sharing in anI >unpartitioned Alpha SMP system.  It does, however, sound as if processoroL >cache content needs to be flushed or invalidated on instance failure (whichJ >may already be standard procedure) to ensure that stale updates won't get, >applied to already-recovered shared memory. >rK >So considerate of Fred to have asked such a comprehensive question, and of J >course of the Wizard to have provided such an informative answer.  Again,H >thanks for the pointer, and apologies for raising what I did originallyH >state might be a dumb question (though it did raise some interesting if' >unrelated explorations along the way).i >e >- bill  >  >>L >>   In addition to the discussion of cache coherency, OpenVMS Galaxy brings8 >>   in (at least) two additional design considerations: >>F >>     o the potential that shared memory will survive a system crash.D >>       This isn't specifically a CPU cache issue, this is an issue3 >>       around recovering from interrupted access.r >>F >>       Options include use of the OpenVMS Galaxy event services, theH >>       lock manager, as well as timers and aging and keepalive schemesF >>       in the shared memory, as well as "careful" updates -- updatesF >>       performed in a sequence that closes or reduces the likelyhoodH >>       of problems from crashes.  (RMS uses "careful" updates, and for >>       just this reason...)  >>H >>     o Differing memory access speeds.  This is obviously not an issueI >>       with currently available systems, but will be a consideration onrG >>       certain AlphaServer GS series platforms -- the current OpenVMSoE >>       spinlock design effectively has an assumption of access timebE >>       uniformity, and has been modified to provide better fairnessyI >>       to the protected resource when non-uniform memory configurationsw >>       are in use. >> >>- >>  --------------------------- pure personal-$ >opinion ---------------------------2 >>    Hoff (Stephen) Hoffman   OpenVMS Engineering >hoffman#xdelta.zko.dec.como >> >b >S   ------------------------------  % Date: Mon, 10 Apr 2000 12:07:43 -0400a' From: "Bill Todd" <billtodd@foo.mv.com>r; Subject: Re: Galaxy (or perhaps Wildfire) and shared memoryp( Message-ID: <8csu4v$1ea$1@pyrite.mv.net>  5 Clair Grant <grant@evms.zko.dec.com> wrote in message-* news:38F1A14D.31CD352A@evms.zko.dec.com...   ...t  E > There is no sharing between hard partitions on WildFire - no sharedpG > memory, no common I/O, no nothing, zip, nada. Within a hard partitioneJ > you may choose to run a multiple instance Galaxy (soft partitions) where0 > you have shared memory, CPU reassignment, etc.  B I understand that, and should have made it much clearer that I wasK addressing potential future capabilities (possibly in Marvel - I don't knowML what the specific road map may be) for the use of shared memory between hardJ partitions that had been alluded to (I'm pretty sure) by (perhaps) Fred in discussions months ago.p  J I do find it interesting if your comment above means that CPU reassignmentI between hard partitions is not something at least under consideration for 9 the future, but that's not relevant to the current topic.    Thanks,,   - bill   >: >5 > --
 > Clair Grantm > VMS Exec Group Project Leader  > COMPAQ Computer Corporationm > 110 Spit Brook Rd. > Nashua, NH  03062c   ------------------------------  % Date: Mon, 10 Apr 2000 12:31:50 -0400e' From: "Bill Todd" <billtodd@foo.mv.com>r; Subject: Re: Galaxy (or perhaps Wildfire) and shared memory ( Message-ID: <8csvi5$2ot$1@pyrite.mv.net>  > Fred Kleinsorge <kleinsorge@star.zko.dec.com> wrote in message% news:8csptc$gup$1@lead.zk3.dec.com...* >a< > Bill Todd wrote in message <8cljtd$c28$1@pyrite.mv.net>... > >o > > I > >The difference between normal SMP operation and Galaxy (or perhaps Sun E > >'domain' or S/390 'hardware-partitioned') operation is that when a  hardware> > >fault occurs in a 'normal' SMP, the entire system stops and5 > >cache/main-memory coherence ceases to be an issue.t > >oK > >In a Galaxy, if a fault disables a cache that is holding dirty data thatn noF > >other processor has yet expressed an interest in, then that data is lost -D > >but subsequent use of the related but unupdated memory can occur. > >t >y >h< > The hardware platforms currently using Galaxy, and the oneK > just-to-be-released, provide only the same level of hardware integrity asoH > anSMP system.  If a failure occurs on a processor, which may have beenK > accessing shared memory *and* that processor failure occurs in such a wayh toK > make its cache inaccessable - then the entire Galaxy must reboot, mappingo$ > that CPU out of the available set.  K Thank you!!!  THAT answers the question (both the newer one and my original L one, for that matter).  I assume that a cache failure as well as a processor  failure could cause this result.  L It does leave me curious as to exactly how the system can know for sure thatH such a failed processor or cache was *not* holding dirty data for sharedI memory, unless it just conservatively assumes that any processor that hasc: physical access to the shared memory held such dirty data.   > I > Note that this is a somewhat unusual failure.  It is a catestrophic CPU  > failure during runtime.w  @ Yes - certainly rare enough that rebooting the entire box is notJ unreasonable (unless you've asserted that the box provides fault-isolationD equivalent to a conventional cluster, that is).  *Not* rebooting andI allowing undetected data corruption to occur as a result is what would be8
 unreasonable.c   >FH > It is entirely possible (but I don't think it is in the code) that theL > failure could cause a signal to all processes accessing shared memory, andD > force them to abandon the shared memory as being suspect (possibly	 corrupt).hB > But that may well require all shared memory to be abandoned, and "scrubbed".u >sA > Note that most failures will result in bugchecks, as opposed tow catestrophicJ > CPU failures, in which case well coded applications can take appropriate@ > steps to recover from the failure (as noted in other replies).  ? As long as hardware failure does not make dirty cached data forfG inter-partition-sharded memory inaccessible, then less drastic recovery G mechanisms can be applied (though they must take the possibility of newc4 failure modes into account, as described elsewhere).  L Thanks again, and apologies for the time it took to clarify exactly why this was a non-issue.   - bill   >c >o >i >e   ------------------------------  % Date: Mon, 10 Apr 2000 12:42:54 +0200a0 From: "Perisa Bujosevic" <perisha@yugolot.co.yu># Subject: GBLSECTIONS change probleme1 Message-ID: <200004101044.MAA08796@smtp.EUnet.yu>a  / When I try to start manually Pathworks it says:-  F The system does not have enough free global sections to use PATHWORKS.B Please see the Installation Guide on how to increase the number of> global sections by modifying the SYSGEN parameter GBLSECTIONS.  S When I change GBLSECTIONS variable with SYSGEN (set GBLSECTIONS 1000, WRITE ACTIVE)tT nothing happens since it is not dynamic value. Is there any way to change this value without restarting the system ?b   ------------------------------  # Date: Mon, 10 Apr 2000 11:18:15 GMTx% From: "Jim Tsetsos" <jim@tsetsos.net>a' Subject: Re: GBLSECTIONS change problemn: Message-ID: <XLiI4.550$UB.2535@news-server.bigpond.net.au>   Nope, you need to reboot!i   -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Jim Tsetsosl mailto:jim@tsetsos.net http://www.tsetsos.net/g ~~~~~~~~~~~~~~~~~~~~~~~~~~~~9 Perisa Bujosevic <perisha@yugolot.co.yu> wrote in messagea+ news:200004101044.MAA08796@smtp.EUnet.yu...s1 > When I try to start manually Pathworks it says:  >uH > The system does not have enough free global sections to use PATHWORKS.D > Please see the Installation Guide on how to increase the number of@ > global sections by modifying the SYSGEN parameter GBLSECTIONS. >WG > When I change GBLSECTIONS variable with SYSGEN (set GBLSECTIONS 1000,P
 WRITE ACTIVE) K > nothing happens since it is not dynamic value. Is there any way to changea
 this value! > without restarting the system ?h >    ------------------------------  % Date: Mon, 10 Apr 2000 10:40:49 -0400 5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>m3 Subject: Re: I can't run/detach in my own X sessiont+ Message-ID: <8csp20$g7u$1@lead.zk3.dec.com>a  I It needs to be able to see a DECW$DISPLAY logical, unless the applicationaI has explicitly wired the display in the open display statement.  When yougF run it detached, it no longer sees the logical you currently have set.      E Ildefonso Junquero wrote in message <38EE0B7D.325B45BC@sainsel.es>...y >) >B >Hello!i >lH >    I'm having an extrange problem... I can't run/detach an x client on >my own xserver. >sH >    I've a new CDE and a configured security manager with node * user * >and transport * >aI >    If I run : run decw$utils:xmag (for example), all works ok, but if i) >run : >e, >        run/detach  decw$utils:xmag/out=a.a >o= >    the client says in a.a : xmag: unable to open display ""  >a >    can anyone help me please?  >l >    Thanks! >e
 >    Ilde. >s >--hG >______________________________________________________________________ G >                  Ildefonso Junquero      e@mail : junquero@sainsel.esu$ >    ___    __o    Software Engineer0 >   ____  _'\<,_   Sainsel Sistemas Navales S.A.0 > ______ (*)/ (*)  C/ Manuel Velasco Pando, N. 7( >                  41007 Sevilla (Spain)+ >   Go faster!     Phone : +34-95-493 64 65 + >  Use the byke!   Fax   : +34-95-493 64 33oG >______________________________________________________________________z >d >c   ------------------------------  % Date: Mon, 10 Apr 2000 07:04:29 -0700-5 From: "cstranslations" <cstranslations@email.msn.com> ' Subject: Is sys$dclexh ever not called?.) Message-ID: <#mAYvWvo$GA.303@cpmsnbbsa04>P  I I'm looking at the description for $dclexh in the systems services manualnL and it says "declares an exit handling routine that receives control when anG image exits."  Implied in the description is normal image exit. It alsoeG specifically states the registered routine is called as a result of theaJ image calling sys$exit or someone bumping the process over the head with a sys$forcex.   K What I'm wondering is if the registered routine is called if someone uses a H "big stick" to hit it over the head - say sys$delprc. To a lesser extentJ what happens if there's an unhanded condition while the image is executingK (although in this second case I'm guessing that it's entirely possible that  "all bets are off").   Joeo   ------------------------------  % Date: Mon, 10 Apr 2000 10:02:29 -0700y5 From: "Larry D Bohan, Jr" <LBohan@dbc.spam_less..com>e+ Subject: Re: Is sys$dclexh ever not called?i2 Message-ID: <lAfyOC=CAqgD6CwHG1uyIz8Inxgz@4ax.com>  4 On Mon, 10 Apr 2000 07:04:29 -0700, "cstranslations"% <cstranslations@email.msn.com> wrote:   J >I'm looking at the description for $dclexh in the systems services manualM >and it says "declares an exit handling routine that receives control when an H >image exits."  Implied in the description is normal image exit. It alsoH >specifically states the registered routine is called as a result of theK >image calling sys$exit or someone bumping the process over the head with ac >sys$forcex. >lL >What I'm wondering is if the registered routine is called if someone uses aI >"big stick" to hit it over the head - say sys$delprc. To a lesser extentoK >what happens if there's an unhanded condition while the image is executingsL >(although in this second case I'm guessing that it's entirely possible that >"all bets are off").  >k >Joe >A  + the docs on $FORCEX cover this topic a bit;M  M http://www.openvms.digital.com:8000/72final/4527/4527pro_039.html#index_x_497    ....= The image executing in the target process follows normal exitmB procedures. For example, if any exit handlers have been specified,? they gain control before the actual exit occurs. Use the Deletea< Process ($DELPRC) service if you do not want a normal exit.  ...o  ; but the Black IDS covers this topic in much better detail  t (ch 26, pp737 ...)   ------------------------------  % Date: Mon, 10 Apr 2000 11:04:06 +0400y- From: Valentin Likoum <val_mail@ncc.volga.ru>a' Subject: Re: locations of C libs in VMSt, Message-ID: <38F17CE6.4C7460F4@ncc.volga.ru>   Obakesan wrote:e >  > HiYa > J > is there a normal place to put stdio and the like?  Here I found them inE > a logical that was something like $DKA0:[SYS0.SYSCOMMON.] in a fileo > called DECC$RTLDEF.TLB > - > I guess that the tlb files are librarys ...  > I > I am trying to compile an olde file here that inlcudes a library calledr > nfbdef ... this isn't in:q >  > #include <nfbdef>l >  > and I also tried >  > #include <sys/nfbdef>d > D > given that I think that it's in the TLB called SYS$LIB_C.TLB, what  > should I inlcude to get this?? >    In C programm:    #include <nfbdef>   Compile it as:1    cc my_source.c + sys$library:sys$lib_c.tlb/lib   5 or you could include sys$lib_c in library seach list  * (don't remember exact logical name for it)   -- a     Valentin Likoums     ncc_valent@my-deja.com  0     This page is best viewed with VT52 terminal.   ------------------------------  % Date: Mon, 10 Apr 2000 07:45:17 -0700P5 From: "Larry D Bohan, Jr" <LBohan@dbc.spam_less..com> ' Subject: Re: locations of C libs in VMS 2 Message-ID: <pebxODYkwSSagIoMu7t3iiJ4LZns@4ax.com>  ? On Mon, 10 Apr 2000 03:20:35 GMT, cjundieseastwd@powerup.com.auO (Obakesan) wrote:>  D >given that I think that it's in the TLB called SYS$LIB_C.TLB, what  >should I inlcude to get this??m  ( in the case of SYS$LIBRARY:SYS$LIB_C.TLB' you can reference it in the CC cmd-line   8  $ cc sys$library:sys$lib_c/library + [dir-spec]FOOBAR.C
  $! .. or ... /  $ cc FOOBAR.C + sys$library:sys$lib_c/library I  7 there's probably a better way to reference this library'- via some DECC$* logical, but right off-hand, d at the moment, I don't see it.   ------------------------------  % Date: Mon, 10 Apr 2000 09:31:08 -1300a- From: John Reagan <reagan@hiyall.zko.dec.com> + Subject: Re: Pascal Compile-TIme stack dumpf2 Message-ID: <38F19F5C.543A290F@hiyall.zko.dec.com>   Dave Weatherall wrote: > A >                 yes that would have been a dumb thing to do :-)t9 > However, I had originally responded to a query about :-  > H >       my pascal module has always compiled but now with latest version2 > of the compiler it crashes the compiler instead. >   F If you (or anybody else) has a piece of Pascal that makes the compilerE bugcheck.  Then please send me mail so I can fix it.  If I don't knowe? about it, I can't fix it.  Even if you have Fortran source thataE bugchecks the compiler, send me that also.  The compiler should printu0 lots of error messages, but NEVER EVER bugcheck.   --   John Reaganl Compaq Pascal Project Leader   ------------------------------  % Date: Mon, 10 Apr 2000 11:36:39 -0400h0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>( Subject: Request for Wildfire/Galaxy URL/ Message-ID: <38F1F503.D1B7A89C@vl.videotron.ca>s  I I have a firend who works in a big IBM shop, and while biased towards IBMlT mainframes, does have a certain level of open-mindedness to look at other platforms.  L Compaq will be having a Wildfire announcement in Montreal (via satellite) inK May, and I'd like to get her curious enough to attend that event so she canhM hopefully be very impressed and come back to her IBM shop and spread the word E that IBM mainframes aren't the only things that can handle big loads.l  G In doing so, i'd like to point her to a URL that would give a quick butrL technical description that would enable her to see how Wildfire will compareJ to IBM mainframes, as well as how Galaxy will allow more flexible resource$ management than IBM's sysplex stuff.  9 Is there such a magic URL hidden away in Digital's site ?    ------------------------------  % Date: Mon, 10 Apr 2000 10:23:00 -0700t7 From: "Arthur E. Ragosta" <ragosta@merlin.arc.nasa.gov>t, Subject: Re: Request for Wildfire/Galaxy URL3 Message-ID: <38F20DF4.EB9B9780@merlin.arc.nasa.gov>n   JF Mezei wrote:    > ... I > In doing so, i'd like to point her to a URL that would give a quick butiN > technical description that would enable her to see how Wildfire will compareL > to IBM mainframes, as well as how Galaxy will allow more flexible resource& > management than IBM's sysplex stuff. >s; > Is there such a magic URL hidden away in Digital's site ?l  > http://www.digital.com/alphaserver/announce/oct99_preview.html   ------------------------------  # Date: Mon, 10 Apr 2000 14:16:36 GMTt- From: "Richard D. Piccard" <piccard@ohio.edu>r" Subject: Re: So who will buy VMS ?( Message-ID: <38F1E241.FDE92FA5@ohio.edu>  f I am reminded of the (I think late-1940s) short story by Isaac Asimov, "A Feeling of Power," about theV archaelogist of computers who re-discovers how to do arithmetic with paper and pencil.  #                                 RDP     & "Brian Schenkenberger, VAXman-" wrote:  % > Will future historians look back at K > the beginning of the 21st century and refer to it as the "ingnoramus age"   > and not the "information age"? >a > --P > VAXman- OpenVMS APE certification number: AAA-0001           VAXman@TMESIS.COM   --B ==================================================================B Dick Piccard                           Academic Technology ManagerB piccard@ohio.edu                                 Computer ServicesB http://oak.cats.ohiou.edu/~piccard/                Ohio University   ------------------------------  # Date: Mon, 10 Apr 2000 16:55:58 GMT9" From: Art Rice <arice@ue.itug.org>" Subject: Re: So who will buy VMS ?8 Message-ID: <jf14fs43tufj3iqprnsmrulseblettldo9@4ax.com>  7 On Tue, 04 Apr 2000 21:14:26 -0700, "Larry D Bohan, Jr"1" <LBohan@dbc.spam_less..com> wrote:  A >On Wed, 05 Apr 2000 01:12:17 GMT, StevenU@POBoxes.com (Steven P.h >Underwood) wrote: >sB >>Well, I am in my low thirties, been using VMS continuously sinceH >>freshman year in college and just took delivery (today, in fact) of an) >>ES40 that will be running OpenVMS v7.2.l >e= >I'm in my mid-thirties,  and I've been using VMS ever since  % >freshman year in college as well;     >i8 >Checking up on the alma mater I find they've gone from 7 >1 vax/1 pdp11/70 to 3 vaxes, and to 1 vax && 2 Alphas i9 >currently  (small school maybe ~1200 students at most), e0 >and a real mix of many other  non-DEC systems,  >than they had back in the day.y >e; >and here, at my current job we took delivery of 3 500au's  5 >&& 4 DS20's this year, and we're due to for perhaps i; >5-8 more DS20's, in a few weeks, and probably as many more ; >workstations.    And we're buying rather more VMS systems -A >per year, now than we used to, 10-15 years ago, all of them new.r/ >We used to buy used, but not so much nowadays.y >c: >All that said,  where do Tandem sites get their people?   OJT. >.5 >I knew a handful of Tandem types a few years ago,   i; >and their bgnd was all unix, w/ a small smattering of VMS.a >h> >You certainly can't cut your teeth on Tandem at most schools.  A Actually, the bulk of the Tandem people learned their traded backa@ in the 70's and 80's when the financial world and telcos started@ purchasing the machines.  Especially when ATMs started to becomeC prevelent. Next came the 911 and healthcare industies.  It is theseoC "old-timers" that we all look to because there is so much about theiA inner workings of the kernel that is NOT in the manuals nowadays.t  E I have never heard of any educational institution that runs a Tandem,nD although I'm sure there may be a few around (working on some obscure government grant/project.)       --  
 Art Rice   *#e# Special Data Processing Corporation)& --------------------------------------' All opinions expressed are mine and do e% not reflect the views of my employer.p   ------------------------------  # Date: Mon, 10 Apr 2000 17:46:19 GMTs" From: Art Rice <arice@ue.itug.org>/ Subject: Re: Sun considers VMS a "mainframe" OSf8 Message-ID: <po44fsgcuetba81og4lrc7l4kpi8ghqal5@4ax.com>  3 On Fri, 7 Apr 2000 21:12:35 GMT, "Terry C. Shannon"a <shannon@world.std.com> wrote:   >l> >"JF Mezei" <jfmezei.spamnot@vl.videotron.ca> wrote in message* >news:38ECD343.19D395A3@vl.videotron.ca...J >> Ok, someone please explain to me why VMS/ALPHA is not considered in the >sameu" >> class as IBM/390/MVS machines ? >># >> Isn't Alpha a faster processor ?n >In most cases, yes. >w5 >> Do IBM disk drives have far superior performance ?t >d1 >Perhaps... which is why CPQ buys disks from IBM.   A FYI the Tandem S7000 and S70000 machines also have been using IBMD drives for a while.    >sK >> Is is a question of the bus which is much faster on IBM than on Alphas ?f > 0 >Have never really figured out parallel sysplex. >i >sE >MS is at a performance disadvantage compared to the 1960s's IBM 360,u >> 1- how come ? > J >Ask His Gateness. Advertising might have a lot to do with it. Advertising. >foisted Bill Clinton on us, so why not an OS? >  >> 2- will wildfire fix that ? >: >I fervently hope so!y >i   -- y
 Art Rice   *# # Special Data Processing Corporation & --------------------------------------' All opinions expressed are mine and do e% not reflect the views of my employer.m   ------------------------------  % Date: Mon, 10 Apr 2000 16:02:11 +0100-B From: Andrew Harrison SUNUK Consultancy <andrew.nospam@uk.sun.com>$ Subject: Re: Sun question for Andrew* Message-ID: <38F1ECF3.68AD1992@uk.sun.com>   Scott Vieth wrote:  	 > Andrew:i >LJ > Is there a newsgroup similar to comp.os.vms where folks discuss Solaris? >O? > I found only alt.solaris.x86 (not exactly what I'm after) andc, > comp.unix.solaris (which has no messages). >g	 > Thanks,  >n > -Scott :^)   It is comp.unix.solariso  H The group is active and there are a lot of current posts (10th of April) for example.  5 You need to talk to whoever manages your news server.s  : There is also comp.sys.sun as well though this has a lower volume.r     Regardse Andrew Harrisonb Enterprise IT Architectt   ------------------------------  % Date: Mon, 10 Apr 2000 02:30:25 -07001? From: Mike Price <mike.priceNOmiSPAM@littlewoods.co.uk.invalid>d Subject: UCX port assignmentso9 Message-ID: <2590395b.029860dd@usw-ex0110-075.remarq.com>l  9 SOem time ago we suffered problems with IP programs using ; UCX V4.2 (written using system service calls rather than C),; whenever we had a serious network problem. We ended up with 7 a port being apparently free (from UCX SHO DEV) but anyv; program that attempted to listen on it got a duplicate name : error. The support centre stated that we should always use the UCX$C_REUSEADDR flag.T; This fixed the immediate problem but didn't explain why theb8 port number was used deep in UCX but not apparently used4 from UCX SHO DEV - this would apparently be fixed in< version 5 (i.e. the 'its fixed in the next release' answer).  5 Any comments on this would be welcome but it was juste background really.  8 My problem now is this - last week Oracle came along and9 decided that a good port number to use for itself was one:9 of the ports that another part of the application uses to7: listent for requests on. As we now use the UCX$C_REUSEADDR8 flag then our program started up quite happily. However,2 the connections and subsequent messages  seemed to: dissapear - they probably went to Oracle which also seemed" to be ignoring requests sometimes.9 So - the question is - does anyone know how to get UCX tot; reserve certain ports that are not part of the 'well known' 9 list, I have looked at portmapper and that doesn't appearr% to fit in with how we use the system.g; I have looked at the manuals but any direction to a chapterw5 and voume would be welcome if you think the answer isd there!!e   Thanks in advance     
 Mike Price         * Sent from AltaVista http://www.altavista.com Where you can also find related Web Pages, Images, Audios, Videos, News, and Shopping.  Smart is Beautifuls   ------------------------------  % Date: Mon, 10 Apr 2000 09:54:17 +01000! From: Andy Burns <andy@burns.net> ! Subject: Re: UCX port assignmentsj8 Message-ID: <me53fsshjdk20lkl4hj1vglo4deq5ek2k0@4ax.com>   Mike Price wrote:.  : >SOem time ago we suffered problems with IP programs using< >UCX V4.2 (written using system service calls rather than C)< >whenever we had a serious network problem. We ended up with8 >a port being apparently free (from UCX SHO DEV) but any< >program that attempted to listen on it got a duplicate name; >error. The support centre stated that we should always use  >the UCX$C_REUSEADDR flag.< >This fixed the immediate problem but didn't explain why the9 >port number was used deep in UCX but not apparently usedv5 >from UCX SHO DEV - this would apparently be fixed inh= >version 5 (i.e. the 'its fixed in the next release' answer).a  I if the port had been used recently it would be in a 2MSLWAIT(?) state andt5 would eventually time out and become available again    A this could happen if you had stopped and then restarted a program J listening on a well known port, it's always good practice to use REUSEADDRJ in this case, the use of a different ephemeral port number from the remoteB end will ensure that the localaddr/localport/remoteaddr/remoteport combination is unique ...e     -- c Andy At Burns Dot Nett   ------------------------------  % Date: Mon, 10 Apr 2000 11:29:29 -0600a From: momoney4u@belize.netE Subject: Unique Strategy Trading Secrets                         8667 ; Message-ID: <0000652273cd$000051aa$000021db@195.102.65.170>e  0 ================================================' "Anybody who tells you that The Love ofl/ Money is the Root of all Evil doesn't have any" 0 ================================================    ' Do you have the Yen to be a Millionare?u  ! 100% return in less than 90 days!e  > Unique Strategy Trading in the International Currency Markets!  ! Largest MarketPlace in the World!   < Get our Reports, Charts and Strategies on the U.S. Dollar vs Japanese yen and euro dollar.e   Example:  E A $10,000 Investment in the yen vs the dollar, "properly positioned",n1 on 08/18 could have returned $30,369 on 09/18/99.   ; For a "FREE NO OBLIGATION" information packet, please call:n   1-888-316-3860      A (Must be 21 years of age. Risk capital only. Past performance notx! indictive of future performance.)-          
      \|||/	     (. .)a ---oOoo--------------- (REMOVAL INSTRUCTIONS)H This is a one-time mailing, done by an independent marketing company. WeK respect your online privacy and apologize if you have received this messageg/ in error.  If you would like to be removed fromF4 our mailing list just follow the Instructions below.   mailto:getthescoop@10pins.comi  ? Please be aware that any disruption to our remove link preventsz    C those that want to be removed from our database from being removed.s                                0 ================================================' "Anybody who tells you that The Love ofe/ Money is the Root of all Evil doesn't have any"n0 ================================================   ------------------------------  # Date: Mon, 10 Apr 2000 11:34:59 GMTr) From: "Neil Rieck" <n.rieck@sympatico.ca>l& Subject: Re: VAX BASIC Specifications?: Message-ID: <D%iI4.14411$2D6.393495@news20.bellglobal.com>  + carbon <prodos8@yahoo.com> wrote in messagee, news:090420001850136485%prodos8@yahoo.com...H > I am interested in writing a VAX BASIC compatable BASIC system for theC > Macintosh.  I am wondering if there are any online user's guides,nH > manuals, specifications that relate to VAX BASIC.  I haven't found any! > on Digital's web site.  Thanks.6  L I'm a application developer who uses VAX-BASIC and its sister DEC-BASIC (forK Alpha). I'm not aware of any online documentation for these products, but IdK recently purchased a paper "doc kit" from Compaq that covers both products.eK (I think you can purchase a CD-ROM based "doc kit" but don't know if it cand< be used from Internet Explorer like the OpenVMS Doc CD-ROM).  L If, you're serious about writing a VAX-BASIC compiler for the MAC, you won'tE be able to do it without a "doc kit" and lots of time. I think you'lldK probably run into some problems writing the RMS support (VAX-BASIC can dealeK with sequential, relative and indexed files directly from the language). IfeC I were you, I'd purchase a "single concurrent user" license for the J VAX-BASIC product (or the DEC-BASIC) and get some experience writing BASICL apps on a VAX (or Alpha) before you attempt this. If this isn't a commercialF project, try to get a hobbyist license for the complier through DECUS.K (you'll still need to buy a "doc kit" unless someone is willing to send youa a surplus set)  
 Good luck.  
 Neil Rieck* Kitchener(New Berlin?)/Waterloo/Cambridge, Ontario, Canada.! http://www3.sympatico.ca/n.rieck/r6 http://www3.sympatico.ca/n.rieck/docs/alpha_diary.html   ------------------------------  # Date: Mon, 10 Apr 2000 13:39:00 GMTo! From: danco@pebble.org (Dan Cook)!& Subject: Re: VAX BASIC Specifications?4 Message-ID: <slrn8f3j3s.4i7.danco@pebble.pebble.org>  5 On Sun, 9 Apr 2000 23:48:27 -0400, Richard B. Gilbert  <DRAGON@compuserve.com> wrote:  L >        DEC BASIC replaced VAX BASIC.   (I *think* it's called DEC BASIC). C >  DEC BASIC is available for either the VAX or Alpha architecture.e  G Hmmm, I was under the impression that DEC BASIC, unlike DEC C, is AlphahF only.  Are you honestly running DEC BASIC on VAXen?  Whoa, if so, I'll have to try that...t   - Danf   ------------------------------  + Date: Mon, 10 Apr 2000 14:25:23 +0000 (   ) 3 From: Christopher Smith <chriss@Mufasa.pubserv.com>e' Subject: Re: VAXStation 3100 boot disks J Message-ID: <Pine.LNX.4.05.10004101419090.14625-100000@Mufasa.pubserv.com>  ! On Sun, 9 Apr 2000, Stefan wrote:t  ? > Are there any disks out there with which I can at least get asG > VAXStation 3100 to do SOMETHING ?? Because I have about 4 of them and E > I wanted to install NetBSD on them but I cant boot from disk (yet).n  * Well, the answer is yes... but, see below.  D > So any software that you know of is welcome, even if its just some > simple demo kinda thing.  G What exactly would you like to do with your VAXen, and how can you bootrI them?  Do you have a cd-rom that will work on these?  If not, do you havecJ a tk50, or DAT (DDS-1) tape drive?  If not, then do they have the built-in floppy?   J I could probably come up with VMS on any of those medium listed except theD floppy, and I'm sure there are several people who could do the same.  H There's also the option of netbooting them -- especially if, as you say,H you're interested in running netbsd.  There's a mopd for netbsd that youJ can run on another netbsd machine, and use to get the kernel loaded onto a2 VAX.  Of course, on a VAX, I certainly prefer VMS.  J I suppose what I'm saying is that there are a lot of options for you.  CanD you explain more fully what you'd like to do with these once they're running?   Regards,   Chris0    O ===============================================================================-@ "My two cents"			(http://rootworks.com/twocentsworth.cgi?128562)= Christopher Smith(chriss@pubserv.com)			Prgramer^W Programmert Prime Synergy of Champaign, IL.m% -------------------------------------NI "Where a calculator on the ENIAC is equipped with 18,000 vacuum tubes and H weighs 30 tons, computers in the future may have only 1,000 vacuum tubes; and weigh only 1.5 tons." -- Popular Mechanics, March 1949  O -------------------------------------------------------------------------------J   ------------------------------  % Date: Mon, 10 Apr 2000 00:04:26 -0700e1 From: "Randy Park" <rjpark@mindspring.com.nospam>t& Subject: Re: VMS indexed files to Unix2 Message-ID: <8crues$i4t$1@slb7.atl.mindspring.net>  G Most people would use ftp in binary mode to transfer such a file.  You,e< of course, would need a TCP/IP stack on your VMS machine andA software on your Unix machine that understands RMS Indexed files.u  ; Another solution would be to use the DCL CONVERT command on C VMS to convert the file from indexed to sequential before moving it C to the Unix machine (ftp again) where you write a program or use angF existing utility to process the RMS sequential file records, which mayG be fixed length records or several different formats of variable lengthnB records.  Then of course you may have binary data in your records,D which need to be processed/converted also.  This can be particularly@ problematic if you have VAX native floating point data or packed
 decimal data.   H If I were doing it, I would convert my VMS file to contain all text dataD in an RMS sequential file and then copy it (ftp in text mode) to the
 Unix machine.d  Y "Scott Risdal" <srisdal@ecodev.com> wrote in message news:38EE0F87.598DBBB5@ecodev.com...  > Hello, >0I > I have a need to transfer some VMS indexed files to Unix where I have aRG > program that will extract the data  to a delimited format which I canmE > then manipulate as needed.  We have done this in the past using theeH > Transl8 product, but for a variety of reasons that I won't go into, weD > cannot use it. This is a one LAST time thing so I am looking for a? > cheaper(free?) solution for moving these files.  Suggestions?f >r	 > Thanks,s >o > Scott  >    ------------------------------  # Date: Mon, 10 Apr 2000 07:27:08 GMT 4 From: LESLIE@209-16-45-102.insync.net (Jerry Leslie); Subject: Re: VMS Information Is Online - VMS Isn't Dead Yet@( Message-ID: <gnfI4.2799$eu.54598@insync>  ' bill davidsen (davidsen@tmr.com) wrote:  : F : My apologies to the VMS folks, the original post has many referencesH : useful to VMS folks, you might want to chase out the original post, so : I copied your group. :   E The original article contained information most people on comp.os.vms1 are already aware of.9   : ) : In article <8bma44$94p$1@joe.rice.edu>,:, : Jerry Leslie <leslie@clio.rice.edu> wrote:A : | Pardon the interruption in the Intel vs. Sun debate, for some- : | information on VMS.1 : A :   Does this have some hidden relevance to either Intel, Sun, orM : computer architecture? v  D The original post was intended to supply information for the lack ofD information, or mis-information about VMS, which has been mentioned > frequently in the long-running thread "Why are Suns so slow?".    L : I thought not, so I edited the followups, although I doubt any are needed. :   / I added comp.sys.intel just for this followup. o  & Thanks for trying to read my mind. ;-)   :v6 : 	[  good info but irrelevant to the posted groups  ] :   A The people in comp.os.vms are well aware of Compaq's spending its F marketing resources on low-margin Billie Boxes, while spending almost @ nothing to market VMS, in spite of the HUGE income VMS provides.  E Jerry Leslie  leslie@clio.rice.edu  (my opinions are strictly my own)a   ------------------------------  % Date: Mon, 10 Apr 2000 10:22:52 +0200r= From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com> 1 Subject: Re: Warning: circuitsurgeon is a ripoff! ) Message-ID: <38F18F5C.5C32B815@gtech.com>a   Beyonder wrote:mC > >Lots of name-calling. And very little substance. None of you twoeH > >can say with certainty that the drive was not send out OK but arrived4 > >non-OK due to something happend during transport. > D > Oh that's very easy to say actually. the drive came in the system,> > but on powerup the drive was defective. DOA. that's what DOA > means you know.t   From a previous post:e  C #LIAR! I did NOT screw up the drive, and you know it. You sent me a1	 defectivec #one.u  G Accusing someone of having sent a defective drive and claiming you havem9 received a defective drive is two very different things !   G > >I am not saying that was what happend. But I am saying that it could C > >be possible (based on what I know). So I would recommend you twor > >to stop the name-calling. > O > nope. see above. and I'll stop when he admits the truth. or I'll stop when it  > keeps-K > someone from being ripped off. or I'll stop when he refunds me. whatever.e  H Contact a lawyer and use the proper channels for dealing with a business, transaction where seller and buyer disagree.   And stop wasting our time.  H > And you are wrong. The news-group CAN (and HAS) do something about it.   Nope.q  O >                                                                        It candM > inform other people in the world about the scum like circuitsurgeon. It hasmS > already done this, as I said, others have told me the same thing. And if it keepsmT > even ONE more person from getting ripped off, then it has done all it needs to do.  E You have posted a lot of name-calling, but almost no facts. Facts are  specificC information about dates for ordering, dates for delivery, dates forC complaints, D prices, copies of various email, specifics about delivery conditions	 etc.etc..   F I doubt that many people in this forum gets convinced by name-calling. We are much more into hard facts.  A The above items would be facts. If you can get a court conviction  backing it, then it would be hard facts.6  G Posting name-calling instead of facts just gives us the impression thath the  facts may not be there.i  R > As for the legal remedy, I'm in Canada, he's in the US. two different countries,Q > it's very difficult, if not impossible. The legal entities in the other country T > are not usually interested unless it's something major. But you have just given me
 > an idea.  C If there is beef on the plate, then the courts will take the case !    Arne   ------------------------------  % Date: Mon, 10 Apr 2000 19:08:16 +0010a% From: paddy.o'brien@zzz.tg.nsw.gov.auS1 Subject: Re: Warning: circuitsurgeon is a ripoff!95 Message-ID: <01JO2Y58M5B6004CKW@tgmail.tg.nsw.gov.au>   I Even CJL at his worst in abuse had more VMS content than has this thread O' degenerated to.  (Is that grammatical?)c  O And abuse from someone who appears to post anonymously does not hold any merit s to me.   Regards, Paddy   ------------------------------  # Date: Mon, 10 Apr 2000 13:32:01 GMTe! From: Beyonder <beyonder@vrx.net>l1 Subject: Re: Warning: circuitsurgeon is a ripoff! ' Message-ID: <38F1D855.5F4D7A95@vrx.net>   & At 1022 AM 04/10/2000 +0200, you wroteH >Accusing someone of having sent a defective drive and claiming you have: >received a defective drive is two very different things !  V no it's not. if someone sent you a defective drive, then you would receive a defective drive.  I >Contact a lawyer and use the proper channels for dealing with a businesss- >transaction where seller and buyer disagree.4 >And stop wasting our time.i  V maybe its wasting your time because you don't care. but it hasn't wasted other peoples+ time. in fact, it's already done some good.   I >> And you are wrong. The news-group CAN (and HAS) do something about it..   >Nope.  - here you are wrong. it has already done this.C  F >You have posted a lot of name-calling, but almost no facts. Facts areM >specific information about dates for ordering, dates for delivery, dates forOQ >complaints, prices, copies of various email, specifics about delivery conditionst
 >etc.etc..  , well then you haven't been paying attention!5 All of this is posted in the thread on the newsgroup.iR yes, all of it - except the copies of email. there is no point in posting email he= ignored. There would be no "response". rather pointless, yes.t  G >I doubt that many people in this forum gets convinced by name-calling.l" >We are much more into hard facts.  3 wrong again. please pay attention. it is all there.aV and it has already done this, and in fact "many people" have already emailed me to say! so. the hard facts are all there.e  B >The above items would be facts. If you can get a court conviction) >backing it, then it would be hard facts.t  # this information is already posted.d  H >Posting name-calling instead of facts just gives us the impression that >the facts may not be there.  - this is already posted. please learn to read.n  D >If there is beef on the plate, then the courts will take the case !  V Hah! you've got to be kidding! You are very very naive if you think this is true. I'll! give you a really good example...   U A few months ago a lawyer friend of mine got hacked on the internet. over 600 megs of0W data were erased. He had logs showing the attacker, ip address, time, date, everything.   V So he went to the police (it's criminal, not civil) but they've just told him they areX "too busy", and "not interested". This happens every day, and they have better things toU do. Oh yes, and BTW the law says that "port scans" are not illegal at all. But that'sa just an aside.   B.   ------------------------------  # Date: Mon, 10 Apr 2000 13:34:18 GMT ! From: Beyonder <beyonder@vrx.net>u1 Subject: Re: Warning: circuitsurgeon is a ripoff!e' Message-ID: <38F1D8E0.74F3F92D@vrx.net>m  6 It's not anonymous at all. it's my real email address.: and I've posted my real name, even in full, several times.B Also note my sig stays the same, pretty much. and most of the time? sign my real name, even though my email address shows my alias. : no big deal. most people choose an alias most of the time. I'm not hiding anything.  4 the content is aimed at other potential "customers".   Dant  & paddy.o'brien@zzz.tg.nsw.gov.au wrote:  J > Even CJL at his worst in abuse had more VMS content than has this thread) > degenerated to.  (Is that grammatical?)o >hP > And abuse from someone who appears to post anonymously does not hold any merit > to me. >i > Regards, Paddy   ------------------------------  # Date: Mon, 10 Apr 2000 13:47:28 GMT1! From: Beyonder <beyonder@vrx.net>o1 Subject: Re: Warning: circuitsurgeon is a ripoff!g' Message-ID: <38F1DBF1.1BECEB2A@vrx.net>e   cstranslations wrote:0  N > I haven't read through the posts in this thread in any great detail nor am I > inclined to do so. >uE > A RZ26 can be had for $25 to $30 bucks (and no - I am not condoning=' > *anyone's* actions, real or implied).= >=J > *INDEPENDENTLY* of the above, if the board went bad six months later howN > much is someone suppose to be accountable six months afterwards on something > already, what, 8+ years old.  ) NO, the board went bad TWO *DAYS* later!! D not six to eight months as circuitsurgeon would like you to believe.2 and BTW there is a GOOD reason it took two days...  @ The hard drive was DOA! No hard drive, no way to test the board.C and BTW that's not two consequitive days. That's two days of use...8  & Day 1 (arrival): dead hard drive (DOA)N Returned hard drive, waited for replacement. cost of shipping $100 (both ways)L Several days later: replacement hard drive arrives, then discovered board is bad.I The next 7+ months: attempt to contact circuitsurgeon, all ignored. Whilep looking for other remedies.pJ 8 months later: final contact with circuitsurgeon: refusal to do anything.N circuitsurgeon "offers" to rip me off again by selling me an io board for $200J Take unit into compaq/dec on insistance that "nothing is wrong with it" byO circuitsurgeon. result: motherboard is confirmed defective and adapter cable isrA also confirmed defective. cost: $100 for testing including labor.=G cost of replacement cable/adapter: $22 - from compaq/dec service centersN circuitsurgeon again "offers" to rip me off by having me send back whatever isP "bad" and he'll charge me for the replacement, cost will be above $200, plus the $200 for the io board.O I contact a dealer in california who sells me a replacement board, with cpu and4 all for a total of $75.b  7 I figure $75 is a lot better than $400 from mr. ripoff.g  N and let's get something straight, I don't consider calling someone who rips meM off a "ripoff" to be name-calling. I have no idea where Arne gets these weird M ideas. I also don't think calling someone like that "scum" to be name-callingS either.v   Dan.   ------------------------------  # Date: Mon, 10 Apr 2000 14:15:26 GMT ( From: Terry Kennedy <terry@gate.tmk.com>1 Subject: Re: Warning: circuitsurgeon is a ripoff!t' Message-ID: <Fst0xq.GI1@spcuna.spc.edu>y  4 In comp.org.decus Beyonder <beyonder@vrx.net> wrote:+ > NO, the board went bad TWO *DAYS* later!!R  N > Several days later: replacement hard drive arrives, then discovered board is > bad.  J   Was bad, or went bad? I'm sorry, this story has too many holes in it for me to believe you.  J   I also don't find it credible that you waited 8 months to have the board7 evaluated by Compaq and then to purchase a replacement.S  '   Hopefully you've learned some things:   M   1) Deal with people you've heard are reputable (I can't say anything eithereK      way about circuitsurgeon, I've never heard of him/her/them/it before).g  J   2) Use a payment method where you can dispute the charges (like a credit      card).   K   3) If there is a problem, get it addressed early. At least in the US, youoI      will lose some of your rights to dispute/complain/etc. by waiting. 2 J      weeks is a good maximum time to wait before disputing, and you should'      have it wrapped up within a month..  G   4) Acting as your own importer/exporter is a pain. Avoid it if at all/H      possible by buying/selling within your own country, or at least useJ      the services of a good customers expediter, so you don't wind up pay-L      ing multiple duty/tax on stuff going back and forth for repair/replace-
      ment.  - 	Terry Kennedy             http://www.tmk.como5         terry@tmk.com             Jersey City, NJ USA    ------------------------------  % Date: Mon, 10 Apr 2000 16:57:26 +0200 = From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com> 1 Subject: Re: Warning: circuitsurgeon is a ripoff! ) Message-ID: <38F1EBD6.6BD6B45B@gtech.com>n   Beyonder wrote:oP > and let's get something straight, I don't consider calling someone who rips meO > off a "ripoff" to be name-calling. I have no idea where Arne gets these weirdbO > ideas. I also don't think calling someone like that "scum" to be name-calling1	 > either.9   It is name-calling.u  + Foul language, no proof and very few facts.t   That is name-calling.    Arne   ------------------------------  % Date: Mon, 10 Apr 2000 17:02:32 +0200c= From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com>o1 Subject: Re: Warning: circuitsurgeon is a ripoff!h) Message-ID: <38F1ED08.A028A3CD@gtech.com>a   Beyonder wrote:u8 > It's not anonymous at all. it's my real email address.< > and I've posted my real name, even in full, several times.D > Also note my sig stays the same, pretty much. and most of the timeA > sign my real name, even though my email address shows my alias.?< > no big deal. most people choose an alias most of the time. > I'm not hiding anything.   ????  B I just went through all your posts to comp.os.vms in this thread !   No sig in any of the posts.  One post signed with B. % All the rest of the posts signed Dan.   A It is difficult not to get the impression that you have a relaxedt attitude to "facts".   Arne   ------------------------------  % Date: Mon, 10 Apr 2000 16:55:19 +0200,= From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com> 1 Subject: Re: Warning: circuitsurgeon is a ripoff!o) Message-ID: <38F1EB57.3CB1C41B@gtech.com>w   Beyonder wrote:'( > At 1022 AM 04/10/2000 +0200, you wroteJ > >Accusing someone of having sent a defective drive and claiming you have< > >received a defective drive is two very different things ! > X > no it's not. if someone sent you a defective drive, then you would receive a defective > drive.  D Yes. But not the other way around. If you receive a defective drive,
 then it is/ not proved that someone send a defective drive.   H > >You have posted a lot of name-calling, but almost no facts. Facts areO > >specific information about dates for ordering, dates for delivery, dates for-S > >complaints, prices, copies of various email, specifics about delivery conditions  > >etc.etc.. > . > well then you haven't been paying attention!7 > All of this is posted in the thread on the newsgroup.xT > yes, all of it - except the copies of email. there is no point in posting email he? > ignored. There would be no "response". rather pointless, yes.   * Are you f.ex. referring to something like:  % #I think I paid about $600 US for it.R  7 #lets not forget customs which cost me about $200-$300.D  B That is not specific facts. It just gives the impression, that you5 actually do not know exactly how much you have paid !   I > >I doubt that many people in this forum gets convinced by name-calling.u$ > >We are much more into hard facts. > 5 > wrong again. please pay attention. it is all there.lX > and it has already done this, and in fact "many people" have already emailed me to say# > so. the hard facts are all there.n  H It must be nice to receive supporting email from "many people", when all thedE public posts are more or less skeptic about the relevance and/or yourc claims !  J > >Posting name-calling instead of facts just gives us the impression that > >the facts may not be there. > / > this is already posted. please learn to read.F  E Can you recommend any good schools to attend to learn to read ?   :-)e  F > >If there is beef on the plate, then the courts will take the case ! > X > Hah! you've got to be kidding! You are very very naive if you think this is true. I'll# > give you a really good example...c > W > A few months ago a lawyer friend of mine got hacked on the internet. over 600 megs of Y > data were erased. He had logs showing the attacker, ip address, time, date, everything.t > X > So he went to the police (it's criminal, not civil) but they've just told him they areZ > "too busy", and "not interested". This happens every day, and they have better things toW > do. Oh yes, and BTW the law says that "port scans" are not illegal at all. But that's  > just an aside.   Hmmm....  B Civil suits in court over a business deal and reporting a criminal hackinge- case to the police has very few similarities.o  F Based on the facts here, then the polices decision sound very strange, butt' there may be other facts unknown to us.    Arne   ------------------------------  % Date: Mon, 10 Apr 2000 12:53:00 -0400> From: Dan <dan@vrx.net>e1 Subject: Re: Warning: circuitsurgeon is a ripoff! & Message-ID: <38F206EB.BF5D9B8@vrx.net>   Arne Vajhj wrote:   > Beyonder wrote: R > > and let's get something straight, I don't consider calling someone who rips meQ > > off a "ripoff" to be name-calling. I have no idea where Arne gets these weirdoQ > > ideas. I also don't think calling someone like that "scum" to be name-callingi > > either.y >e > It is name-calling.   C calling someone a "ripoff" when they rip you off is name-calling???e  - > Foul language, no proof and very few facts.l   > That is name-calling.n  A foul language? what foul language? what no proof? what few facts?o$ what the heck are you talking about?  H what drugs are you on? can I get some? You're in some weird fantasy landP that is ignoring the posted facts, and information for your own convenience justB to support this guy that has ripped off many people (not just me).  B I'd like to know why you are so adamant about protecting this guy?! Are you two in business together?    Dan.   ------------------------------  & Date: 10 Apr 2000 13:05:59 -0400 (EDT) From: KAPLANSKYM@NCCCOT.AGR.CAX Subject: Warning: Circumcision is a ripoff!  (Was: Warning: circuitsurgeon is a ripoff!), Message-ID: <01JO2LFXUN9U000DWW@VMGW.AGR.CA>   Enough!n   Marvin Kaplansky   ------------------------------  % Date: Mon, 10 Apr 2000 10:33:47 -0400h5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>s. Subject: Re: Wierd Elsa Gloria Synergy Problem+ Message-ID: <8csokq$g1b$1@lead.zk3.dec.com>a  E Tom Leitner wrote in message <8cpaeh$mlh$1@fstgss02.tu-graz.ac.at>...d  L >I'm running mine at 1280x768, 75Hz and 16-bit (!!!) now which is sufficientJ >for all color intensive applications and where the wobble is almost gone." >This is under Tru64 Unix, though. >f    I I have full 16-bit support in the rewritten VMS server.  Doesn't buy muchh over 24-bit,H but it may be a bit faster... the P2 only deals with pixel sizes of 8 orH 32 - so 16 bit color still ends up taking the same amount of memory (theA "other" 16-bits can be used for a back buffer if we did a real 3Dg implementation).   ------------------------------  # Date: Mon, 10 Apr 2000 09:24:59 GMTo+ From: "Andy Proctor" <aproctor@hotmail.com>n0 Subject: Will COMPAQ Australia please contact me4 Message-ID: <955358858.792684@talita.eclipse.net.uk>   All,  L If this is not the appropriate newsgroup, please accept my apologies. I haveL been trying to contact COMPAQ in Australia for about a week now (I am in theL UK) I've left messages/Sent Faxes, even had COMPAQ UK trying to get them for me - but no response.m  D Will someone from COMPAQ Australia please contact me. I would like aG quotation for Hardware support on some PWS500au, and some PC's that our I customer (in Australia) has purchased from us in the UK. (We are a COMPAQuJ OEM) The Alpha's run VMS (The best!). I want to place support contracts onL behalf of our customers (it's the way we do things) but I cannot get someoneK in Aus to call or E Mail me back! I know that CPQ Aus may read this NG, and I certainly that CPQ staff read it. Also for CPQ staff, please post correcttI contact details on the web site (the phone numbers are not correct - as I ' was told by a security guy in CPQ Aus!)u  K I have not given my work E Mail, but an alternative (I get too many E mailsmH anyway and would not like to receive spam from MLM etc) is given. Once I6 have a contact, I will give all the necessary details!   Thanks in advance for the help   Andy Proctor( Global Simulation Systems www.gssl.co.uk! E Mail me at aproctor@hotmail.com    ------------------------------   End of INFO-VAX 2000.201 ************************a4aK  0             Z    JZ  g h            7 g h            7                 JZ ah                ߫4   uav`a `Y`aEaC  .   `W                                   saA       N1@       Rv h 7           ߫4   uav``p`W8aEaD  F   `Vawaa&\                a8a8                               V<                  < X                          ߫4   uav`a `Y aF,aE  $   `2X     DAYNOVB                                               Q ߫4        ``Y aFtaE       



ޭ











aɣaLt







4   _ai`E`F aGaF@  8   `YaJ aC@ (                 >                                                          











߫4   uav`a `Y`aGaF  .   `W                                   saA       N1@       < h            ߫4   uav``B`JaHDaG   F   `JT        aaB  /X            `    4  Rv h             Rv h                             4 aՄ                ߫4   aFnd`%`2aHaG  H   `2X    aH aHapx        a* !@   aHla!T     aE$a/     
p  
              `x̪  0m  .m        ߫4   OaZ`y `W
|aIaHX  8   `L!    aDF            Ra                                      	i  











aUaX











߫4   uav`a `Y`aJaI  .   `W                                   saA       N1@       Rv h 	           ߫4   vaW`B`JaJaI  F   `Ja    ana  /            `    ^  f  LK            
 f  LK            
                  ^ aD                ߫4   aFnd`$`)xaK aJ,     `Ja
<   aB4T   ˶x         ߫4   a`2<`8aKhaJ     `La~a`L



aa"=



߫4   uav``B`JaLaK4  F   `VaC    aD aLp  0             :    b  < h             < h                             b ah                ߫4   uav`a `Y`aLaK|  .   `W                                   saA       N1@       Rv h 7           ߫4   uav``B`JaMTaL0  F   `JTa{    a},aQ  /\            `x    4  Rv h            4 Rv h            4                 4 aT                ߫4   uav``B`JޜaMaL  :   `K     a}.                    @      N1@        N1@avK\_r _ _  }n h            8       `a,߫4   aFnd`%`2aNaMh  F   `J    aN4 aNgaT        a?    aNaF(     #aFa/@     	                    `v,  ` `     

߫4   uav`a `Y`aO(aN  .   `W                                   saA       N1@       Rv h 7           ߫4   aFnd`%`2aOaN  F   `J    aOl aOapx        a !@   a=paOP     x`aEa/      y  y            `v,  [)  .m     

߫4   uav``p`W8aPaO<  F   