0 INFO-VAX	Sat, 04 Feb 2006	Volume 2006 : Issue 70      Contents: Re: Alpha for XP Re: Alpha for XP) C  History (was: null terminated strings)  FaxSR 	 Re: FaxSR A Re: FOR070.DAT files appearing - (Crow doesn't taste so bad after & Re: Hobbyist kit installation problems Re: Memory allocation query  Re: Phaser 560 and DCPS problem  Re: Selling Alphaservers!   F ----------------------------------------------------------------------  % Date: Sat, 04 Feb 2006 16:48:44 +0100 ( From: Paul Sture <paul.sture@bluewin.ch> Subject: Re: Alpha for XP + Message-ID: <44k0mtF2hpniU1@individual.net>    Wilm Boerhout wrote:) > Bob Koehler mailde op 3-2-2006 15:43...  >  >>    Yes, a real bargain! >>. >>    Considering I paid $50 for a real Alpha. >  > K > True, for acquiring the Alpha. Now add your power bill, floorspace costs   > and calculate again... >   I Excuse me, but you are talking Hobbyist here. Your mention of floorspace  7 isn't relevant if you have a spare corner in your home.   8 IMHO it should not be more expensive than VMware. Punkt.   ------------------------------  $ Date: Sat, 4 Feb 2006 10:36:46 -0500% From: "Chris Moore" <no.one@no.where>  Subject: Re: Alpha for XP : Message-ID: <jS3Ff.19273$Sk1.384270@news20.bellglobal.com>  ? "Malcolm Dunnett" <nothome@spammers.are.scum> wrote in message  & news:npPJcXLlwjA2@malvm9.mala.bc.ca...< > In article <GXSEf.18301$Sk1.347890@news20.bellglobal.com>,+ >   "Chris Moore" <no.one@no.where> writes:  >>K >> Alphas are soon to follow the same path to oblivion.  Last Order Date is G >> fast approaching, and how do I keep Alpha 1000 apps running without  	 >> Appl'n  >> Support people, > H >   If you haven't got application support people then you can't supportD > the application on an emulator either. The emulator just makes theE > underlying hardware look like an alpha, everything above that ( vms D > and application support ) would be the same as on a real alpha and% > require staff with the same skills.   L No disagreement from me, but for an enterprise where the "legacy apps" have L been stable for many years, and thus most of the experienced support people M have left through attrition (and otherwise) nobody wants to rock the boat by  M suggesting any kind of migration that would require programming skill -- I'm  J not supporting this view, just reporting it, and I'm sure many places are L doing the same stupid things.  That leaves a sysadmin trying to keep things D going for a few more years on aging hardware that might not even be 2 supported by hardware maintenance vendors anymore.   > H >   If you have an Alphaserver 1000 that you want to replace I'd suggestE > looking at a DS10 or DS10L before considering moving it to an Alpha D > emulator on a PC. ( of course if there are benchmarks showing thatI > an emulator on a PC can outperform current Alpha CPUs then the emulator D > might be a better answer ). I was kind of underwhelmed with what IE > saw in the announcement here ( you need 1GB of memory on the PC and 0 > the "virtual alpha" can only use 96MB of it? ) >   F I'm not saying this product as it exists today is the answer, but I'm M pleased that someone is looking towards potential future markets - hopefully  8 by the time I need it, the product will be fully viable.  < >> or the ability to update to a new system unit?  Answer is2 >> this emulator -- maybe not today, but soon .... >> > > >  HP would like to think the answer is to move to Itanium :-) >   M They can think what they will, but the penny-pinching customers of the world  J will be looking for longevity at the lowest possible cost. (both hardware H and software support)  Besides, the Windows-centric types will FAR more C readily accept an emulator running on what they perceive to be the  K "universal solution", than another of those "legacy boxes" -- and at least  ; in my enterprise, they are the guys holding the cheque book   I >  You're right, if Itanium flops and HP stops selling Alphas on schedule E > and various other dire things come to pass then this emulator might ; > be the only option, but hopefully that won't happen SOON.  >   D In my 20+ years of VMS duty, "dire things" have become the everyday F watchword, thanks to Digital and Compaq mismanagement of the product.    ------------------------------  % Date: Sat, 04 Feb 2006 05:54:58 -0800 # From: "Tom Linden" <tom@kednos.com> 2 Subject: C  History (was: null terminated strings)( Message-ID: <ops4gblwiyzgicya@hyrrokkin>   (Courtesy Thor Olson)   $ History of the C family of languages  E 1972 - The precursor to C, the language B, is developed at Bell Labs. E The B language is fast, easy to maintain, and useful for all kinds of G development from systems to applications. The entire team that designed F the language is immediately fired for behavior unbefitting a telephoneH company employee, and the project is handed to Dennis Ritchie. He altersD the language to be incomprehensible, difficult to maintain, and onlyC useful for systems development. He also designs in a pointer system B guaranteed to give every program over 500 lines a pointer into the operating system.   F 1982 - It is discovered that 97% of all C routine calls are subject to< buffer overrun exploits. C programmers begin to realize thatH initializing a variable to whatever happens to be lying around in memoryA is not necessarily a good idea. However, since enforcing sensible G variable initialization would break 97% of all C programs in existence,  nothing is done about it.   G 1984 - The number of operating systems bad pointers can get to has been  dramatically increased.   F 1985 - A variant of C with object oriented capabilities, called C WithG Classes, is ready to go commercial. However, the name C With Classes is A considered too clear and easy for outsiders to understand, so the ! commercial version is called C++.   D 1986 - C becomes so popular that industry analysts recommend writingF business applications in it. They argue that applications written in CB will be portable to many different systems. Many of these industryE analysts are suspected of being under the influence of hallucinogens.   < 1988 - Industry analysts finally run out of LSD. After theirE hallucinations fade, they notice that business apps written in C take C five times longer to produce, and are still not portable. They stop F recommending that business apps be written in C, except for a minorityD that switch to crack cocaine and start recommending business apps beF written in C++ because "object orientation will result in code reuse".  D 1990 - By this time, all C compilers have turned into C++ compilers.B But, since most C++ programs do not use any of the object orientedD features of the language, this means in practical terms that bloatedE code structures with pointers into the operating system are now being * compiled with an object-oriented compiler.  F 1990 - After hiring some industry analysts that switched from crack toE sniffing glue, Sun decides to create a language called Oak to program H set-top television boxes. Since all their programmers have had stilted CF syntax imprinted into their DNA by this time, the new language borrowsF heavily from C and C++ syntax. However the set-top boxes don't have anG operating system for bad pointers to get to, so pointers are eliminated   from the language.   C 1994 - Someone at Sun finally realizes what a stupid idea it was to A develop a special language just for set-top television boxes. The H language is renamed Java and repositioned as an "Internet" language thatB is supposed to be portable to many platforms. This works well as aH marketing campaign, since less than 3% of people in the industry at thisC time realize what the Internet is, and since hallucinating industry H analysts continue to be suckers for the mythical idea of "portability to different platforms".   F 1995 - Sun offers free psychedelic mushrooms to industry analysts, whoB immediately start writing articles about how Java is the future of? programming because of its portability and integration with the 	 Internet.   E Mid 1996 - 17,468,972 articles appear about how Java is the future of 9 programming. The age of Java applets in web pages begins.   G Late 1996 - Programmers trying to produce actual web pages with applets G that really work commit mass suicide out of frustration and depression. G Industry analysts increase their dosage of hallucinogens to compensate.   B 1997 - Taking the advice of hallucinating industry analysts, CorelD decides to rewrite all their applications, including WordPerfect, inH Java. The end result is the first known word processor that is slower to use than a typewriter.  F 1998 - Realizing that the applet thing is fading fast, Sun repositionsD Java again, this time as a server language. They steal the design ofB Microsoft Transaction Server and convince everyone to pretend they created the design.   E 1999 - Java 2 Enterprise Edition is introduced to the rave reviews of C drunk and stoned industry analysts. 21,499,512 articles are written ? about it, but no one actually uses it because it's immature and 
 expensive.  D 2000 - J2EE finally works, sort of. Just about the time all the JavaH vendors are ready to start making money on it, Microsoft announces .NET,D which includes almost all the features of J2EE except the outrageousC cost. In fact, Microsoft decides to give .NET away free for Windows G users. Scott McNealy is so outraged he files another irrational lawsuit  against Microsoft.  @ .NET includes a new C-family language, C#, pronounced "C-pound",B continuing the tradition of languages in this family having stupid names.  ? 2001 - Microsoft's marketing department realizes that no one in E marketing has ever talked to a live Microsoft product developer. They C have lunch with one and discover that the pronunciation is actually  supposed to be "C sharp".   C 2002 - C# is introduced as part of the release version of Microsoft G .NET. C++ developers on the Microsoft platform rejoice over the concept F of "managed code", which means they finally receive the same automaticH memory management features that Visual Basic has had since 1991 and Java has had since 1995.   @ copyright (C) 1996-2006 by Billy S. Hollis, originally posted on! dotnetmasters.com 13 January 2006 + Please do not remove this copyright message    ------------------------------  % Date: Sat, 04 Feb 2006 06:59:58 -0500 ! From: Hal Kuff <kuff@comcast.net>  Subject: FaxSRB Message-ID: <kuff-0E3E12.06595804022006@ia.vsrv-sjc.supernews.net>  G Anyone out there still using fgaxSR on Alpha, what are you thinking in   terms of migration... pmdf-fax?    ------------------------------  $ Date: Sat, 4 Feb 2006 09:48:23 -05007 From: "Tom Simpson" <thomas.simpson1@fubar.comcast.net>  Subject: Re: FaxSR0 Message-ID: <2rGdnS_aLYdfI3neRVn-vw@comcast.com>  K Check out Networking Dynamics Compufax.  It looked like a good replacement  = candidate when I was in the market for a replacement product.   E Eventually we were able to replace the documents we were faxing with  K e-mails, so we ended up not replacing FAXSR at all.  You may not have that   option.    Regards, Tom     / "Hal Kuff" <kuff@comcast.net> wrote in message  < news:kuff-0E3E12.06595804022006@ia.vsrv-sjc.supernews.net...H > Anyone out there still using fgaxSR on Alpha, what are you thinking in" > terms of migration... pmdf-fax?    ------------------------------  % Date: Sat, 04 Feb 2006 15:45:39 +0100 ( From: Paul Sture <paul.sture@bluewin.ch>J Subject: Re: FOR070.DAT files appearing - (Crow doesn't taste so bad after+ Message-ID: <44jt0kF2hpfsU1@individual.net>    briggs@encompasserve.org wrote: X > In article <44bvpnF1ef83U1@individual.net>, Paul Sture <paul.sture@bluewin.ch> writes: >  >>Jeff Cameron wrote:  >> >><snip> >>M >>>I distinctly remember that this was the behavior back in version 5.x, that O >>>if you had rights (that is granted by file protections or ACLs) to writ in a O >>>directory the owner of the file would be that of the creator, however if you O >>>used a privilege where a right (protection/ACL) was refused, the owner would ; >>>be the owner of the parent directory. I stand corrected.  >>> O >>>Do you remember this behavior in the past? Or are my grey matter cells truly 
 >>>misfiring?  >>> 2 >>>Boy! I'm beginning to acquire a taste for crow. >>>   I With all due respect Jeff, in this thread you came across to me as a bit  I high handed, but at that time I didn't have a system of my own available   to prove it.  H The ability to admit you have been wrong when you _are_ wrong is No Bad  Thing in my eyes.    >>J >>I don't remember all the gory details, but at one customer they noticed G >>a change in behaviour in this area with V6.1 (they had skipped V6.0).  >>% >>IIRC the following change bit them:  >>D >>The V5.n behaviour allowed users to create a file under their own F >>ownership in a world writable directory, even if a previous version J >>existed and was owned by another user. You can see the obvious security 
 >>risk there.  >  > C > My recollection on this is not solid, but is it possible that you F > are referring to the change where the rights to create a new versionB > of an existing file were changed to require delete access to the< > previous version as well as write access to the directory?  E My recollection isn't solid either, but thats sounds correct, as the  D temporary workaround was to set the default protection for affected  users as wo:wred.   H Yuk, you may say, but it got "the show on the road" with the upgrade to F V6.1. The affected users were well locked down - captive accounts and , even the Setup keys on their VTs removed :-)  C > Ownership is not the key to the security risk in the scenario you C > have presented.  Access rights are.  You have effectively deleted C > the previous version without having delete access to the previous 
 > version.  D Exactly. But the point I was making was that no ACLs were necessary  before the upgrade.   D > The fix for this is not an ownership test but rather a file access. > rights test.  And that is what was deployed. >    Agreed.    ------------------------------  % Date: Sat, 04 Feb 2006 02:00:29 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> / Subject: Re: Hobbyist kit installation problems , Message-ID: <43E45106.F45A59D3@teksavvy.com>   "greg.chabala@gmail.com" wrote: E > Speaking of which, now that I plan to take off DECnet Plus, I'm not I > sure I want it installed at all. I don't have anything else that speaks  > DECnet, so why bother?    G DECNET is also usefull for intra-node communications. Some applications E use decnet to talk to themselves.  Also, DECNET is used by web server C scripting. And DECNET gives you a great way to have one application B running under one usename speak or start an application on another# username (with the use of proxies).   8 DECNET4 is really a no brainer to install and configure.  C You need to decide on a node area, node number , node name (6 chars + max). Run @SYS$MANAGER:NETCONFIG and voila.   G Caveat: you will wish to have the SYSGEN paremete SCSNODENAME match the B DECNET node name, and will be required to have SCSNODEID match the decnet area/number.   H ( for instance, area of 1 and node number 9 would result in 1 * 1024 + 9 = 1033 )  D DECNET 4 is one of those thing where once it is installed, you don't5 worry about it. and it is there whenever you need it.    ------------------------------  % Date: Sat, 04 Feb 2006 16:29:17 +0100 ( From: Paul Sture <paul.sture@bluewin.ch>$ Subject: Re: Memory allocation query+ Message-ID: <44jvieF26dm2U1@individual.net>    Ken Fairfield wrote: > Paul Sture wrote:  >  >> ade@nowhere.com wrote:  >> >>> Hi,  >>>  >>> Vax 4000-400, VMS 6.1  >>> H >>> This system has recently crashed citing         "NOPRCPGFL, Failure  >>> to assign process pagefile"  >>> J >>> The reason for the crash is apparently because PHD$L_PRCPGFLREFS is 0. >>> C >>> My query is about the process that this crash is occurring to.  H >>> Someone has stated that this process must have page a file quota of F >>> 900,000. The system has one page file and no swap file, the total H >>> size of which is about 140,000 blocks and the system has 128MB main H >>> memory. I was wondering what the assembled gurus would recommend in D >>> this instance. Is it indeed 'normal' to specify a quota well in + >>> excess of what the system could handle?  >>>  >>> Your thoughts please.  >>>  >>J >> Without knowing what the process does, 900,000 for Pgflquo sounds like 6 >> it's suitable for an Alpha version of that process. >  > @ > Yes, but he's not on an Alpha, he's on a VAX 4000-400 (perhaps; > that was your point since you say, "suitable for an Alpha  > *version* of that process").  ! Yes, that was *exactly* my point.    Unfortunately,  H HELP/MESSAGE NOPRCPGFL doesn't give me anything on either Alpha or VAX, # with the versions I have access to.      > = > Since this is a VAX, the SYSGEN parameter VIRTUALPAGECNT is = > relevant.  I don't recall what happens if PGFLQUO is larger > > than VIRTUALPAGECNT, but I do know that parameter limits the4 > amount of virtual memory any one process can have. > > > What's more worrisome, especially in the context of a systemB > crash, is that you may very well have problems if VIRTUALPAGECNT@ > is set *larger* than the sum of (pages of memory) + (blocks ofD > page file).  It needs to be less than that sum by some comfortable	 > margin.  >      Agreed.   ? > Note that the 128MB of memory plus 140000 blocks of page file C > is only a little over 400K pages, much less than the 900K PGFLQUO C > referenced above...  Again, I would expect running out of PGFLQUO @ > to crash a process, not the system, but we don't know anything@ > about this process (privileged? inner-mode?) and we don't know > what VIRTUALPAGECNT is.  > C There's a fundamental problem here, which is why I homed in on the  H possibility that the application developer was quoting figures suitable 
 for an Alpha.    ------------------------------   Date: 4 Feb 2006 08:45:49 -0800 , From: "rcyoung" <rcyoung@aliconsultants.com>( Subject: Re: Phaser 560 and DCPS problemC Message-ID: <1139071549.067680.245530@f14g2000cwb.googlegroups.com>   G Here are the logicals, and the settings I tried per your last response.  After "busy", it just sits. E The PHASER560_DEVCTL has the  Phaser 750 files extracted from the std  libarary, and then renamed (per B http://www.hoffmanlabs.org/vmsfaq/vmsfaq.txt  section 5.34 which I, found on the internet) as "unrecognized" ...  F Directory of TEXT library SYS$COMMON:[SYSLIB]PHASER560_DEVCTL.TLB;1 on 4-FEB-2006 11:40:36 D Creation date:   1-FEB-2006 15:06:40      Creator:  Librarian T09-20? Revision date:   1-FEB-2006 15:06:40      Library format:   3.0 > Number of modules:      3                 Max. key length:  39D Other entries:          0                 Preallocated index blocks:  11 B Recoverable deleted blocks:      0        Total index blocks used:   1 B Max. Number history records:      20      Library history records:   0    LPS$$UNRECOGNIZED_INITPSDEVICE LPS$$UNRECOGNIZED_SETINPUTTRAY LPS$$UNRECOGNIZED_SETOUTPUTTRAY     C Here is the print attempt with the DCPS_LIB set to only the default        (LNM$SYSTEM_TABLE)      "DCPS$PHASER_PID" = "00000137"   "DCPS$TEST" = "1"    "DCPS_LIB" = "DCPS$DEVCTL"   (LNM$SYSCLUSTER_TABLE)9 $ PRINT /PARAMETERS=DIAG=FILE=BOTH/queue=phaser login.com 4 Job LOGIN (queue PHASER, entry 24) started on PHASER $ sh queue/all phaser F Printer queue PHASER, busy, on EMAC73::"IP_RAWTCP/208.246.88.20:9100",) mounted form DCPS$DEFAULT (stock=DEFAULT)   4   Entry  Jobname         Username     Blocks  Status4   -----  -------         --------     ------  ------6      24  LOGIN           YOUNG             1  Starting  8 %%%%%%%%%%%  OPCOM   4-FEB-2006 11:36:19.37  %%%%%%%%%%%" Message from user SYSTEM on EMAC735 Queue PHASER: %DCPS-W-NOT_READY, Printer is not ready  $ sh queue/all phaser F Printer queue PHASER, busy, on EMAC73::"IP_RAWTCP/208.246.88.20:9100",) mounted form DCPS$DEFAULT (stock=DEFAULT)   4   Entry  Jobname         Username     Blocks  Status4   -----  -------         --------     ------  ------6      24  LOGIN           YOUNG             1  Starting   ------------------------------  % Date: Sat, 04 Feb 2006 05:57:20 -0600 ! From: John <norad869@comcast.net> " Subject: Re: Selling Alphaservers!* Message-ID: <43E496A0.5060006@comcast.net>  7 Well - the RAID set must be running degraded then!  LOL    Jiri Kulhan wrote:   > vanjkos@gmail.com wrote: > [...]  > @ >> -2 Seagate Cheetah 18.2GB Ultra SCSI 80 pin in raid-5 config; >  > G > Until now I thought that number of discs required for RAID 5 must be  I > an odd number, starting with three... is there some el cheapo version?   > :-)) >  > J. >    ------------------------------   End of INFO-VAX 2006.070 ************************