1 INFO-VAX	Mon, 10 Dec 2001	Volume 2001 : Issue 685       Contents:- Re: a few ?'s for the VAXperts re 6k machines P Building a business-quality desktop.	Re: Packard Foundation Tells Merger Urgers  Re: Compaq after merger-failure  Re: Compaq after merger-failure  Re: Compaq after merger-failure   Re: Compaq now a takeover target  Re: Compaq now a takeover target  Re: Compaq now a takeover target  Re: Compaq now a takeover target  Re: Compaq now a takeover target  Re: Compaq now a takeover target  Re: Compaq now a takeover target  Re: Compaq now a takeover target  Re: Compaq now a takeover target  Re: Compaq now a takeover target  Re: Compaq now a takeover target  Re: Compaq now a takeover target  Re: Compaq now a takeover target  Re: Compaq now a takeover target  Re: Compaq now a takeover target Re: disk shadowing0 Re: Encompass Board Election Countdown Continues0 Re: Encompass Board Election Countdown Continues0 Re: Encompass Board Election Countdown Continues Re: gnu tar for VMS 4 Re: How to implement Login Fails for Open VMS on VAX: Re: How to tell if foreign command (SET COMMAND) is known? Re: IBM VMS and Oracle Re: IBM VMS and Oracle Re: IBM VMS and Oracle Re: IBM VMS and Oracle! Lost Programs - FUser & SHOW_USER 9 Re: Packard Foundation Tells Merger Urgers to "Paq it In" 9 Re: Packard Foundation Tells Merger Urgers to "Paq it In" 9 Re: Packard Foundation Tells Merger Urgers to "Paq it In"  Re: Secure Telnet  Re: Socket limits  Re: The demise of compaq Re: The demise of compaq5 Re: Tru64.org Flash Poll on Merger "Pearl Harbor Day" ; Writing a device driver: virtual/physical address questions   F ----------------------------------------------------------------------  % Date: Sun, 09 Dec 2001 19:57:35 +0000 % From: "a.carlini" <arcarlini@iee.org> 6 Subject: Re: a few ?'s for the VAXperts re 6k machines' Message-ID: <3C13C22F.35524B66@iee.org>    Robert Schaefer wrote: > K > Hi there!  I've got a handful of questions about a 6320 machine I have in G > the basement.  My machine has processors from different sources, with N > different serial numbers and eeprom versions.  I know I can change them, andG > it boots just fine (aside form the error msg) as is, but is there any N > advantage to one eeprom version or another?  Should I just upgrade them bothI > to the latest version, or is it tied to a hardware revision?  Is anyone   ' The latest version is generally best. I * used to have a revision matrix that listed( required HW ECOs and console SW revs for' the XMI machines and their peripherals, ' but somewhere along the line I lost it.   6 > keeping an archive of the various eeprom versions?     No idea.   > Also, will differentK > class processors work in the same computer?  As in if I pop two '400 cpus L > into my 6320 will I get a bastard 632420 with possibly only two processors? > eligable to be the primary, or will I get a non-booting box?    * A non-booting box would be my first guess.  	 > Another M > sort-of related question-- is it possible to make a bit-identical image of, H > say, a TK70 tape; something like `dd' with unix?  I'd like to save theK > eeprom image I end up overwriting just to keep the bits from fading away.   ' TPC (TAPECOPY?) should do this for you. - It can make an image of a tape in a disk file + and can (IIRC) reconstruct a tape from that  image file.   N > One last question before I quit bugging you...  Will one hobbyist PAK boot a0 > cluster, or will I need a PAK for each member?   No idea.   Antonio    --     --------------- - Antonio Carlini             arcarlini@iee.org    ------------------------------   Date: 10 Dec 2001 00:44:29 GMT& From: peter@abbnm.com (Peter da Silva)Y Subject: Building a business-quality desktop.	Re: Packard Foundation Tells Merger Urgers  % Message-ID: <9v10hd$d4m@web.nmti.com>   , In article <3C13BDCF.9BE08AED@videotron.ca>,/ JF Mezei  <jfmezei.spamnot@videotron.ca> wrote: M > Corporation needs 250 desktops, and 25 servers. If Compaq drops  the access K > business because it doesn't make money,  a customer will more than likely N > order his 250 desktops from dell and while he is at it, order the 25 servers > from dell too.  K Why would a corporation need that many servers for only 250 desktops... oh, 0 you mean those rack-mounted PCs running Windows.  I That also explains why someone could buy them from Dell, since AFAIK Dell 8 doesn't actually have a server operating system to sell.  N > What compaq should have done is to lump all wintel products together (server > and desktops),  G Well, yes, that's bloody obvious. It's not like there's anything in the L rack-mounted units that's all that different from the mini-towers and stuff.  N > Ahhh, but what to do about the consumer vs business desktops ???? Accept theN > fact that they are one of the same and price all of them competitively. Dell5 > does, and if Compaq doesn't, then Compaq will lose.   K Either that, or actually *sell* business quality desktops. Something like a . beefier Multia or a compact form factor DL360.   The high end unit:  K On the front, two slots for those nice low profile removable drives under a M locking cover, over a low profile laptop-style CDROM and floppy, plus USB and  sound jacks and a PC-Card slot.   E On the back, ethernet, SCSI, video, mouse, keyboard, and serial port. F Duplicate USB and sound jacks. And a lock that lets you pull the whole< motherboard out of the slab the way you could on the Multia.  J Inside, at most one riser that can fit one PCI card, or one ISA card, room for one IDE drive.  J Not a bloody iPaq/Evo legacy-free netcomputer. A real PC that you can load& up with some beefiness if you need it.  K The low end unit skips the SCSI and SCSI drives, and the riser. As a result J it's about 1.5" high. Same motherboard otherwise. Maybe a couple of PCMCIA slots for PC-Card expansion.  > Build it reasonably solidly, and price it fairly. We'd buy it.   --  +  `-_-'   In hoc signo hack, Peter da Silva. E   'U`    "A well-rounded geek should be able to geek about anything." L                                                        -- nicolai@esperi.org          Disclaimer: WWFD?   ------------------------------  % Date: Sun, 09 Dec 2001 14:31:33 -0500 - From: JF Mezei <jfmezei.spamnot@videotron.ca> ( Subject: Re: Compaq after merger-failure, Message-ID: <3C13BC0C.C07036F4@videotron.ca>   John McLean wrote:E > Up to the end of September Compaq was losing about $1 billion every J > three weeks in the PC area.  Compaq could give the new PC group a fairlyI > generous $2 billion to get them started and then just walk away knowing E > that what they have "donated" would be lost in just 6 weeks anyway.   L Nop, won't happen. You see Winkler controls both the enterprise side (though2 the wintel server) and the access side (PCs etc).   J > I suspect that there will be a power-struggle of substantial proportions > within Compaq.  L Why should there be ? Marcello and his underlings are the only ones who knowJ about the true enterprise servers. They have had their legs cut from underN them when Alpha was murdered and will now depend on Winkler giving them accessJ to that IA64 thing if they want to survive. Marcello was in no position toP argue/battle with Winkler prior to June 25 and is in a much weaker position now.  K With the murder of Alpha and now Tru64, what Marcello has left is very weak F and it woudln't surprise me to see Winkler point fingers at Marcello'sJ business to exclaim "see I told you so, we're now seing the final downwardM spiral of the old legacy mainframes and we'll see wintel servers rise to meet  the challenge".   M The guy who is in control of the stuff which has the image of being the shape M of things to come will always have the benefit of the doubt and be given some L latitude as to when his predictions will come true. The guy in charge of theK older stuff that has the image of being dead will not be given such chances 0 and will be killed at the first sign of disease.  H Marcello was given his chance last year, and while he did achieve a turnM around from negative growth to positive growth, I guess that wasn't enough to L impress the folks at Compaq and VMS was brought back to the basement without any more marketing.    ------------------------------  # Date: Mon, 10 Dec 2001 03:48:36 GMT 4 From: "Terry C. Shannon" <terryshannon@mediaone.net>( Subject: Re: Compaq after merger-failure- Message-ID: <ogWQ7.22691$wL4.56876@rwcrnsc51>   > "John McLean" <mcleanj@swissonline.delete.ch> wrote in message/ news:3C13AEAF.5114ED5F@swissonline.delete.ch...  >  >  > "Terry C. Shannon" wrote:  > >  > ...  > K > > It's at least possible. From a technical standpoint it's still the best J > > UNIX. That said, CPQ and HWP may continue down the condolidation path. We'll 
 > > see... > E > Oh !  So that's another technically superior product that they keep I > hidden is it ?  I must be honest and say that I hadn't realised that it J > was significantly better.  When you work on VMS you really don't look at > other OSs that much...  :-)   I Most of the big analyst firms who track UNIX have very good things to say K about Tru64. Tru64 is one of the more recent UNIX implementations, hence it J has excellent scalability. The RAS features from VMS and NSK help, too. AsK does the best UNIX clustering in the business, and we all know what OS that  clustering stuff came from!    > C > > Oh, hardly! Assuming CPQ can overcome the wishes of the Houston  loyalists, aI > > de-emphasis of the consumer peecees (wireless and Evo and iPAQ should  remainG > > keepers) and a tighter focus on enterprise stuff seems more likely.  > G > This is the $64 million question, isn't it ?  Will the Houston people I > finally see that light now that several thousand watts are being used ?   I No disrespect intended to the Houston crowd, but I fervently hope they DO $ get the message and act accordingly.  D > If Compaq's own financial statements weren't enough, the fact thatJ > independent evaluators for Walter Hewlett said that high exposure to the= > PC market was a very bad thing should sink home.... I hope.    Agreed.    > I > For the rest of Compaq I hope that the PCs are cut lose with Winkler in J > charge of them.  He seems to be one of the main pro-PC people so if he's9 > going to make big statements let's give him his chance.  > E > Up to the end of September Compaq was losing about $1 billion every J > three weeks in the PC area.  Compaq could give the new PC group a fairlyI > generous $2 billion to get them started and then just walk away knowing E > that what they have "donated" would be lost in just 6 weeks anyway.  > G > The only downside to this would be the staff who might suffer bigtime $ > when the new PC company collapsed. >  > : > > Stay tuned, the next few months should be interesting! > J > I suspect that there will be a power-struggle of substantial proportions > within Compaq.  E Yep. Which will give me plenty of grist for Shannon Knows Mergers, er  Compaq! ;-}    ------------------------------  # Date: Mon, 10 Dec 2001 03:54:46 GMT 4 From: "Terry C. Shannon" <terryshannon@mediaone.net>( Subject: Re: Compaq after merger-failure- Message-ID: <amWQ7.22712$wL4.57005@rwcrnsc51>   : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:3C13BC0C.C07036F4@videotron.ca... > John McLean wrote:G > > Up to the end of September Compaq was losing about $1 billion every L > > three weeks in the PC area.  Compaq could give the new PC group a fairlyK > > generous $2 billion to get them started and then just walk away knowing G > > that what they have "donated" would be lost in just 6 weeks anyway.  > F > Nop, won't happen. You see Winkler controls both the enterprise side (though 3 > the wintel server) and the access side (PCs etc).   I Mary McDowell is the direct report for ISSG stuff, which in general isn't C doing all that poorly. The eight-ways are quite profitable, and the @ forthcoming QuickBlade stuff should fall into the same category.    J > Marcello was given his chance last year, and while he did achieve a turnL > around from negative growth to positive growth, I guess that wasn't enough toF > impress the folks at Compaq and VMS was brought back to the basement without  > any more marketing.   L IMHO you are being unfair to Rich Marcello and Mark Gorham. Rich now has the7 entire High Performance Systems Division; Mark has VMS.   I Rich and Mark are not the problem. If you believe there is a problem with H VMS marketing, by all means find out who bears direct responsibility for* marketing and take up the issue with them.   ------------------------------  % Date: Sun, 09 Dec 2001 14:23:58 -0500 - From: JF Mezei <jfmezei.spamnot@videotron.ca> ) Subject: Re: Compaq now a takeover target , Message-ID: <3C13BA46.734DD8EC@videotron.ca>   JoAnn DiFrancesco wrote:M > Yeah, that's the ticket! Become a leader selling Wintel desktops at a loss!  > What a concept! + > What a load of ill-considered codswallop! J > While I appreciate the sophomoric humour in this suggestion, I'd sure as, > heck like to see some facts to back it up.  M Have you ever listened to Winkler speak publicly ? Take the time to listen to N him during financial analysts teleconferences. They are generally available onE the web. Listen to them live before they edit those portions out. Use G Deja/ggoogle to look up this newsgroup for comments about previous such I conferences, you will probably find Alan Grieg quoting quite a bit of the  stuff said by Winkler.  J > Oh, purposefully hiding? I would assume that anyone who could PROVE thisJ > allegation could have a hell of a lot of fun with the SEC, et al. In the? > absence of proof, it's nothing but Blovation Without a Cause!   J By combining the true enterprise stuff with the "toy" enterprise stuff andH producing just a single number, you are effectively hiding what the trueL enterprise stuff generates because the total number is tainted by the wintelE server numbers which have totally different margins and profit/costs.   E IBM is proud to brag about MVS/mainframes generating the profits that B subsidize its PC business. But Compaq works hard not to admit thatI VMS/Tandem/Tru64 are saving the day and prefers instead to brag about how H Wintel will rule the enterprise and will upscale to displace mainframes.   ------------------------------  $ Date: Sun, 9 Dec 2001 19:24:27 -0600$ From: "del cecchi" <dcecchi@msn.com>) Subject: Re: Compaq now a takeover target 4 Message-ID: <x9UQ7.1934$Lk3.50070@eagle.america.net>  - "cjt" <cheljuba@prodigy.net> wrote in message % news:3C131B60.297BC55B@prodigy.net...  > IsraelRT wrote:  > > ; > > >"The family foundation that is the largest shareholder 7 > > >of Hewlett-Packard, said it would vote against the & > > >proposal to buy Compaq Computer." > > G > > Once the merger is cancelled, Compaq is likely to go belly-up or at  > > least be taken over .  > > E > > The takeover could even be  by Dell if the anti-trust legislators  can  > > be bought off .  > H > If they're willing to drop the Microsoft suit, I can't imagine the DOJ would  > get in the way of Dell.   G Cool, a monopoly in desktops.  :-)  Them and about 50,000 garages?  Can 9 one carry computer shopper without a forklift these days?   
 del cecchi   ------------------------------  $ Date: Sun, 9 Dec 2001 20:28:37 -0500% From: "John Vottero" <John@mvpsi.com> ) Subject: Re: Compaq now a takeover target / Message-ID: <u183u7j5k1m74e@news.supernews.com>   7 "IsraelRT" <israelrt@optushome.com.au> wrote in message 2 news:lud51u4fpgt7g0m4nl0e26rcii7c45olog@4ax.com...9 > >"The family foundation that is the largest shareholder 5 > >of Hewlett-Packard, said it would vote against the $ > >proposal to buy Compaq Computer." > E > Once the merger is cancelled, Compaq is likely to go belly-up or at  > least be taken over .  >   D Why would Compaq go belly up?  They billions of dollars in the bank.  G > The takeover could even be  by Dell if the anti-trust legislators can  > be bought off .    Why would Dell want Compaq?    ------------------------------  # Date: Mon, 10 Dec 2001 03:18:12 GMT   From: cjt <cheljuba@prodigy.net>) Subject: Re: Compaq now a takeover target + Message-ID: <3C14295D.E5E7061F@prodigy.net>    John Vottero wrote:  >  <snip> > Why would Dell want Compaq?u   Good point.    ------------------------------  # Date: Mon, 10 Dec 2001 03:17:09 GMTl  From: cjt <cheljuba@prodigy.net>) Subject: Re: Compaq now a takeover target + Message-ID: <3C14291E.23125DF9@prodigy.net>a   del cecchi wrote:  > / > "cjt" <cheljuba@prodigy.net> wrote in messagek' > news:3C131B60.297BC55B@prodigy.net...a > > IsraelRT wrote:a > > >e= > > > >"The family foundation that is the largest shareholderk9 > > > >of Hewlett-Packard, said it would vote against then( > > > >proposal to buy Compaq Computer." > > >aI > > > Once the merger is cancelled, Compaq is likely to go belly-up or ate > > > least be taken over .t > > >oG > > > The takeover could even be  by Dell if the anti-trust legislatorsn > cank > > > be bought off .q > >aJ > > If they're willing to drop the Microsoft suit, I can't imagine the DOJ > woulds > > get in the way of Dell.n > I > Cool, a monopoly in desktops.  :-)  Them and about 50,000 garages?  Can ; > one carry computer shopper without a forklift these days?n >  > del cecchi  O You obviously haven't looked at Computer Shopper lately.  It's a dead industry.    ------------------------------  % Date: Sun, 09 Dec 2001 22:32:37 -0500 - From: JF Mezei <jfmezei.spamnot@videotron.ca>s) Subject: Re: Compaq now a takeover target , Message-ID: <3C142CD3.B850F3A0@videotron.ca>   John Vottero wrote:nF > Why would Compaq go belly up?  They billions of dollars in the bank.  J The killing of Alpha and subsequent rapid admission that Tru64 was a gonerM even before the merger with Carly was consumed has sent a strong message thatd8 will repell a certain number of customers for some time.  M Combine this with the void between June 25 and the time when VMS can reliably N serve on that IA64 thing with performance that is better than  Alpha,  and you6 have a long time without substantial enterprise sales.  K If discussions about the future of Compaq continue for much longer, it will6N further impact sales to customers for whom long term relationship is importantN for support. If there is speculation about Compaq being target for a takeover,K customers will hesitate before committing millions for an enterprise systemi/ with a company whose future is bound to change.r    K If Compaq doesn't downsize to match lowered sales, its billions in the bankdK will vanish rather quickly. And if Compaq downsizes, it will be made a true G small niche player in the industry and Carly, Gates and Grove will haveoG achieved what they wanted: destroy Alpha, VMS and Tru64 and remove a PClN manufacturer from an overcrowded field. And it will be a big blow to Winkler's9 ego if he must downsize his money losing wintel business.s    M If the merger is not consumed, nothing short of a big clean sweep of Compaq'seK PC-oriented top management (especially Winkler) will give Compaq sufficientiI credibility with potential enterprise customers to rebuild its enterprisei
 business.   N So the big question is whether Compaq management will be thrown out or whether? they will stay on and given a chance to fix up the PC business.'  M Meanwhile, Sony is slowly ramping up its PC business. I have said before thatmL eventually, I believe that "Access" devices to use Compaq's terminology will be made in asia. i  F Sony has high volume distribution networks around the world that carryI everything from walkmans to professional video equipment. It can use thataG existing network to distribute PCs to electronics stores. It can send a H shipful of electrinics anywhere in the world, so its PCs travel for very
 little money.-  J Compaq can't afford a similar distribution network because it only sells aN relatively low number of units compared to Sony's vast number of devices. DellP is the only model that can survive in north america for desktops, in my opinion.   ------------------------------  # Date: Mon, 10 Dec 2001 03:40:20 GMTc4 From: "Terry C. Shannon" <terryshannon@mediaone.net>) Subject: Re: Compaq now a takeover targeti) Message-ID: <E8WQ7.335$7y.2026@rwcrnsc54>o  : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:3C13BA46.734DD8EC@videotron.ca... > JoAnn DiFrancesco wrote:I > > Yeah, that's the ticket! Become a leader selling Wintel desktops at a  loss!n > > What a concept!C- > > What a load of ill-considered codswallop!mL > > While I appreciate the sophomoric humour in this suggestion, I'd sure as. > > heck like to see some facts to back it up. >dL > Have you ever listened to Winkler speak publicly ? Take the time to listen toC > him during financial analysts teleconferences. They are generally  available onG > the web. Listen to them live before they edit those portions out. UsehI > Deja/ggoogle to look up this newsgroup for comments about previous such K > conferences, you will probably find Alan Grieg quoting quite a bit of the  > stuff said by Winkler. >,L > > Oh, purposefully hiding? I would assume that anyone who could PROVE thisL > > allegation could have a hell of a lot of fun with the SEC, et al. In theA > > absence of proof, it's nothing but Blovation Without a Cause!o >eL > By combining the true enterprise stuff with the "toy" enterprise stuff andJ > producing just a single number, you are effectively hiding what the trueG > enterprise stuff generates because the total number is tainted by theS wintelG > server numbers which have totally different margins and profit/costs.e >aG > IBM is proud to brag about MVS/mainframes generating the profits thatfD > subsidize its PC business. But Compaq works hard not to admit thatK > VMS/Tandem/Tru64 are saving the day and prefers instead to brag about howoJ > Wintel will rule the enterprise and will upscale to displace mainframes.  F IBM (a company for whom I have a lot of respect) does an excellent jobH marketing its products. As for Compaq, the fact that they are a big-timeL Windoze OEM should not get in the way of the fact that Windoze DOES NOT ruleJ the enterprise, and almost certainly not will do so for quite some time to come.e  G I suspect that if Rich Marcello and Mark Gorham were not muzzled by the ? Houston crowd, their advertising and marketing efforts would be- substantially more aggressive.  L Why, they might even try to do some VMS marketing in the same manner that VX3 Company recently (and very successfully) undertook.3   ------------------------------  # Date: Mon, 10 Dec 2001 03:43:21 GMTS4 From: "Terry C. Shannon" <terryshannon@mediaone.net>) Subject: Re: Compaq now a takeover targeti) Message-ID: <sbWQ7.355$7y.1672@rwcrnsc54>   : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:3C142CD3.B850F3A0@videotron.ca... > John Vottero wrote:nH > > Why would Compaq go belly up?  They billions of dollars in the bank. >eL > The killing of Alpha and subsequent rapid admission that Tru64 was a gonerJ > even before the merger with Carly was consumed has sent a strong message that: > will repell a certain number of customers for some time.  K Alpha is dead and I believe Compaq can make a defendable financial case foruL terminating Alpha beyond EV7 (the 21364 will be publicly unveiled in just 48 hours in Dallas, TX).   K Tru64 UNIX is a horse of a different color. Compaq has an, umm, interestinge" issue to deal with in this regard!   ------------------------------  # Date: Mon, 10 Dec 2001 03:54:04 GMT4$ From: Ric Werme <werme@mediaone.net>) Subject: Re: Compaq now a takeover targetr< Message-ID: <wlWQ7.3024$zX1.5042353@typhoon.ne.mediaone.net>  & "del cecchi" <dcecchi@msn.com> writes:  H >Cool, a monopoly in desktops.  :-)  Them and about 50,000 garages?  Can: >one carry computer shopper without a forklift these days?  J Have you seen Computer Shopper lately?  A mere ghost of what it was in theB pre-WWW days.  Or whatever it is that made the glory days history.   	-Ric Werme  --K "When we allow fundamental freedoms to be sacrificed in the name of real or D perceived emergency, we invariably regret it.   -- Thurgood MarshallC    Ric Werme                            | werme@nospam.mediaone.nete>    http://people.ne.mediaone.net/werme  |       ^^^^^^^ delete   ------------------------------  % Date: Sun, 09 Dec 2001 23:39:54 -0500 - From: JF Mezei <jfmezei.spamnot@videotron.ca>o) Subject: Re: Compaq now a takeover target-, Message-ID: <3C143C93.DB04A96B@videotron.ca>   "Terry C. Shannon" wrote:rM > Tru64 UNIX is a horse of a different color. Compaq has an, umm, interestingo$ > issue to deal with in this regard!  N Compaq has burned its bridges. Considering its past ineptude to fix enterpriseN image problems, is there any reason why we could expect winkler and friends toC change their tunes and make a convincing gesture to give tru64 somei credibility again ?X    M What about Winkler and friends starting to court Tru64 customers and pitching.E them IA64 Windows solutions to replace their old legacy Tru64 stuff ?y   ------------------------------  # Date: Mon, 10 Dec 2001 05:05:15 GMT   From: cjt <cheljuba@prodigy.net>) Subject: Re: Compaq now a takeover targetl+ Message-ID: <3C144277.CE2B5CE8@prodigy.net>t   JF Mezei wrote:  >  > "Terry C. Shannon" wrote:oO > > Tru64 UNIX is a horse of a different color. Compaq has an, umm, interestingn& > > issue to deal with in this regard! > P > Compaq has burned its bridges. Considering its past ineptude to fix enterpriseP > image problems, is there any reason why we could expect winkler and friends toE > change their tunes and make a convincing gesture to give tru64 somei > credibility again ?  > O > What about Winkler and friends starting to court Tru64 customers and pitchingaG > them IA64 Windows solutions to replace their old legacy Tru64 stuff ?   & I think Sun will clean up as a result.   ------------------------------  # Date: Mon, 10 Dec 2001 05:21:10 GMTsL From: winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr")) Subject: Re: Compaq now a takeover target 8 Message-ID: <00A0647F.E2CAD561@SSRL04.SLAC.STANFORD.EDU>  c In article <wlWQ7.3024$zX1.5042353@typhoon.ne.mediaone.net>, Ric Werme <werme@mediaone.net> writes: ' >"del cecchi" <dcecchi@msn.com> writes:i >hI >>Cool, a monopoly in desktops.  :-)  Them and about 50,000 garages?  Cane; >>one carry computer shopper without a forklift these days?m >oK >Have you seen Computer Shopper lately?  A mere ghost of what it was in thelC >pre-WWW days.  Or whatever it is that made the glory days history.a  O Most of the advertising for MicroTimes, the magazine that I wrote a column  fortN in 1998-2001, came from garage or storefront PC operations.  MicroTimes is nowL out of business (after over 20 years).  Name-brand PCs got so cheap (or cameK with such  big rebates from MSN or AOL) that the garage operations couldn'tsJ compete. Not enough big vendors are left to buy many ads either.  No ads,  no magazine.  K (I was writing about freeware, mostly Linux/xBSD, but I'd slip in plugs forgL VMS and appropriate slams against Microsoft whenever I could.  No more bully pulpit.  *sigh* )h   -- Alan     O ===============================================================================10  Alan Winston --- WINSTON@SSRL.SLAC.STANFORD.EDUM  Disclaimer: I speak only for myself, not SLAC or SSRL   Phone:  650/926-30567M  Physical mail to: SSRL -- SLAC BIN 69, PO BOX 4349, STANFORD, CA  94309-0210 O ===============================================================================t   ------------------------------  # Date: Mon, 10 Dec 2001 06:07:35 GMTe. From: "aaron spink" <aaronspink@earthlink.net>) Subject: Re: Compaq now a takeover targeto? Message-ID: <HiYQ7.3$vK1.2932@newsread2.prod.itd.earthlink.net>w  1 "Ric Werme" <werme@mediaone.net> wrote in message 6 news:wlWQ7.3024$zX1.5042353@typhoon.ne.mediaone.net...L > Have you seen Computer Shopper lately?  A mere ghost of what it was in theD > pre-WWW days.  Or whatever it is that made the glory days history. >hD Computer Shopper was simply a price guide/index.  What killed it wasF websites like Pricewatch.  All I ever used Computer Shopper for was to< compare prices.  Now I can do that with a click of a button.   Aaron Spink  speaking for myselfu   ------------------------------  % Date: Mon, 10 Dec 2001 01:31:14 -0500 - From: JF Mezei <jfmezei.spamnot@videotron.ca>h) Subject: Re: Compaq now a takeover targete, Message-ID: <3C1456A2.5BD01582@videotron.ca>  * Alan Winston - SSRL Admin Cmptg Mgr wrote:N > out of business (after over 20 years).  Name-brand PCs got so cheap (or cameM > with such  big rebates from MSN or AOL) that the garage operations couldn't K > compete. Not enough big vendors are left to buy many ads either.  No ads,  > no magazine.  L Don't forget that garage operations almost always provided bootlegged copiesL of Windows with the poor owner stuck without a CD to reinstall Windows. OnceL owners realise the importance of having a real windows CD to boot from, that2 garage operation PC doesn't look so great anymore.  F And I think that with XP, it becomes harder to run a garage operation.   ------------------------------  # Date: Mon, 10 Dec 2001 06:46:18 GMT   From: cjt <cheljuba@prodigy.net>) Subject: Re: Compaq now a takeover target + Message-ID: <3C145A26.DA939EA6@prodigy.net>c   JF Mezei wrote:t > , > Alan Winston - SSRL Admin Cmptg Mgr wrote:P > > out of business (after over 20 years).  Name-brand PCs got so cheap (or cameO > > with such  big rebates from MSN or AOL) that the garage operations couldn'teM > > compete. Not enough big vendors are left to buy many ads either.  No ads,g > > no magazine. > N > Don't forget that garage operations almost always provided bootlegged copies  I I think that's an assumption on your part that needs to be substantiated.   N > of Windows with the poor owner stuck without a CD to reinstall Windows. OnceN > owners realise the importance of having a real windows CD to boot from, that4 > garage operation PC doesn't look so great anymore. > H > And I think that with XP, it becomes harder to run a garage operation.   ------------------------------  # Date: Mon, 10 Dec 2001 02:45:54 GMT ) From: rob.buxton@wcc.govt.nz (Rob Buxton)a Subject: Re: disk shadowingg1 Message-ID: <3c14218c.618957023@news.wcc.govt.nz>C  F Also, try and close down all your apps so there's no files open on the disks at close down.C You might want to do a show device /files on all of the shadow setsiB during the close down and then work out what else needs to be shut down, images deinstalled etc.   1 Managed to stop most of the merges etc. that way.o      A On Sun, 9 Dec 2001 23:27:32 +0800, "Kenneth" <chehon@net-yan.com>> wrote:  I >I am using 2 X AS8400 running VMS7.2-1 cluster with host base shadowing.tH >When I shudown one node, the other node will perform a disk merge. ThisM >would not be a prblem if it only perform the minimerge, however, some of the M >disks will perform a full merge and will create extensive addtional IO. I amoK >using HSJ50 for all disks and have checked the characteristic and flag forh9 >the device and all the disk are have the "wlg" flag set.t > J >The problem is why one disks can perform minimerge sometime but sometimesH >not? Is there any thing I can do to make them do minimerge everything IK >shutdown the system? Can I stop the merging in the middle and schedule ther >merge off the peak hours? >r >f   ------------------------------  % Date: Sun, 09 Dec 2001 20:38:33 -0500,' From: Jim Becker <jbecker@ui.urban.org>o9 Subject: Re: Encompass Board Election Countdown Continuesr, Message-ID: <3C141219.443D249D@ui.urban.org>   Larry Kilgallen wrote:= > The paper ballot said it must be postmarked by December 12.t  + Yup. Here's an excerpt from the web page...i    B All ballots should be sent into Encompass Headquarters by DecemberE 12th (postmarked date is fine). Ballots can be faxed to 312-673-4609.   B Your membership ID number is not necessary - as in previous year's? elections, all ballots are being validated and are tabulated in  strictest confidence.   # Thank you again for participating. e     --
 Jim Becker+ The Urban Institute (http://www.urban.org/)f' Encompass (http://www.encompassus.org/)-. ESILUG (http://encompasserve.org/lugs/esilug/)   ------------------------------   Date: 9 Dec 2001 20:05:06 -0600C9 From: kaplow_r@eisner.encompasserve.org.mars (Bob Kaplow)F9 Subject: Re: Encompass Board Election Countdown Continueso3 Message-ID: <rOoLGjMiWjV7@eisner.encompasserve.org>-  e In article <PfNQ7.22247$ER5.279152@rwcrnsc52>, "Terry C. Shannon" <terryshannon@mediaone.net> writes:: > Folks, > L > I've heard from a number of Encompass members who claim that they have notK > received election material as of yet. Thanks to the crufty and squatulent.J > MSN Search facility that returns "THIS PAGE CANNOT BE DISPLAYED" I can'tJ > help from where I'm at right now. Perhaps you will have better luck from > your own browser.  > M > This is an important election... regardless of who you'd like to see on thenA > Board, PLEASE take the opportunity to vote in this cycle. VoternE > participation in recent BoD elections has been disappointingly low.- > J > Votes must be received by snailmail or fax to SBA in Chicago by close of > business on 12 December,  J I FAXed my ballot in last week (along with several others collected at theI CARTS meeting where I brought a stack of them). But it's already too lateaI for any one to mail them in outside of the Chicago metropolan area if theoI deadline is RECEIVED by the 12th. Didn't it say POSTMARKED by the 12th???y    B         The problem with governments is that citizens need to keepD         them on a short leash; unfortunately the nature of the beastD         is such that governments can usually arrange it so that only"         they hold their own leash.  1 	26-October, 2001: A day that will live in infamyn4 	Support Freedom: http://www.indefenseoffreedom.org/   ------------------------------  # Date: Mon, 10 Dec 2001 03:44:35 GMT 4 From: "Terry C. Shannon" <terryshannon@mediaone.net>9 Subject: Re: Encompass Board Election Countdown Continuess- Message-ID: <DcWQ7.22659$wL4.56000@rwcrnsc51>n  F "Bob Kaplow" <kaplow_r@eisner.encompasserve.org.mars> wrote in message- news:rOoLGjMiWjV7@eisner.encompasserve.org...oC > In article <PfNQ7.22247$ER5.279152@rwcrnsc52>, "Terry C. Shannon"f# <terryshannon@mediaone.net> writes:w
 > > Folks, > >aJ > > I've heard from a number of Encompass members who claim that they have notIB > > received election material as of yet. Thanks to the crufty and
 squatulentL > > MSN Search facility that returns "THIS PAGE CANNOT BE DISPLAYED" I can'tL > > help from where I'm at right now. Perhaps you will have better luck from > > your own browser.n > >tK > > This is an important election... regardless of who you'd like to see one theuC > > Board, PLEASE take the opportunity to vote in this cycle. VoterlG > > participation in recent BoD elections has been disappointingly low.e > >lL > > Votes must be received by snailmail or fax to SBA in Chicago by close of > > business on 12 Decemberi >oL > I FAXed my ballot in last week (along with several others collected at theK > CARTS meeting where I brought a stack of them). But it's already too latenK > for any one to mail them in outside of the Chicago metropolan area if thetK > deadline is RECEIVED by the 12th. Didn't it say POSTMARKED by the 12th???r >m  ( Yes indeed it did! Sorry for the errata!   terry s    ------------------------------  % Date: Sun, 09 Dec 2001 21:31:01 -0500s From: gce <ge@gce.com> Subject: Re: gnu tar for VMS' Message-ID: <3C141E65.CF5FB1D2@gce.com>s  J I got a port of gtar for VMS working some time ago under VMS Posix. Worked< ok and was put on the sigtapes. Looks like it was fall 1992.# Source was in [vax92b.gce92b.gtar].l     David Schwartz wrote:e > H > I have some very large tarfiles that are not readable by VMS tar. I'veC > tried downloading a "new" vmstar from the Compaq OpenVMS freewareoD > site, and also gotten the latest VMS tar (V3.4-1) from the ProcessG > Software VMS archive. All fail with the same error, when doing eitheroC > a test (-t) or extract (-x), after partially processing the file:i > B > ----------     0/0        101 Dec 31 16:00:00 1969 ././@LongLink4 > tar: directory checksum error for <long file path> > D > The file I'm trying to extract was created with gnu tar on a LinuxC > system. I'm told this is a known incompatibility, and that tar onJE > Solaris has the same problem.  I copied the file to a Linux system,.G > and gnu tar can read it, so the file is not corrupted. Recreating theeC > file as several smaller files has proved to be time consuming andh< > cumbersome, and it won't help me with the next large file. > ! > Is there a VMS port of Gnu tar?    ------------------------------  # Date: Mon, 10 Dec 2001 03:14:04 GMTo5 From: "Ken and Kelley Coleman" <knkcoleman@attbi.com>p= Subject: Re: How to implement Login Fails for Open VMS on VAXi- Message-ID: <0MVQ7.27334$Yy.327376@rwcrnsc53>   F Is there a tool to proactively contact the sys admin should there be a- suspected intrusion security breach going on?n  < (Be kind if this is obvious. I'm relatively new to OpenVMS.)      : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:3C10EB1D.848FF61F@videotron.ca... > Larry Kilgallen wrote:H > > SYSTEM can log in at the console under a lot of otherwise-prohibitedH > > circumstances.  Have you verified your claim with a test, or are you > > just guessing ?a >uF > Note that if you have an operator which is constantly logged in, DOS becomesfL > less of an issue since he can then delete/intrusion from his terminal that is > already logged in.   ------------------------------  $ Date: Sun, 9 Dec 2001 20:26:13 -0500% From: "John Vottero" <John@mvpsi.com> C Subject: Re: How to tell if foreign command (SET COMMAND) is known?e/ Message-ID: <u183pm8njqkj23@news.supernews.com>s  J If you want to check and see if MYCMD is a verb defined in the current DCL tables do something like this:  ( $ MYCMXYZ1JUKWO/KDJU/IDUEK/KDISED/JKSKDI  5 and see if $STATUS is %DCL-W-IVVERB or %DCL-W-IVQUAL.c  L IVVERB would mean that the MYCM... verb is NOT defined and IVQUAL would mean that it IS defined.,  H Note that in current versions of VMS, only the first 4 characters of theJ verb count.  Use only the first 4 characters of your verb followed by junkI to reduce the possibility of matching a DCL symbol.  Follow the verb witheI plenty of garbage qualifiers to make sure you don't have a valid command.i    < "Michael Zarlenga" <zarlenga@conan.ids.net> wrote in message) news:u159e75t3geg61@corp.supernews.com... ? > I define several foreign commands via SET COMMAND .CLD files.s >eF > Aside from actually trying it, how can I tell if a command is known? > = > Is there an equivalent to SHOW SYMBOL for foreign commands?a >. > -- > -- Mike Zarlenga   ------------------------------   Date: 9 Dec 2001 13:08:29 -0600 + From: young_r@encompasserve.org (Rob Young)t Subject: Re: IBM VMS and Oracle33 Message-ID: <QwORu1pRCbrM@eisner.encompasserve.org>o  \ In article <3C13A353.A7D7C1D3@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes: > M > But what would happen with RDB and Oracle, two products fairly key to VMS ?lP > Could IBM maintain a cordial relationaship with Oracle for those 2 products onJ > VMS, or would this mean war, with IBM porting DB2 to VMS and abandonning > Oracle on VMS ?c  C 	Yes.  Oracle on AIX is strong today.  Oracle is a big back-end foriG 	quite a few products out there.  It is important for that relationshipe. 	to continue for a lot of folks, not just IBM.   				Rob    ------------------------------  # Date: Sun, 09 Dec 2001 22:09:16 GMTeL From: winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr") Subject: Re: IBM VMS and Oracle-8 Message-ID: <00A06443.8CF39D0D@SSRL04.SLAC.STANFORD.EDU>  \ In article <3C13A353.A7D7C1D3@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes:N >Lets assume for a minute that IBM buys VMS and the alpha *systems* designers. >eI >VMS is ported to Power4 and the ex-dec engineers build the equivalent of-L >wildfires with the power architecture, relegating IA64 to bloated-ware that >runs only wintel stuff. >tO >VMS customer are happy to finally be in the hands of a corporation that has noaJ >plans to get rid of their product, and a port to a platform that has more5 >credibility in the arena where quality is important.. >iL >But what would happen with RDB and Oracle, two products fairly key to VMS ?O >Could IBM maintain a cordial relationaship with Oracle for those 2 products onaI >VMS, or would this mean war, with IBM porting DB2 to VMS and abandonning9 >Oracle on VMS ?  J Oracle's perfectly happy to sell software for the IBM mainframes on which K DB2 was first developed.  Why should they be terminallpy offended at an IBMqF DB2 port to VMS?  And since the Rdb/DBMS group keeps making money, whyF should they stop that?  If current VMS customers moved to an (entirelyH notional) IBM Power4 VMS, they'd be more interested in continuing to runI Rdb than in switching to  DB2, and Oracle could sell them new licenses.  cJ Oracle doesn't seem to make strategic decisions based entirely on pique.    K [Also, it would probably be easier for IBM to reactivate the VMS version of1J Informix than to port DB2 - don't know for sure.  However, if they had DB2J customers who wanted VMS clustering, it might be worth their while, in theJ so-entirely-speculative-it-isn't-really-worth-talking-about realm in which  this conversation is operating.]   -- Alann    O ===============================================================================e0  Alan Winston --- WINSTON@SSRL.SLAC.STANFORD.EDUM  Disclaimer: I speak only for myself, not SLAC or SSRL   Phone:  650/926-3056 M  Physical mail to: SSRL -- SLAC BIN 69, PO BOX 4349, STANFORD, CA  94309-0210rO ===============================================================================c   ------------------------------  $ Date: Sun, 9 Dec 2001 20:50:08 -0500% From: "John Vottero" <John@mvpsi.com>i Subject: Re: IBM VMS and Oracle-/ Message-ID: <u1856ma2jbjk98@news.supernews.com>1  J It might make more sense for Oracle to buy VMS.  I think that Oracle wantsK to sell people orange boxes that are just "database servers".  Why not have6 VMS under the skin?L  : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:3C13A353.A7D7C1D3@videotron.ca...D > Lets assume for a minute that IBM buys VMS and the alpha *systems*
 designers. >iJ > VMS is ported to Power4 and the ex-dec engineers build the equivalent ofH > wildfires with the power architecture, relegating IA64 to bloated-ware that > runs only wintel stuff.  >wI > VMS customer are happy to finally be in the hands of a corporation thato has noK > plans to get rid of their product, and a port to a platform that has moreo6 > credibility in the arena where quality is important. >oK > But what would happen with RDB and Oracle, two products fairly key to VMSs ?tD > Could IBM maintain a cordial relationaship with Oracle for those 2 products onoJ > VMS, or would this mean war, with IBM porting DB2 to VMS and abandonning > Oracle on VMS ?g   ------------------------------   Date: 9 Dec 2001 20:29:38 -0600l- From: Kilgallen@SpamCop.net (Larry Kilgallen)i Subject: Re: IBM VMS and OracleN3 Message-ID: <$kF+EgtuhvR$@eisner.encompasserve.org>=  W In article <u1856ma2jbjk98@news.supernews.com>, "John Vottero" <John@mvpsi.com> writes: L > It might make more sense for Oracle to buy VMS.  I think that Oracle wantsM > to sell people orange boxes that are just "database servers".  Why not haveO > VMS under the skin?   F Because they don't want to alienate all their other platform partners.   ------------------------------  # Date: Mon, 10 Dec 2001 03:25:40 GMTe) From: rob.buxton@wcc.govt.nz (Rob Buxton)a* Subject: Lost Programs - FUser & SHOW_USER1 Message-ID: <3c142597.619992061@news.wcc.govt.nz>2   Hi All,d  > The Cluster I inherited some time ago has a couple of VAX only) utilities. I do not have the source code.@A Where these came from is anyones guess. I'd really quite like the? Alpha variants or source code.  : FUSER has two options (that I know about) -list and -kill.  D The first will list any processes cluster wide that have access to a( specified file. e.g fuser file.dat -list  C The -kill will kill any processes that have access to that file. Itn works Cluster wide. B It's used at the start of our overnights to kill off any processes2 that have files open that will need to be updated.  C SHOW_USER is a program that just displays a given Users UAF Record.m@ The image is installed with SYSPRV and controlled by ACL Access.  0 Has anyone seen any utilities like these around?  C I've seen a File utility that lists who has a file open. The one we 0 have is just a bit more powerful (or dangerous!)  ( Tried looking through the Freeware CDs.   C Alas, on Google FUser returns info about either Toner Cartridges or  the unix fuser command.    TIAt   Rob.   ------------------------------  % Date: Sun, 09 Dec 2001 14:39:05 -0500 - From: JF Mezei <jfmezei.spamnot@videotron.ca>oB Subject: Re: Packard Foundation Tells Merger Urgers to "Paq it In", Message-ID: <3C13BDCF.9BE08AED@videotron.ca>  K > > IMHO the consumer products are nothing but a big albatross 'round CPQ'sa	 > > neck.r  ! Consider the following scenario: -  K Corporation needs 250 desktops, and 25 servers. If Compaq drops  the accessuI business because it doesn't make money,  a customer will more than likelyoL order his 250 desktops from dell and while he is at it, order the 25 servers from dell too.  E One can use the same arguments about desktops in the wintel market as F compilers or Alpha in the VMS market. They may cost money but they are5 necessary evil to generate the more profitable sales.u  L What compaq should have done is to lump all wintel products together (serverL and desktops), at which point it becomes easier to hide the costs of sellingG desktops since they aere now part of the cost of selling the profitablehN servers, just as Alpha is part of the cost of selling profitable VMS machines.    M Ahhh, but what to do about the consumer vs business desktops ????  Accept theoL fact that they are one of the same and price all of them competitively. Dell3 does, and if Compaq doesn't, then Compaq will lose.-   ------------------------------  $ Date: Sun, 9 Dec 2001 20:44:16 -0500% From: "John Vottero" <John@mvpsi.com>hB Subject: Re: Packard Foundation Tells Merger Urgers to "Paq it In"/ Message-ID: <u184rhkrtdre43@news.supernews.com>f  : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:3C13BDCF.9BE08AED@videotron.ca...G > > > IMHO the consumer products are nothing but a big albatross 'roundt CPQ's  > > > neck.a >d" > Consider the following scenario: >eF > Corporation needs 250 desktops, and 25 servers. If Compaq drops  the accessK > business because it doesn't make money,  a customer will more than likelyhF > order his 250 desktops from dell and while he is at it, order the 25 serversi > from dell too. >d  J You're confusing the access business with consumer products.  They are twoF very different things.  I believe that Compaq should dump the consumerE products but they should still be selling desktop systems targeted atf businesses.e  K You may not have noticed but Compaq is now selling stereo components (as is  HP).   ------------------------------  # Date: Mon, 10 Dec 2001 03:50:58 GMTg4 From: "Terry C. Shannon" <terryshannon@mediaone.net>B Subject: Re: Packard Foundation Tells Merger Urgers to "Paq it In") Message-ID: <CiWQ7.376$7y.2234@rwcrnsc54>   > "John McLean" <mcleanj@swissonline.delete.ch> wrote in message/ news:3C13B295.DD9AC6FC@swissonline.delete.ch...  >j >: > "Terry C. Shannon" wrote:  > >d > ...t > >rK > > IMHO the consumer products are nothing but a big albatross 'round CPQ'sh	 > > neck.l >s* > That's stating the bleedin' obvious  ;-) >uI > I'm starting to wonder if a "revised offer" might appear given that oneiD > of the biggest stumbling blocks was that the estimates of Compaq'sC > earnings for 2002 have taken a nosedive since the original offer.  >nI > If Capellas didn't find a way to get a large chunk of Alpha developmentdD > costs out of the picture before Sep 30th accounts then I bet he is > kicking himself now. >-  E Another interesting point, John. While Compaq *did* manage to use thetJ Alphacide to improve its numbers in 2FQ (note the big one-time charge thatF enabled CPQ to make its earnings projection despite a hefty decline inI revenue), little light has been shed on how much money has been saved/hasm changed hands since then!w   ------------------------------   Date: 9 Dec 2001 16:51:39 -0800 ( From: bob@instantwhip.com (Bob Ceculski) Subject: Re: Secure Telnet= Message-ID: <d7791aa1.0112091651.5aeb8daa@posting.google.com>o  T Robin <rlb@austin.rr.com> wrote in message news:<3C138236.7F28156B@austin.rr.com>...< > Anyone have any recommendations for running secure telnet?  H Tcpware has SSH built in as does multinet ... another option is to run a Decnet over IP tunnel ...e   ------------------------------  # Date: Mon, 10 Dec 2001 02:35:26 GMT 4 From: "Matt Muggeridge" <Matt.Muggeridge@compaq.com> Subject: Re: Socket limitsA Message-ID: <ObVQ7.438746$bY5.1818235@news-server.bigpond.net.au>t  F Assuming you have a suitably configured and tuned system with accounts@ configured with appropriate quotas... TCP/IP on OpenVMS supports  D     sockets - 10,000 today, (more than that in some future release).  J The listen backlog is limited in the same way (not that I've ever cared to
 try that).   Matt.t  2 "C.W.Holeman II" <cwhii@mail.com> wrote in message! news:3C11863C.2584804@mail.com...dG > It appears to me that the VAXELN TCP/IP socket implementation has the  > following limits:  >h3 >     o listen() has a maximum queue length of one.h= >     o A maximum of four UDP messages are buffered per port. > >     o There is a maximum of 64 sockets & files open per job. > H > I have not used any other socket implementations. How do these compare	 > to VMS?n >S > -- > C.W.Holeman II > cwhii@mail.com > http://also.as/cwhii   ------------------------------  % Date: Sat, 08 Dec 2001 23:39:52 +0100 ( From: Bernd Paysan <bernd.paysan@gmx.de>! Subject: Re: The demise of compaqo& Message-ID: <3C1296B8.3750CB78@gmx.de>   Jerry Leslie schrieb:nI >   "SMS: Joel, what, in your opinion, is the single greatest development ' >    sin a software company can commit?g > K >    Joel: Deciding to completely rewrite your product from scratch, on theyH >    theory that all your code is messy and bug prone and is bloated andD >    needs to be completely rethought and rebuild from ground zero." >  > I think I'll stick with: > - > "Plan to throw one away; you will, anyhow."f' > "The Mythical Man-Month", Fred Brooks   B These two things should not be read without context. For a typicalB Microsoft-like company, a complete rewrite just produces new crap,D rather than the old and settled crap they are already shipping. Why?E Because a project takes considerably longer than the average turnoverlE rate. I.e. if you start from scratch, all the knowledge about the oldwE code is already gone. And without that knowledge, you *can't* benefita from the rewrite.   G And the way he describes the number of quirks and incompatibilities yousF have to work around when programming Windows is wonderful. It's reallyG that way: loads of junk written to work around bugs in the library thatd? should have been fixed *there*, not in the application, damned!b  H So while I hate to agree with a pointy haired Microsoft boss, he's rightH - his way. For a development team, it's good to start over from scratch,E but for a company that doesn't maintain a real "development team" but:H rather a bunch of people crowded together, it is a disaster. Note that ID think crowding people together instead of forming a real development7 team is how the disaster starts in the first place ;-)..  G I think the Mozilla example is a very typical one. This project started H by open sourcing Netscape. Actually, they couldn't open source Netscape,D because major parts were third party libraries. So they open sourcedE something that couldn't be compiled to something useful. While peopletB looked through the monstrous code (created by those people crowdedH together in the Netscape office), they decided to rewrite it rather thanH to fix it. Eventually, they rewrote everything. But in the meantime, theE people at Netscape fought against the rewrites. Really, it would haven@ been better if these people had moved the original (proprietary)H Netscape forward. But then, noone of the outsider developers (who reallyD did most of the rewrites) would have been inclined to put their code
 into Mozilla.   C A well organized project doesn't notice rewrites. They just happen. H There might be some minor hiccups, but a rewrite of one critical part isH not a big deal. The Linux VM e.g. was rewritten several times. Recently,E it happened right in the middle of a stable kernel maintenance phase. B Joel would have gone up the wall on a decision like that. In fact,= rewriting the VM of Windows NT could well be something that's7E impossible. Even if it's broken (ok, it is broken ;-), it's somethingtG that won't happen, because the VM division never would admit that there D really are problems. You would add code to your applications to work around the problems.   -- m Bernd Paysan7 "If you want it done right, you have to do it yourself". http://www.jwdt.com/~paysan/   ------------------------------   Date: 10 Dec 2001 00:50:29 GMT& From: peter@abbnm.com (Peter da Silva)! Subject: Re: The demise of compaqy% Message-ID: <9v10sl$dpo@web.nmti.com>a  & In article <3C1296B8.3750CB78@gmx.de>,* Bernd Paysan  <bernd.paysan@gmx.de> wrote:J > to fix it. Eventually, they rewrote everything. But in the meantime, theG > people at Netscape fought against the rewrites. Really, it would have0B > been better if these people had moved the original (proprietary)J > Netscape forward. But then, noone of the outsider developers (who reallyF > did most of the rewrites) would have been inclined to put their code > into Mozilla.n  G Well, you don't throw the old one away until the new one's ready or youU don't need it any more.e  0 (sort of like Alpha and IA64, now I think of it)   -- w+  `-_-'   In hoc signo hack, Peter da Silva.dE   'U`    "A well-rounded geek should be able to geek about anything."rL                                                        -- nicolai@esperi.org          Disclaimer: WWFD?   ------------------------------   Date: 9 Dec 2001 20:01:48 -0600l9 From: kaplow_r@eisner.encompasserve.org.mars (Bob Kaplow)a> Subject: Re: Tru64.org Flash Poll on Merger "Pearl Harbor Day"3 Message-ID: <w1sP7i9h78OD@eisner.encompasserve.org>e  s In article <QgqQ7.7497$Sj1.3002542@typhoon.ne.mediaone.net>, "Terry C. Shannon" <terryshannon@mediaone.net> writes:tN > Got an opinion about the December 7 Development in the HWP-CPQ Merger story? > Why not pay a visit to5 > www.tru64.org and weigh in on the following survey:- > K > In light of the announcement that the David and Lucile Packard Foundationp > will not support HP's E > proposed acquisition of Compaq, do you believe the proposed merger:m >  > -  Is likely to succeedC >  > - May or may not succeed >  > - It's toast!n  H What is needed for it to pass? IIRC the Compaq/DEC merger required a 2/3G vote of all outstanding shares. While it got a huge percentage positiveaL vote, ther were enough "no responses" that the total at the end came to justJ under 70% of the shares voted YES. So in reality it only passed by about 3 percent.  G If this one also requires a 2/3 vote of all outstanding shares, and yourJ start with a large chunk of NO, you're fighting a major uphill battle. NowH if the Hewletts and Packards start soliciting proxies for others to vote against, it's toast.  I Of course, it may be toasted by the government before it gets that far...M  B         The problem with governments is that citizens need to keepD         them on a short leash; unfortunately the nature of the beastD         is such that governments can usually arrange it so that only"         they hold their own leash.  1 	26-October, 2001: A day that will live in infamye4 	Support Freedom: http://www.indefenseoffreedom.org/   ------------------------------  # Date: Sun, 09 Dec 2001 20:32:02 GMThG From: Simon Clubley <simon_clubley@remove_me.altavista.co.uk-Earth.UFP>nD Subject: Writing a device driver: virtual/physical address questions6 Message-ID: <6TPQ7.55296$xS6.88962@www.newsranger.com>  E This is on an Alpha, VMS V7.2. I am using C instead of MACRO-32, as In know C better.  E [Note: Although I am well experienced in other aspects of VMS, I haveaD recently resumed writing my first ever VMS device driver. This meansD that I may simply need pointing to a section in the manuals. It alsoC means that I may not understand some VMS driver concepts as well aso I think I do. :-)]  A I am working with a PCI device which uses DMA for reading/writingBB process/device driver buffers. This device does this by writing toE PCI memory space, which is of course mapped to Alpha physical memory.y  G This bit I have no problem with; the problem is translating between VMS I virtual addresses and physical addresses as well as allocating physically  contiguous memory.   Question 1:i  E EXE$STD_READ loads IRP$L_SVAPTE with the address of the PTE that mapsiF to the first page of the buffer; I am assuming that this is the L3 PTEG of the locked down buffer. The problem is, assuming a buffer that spansSG several pages, how do you get, from this PTE, the physical addresses ofv( the other pages assigned to the buffer ?  G The problem I am seeing is that you may have to return to the L1/L2 PTE=G tables in order to get this information as the L3 PTE may be located at  the end of the L3 page.S  M The method I've come up with (untested as yet, I'm still writing the driver!) E is to get the original virtual address using the macros in VMS_MACROSAM and then use the other macros to get the PTE for each page. Is this correct ?=E [Of course you need to do this while the driver has process context.],  * I am wondering what I am missing here. :-)   Question 2:c  N The driver itself needs to allocate a chunk of physically (not just virtually)E contiguous memory, which it will pass the physical base address of to M the device's DMA controller. [The memory contains instructions, written usingyI the device's own opcodes, for controlling the data transfer to the user'sa buffer.]  I I couldn't find any reference to allocating physical memory, just virtualtJ memory, within the driver manuals. I found a pointer (in the VAXBI sectionD of the VAX driver manuals!) which led me to the following prototype:  1 int exe_std$alophycntg(int npages, void **sva_p);e  J The problem is that I can't find any documentation for using this routine.L For example, does it have to be called at a minimum IPL ? At a maximum IPL ?  L Assuming that it's the right one, where do you find it formally documented ?   Question 3:   K Just curious, but do DEC/CPQ devices generally require the driver writer toyF just load a PTE or system virtual address into the device, or does theE driver writer have to translate to physical addresses as I am doing ?o  H Finally, I would like to comment that the driver manual is well written;E the questions above are about material that I cannot find documented..   Thanks for any information,:   Simon.   -- O@ Simon Clubley, simon_clubley@remove_me.altavista.co.uk-Earth.UFPK In the task of removing Microsoft from the marketplace, I have discovered a E truly remarkable plan, but this signature is too small to contain it.a   ------------------------------   End of INFO-VAX 2001.685 ************************