1 INFO-VAX	Tue, 18 Jun 2002	Volume 2002 : Issue 336       Contents: Re: "Tru64 and OpenVMS Times" @ Re: "We stay with VMS because we have forgotten that we have it"@ Re: "We stay with VMS because we have forgotten that we have it"@ Re: "We stay with VMS because we have forgotten that we have it" A PL/I question ...  RE: A PL/I question ... ) Re: A Proposal for All C.O.V. to Consider ) Re: A Proposal for All C.O.V. to Consider ) Re: A Proposal for All C.O.V. to Consider ) RE: A Proposal for All C.O.V. to Consider ) Re: A Proposal for All C.O.V. to Consider # Re: A [departure from] VMS Disaster  Re: Affinity and Itanic  Re: ALL-IN-1 anniversary Re: ANN: Updated JUMP - Re: Another C and PL/I question - Descriptors - RE: Another C and PL/I question - Descriptors ! Re: Apache bug affect VMS Apache? ! Re: Apache bug affect VMS Apache? 0 can VMS mail block e-mail from certian address'?  CONNECT:Direct wildcard transfer Re: Copying a file via FID Re: Could linux become VMS?  Re: Could linux become VMS?  Re: Could linux become VMS?  Re: Could linux become VMS?  Re: Could linux become VMS?  Re: Could linux become VMS?  Re: Could linux become VMS?  Re: Could linux become VMS?  Re: Could linux become VMS? 	 CSA->DSPP > Re: DCL Puzzle: Can you produce this DIRECTORY command output?4 Disaster Tolerant advertising using Fortune magazine8 Re: Disaster Tolerant advertising using Fortune magazine9 Re: How can I tell which version of DECNet I am running ? ? Re: Interesting ZDnet post about Tandem -- Can VMS do the same? ? Re: Interesting ZDnet post about Tandem -- Can VMS do the same? ? Re: Interesting ZDnet post about Tandem -- Can VMS do the same? ? Re: Interesting ZDnet post about Tandem -- Can VMS do the same? ? Re: Interesting ZDnet post about Tandem -- Can VMS do the same?  Is list working? Re: Linus' comments about VMS  Re: Linus' comments about VMS  Re: Linus' comments about VMS  Re: Linus' comments about VMS  Re: Linus' comments about VMS  Re: Linus' comments about VMS  Re: Linus' comments about VMS  Re: Linus' comments about VMS  Re: Linus' comments about VMS  Re: Linus' comments about VMS  Re: Linus' comments about VMS  Re: Linus' comments about VMS  Re: Linus' comments about VMS < Location of DLM papers ?, was: Re: Linus' comments about VMS@ Re: Location of DLM papers ?, was: Re: Linus' comments about VMS misc Qns, Re: My conversation with Linus about VMS ..., Re: My conversation with Linus about VMS ..., Re: My conversation with Linus about VMS ..., Re: My conversation with Linus about VMS ..., Re: My conversation with Linus about VMS ..., Re: My conversation with Linus about VMS ..., Re: My conversation with Linus about VMS ..., Re: My conversation with Linus about VMS ..., Re: My conversation with Linus about VMS ..." Re: Need DECNet packet information" Re: Need DECNet packet information. Re: One of Terry's shoes seems to have droppedG RE: OpenVMS, Volume Desktop OS (Re: Mark Gorham's Beer Bash in Reading) G RE: OpenVMS, Volume Desktop OS (Re: Mark Gorham's Beer Bash in Reading) G RE: OpenVMS, Volume Desktop OS (Re: Mark Gorham's Beer Bash in Reading)  PLUG: txt2pdf 5.8  Port of SAMBA 2.2.4 8 Re: Request for path balancing (Was: Re: Carly was here)8 Re: Request for path balancing (Was: Re: Carly was here)8 Re: Request for path balancing (Was: Re: Carly was here)8 Re: Request for path balancing (Was: Re: Carly was here)" Re: rooted devices and /TRANS=CONC2 Re: SSH v2 (was Re: OpenVMS, Volume Desktop OS...)1 Re: Sun is on the way out!  Only HP and IBM left! ' Re: Sun Setting, Not Rising, says eWeek ! Re: Sun/Slowaris future in doubt! ! Re: Sun/Slowaris future in doubt! ! Re: Sun/Slowaris future in doubt! ! RE: Sun/Slowaris future in doubt! ! Re: Sun/Slowaris future in doubt!  Re: unix history Re: unix history Re: unix history Re: unix history RE: unix history5 Re: US FDA ponders Slowaris over VMS ... what morons! , Re: What kind of morons work at the FDA now?* Re: Why is TCPIP$ROUTE.DAT in SYS$COMMON ?* Re: Why is TCPIP$ROUTE.DAT in SYS$COMMON ?9 RE: Will AIT-2 or AIT-3 media/changers work with OpenVMS?  Re: __asm__ in DEC CXX  F ----------------------------------------------------------------------  % Date: Tue, 18 Jun 2002 09:41:13 +0200 E From: Jan C. =?iso-8859-1?Q?Vorbr=FCggen?= <jvorbrueggen@mediasec.de> & Subject: Re: "Tru64 and OpenVMS Times"* Message-ID: <3D0EE419.FEEF6C5@mediasec.de>  F > Now that's just dumb, imho.  If I want to make a TLA plural, I use aD > lower-case "s".  That allows me to reference many TLAs with usefulF > separation of the acronym from the "number".  The plural is NOT part > of the acronym.   K That is not the reason. Every registered mark as an umbra, as it were, that J is also protected by the mark - not only the literal meaning as written inI a certain font and character size, for instance. It gets interesting when K two such umbrae collide - and the IP and litigation lawyers get rich on it.    	Jan   ------------------------------  % Date: Tue, 18 Jun 2002 10:52:21 -0400 ; From: "Brian Tillman" <tillman_brian@notnoone.notnohow.com> I Subject: Re: "We stay with VMS because we have forgotten that we have it" $ Message-ID: <3d0f4919$1@news.si.com>  F >I recently had a talk with such a guy, and use of Charon-VAX to driveJ >their billion dollar mfkt plant was not seen as appropriate (eg. cause of( >the uptime problems of the host opsys).  L SRI will soon be introducing CHARON-VAX/AXP PLUS that will run under OpenVMS, Alpha.  No worry about uptime problems then. --A Brian Tillman                   Internet: tillman_brian at si.com A Smiths Aerospace                          tillman at swdev.si.com = 3290 Patterson Ave. SE, MS      Addresses modified to prevent < Grand Rapids, MI 49512-1991     SPAM.  Replace "at" with "@"8        This opinion doesn't represent that of my company   ------------------------------  % Date: Tue, 18 Jun 2002 10:56:50 -0400 ; From: "Brian Tillman" <tillman_brian@notnoone.notnohow.com> I Subject: Re: "We stay with VMS because we have forgotten that we have it" $ Message-ID: <3d0f4a27$1@news.si.com>  G >Probably longer than that as he said it was an RSX application running F >under emulation. I'm surprised they didn't just stick with the PDP-11 >in the first place.  3 CHARON-11 might be a suitable answer, in that case.  --  A Brian Tillman                   Internet: tillman_brian at si.com A Smiths Aerospace                          tillman at swdev.si.com = 3290 Patterson Ave. SE, MS      Addresses modified to prevent < Grand Rapids, MI 49512-1991     SPAM.  Replace "at" with "@"8        This opinion doesn't represent that of my company   ------------------------------  % Date: Tue, 18 Jun 2002 11:34:10 -0400 ! From: Jim Agnew <jpagnew@vcu.edu> I Subject: Re: "We stay with VMS because we have forgotten that we have it" & Message-ID: <3D0F52F2.6651681@vcu.edu>  D Probably they wanted the extra speed on the application...  a vax inD emulation mode would probably have given them extra speed, and lower maintenance to boot.   Jim    Bill Gunshannon wrote: > 7 > In article <SKEO8.249293$305.3479967@news.chello.at>, 3 >  peter@langstoeger.at (Peter LANGSTOEGER) writes:  > |>J > |> Am I the only one, who doesn't get a good feeling when reading this ?K > |> For me this is a dead system, because it wasn't improved for a decade. L > |> And when the time of this VAX h/w is coming (more likely this year thanN > |> later) and you won't get spare parts any longer, then you HAVE TO discardO > |> the whole system alltogether (like it or not) and replacing it with what ? C > |> The application probably wasn't touched/improved for a decade,  > H > Probably longer than that as he said it was an RSX application runningG > under emulation. I'm surprised they didn't just stick with the PDP-11  > in the first place.  >  > bill >  > --L > Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesF > bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. > University of Scranton   |@ > Scranton, Pennsylvania   |         #include <std.disclaimer.h>   ------------------------------    Date: 18 Jun 2002 07:42:27 -0700, From: bubbapig@hotmail.com (Jeffrey Cameron) Subject: A PL/I question ...< Message-ID: <b22333b7.0206180642.26e1366@posting.google.com>  * I have two procedures which I call from C:  . /************ C CODE ************************/   #include <descrip.h>  . extern "C" char *get_filename(void *, void *);+ extern "C" void *print_out(void *, void *);   ; /* where I use the void * pointers rather then arguments as  declarators */   .  .  .   
 int num = 12;  short val = 1; struct dsc$descriptor_ubs desc; @ /* there are three other fields which i fill in correctly for an6 unaligned bit string but do not have right on me :) */  desc.dsc$a_base = (char*)(&val); desc.dsc$l_pos = 0L;   print_out(&num, &desc); ' char *temp = get_filename(&num, &desc);   5 /****************** PL/I Code **********************/   = get_filename: proc (indx, in_temparg) returns (char(80) var);    dcl indx         fixed bin(31);  dcl in_temparg   bit(1); dcl filename     char(80) var;   put skip list (indx);  put skip list (in_temparg);    return (filename)    end get_filename;   . /*------------------------------------------*/  ! print_out: proc (param1, param2);    dcl param1       fixed bin(31);  dcl param2       bit(1);   put skip list (param1);  put skip list (param2);    end print_out;  D /***************************** END CODE ***************************/  B The problem is this, the print_out procedure is a test stub I madeB which works fine, it prints out both the bit and the integer valueE correctly which are passed into it. You'll notice I do NOT modify the  structure in any way.   F When I call the get_filename procedure though the addresses and valuesC for both parameters are screwed and lead to access violation when I 
 try to print.   D THEY ARE THE EXACT SAME PARAMETER LISTS!!! The only difference is inE the names and the fact that one procedure returns something while the A other does not. What the heck is going on here guys? Is this some E obscure bug? Or a piece of the calling standard I am unfamiliar with?  How do I get around this?   C p.s. I am compiling the C with all the compiler options for sharing E code with other psects /EXTERN=COMMOM/nomember_alignment/noopt ... et  al. IF that helps at all?    Thanks again in advance! Jeff Cameron   ------------------------------  % Date: Tue, 18 Jun 2002 08:18:48 -0700 # From: "Tom Linden" <tom@kednos.com>   Subject: RE: A PL/I question ...9 Message-ID: <CIEJLCMNHNNDLLOOGNJIGEFDFDAA.tom@kednos.com>   A get_filename is not a char, it is a char var, you need to declare # it with a descriptor in you C code.    >-----Original Message----- 4 >From: Jeffrey Cameron [mailto:bubbapig@hotmail.com]% >Sent: Tuesday, June 18, 2002 7:42 AM  >To: Info-VAX@Mvb.Saic.Com >Subject: A PL/I question ...  >  > + >I have two procedures which I call from C:  > / >/************ C CODE ************************/  >  >#include <descrip.h>  > / >extern "C" char *get_filename(void *, void *); , >extern "C" void *print_out(void *, void *); > < >/* where I use the void * pointers rather then arguments as >declarators */  >  >..  >..  >..  >  >int num = 12; >short val = 1;   >struct dsc$descriptor_ubs desc;A >/* there are three other fields which i fill in correctly for an 7 >unaligned bit string but do not have right on me :) */ ! >desc.dsc$a_base = (char*)(&val);  >desc.dsc$l_pos = 0L;  >  >print_out(&num, &desc);( >char *temp = get_filename(&num, &desc); > 6 >/****************** PL/I Code **********************/ > > >get_filename: proc (indx, in_temparg) returns (char(80) var); >   >dcl indx         fixed bin(31); >dcl in_temparg   bit(1);  >dcl filename     char(80) var;  >  >put skip list (indx); >put skip list (in_temparg); >  >return (filename) >  >end get_filename; > / >/*------------------------------------------*/  > " >print_out: proc (param1, param2); >   >dcl param1       fixed bin(31); >dcl param2       bit(1);  >  >put skip list (param1); >put skip list (param2); >  >end print_out;  > E >/***************************** END CODE ***************************/  > C >The problem is this, the print_out procedure is a test stub I made C >which works fine, it prints out both the bit and the integer value F >correctly which are passed into it. You'll notice I do NOT modify the >structure in any way. > G >When I call the get_filename procedure though the addresses and values D >for both parameters are screwed and lead to access violation when I >try to print. > E >THEY ARE THE EXACT SAME PARAMETER LISTS!!! The only difference is in F >the names and the fact that one procedure returns something while theB >other does not. What the heck is going on here guys? Is this someF >obscure bug? Or a piece of the calling standard I am unfamiliar with? >How do I get around this? > D >p.s. I am compiling the C with all the compiler options for sharingF >code with other psects /EXTERN=COMMOM/nomember_alignment/noopt ... et >al. IF that helps at all? >  >Thanks again in advance! 
 >Jeff Cameron  >  >---' >Incoming mail is certified Virus Free. ; >Checked by AVG anti-virus system (http://www.grisoft.com). @ >Version: 6.0.370 / Virus Database: 205 - Release Date: 6/5/2002 >  --- & Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).? Version: 6.0.370 / Virus Database: 205 - Release Date: 6/5/2002    ------------------------------  # Date: Tue, 18 Jun 2002 15:09:39 GMT 1 From: "Terry C. Shannon" <terryshannon@attbi.com> 2 Subject: Re: A Proposal for All C.O.V. to Consider= Message-ID: <T2IP8.88650$6m5.74291@rwcrnsc51.ops.asp.att.net>   K There is a cheaper way to do this, and one that tends to embarrass the hell  out of marketeers.  H I went to VISTAPRINT.COM and got some business cards made up. They read:   VMS Marketing Volunteers, LTD. VMS: The Secure Business OS    Terry C. Shannon VMS Evangelist   address and contact info.   K Cost of cards: ten bucks. Look on face of marketeer or senior HPQ exec when * they are presented with a card: PRICELESS!    0 "John Vottero" <John@mvpsi.com> wrote in message) news:ugt87bieo4ue13@news.supernews.com... K > There is no way that I'm going to pay anything to get HP to tell me about  > it's products! > K > While I would love to see this happen, it should be HP asking customer to I > participate, not the other way around.  And it should be HP footing the  > bill.  > 0 > "John Smith" <a@nonymous.com> wrote in messageD > news:GAsP8.54078$831.18273@news01.bloor.is.net.cable.rogers.com...) > > A Proposal for All C.O.V. to Consider ) > > *************************************  > >  > >  > > The Thesis > > --------------- K > > Certainly for the past year, and in reality much longer than that, most  ofF > > us have been collectively less than pleased with the commitment to > marketing $ > > of OpenVMS by Compaq and now HP. > > H > > We seem to be able to get some measure of response from HP recently,	 > perhaps H > > even more than Compaq has given, but even so it is less complete andH > > forthright than we would like to hear. HP seems to be reticent aboutI > > discussing their plans for OpenVMS in the kind of language that would  giveK > > the OpenVMS community comfort, though in fairness to them they may feel  > that > > they are being frank.  > > H > > We have a wide variety of concerns that we as a community would likeL > > addressed, and by addressed I mean that we would like to hear from Carly
 > > about. > >  > >  > >  > >  > > The Proposal > > ------------------G > > Let's band together (Encompass, LUG's, c.o.v.) and host our own web G > > conference and invite Carly to join us for an hour to discuss where 	 > OpenVMS I > > is going, in her own words. But the key to the whole exercise is that  > Carly J > > personally participates. Stallard, Gorham, Marcello, et al. could join in > ifL > > they are available, but I for one wouldn't care to hear from Capellas or > > Winkler. > > H > > I think that creating an 'event' such as this would carry additional > weightD > > to our arguments with HP about why a serious marketing effort is	 important I > > to the future of VMS, even more than the individual email and letterse some) > > have written to the executives at HP.y > >  > >e > >  > >v > > How Do We Do Thise  > > ----------------------------H > > If we as a group decide to do something like this, we need to have a smalltK > > group of c.o.v. participants/OpenVMS users with sufficient 'clout' withr HPC > > to extend the formal invitation to Carly on behalf of the groupi
 > (Terry(??),pJ > > Encompass, some LUG's, some of the stock exchanges, major/not-so-majorH > > corporate customers, gov't agencies, ISV's, large integrators (SAIC, CSC,G > > etc..., educational institutions). I'd expect that we could have atJ leasteC > > 300 companies (with perhaps a total of 1200 decision-makers andq > influencers)J > > represented in the conference. It would be a marvelous opportunity for > CarlyMD > > to reach out and speak with a lot of real customers for an hour. > >rJ > > I know that this would represent some perceived risk on the part of HP andeH > > Carly in particular to agree to participate in a conference where HP > didn'tJ > > control the agenda. But I also think that this is a golden opportunity forlJ > > us to be proactive and for Carly to understand the strength of the VMS9 > > market if they'd only give us some public commitment.r > >r > >d > >s > >s
 > > Logisticsa
 > > ---------tL > > From a strictly technical point of view, I have recently participated in ao: > > number of web-based audio-visual conferences hosted by www.placeware.com, > asK > > part of a software beta I am participating in for a very large softwarenK > > vendor. The Placeware conferencing seems to be a very good way to get aD > lottL > > of people together rather inexpensively for such purposes. I'm sure that< > > there are other, just as capable services available too. > >bK > > The Placeware service (http://www.placeware.com/demos/demos.cfm) offers  > allcL > > the audio-visual goodies one would expect, and just as importantly in myH > > opinion, the services of a live moderator to maintain some order and flow > toJ > > the process. I think this would be useful to have in order to convince > CarlyiD > > that her participation would not have the conference turn into a  > > free-for-all at her expense. > >a > >V > >e > >a > > The Cost Details > > ----------------------8 > > The costs are approximately as follows (all in USD):< > > $50 per connection/location in North America (toll-free)= > > $1500 per hour for moderation services (+some other bits)nA > > $500 for a web deliverable audio/visual archive (.WMV format)  > >nK > > The per connection cost is for a single connection so to minimize costs. > you J > > can put a number of people from each firm participating in a boardroom > withH > > a PC connected to a large monitor/projector, and have a speakerphoneL > > available in the boardroom which all could share. Placeware can handle a1 > > conference with up to 2500 phone connections.p > >,> > > There are toll-free services that can be made available inL > > Europe/Australia/elsewhere if sufficient numbers of participants warrant > it, K > > but regular long-distance dial-in for the audio portion is available. IsK > > suspect that most of the regular non-North American c.o.v. participants  > forhG > > the most part have access to discounted long-distance services from: theirrK > > countries, so it could be a user pay for these companies (sorry to take, alJ > > decidedly North American view on this, but it simplifies things just a bite > > for discussion purposes).  > >uI > > So to work through a hypothetical conference of 200 connections wherel thesF > > North American users are calling toll-free and those outside North American6 > > are paying their own telephone connection, we get: > >y$ > > Connections: 200 x $50 = $10,000$ > > Moderator/tech support   $ 1,500 > > .WMV Archive:   $   500n > >m$ > > Approx. Gross Total      $12,000 > >P$ > > Per 'firm' = 12,000/200= $    60 > >PK > > It's a lot cheaper than a plane ticket and hotel, or even the taxi rideO toL > > the airport. And the cost per participant drops when you have 4-5 people > inH > > the room taking part. It would cost somewhat more if toll-free links wereE > > added outside of North America. Sorry, I don't have a hard figure6I > > available - a WAG might be double the price outside of North America,h butj > atG > > that price you'd probably be better off calling directly yourself).a > >o > >  > >f > > Cost Sharing > > ------------I > > Somebody has to go 'on-the-hook' to pay the bill for a web conferencea suchJ > > as this, ie. Placeware wants one place to send the invoice to. To that > end,H > > perhaps one c.o.v. participant who might have a Visa/Mastercard/AmexH > > merchant account could put the effort into creating a secure on-line	 > paymenthI > > and registration process into place to collect funds from those of usnK > > interested in participating, and then pay Placeware (or whichever other L > > service) for the conference. Sorry guys, I don't have a merchant Visa et > al.i > > account to do this.r > >nK > > I know that this would require some discussion with your own management  toI > > do something like this and it would also have to be able to deal withtI > > processing refunds (less some small handling/administrative amount toh > offsetK > > any Visa/Mastercard chargebacks) in the event that the conference can'tO be! > > done (Carly declines, etc..).r > >sJ > > Such a site should also be capable of collecting the names & titles of allmE > > those from an organization that would be participating in the webPJ > > conference, the list of which could be presented to HP - I wouldn't beK > > surprised if it would represent better intelligence about their OpenVMSs+ > > customers than they currently have now.- > >- > >  > >eJ > > Maybe I'm being too naive to think that we could get Carly to join us, but.L > > if there were sufficient numbers of large corporations involved, and youK > > could get your CIO/CTO/CFO/COO/CEO to sit in for an hour, then maybe wehJ > > could get HP/Carly to join us. Of course the flip side to this is that ifG > > she declines to participate, we all get to draw our own conclusions1 about  > > her commitment to OpenVMS. > > F > > The big question is: Is it better for her to standup in front of a > skepticalsK > > audience and make her case, or to be judged in absentia. From what I'veE > seenL > > of her, I'd bet she'd come to our party. Either way, we come away with a+ > > better sense of what to expect from HP.  > >s > >h > > Any thoughts?  > >- > >- >- >-   ------------------------------  # Date: Tue, 18 Jun 2002 15:56:11 GMTa# From: "John Smith" <a@nonymous.com> 2 Subject: Re: A Proposal for All C.O.V. to ConsiderJ Message-ID: <vKIP8.384011$t8_.309346@news01.bloor.is.net.cable.rogers.com>  < "Terry C. Shannon" <terryshannon@attbi.com> wrote in message7 news:T2IP8.88650$6m5.74291@rwcrnsc51.ops.asp.att.net...oH > There is a cheaper way to do this, and one that tends to embarrass the hell > out of marketeers. > J > I went to VISTAPRINT.COM and got some business cards made up. They read: >e  > VMS Marketing Volunteers, LTD. > VMS: The Secure Business OS  >s > Terry C. Shannon > VMS Evangelist >o > address and contact info.w > H > Cost of cards: ten bucks. Look on face of marketeer or senior HPQ exec when, > they are presented with a card: PRICELESS!    K The problem is that they are embarrassed about this only if they think thatgK you are doing something that they *should* be doing. Chances are they think A what they are doing/not doing is just the right amount of effort.a   ------------------------------  % Date: Tue, 18 Jun 2002 11:35:01 -0400 ! From: Jim Agnew <jpagnew@vcu.edu>a2 Subject: Re: A Proposal for All C.O.V. to Consider' Message-ID: <3D0F5325.87EB7AD3@vcu.edu>h   oooo.....  HARDBALL....t   Great, Terry!!!    "Terry C. Shannon" wrote:n > M > There is a cheaper way to do this, and one that tends to embarrass the hells > out of marketeers. > J > I went to VISTAPRINT.COM and got some business cards made up. They read: >   > VMS Marketing Volunteers, LTD. > VMS: The Secure Business OS- >  > Terry C. Shannon > VMS Evangelist >  > address and contact info.e > M > Cost of cards: ten bucks. Look on face of marketeer or senior HPQ exec whene, > they are presented with a card: PRICELESS! > 2 > "John Vottero" <John@mvpsi.com> wrote in message+ > news:ugt87bieo4ue13@news.supernews.com...rM > > There is no way that I'm going to pay anything to get HP to tell me about  > > it's products! > >rM > > While I would love to see this happen, it should be HP asking customer to K > > participate, not the other way around.  And it should be HP footing the 	 > > bill.l > >e2 > > "John Smith" <a@nonymous.com> wrote in messageF > > news:GAsP8.54078$831.18273@news01.bloor.is.net.cable.rogers.com...+ > > > A Proposal for All C.O.V. to Consider(+ > > > *************************************  > > >x > > >  > > > The Thesis > > > ---------------;M > > > Certainly for the past year, and in reality much longer than that, moste > ofH > > > us have been collectively less than pleased with the commitment to
 > > marketing-& > > > of OpenVMS by Compaq and now HP. > > > J > > > We seem to be able to get some measure of response from HP recently, > > perhapsuJ > > > even more than Compaq has given, but even so it is less complete andJ > > > forthright than we would like to hear. HP seems to be reticent aboutK > > > discussing their plans for OpenVMS in the kind of language that woulda > giveM > > > the OpenVMS community comfort, though in fairness to them they may feel  > > that > > > they are being frank.c > > > J > > > We have a wide variety of concerns that we as a community would likeN > > > addressed, and by addressed I mean that we would like to hear from Carly > > > about. > > >  > > >! > > >l > > >r > > > The Proposal > > > ------------------I > > > Let's band together (Encompass, LUG's, c.o.v.) and host our own webgI > > > conference and invite Carly to join us for an hour to discuss wherer > > OpenVMSlK > > > is going, in her own words. But the key to the whole exercise is thatt	 > > Carly L > > > personally participates. Stallard, Gorham, Marcello, et al. could join > in > > ifN > > > they are available, but I for one wouldn't care to hear from Capellas or > > > Winkler. > > >-J > > > I think that creating an 'event' such as this would carry additional
 > > weightF > > > to our arguments with HP about why a serious marketing effort is > important K > > > to the future of VMS, even more than the individual email and letterst > some+ > > > have written to the executives at HP.a > > >  > > >: > > >y > > >n > > > How Do We Do This." > > > ----------------------------J > > > If we as a group decide to do something like this, we need to have a > small M > > > group of c.o.v. participants/OpenVMS users with sufficient 'clout' with* > HPE > > > to extend the formal invitation to Carly on behalf of the groupi > > (Terry(??),eL > > > Encompass, some LUG's, some of the stock exchanges, major/not-so-majorJ > > > corporate customers, gov't agencies, ISV's, large integrators (SAIC, > CSC,I > > > etc..., educational institutions). I'd expect that we could have ath > least E > > > 300 companies (with perhaps a total of 1200 decision-makers andn > > influencers)L > > > represented in the conference. It would be a marvelous opportunity for	 > > Carly,F > > > to reach out and speak with a lot of real customers for an hour. > > >/L > > > I know that this would represent some perceived risk on the part of HP > andhJ > > > Carly in particular to agree to participate in a conference where HP
 > > didn'tL > > > control the agenda. But I also think that this is a golden opportunity > for L > > > us to be proactive and for Carly to understand the strength of the VMS; > > > market if they'd only give us some public commitment.  > > >l > > >  > > >i > > >( > > > Logistics  > > > --------- N > > > From a strictly technical point of view, I have recently participated in > a*< > > > number of web-based audio-visual conferences hosted by > www.placeware.com, > > asM > > > part of a software beta I am participating in for a very large softwaretM > > > vendor. The Placeware conferencing seems to be a very good way to get al > > lotIN > > > of people together rather inexpensively for such purposes. I'm sure that> > > > there are other, just as capable services available too. > > >wM > > > The Placeware service (http://www.placeware.com/demos/demos.cfm) offers  > > allMN > > > the audio-visual goodies one would expect, and just as importantly in myJ > > > opinion, the services of a live moderator to maintain some order and > flow > > toL > > > the process. I think this would be useful to have in order to convince	 > > Carly F > > > that her participation would not have the conference turn into a" > > > free-for-all at her expense. > > >t > > >R > > >M > > >e > > > The Cost Details > > > ----------------------: > > > The costs are approximately as follows (all in USD):> > > > $50 per connection/location in North America (toll-free)? > > > $1500 per hour for moderation services (+some other bits)aC > > > $500 for a web deliverable audio/visual archive (.WMV format)i > > >rM > > > The per connection cost is for a single connection so to minimize costs6 > > youVL > > > can put a number of people from each firm participating in a boardroom > > withJ > > > a PC connected to a large monitor/projector, and have a speakerphoneN > > > available in the boardroom which all could share. Placeware can handle a3 > > > conference with up to 2500 phone connections.s > > >p@ > > > There are toll-free services that can be made available inN > > > Europe/Australia/elsewhere if sufficient numbers of participants warrant > > it,:M > > > but regular long-distance dial-in for the audio portion is available. IaM > > > suspect that most of the regular non-North American c.o.v. participants  > > foraI > > > the most part have access to discounted long-distance services fromR > their M > > > countries, so it could be a user pay for these companies (sorry to takes > aeL > > > decidedly North American view on this, but it simplifies things just a > bitu > > > for discussion purposes).  > > > K > > > So to work through a hypothetical conference of 200 connections wheren > theoH > > > North American users are calling toll-free and those outside North	 > Americao8 > > > are paying their own telephone connection, we get: > > >.& > > > Connections: 200 x $50 = $10,000& > > > Moderator/tech support   $ 1,500 > > > .WMV Archive:   $   500* > > > & > > > Approx. Gross Total      $12,000 > > >-& > > > Per 'firm' = 12,000/200= $    60 > > >lM > > > It's a lot cheaper than a plane ticket and hotel, or even the taxi ridel > toN > > > the airport. And the cost per participant drops when you have 4-5 people > > inJ > > > the room taking part. It would cost somewhat more if toll-free links > wereG > > > added outside of North America. Sorry, I don't have a hard figure K > > > available - a WAG might be double the price outside of North America,c > butt > > atI > > > that price you'd probably be better off calling directly yourself).o > > >  > > >, > > >  > > > Cost Sharing > > > ------------K > > > Somebody has to go 'on-the-hook' to pay the bill for a web conferencet > suchL > > > as this, ie. Placeware wants one place to send the invoice to. To that > > end,J > > > perhaps one c.o.v. participant who might have a Visa/Mastercard/AmexJ > > > merchant account could put the effort into creating a secure on-line > > payment K > > > and registration process into place to collect funds from those of us M > > > interested in participating, and then pay Placeware (or whichever othertN > > > service) for the conference. Sorry guys, I don't have a merchant Visa et > > al.n > > > account to do this.a > > > M > > > I know that this would require some discussion with your own management  > toK > > > do something like this and it would also have to be able to deal withoK > > > processing refunds (less some small handling/administrative amount to 
 > > offsetM > > > any Visa/Mastercard chargebacks) in the event that the conference can'ti > be# > > > done (Carly declines, etc..).  > > > L > > > Such a site should also be capable of collecting the names & titles of > allcG > > > those from an organization that would be participating in the weboL > > > conference, the list of which could be presented to HP - I wouldn't beM > > > surprised if it would represent better intelligence about their OpenVMSn- > > > customers than they currently have now.s > > >/ > > >m > > > L > > > Maybe I'm being too naive to think that we could get Carly to join us, > but.N > > > if there were sufficient numbers of large corporations involved, and youM > > > could get your CIO/CTO/CFO/COO/CEO to sit in for an hour, then maybe we L > > > could get HP/Carly to join us. Of course the flip side to this is that > ifI > > > she declines to participate, we all get to draw our own conclusionsr > aboutn  > > > her commitment to OpenVMS. > > >eH > > > The big question is: Is it better for her to standup in front of a
 > > skepticaleM > > > audience and make her case, or to be judged in absentia. From what I've  > > seenN > > > of her, I'd bet she'd come to our party. Either way, we come away with a- > > > better sense of what to expect from HP.  > > >b > > >m > > > Any thoughts?  > > >  > > >  > >t > >    ------------------------------  % Date: Tue, 18 Jun 2002 12:33:28 -0400y* From: WILLIAM WEBB <WWEBB1@email.usps.gov>2 Subject: RE: A Proposal for All C.O.V. to Consider- Message-ID: <0033000068530824000002L042*@MHS>r  7 =0AOnce upon a time I saw a Microsoft job posting for al) position entitled "Technical Evangelist".d   WWWebb   -----Original Message-----/ From: Info-VAX-Request@Mvb.Saic.Com at INTERNETm% Sent: Tuesday, June 18, 2002 11:56 AMsB To: Webb, William W Raleigh, NC; Info-VAX@Mvb.Saic.Com at INTERNET2 Subject: RE: A Proposal for All C.O.V. to Consider    < "Terry C. Shannon" <terryshannon@attbi.com> wrote in message7 news:T2IP8.88650$6m5.74291@rwcrnsc51.ops.asp.att.net...eH > There is a cheaper way to do this, and one that tends to embarrass th= eo hell > out of marketeers. >tH > I went to VISTAPRINT.COM and got some business cards made up. They re= ad:t >   > VMS Marketing Volunteers, LTD. > VMS: The Secure Business OS  >  > Terry C. Shannon > VMS Evangelist >- > address and contact info.m >lH > Cost of cards: ten bucks. Look on face of marketeer or senior HPQ exe= c( when, > they are presented with a card: PRICELESS!    H The problem is that they are embarrassed about this only if they think = thatH you are doing something that they *should* be doing. Chances are they t= hinkB what they are doing/not doing is just the right amount of effort.=   ------------------------------  % Date: Tue, 18 Jun 2002 12:44:41 -0500r1 From: "Dave Gudewicz" <david.gudewicz@abbott.com>t2 Subject: Re: A Proposal for All C.O.V. to Consider1 Message-ID: <aenrl3$abp$1@fizban.pprd.abbott.com>t  H Maybe we should throw in a "can you hear me now" for good measure.  I do like the PRICELESS thing btw.e  L I tuned in to the server web cast today and VMS was mentioned several times,G but when I asked an on-line question about its absense in the first fewsG slides, I got the "VMS is strategic, but not a growth platform" answer. K What would make it a growth platform you might ask?  Lots of sales followed  by more of the same I'd guess. -- Dave...e  3 More than one cigar at a time is excessive smoking.r -----Mark Twain,  < "Terry C. Shannon" <terryshannon@attbi.com> wrote in message7 news:T2IP8.88650$6m5.74291@rwcrnsc51.ops.asp.att.net...rH > There is a cheaper way to do this, and one that tends to embarrass the hell > out of marketeers. >rJ > I went to VISTAPRINT.COM and got some business cards made up. They read: >   > VMS Marketing Volunteers, LTD. > VMS: The Secure Business OS1 >0 > Terry C. Shannon > VMS Evangelist >  > address and contact info.0 > H > Cost of cards: ten bucks. Look on face of marketeer or senior HPQ exec when, > they are presented with a card: PRICELESS! >o > 2 > "John Vottero" <John@mvpsi.com> wrote in message+ > news:ugt87bieo4ue13@news.supernews.com...kG > > There is no way that I'm going to pay anything to get HP to tell me  about  > > it's products! > >IJ > > While I would love to see this happen, it should be HP asking customer toK > > participate, not the other way around.  And it should be HP footing thef	 > > bill.i > >y2 > > "John Smith" <a@nonymous.com> wrote in messageF > > news:GAsP8.54078$831.18273@news01.bloor.is.net.cable.rogers.com...+ > > > A Proposal for All C.O.V. to Considera+ > > > *************************************. > > >t > > >, > > > The Thesis > > > ---------------mH > > > Certainly for the past year, and in reality much longer than that, most > ofH > > > us have been collectively less than pleased with the commitment to
 > > marketings& > > > of OpenVMS by Compaq and now HP. > > >tJ > > > We seem to be able to get some measure of response from HP recently, > > perhapsgJ > > > even more than Compaq has given, but even so it is less complete andJ > > > forthright than we would like to hear. HP seems to be reticent aboutK > > > discussing their plans for OpenVMS in the kind of language that wouldw > giveH > > > the OpenVMS community comfort, though in fairness to them they may feel > > that > > > they are being frank.) > > > J > > > We have a wide variety of concerns that we as a community would likeH > > > addressed, and by addressed I mean that we would like to hear from Carlyl > > > about. > > >  > > >a > > >a > > >  > > > The Proposal > > > ------------------I > > > Let's band together (Encompass, LUG's, c.o.v.) and host our own webfI > > > conference and invite Carly to join us for an hour to discuss where  > > OpenVMSrK > > > is going, in her own words. But the key to the whole exercise is thata	 > > CarlygL > > > personally participates. Stallard, Gorham, Marcello, et al. could join > in > > ifK > > > they are available, but I for one wouldn't care to hear from Capellash or > > > Winkler. > > > J > > > I think that creating an 'event' such as this would carry additional
 > > weightF > > > to our arguments with HP about why a serious marketing effort is > importantnK > > > to the future of VMS, even more than the individual email and letters  > some+ > > > have written to the executives at HP.r > > >j > > >n > > >i > > >  > > > How Do We Do Thisf" > > > ----------------------------J > > > If we as a group decide to do something like this, we need to have a > smallsH > > > group of c.o.v. participants/OpenVMS users with sufficient 'clout' with > HPE > > > to extend the formal invitation to Carly on behalf of the group  > > (Terry(??),PL > > > Encompass, some LUG's, some of the stock exchanges, major/not-so-majorJ > > > corporate customers, gov't agencies, ISV's, large integrators (SAIC, > CSC,I > > > etc..., educational institutions). I'd expect that we could have ath > leastcE > > > 300 companies (with perhaps a total of 1200 decision-makers and  > > influencers)L > > > represented in the conference. It would be a marvelous opportunity for	 > > CarlyaF > > > to reach out and speak with a lot of real customers for an hour. > > >nL > > > I know that this would represent some perceived risk on the part of HP > and J > > > Carly in particular to agree to participate in a conference where HP
 > > didn'tL > > > control the agenda. But I also think that this is a golden opportunity > fortL > > > us to be proactive and for Carly to understand the strength of the VMS; > > > market if they'd only give us some public commitment.o > > >t > > >- > > >- > > >- > > > Logisticsa > > > ---------0K > > > From a strictly technical point of view, I have recently participateda in > a < > > > number of web-based audio-visual conferences hosted by > www.placeware.com, > > asD > > > part of a software beta I am participating in for a very large softwareK > > > vendor. The Placeware conferencing seems to be a very good way to get  a  > > lotII > > > of people together rather inexpensively for such purposes. I'm surek that> > > > there are other, just as capable services available too. > > >nF > > > The Placeware service (http://www.placeware.com/demos/demos.cfm) offers > > allkK > > > the audio-visual goodies one would expect, and just as importantly ina myJ > > > opinion, the services of a live moderator to maintain some order and > flow > > toL > > > the process. I think this would be useful to have in order to convince	 > > CarlylF > > > that her participation would not have the conference turn into a" > > > free-for-all at her expense. > > >  > > >a > > >t > > >t > > > The Cost Details > > > ----------------------: > > > The costs are approximately as follows (all in USD):> > > > $50 per connection/location in North America (toll-free)? > > > $1500 per hour for moderation services (+some other bits) C > > > $500 for a web deliverable audio/visual archive (.WMV format)h > > >sG > > > The per connection cost is for a single connection so to minimizet costsl > > younL > > > can put a number of people from each firm participating in a boardroom > > withJ > > > a PC connected to a large monitor/projector, and have a speakerphoneL > > > available in the boardroom which all could share. Placeware can handle au3 > > > conference with up to 2500 phone connections.i > > >s@ > > > There are toll-free services that can be made available inF > > > Europe/Australia/elsewhere if sufficient numbers of participants warrant  > > it,tK > > > but regular long-distance dial-in for the audio portion is available.o Ie@ > > > suspect that most of the regular non-North American c.o.v. participants > > forrI > > > the most part have access to discounted long-distance services from  > their H > > > countries, so it could be a user pay for these companies (sorry to take > a L > > > decidedly North American view on this, but it simplifies things just a > bita > > > for discussion purposes).w > > > K > > > So to work through a hypothetical conference of 200 connections where  > theMH > > > North American users are calling toll-free and those outside North	 > America 8 > > > are paying their own telephone connection, we get: > > >,& > > > Connections: 200 x $50 = $10,000& > > > Moderator/tech support   $ 1,500 > > > .WMV Archive:   $   500r > > > & > > > Approx. Gross Total      $12,000 > > >n& > > > Per 'firm' = 12,000/200= $    60 > > > H > > > It's a lot cheaper than a plane ticket and hotel, or even the taxi ride > toG > > > the airport. And the cost per participant drops when you have 4-5  people > > inJ > > > the room taking part. It would cost somewhat more if toll-free links > wereG > > > added outside of North America. Sorry, I don't have a hard figure K > > > available - a WAG might be double the price outside of North America,  > but  > > atI > > > that price you'd probably be better off calling directly yourself).e > > >h > > >l > > >t > > > Cost Sharing > > > ------------K > > > Somebody has to go 'on-the-hook' to pay the bill for a web conferenceo > suchL > > > as this, ie. Placeware wants one place to send the invoice to. To that   > > end,J > > > perhaps one c.o.v. participant who might have a Visa/Mastercard/AmexJ > > > merchant account could put the effort into creating a secure on-line > > paymentnK > > > and registration process into place to collect funds from those of usrG > > > interested in participating, and then pay Placeware (or whicheveru other K > > > service) for the conference. Sorry guys, I don't have a merchant Visai et > > al.t > > > account to do this.  > > >iB > > > I know that this would require some discussion with your own
 management > toK > > > do something like this and it would also have to be able to deal withtK > > > processing refunds (less some small handling/administrative amount tok
 > > offsetG > > > any Visa/Mastercard chargebacks) in the event that the conference  can'ti > be# > > > done (Carly declines, etc..).c > > >hL > > > Such a site should also be capable of collecting the names & titles of > all G > > > those from an organization that would be participating in the webaL > > > conference, the list of which could be presented to HP - I wouldn't beE > > > surprised if it would represent better intelligence about theirl OpenVMSI- > > > customers than they currently have now.  > > >h > > >e > > >sL > > > Maybe I'm being too naive to think that we could get Carly to join us, > but J > > > if there were sufficient numbers of large corporations involved, and youlJ > > > could get your CIO/CTO/CFO/COO/CEO to sit in for an hour, then maybe weL > > > could get HP/Carly to join us. Of course the flip side to this is that > ifI > > > she declines to participate, we all get to draw our own conclusionsr > aboutn  > > > her commitment to OpenVMS. > > >iH > > > The big question is: Is it better for her to standup in front of a
 > > skeptical H > > > audience and make her case, or to be judged in absentia. From what I've > > seenL > > > of her, I'd bet she'd come to our party. Either way, we come away with a - > > > better sense of what to expect from HP.o > > >m > > >o > > > Any thoughts?e > > >a > > >  > >f > >w >d >v   ------------------------------  + Date: Tue, 18 Jun 2002 07:06:00 -0500 (CDT) ? From: "Doc.Cypher" <Use-Author-Supplied-Address-Header@[127.1]> , Subject: Re: A [departure from] VMS Disaster< Message-ID: <200206181206.g5IC60nX083992@cryptofortress.com>  J On Mon, 17 Jun 2002, "Brian Tillman" <tillman_brian@notnoone.notnohow.com> wrote:4 >John Eisenschmidt (jweisen@eisenschmidt.org) wrote: >n
 ><...snip...>5 >pF >What I want to know is why do your messages all show up as blank with
 >attachments?l  F It's the mime encoding. Which application would you like to blame? :-)  + John's software for posting mime to Usenet.s  . Or yours for not showing Usenet as plain text.     Doc. -- c6 The bigger the humbug, the better people will like it.K ~ Phineas Taylor Barnum.                             https://vmsbox.cjb.neta   ------------------------------   Date: 18 Jun 2002 16:03:23 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)   Subject: Re: Affinity and Itanic, Message-ID: <aenlkb$24rn$2@info.cs.uofs.edu>  , In article <3D0BDD03.522A5B3A@videotron.ca>,0  JF Mezei <jfmezei.spamnot@videotron.ca> writes: |>P |> Consider the VAX at the end of the 1980s. Had Digital priced and marketed theP |> "slower" VAXes very competitively against the new RISC kids on the block,  it; |> would have been able to do to Sun what Sun did to Alpha.  |> a  H Consider that while I still run VAXen in production (as well as at home)J I have a stack of old Sparcstations piled in the corner that are not worthG the effort of trying to use.  Last thing run ont hem was Amoeba.  TheseeI machines were contemporaries and the VAX could easily have held it's own.p   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   r   ------------------------------  % Date: Tue, 18 Jun 2002 13:36:48 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca> ! Subject: Re: ALL-IN-1 anniversary-, Message-ID: <3D0F6FAE.B68C42C5@videotron.ca>   Bob Knowles wrote:F >             If you think that you know who this might be then please > complete the form atL > http://www.openvms.compaq.com/commercial/a1_anniversary.html The first 200M > replies received by 13th July will be sent a commemorative item, regardlesse5 > of whether they qualify as longest-standing or not.e    L Well, I wish I could fill out your form, but you only allow those sites thatJ started up to 1985 to fill the form. I doubt you have 200 customers dating* back that far worldwide that still use A1.   ------------------------------  # Date: Tue, 18 Jun 2002 07:10:01 GMTh4 From: "Jonathan Ridler" <jonathan.ridler@compaq.com> Subject: Re: ANN: Updated JUMP0 Message-ID: <d1BP8.1$Va1.94301@news.cpqcorp.net>  G JUMP can do something similar to SETUSER (with various qualifiers), butm= the most common way to use JUMP is in EXACT mode which uses a H pseudo-terminal actually logged in as the user (without the need for theB password) -- this is NOT an impersonation: it is EXACTLY the user.  D I suggest you download the kit and read the documentation -- JUMP isE very powerful and flexible, with good security and auditing features. 2 I am not aware of it causing any systems to crash.  	 Jonathan.o  > "Michael Austin" <maustin@firstdbasource.com> wrote in message, news:3D0E5898.799936E4@firstdbasource.com... > Hunter Goatley wrote:b > > ? > > JUMP V4.1 has been released by its author, Jonathan Ridler.u > > L > > JUMP allows a user to login exactly  as another user without a password.L > > It  also  allows a more restricted (non-exact) impersonation of  anotherJ > > user. Use of JUMP  is  restricted  to  specific  categories  of users:I > > Systems Programmers, Operators and any specifically authorised users.c > >nK > > You can find out more about JUMP by downloading the JUMP.ZIP file from:n > >I# > > http://www.process.com/openvms/n > >a8 > > ftp://ftp.process.com/vms-freeware/fileserv/jump.zip= > > http://vms.process.com/ftp/vms-freeware/fileserv/jump.zip 4 > > ftp://ftp.tmk.com/vms-freeware/fileserv/jump.zip9 > > http://www.tmk.com/ftp/vms-freeware/fileserv/jump.zip  > >b2 > > and on the other mirrors in the next 24 hours. > > 
 > > Hunter
 > > ------= > > Hunter Goatley, Process Software, http://www.process.com/L< > > goathunter@goatley.com    http://www.goatley.com/hunter/@ > > New Robert R. McCammon site: http://www.RobertRMcCammon.com/ >rF > Just curious, but what is the difference in this vs. SETUSER? (AlphaH > version changes all logicals including sys$login and by resetting yourI > privs (set proc/priv) to the users privs can "impersonate" that user as0 > well.A > C > Really nice when trying to find out why something works with youruI > account and not user abc.  Just make sure you use the command procedureeJ > to link it or use the exact commands in link.com -- the result of havingJ > undefined symbols will cause the system to crash in Kernel mode -- not a > very pleasant experience...  >  > --
 > Regards, > 9 > Michael Austin            Registered Linux User #261163p9 > First DBA Source, Inc.    http://www.firstdbasource.comM > Sr. Consultant > 704-947-1089 (Office)S > 704-236-4377 (Mobile)  >    ------------------------------  + Date: Tue, 18 Jun 2002 13:43:56 +0000 (UTC)k- From: lewis@spyder.mitre.org (Keith A. Lewis)p6 Subject: Re: Another C and PL/I question - Descriptors. Message-ID: <aendes$567$1@newslocal.mitre.org>   "Jeffrey Cameron" <bubbapig@hotmail.com> writes in article <XStP8.1484$Rf7.529909@news20.bellglobal.com> dated Mon, 17 Jun 2002 19:01:08 -0400:tL >1. My C++ modules must interact with certain PL/I code and quite often thseG >PL/I modules either return or take as a parameter a bit (usually as anR' >unaligned bit string), for example ...  >s3 >/***************** PL/I CODE ********************/b8 >get_filename: proc (indx, in_temparg) returns (bit(1)); >. >dcl in_temparg bit(1);S >dcl indx fixed bin(15); >  >/* ... snip ... */o >  >end get_filename; >e >aM >But how do I declare this in C so that it knows to pass in and return a bit,eM >there are obviously no atomic data types for this operation in C/C++. I havehK >found a lot of litaerature on passing by descriptor and have tried passingyL >in the address of an unaligned bit string descriptor to the PL/I program to >no avail. e.g.:  J I would *guess* that PL/I passes the bit(1) parameter by reference and theH return value by value.  You could experiment to determine if my guess is& right or if somebody else's is, but...  J I'm assuming you can't modify the existing PL/I code for some reason.  CanI you write more?  Write a wrapper function which uses parameter types that & integrate easily with C.  For example:  @ get_filename_c: proc (indx, in_temparg) returns (fixed bin (7));   dcl indx fixed bin(15);t# dcl in_temparg fixed bin (7) value;    dcl final_temparg	bit; dcl result		bit;  ! final_temparg = (in_temparg = 1); + result = get_filename(indx, final_temparg);t if result then e     return (1);@ else p     return (0);    end;  2 /***************** C CODE ***********************/, extern "C" char get_filename(short *, char);  + --Keith Lewis              klewis$mitre.orgu> The above may not (yet) represent the opinions of my employer.   ------------------------------  % Date: Tue, 18 Jun 2002 06:44:19 -0700t# From: "Tom Linden" <tom@kednos.com> 6 Subject: RE: Another C and PL/I question - Descriptors9 Message-ID: <CIEJLCMNHNNDLLOOGNJIOEFAFDAA.tom@kednos.com>a  C I posted a reply to this yesterday but it never made it to the listhA for some reason.  I think the wrapper suggestion is workable, but,A it should work directly.  I don't know about C++, but C certainlyeF tries to align things unless you tell it not to, owing to a deficiency in the alpha architecture.  C Unaligned bit strings are passed as descriptors.  See section 7.8.3e@ in the reference manual, which if you don't have, you may freely? download from freja.kednos.com.  Sounds like you need a supportt- agreement.  Mail me offline if you need help.i   Tomo   >-----Original Message-----s5 >From: Keith A. Lewis [mailto:lewis@spyder.mitre.org]e% >Sent: Tuesday, June 18, 2002 6:44 AMt >To: Info-VAX@Mvb.Saic.Com7 >Subject: Re: Another C and PL/I question - Descriptors  >  > ; >"Jeffrey Cameron" <bubbapig@hotmail.com> writes in articleo@ ><XStP8.1484$Rf7.529909@news20.bellglobal.com> dated Mon, 17 Jun >2002 19:01:08 -0400:lB >>1. My C++ modules must interact with certain PL/I code and quite >often thse-H >>PL/I modules either return or take as a parameter a bit (usually as an( >>unaligned bit string), for example ... >>4 >>/***************** PL/I CODE ********************/9 >>get_filename: proc (indx, in_temparg) returns (bit(1));o >> >>dcl in_temparg bit(1); >>dcl indx fixed bin(15);i >> >>/* ... snip ... */ >> >>end get_filename;n >> >>@ >>But how do I declare this in C so that it knows to pass in and >return a bit,@ >>there are obviously no atomic data types for this operation in >C/C++. I haveL >>found a lot of litaerature on passing by descriptor and have tried passingB >>in the address of an unaligned bit string descriptor to the PL/I >program tor >>no avail. e.g.:a >mK >I would *guess* that PL/I passes the bit(1) parameter by reference and thejI >return value by value.  You could experiment to determine if my guess is-' >right or if somebody else's is, but...  >dK >I'm assuming you can't modify the existing PL/I code for some reason.  CantJ >you write more?  Write a wrapper function which uses parameter types that' >integrate easily with C.  For example:o >hA >get_filename_c: proc (indx, in_temparg) returns (fixed bin (7));e >o >dcl indx fixed bin(15);$ >dcl in_temparg fixed bin (7) value; >a >dcl final_temparg	bit;e >dcl result		bit;S > " >final_temparg = (in_temparg = 1);, >result = get_filename(indx, final_temparg); >if result thenp >    return (1); >elsea >    return (0); >k >end;  > 3 >/***************** C CODE ***********************/I- >extern "C" char get_filename(short *, char);l > , >--Keith Lewis              klewis$mitre.org? >The above may not (yet) represent the opinions of my employer.p >e >---' >Incoming mail is certified Virus Free. ; >Checked by AVG anti-virus system (http://www.grisoft.com).e@ >Version: 6.0.370 / Virus Database: 205 - Release Date: 6/5/2002 >k --- & Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).? Version: 6.0.370 / Virus Database: 205 - Release Date: 6/5/2002o   ------------------------------  # Date: Tue, 18 Jun 2002 13:31:33 GMT-+ From: "Rick Barry" <barry@star.zko.dec.com>f* Subject: Re: Apache bug affect VMS Apache?2 Message-ID: <VCGP8.20$wi1.263291@news.cpqcorp.net>  F We're investigating the problem and will release a patch if there is a security% vulnerability with Apache on OpenVMS.   
 Rick Barry) Compaq Secure Web Server Development Teamr OpenVMS Systems Software Group Hewlett Packard Companyr
 Nashua, NH  5 "Bob Ceculski" <bob@instantwhip.com> wrote in message 6 news:d7791aa1.0206171906.25524c8@posting.google.com...A > Does this affect VMS and will there be patches for prior Apacheo > versions?s >  >l > June 17, 2002i" > Flaw Found in Apache HTTP Server > By  Dennis Fishery >aH > A buffer overrun vulnerability in the Apache HTTP server included withA > many popular Web servers enables an attacker to execute code one > vulnerable machines. >vH > The flaw lies in the way that the server handles data transmissions ofG > unknown size. Typically, these transmissions are broken into "chunks"tF > for easier handling. But Apache's HTTP server misinterprets the sizeH > of the chunks, which leads to an overrun of the heap memory, accordingE > to an advisory published Monday by Internet Security Systems Inc.'sr > X-Force research team. >iC > The vulnerability can be exploited remotely by way of a carefullymH > crafted invalid request to the server, and the flawed functionality isG > enabled by default. Exploiting the flaw could either lead to a denialh? > of service on the machine or the execution of malicious code.d > ? > The Apache Software Foundation's Apache Server Project, whichlG > maintains the open-source HTTP server, also issued a bulletin warning B > that all versions of Apache 1.3 are vulnerable, as are copies of > version 2 up to 2.0.39.  > B > However, versions 2.0 and later are not vulnerable to the remote, > execution of code, Apache's advisory said. > G > Apache is by far the most widely deployed Web server on the Internet, G > running on more than 50 percent of machines surveyed by Netcraft Ltd.tC > Its HTTP server is included in products such as Oracle Corp.'s 9ig/ > Application Server and IBM Corp.'s WebSphere.e >sE > The Foundation said it was forced to release its advisory early-and D > without an available patch-because of ISS' decision to publish its > bulletin.    ------------------------------  % Date: Tue, 18 Jun 2002 14:11:20 +0000i2 From: John Eisenschmidt <jweisen@eisenschmidt.org>* Subject: Re: Apache bug affect VMS Apache?4 Message-ID: <20020618141120.B16340@eisenschmidt.org>  X ::silently praying this doesn't start a "why Apache should be written in Bliss thread"::   Rick,e  R Glad to see you guys are on top of this. We look forward to hearing back from you.  J Unless the Voices are Mistaken, Rick Barry (barry@star.zko.dec.com) Wrote:H > We're investigating the problem and will release a patch if there is a
 > security' > vulnerability with Apache on OpenVMS.t >  > Rick Barry+ > Compaq Secure Web Server Development Teamm  > OpenVMS Systems Software Group > Hewlett Packard Company  > Nashua, NH > 7 > "Bob Ceculski" <bob@instantwhip.com> wrote in messageg8 > news:d7791aa1.0206171906.25524c8@posting.google.com...C > > Does this affect VMS and will there be patches for prior Apachea
 > > versions?f > >l > >i > > June 17, 2002 $ > > Flaw Found in Apache HTTP Server > > By  Dennis Fisher- > > J > > A buffer overrun vulnerability in the Apache HTTP server included withC > > many popular Web servers enables an attacker to execute code one > > vulnerable machines. > > J > > The flaw lies in the way that the server handles data transmissions ofI > > unknown size. Typically, these transmissions are broken into "chunks"aH > > for easier handling. But Apache's HTTP server misinterprets the sizeJ > > of the chunks, which leads to an overrun of the heap memory, accordingG > > to an advisory published Monday by Internet Security Systems Inc.'sh > > X-Force research team. > >tE > > The vulnerability can be exploited remotely by way of a carefullyiJ > > crafted invalid request to the server, and the flawed functionality isI > > enabled by default. Exploiting the flaw could either lead to a denialnA > > of service on the machine or the execution of malicious code.  > > A > > The Apache Software Foundation's Apache Server Project, which I > > maintains the open-source HTTP server, also issued a bulletin warningtD > > that all versions of Apache 1.3 are vulnerable, as are copies of > > version 2 up to 2.0.39.i > >aD > > However, versions 2.0 and later are not vulnerable to the remote. > > execution of code, Apache's advisory said. > >uI > > Apache is by far the most widely deployed Web server on the Internet, I > > running on more than 50 percent of machines surveyed by Netcraft Ltd. E > > Its HTTP server is included in products such as Oracle Corp.'s 9i 1 > > Application Server and IBM Corp.'s WebSphere.a > >oG > > The Foundation said it was forced to release its advisory early-andtF > > without an available patch-because of ISS' decision to publish its
 > > bulletin.e >    -- t/ John W. Eisenschmidt <jweisen@eisenschmidt.org>c6  Homepage URL    | http://www.eisenschmidt.org/jweisenM  PGP Public Key  | http://www.eisenschmidt.org/jweisen/misc/jeisenschmidt.asctD  PGP Fingerprint | 5F9B F916 5AD1 3295 CF99 BC1E 1F97 E6A3 37E3 BEF2  O FOO MANE PADME HUM: "Our first obligation is to keep the FOO counters turning."h   ------------------------------  % Date: Tue, 18 Jun 2002 12:36:58 -0400 4 From: wpg_michael@esonic.e-sonic.com (Michael Young)9 Subject: can VMS mail block e-mail from certian address'? / Message-ID: <02061812365835@esonic.e-sonic.com>,   Hello experts,  J Can someone walk me through setting up my e-mail acct to block e-mail fromI certian e-mail address'?  I'm using the MAIL program that comes with VMS.0  M Sorry if I haven't given enough info to help you show me where to start.  I'ml new to this.   Regards,
 Michael Young    Electro Sonic, Winnipeg@" Phone (204)783-3105, (800)665-1358 Fax   (204)774-7288 % e-mail:wpg_michael@esonic.e-sonic.coml   ------------------------------    Date: 18 Jun 2002 07:56:22 -07008 From: xavier.nicolovici@citicorp.com (Xavier Nicolovici)) Subject: CONNECT:Direct wildcard transfers= Message-ID: <e033a250.0206180656.768a3fc0@posting.google.com>g   Hi,h  E I'm currently trying to set up a file transfer between a VMS platform @ and an NT one. We have concluded to use, for various reason, the+ CONNECT:Direct tool from Sterling Commerce.   D I would like to transfer a set of file from this VMS platform to theD NT one, but the only information I have about those file is the fileD name pattern. A random generated number is added to the filename, inC order to avoid file overwritting (the file are created from time tod time).  E Now, CONNECT:Direct can not perform wildcard file transfer, and I waseC wondering, for this simple task, if someone has already developed anE small script that loops in a particular VMS area, and search for each : existing file to perform a call to the CONNECT:Dircet API.   Many thanks in advance,r   Xavier Nicolovicih   ------------------------------  # Date: Tue, 18 Jun 2002 10:10:35 GMTe From: system@SendSpamHere.ORGa# Subject: Re: Copying a file via FIDs0 Message-ID: <00A0FA17.52F859F8@SendSpamHere.ORG>  a In article <cflsgu0g6hmmlmd5asug97im1b11hahpjb@4ax.com>, jlsue <jlsuexxxz@screaminet.com> writes:h> >Well, Eigen Corp. had Eigen tools - a collect of lots of neatD >utilities for system and application troubleshooting.  One of these+ >was FIDENTER that would do what you want.S4   http://www.nserver.com/eigen/.   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM2            c5   "Well my son, life is like a beanstalk, isn't it?" a   ------------------------------  % Date: Tue, 18 Jun 2002 08:14:52 -0400w2 From: Atlant Schmidt <atlantnospam@mindspring.com>$ Subject: Re: Could linux become VMS?. Message-ID: <3D0F243C.663BDE9E@mindspring.com>   Frank Sapienza wrote:e  J > You're simply confirming what appears to be the growing suspicion around9 > here: you're an ass.  Or a troll, whichever you prefer.9   Thanks for sharing that.  & And that advances the cause of VMS_in_ the_marketplace exactly how?   Atlant   ------------------------------    Date: 18 Jun 2002 07:52:02 -0600- From: koehler@encompasserve.org (Bob Koehler)o$ Subject: Re: Could linux become VMS?3 Message-ID: <NXaKWAOTuBZN@eisner.encompasserve.org>P  b In article <3D0DE0A3.5E19CFD@mindspring.com>, Atlant Schmidt <atlantnospam@mindspring.com> writes:& > So DCL *STILL* hasn't got this basic > functionality built in.a > # > Yes,  I can really see where it'ss$ > far superior to Unix shells. (NOT)  H    Just because one UNIX shell has it, doesn't mean it needs to be built3    in.  You do have experience with ksh, don't you?h   ------------------------------    Date: 18 Jun 2002 07:53:41 -0600- From: koehler@encompasserve.org (Bob Koehler)A$ Subject: Re: Could linux become VMS?3 Message-ID: <r6ojsp$GueCQ@eisner.encompasserve.org>.  c In article <3D0DE107.F47A8BDE@mindspring.com>, Atlant Schmidt <atlantnospam@mindspring.com> writes:u > + > The filename field from ls *ISN'T* right-a+ > justified; where did you get *THAT* idea?l- > It's just placed as the rightmost column soo1 > that it's trivially easy to accommodate varyingr > filename lengths.c >   G    From your statement.  The filename column on my UNIX systems is onlyuA    on the right for ls -l.  So what?  sort and awk can act on anynE    column.  What's the big advantage of having the names on the irhgtv    instead of the left?.   ------------------------------    Date: 18 Jun 2002 07:57:27 -0600- From: koehler@encompasserve.org (Bob Koehler)/$ Subject: Re: Could linux become VMS?3 Message-ID: <cO7SmdFpwswr@eisner.encompasserve.org>v  c In article <3D0DE8C7.8FE8CD59@mindspring.com>, Atlant Schmidt <atlantnospam@mindspring.com> writes:  > 8 > In the cshell and tcshell, *NOCLOBBER* is your friend.= > Always set it. Yes, you'll curse at the confirmation prompt 9 > when doing something like "ls > temp.tmp", but someday,/& > it'll pay all that frustration back.  C    So the shell has to work around the failure of the OS to providec7    versions in the file system?  I think I'll keep VMS./   ------------------------------    Date: 18 Jun 2002 07:58:23 -0600- From: koehler@encompasserve.org (Bob Koehler)e$ Subject: Re: Could linux become VMS?3 Message-ID: <K9CVZQkK6s4e@eisner.encompasserve.org>c  W In article <3D0E27A2.6040504@tsoft-inc.com>, David Froble <davef@tsoft-inc.com> writes:i >  > S > I seem to remember that user mode is the default for logical definition, thus no e$ > requirement for the specification.  D    You recall wrong.  Supervisor mode is the default for the define,!    assign, and deassign commands..        ------------------------------    Date: 18 Jun 2002 07:55:52 -0600- From: koehler@encompasserve.org (Bob Koehler)u$ Subject: Re: Could linux become VMS?3 Message-ID: <XUix$oZkRsqd@eisner.encompasserve.org>c  c In article <3D0DE80C.F184B0BE@mindspring.com>, Atlant Schmidt <atlantnospam@mindspring.com> writes:f > Dave Weatherall wrote: > F >> And of course if you're executing a DCL procedure there's always :- >>0 >>         $ @my_command /output=sys$login:mylog >>D >> Errors still come out on your screen, I think, 'cos sys$error andF >> sys$output are different but that's the case with Unix too isn't it >> (stderr and stdout)?l >  > Yes, stderr and stdout.   C    Except, of course, csh doesn't like to pipe them separately.  In)F    practice I find separate piping of them in ksh is much more usefull6    than pushd/popd in csh.  But I'll keep DCL any day.   ------------------------------  % Date: Tue, 18 Jun 2002 14:30:13 +0100n4 From: John Laird <john@laird-towers.freeserve.co.uk>$ Subject: Re: Could linux become VMS?8 Message-ID: <q9duguspekkti56530dc36dfscgp4sfoed@4ax.com>  F On 18 Jun 2002 07:58:23 -0600, koehler@encompasserve.org (Bob Koehler) wrote:  X >In article <3D0E27A2.6040504@tsoft-inc.com>, David Froble <davef@tsoft-inc.com> writes: >> n >> dT >> I seem to remember that user mode is the default for logical definition, thus no % >> requirement for the specification.  >uE >   You recall wrong.  Supervisor mode is the default for the define, " >   assign, and deassign commands.  L Just to nit-pick, the mode defaults to that of the caller, so in general DCLK creates supervisor-mode entries and system services called within user-modex< programs create user-mode ones, unless otherwise overridden.  B Must be time someone complained again about the silent fallback ofD def/sys/exec to def/sys without the appropriate priv, so I will :-))     	John@   ------------------------------   Date: 18 Jun 2002 16:50:38 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)s$ Subject: Re: Could linux become VMS?, Message-ID: <aenocu$267e$4@info.cs.uofs.edu>  3 In article <r6ojsp$GueCQ@eisner.encompasserve.org>,-0  koehler@encompasserve.org (Bob Koehler) writes:f |> In article <3D0DE107.F47A8BDE@mindspring.com>, Atlant Schmidt <atlantnospam@mindspring.com> writes: |> > s. |> > The filename field from ls *ISN'T* right-. |> > justified; where did you get *THAT* idea?0 |> > It's just placed as the rightmost column so4 |> > that it's trivially easy to accommodate varying |> > filename lengths. |> > n |> oJ |>    From your statement.  The filename column on my UNIX systems is onlyD |>    on the right for ls -l.  So what?  sort and awk can act on anyH |>    column.  What's the big advantage of having the names on the irhgt |>    instead of the left?  9 because the fileneam is likely to contain entries of fromy: 1 to ? characters, putting it last keeps the other columns; lined up, making it not only easier to read for humans, butA8 also easier to parse if your going to manipulate it with9 other programs.  You can not only parse it as fields, butt! also by absolute column location.    bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   >   ------------------------------   Date: 18 Jun 2002 17:12:03 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)>$ Subject: Re: Could linux become VMS?, Message-ID: <aenpl3$279m$1@info.cs.uofs.edu>  3 In article <cO7SmdFpwswr@eisner.encompasserve.org>,;0  koehler@encompasserve.org (Bob Koehler) writes:f |> In article <3D0DE8C7.8FE8CD59@mindspring.com>, Atlant Schmidt <atlantnospam@mindspring.com> writes: |> > r; |> > In the cshell and tcshell, *NOCLOBBER* is your friend. @ |> > Always set it. Yes, you'll curse at the confirmation prompt< |> > when doing something like "ls > temp.tmp", but someday,) |> > it'll pay all that frustration back.L |> pF |>    So the shell has to work around the failure of the OS to provide: |>    versions in the file system?  I think I'll keep VMS.  B Failure implies they tried and couldn't do it.  The fact is it wasA never wanted and the OS (actually, the file system, of which Unixe? is capable of supporting more than one, meaning a versioning FStA could be added if anyone really wanted it) doesn't provide thingss the users don't want.e  C People here always seem to assume that there is only one way to do )@ anything and if you don't do it that way, you are somehow wrong.  > A rather strange view of the world.  So then, do all VMS users@ the exact same car, eat the exact same diet, read the exact same
 books, etc.??e   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   t   ------------------------------  % Date: Tue, 18 Jun 2002 09:04:51 -0700 # From: "Tom Linden" <tom@kednos.com>s Subject: CSA->DSPP9 Message-ID: <CIEJLCMNHNNDLLOOGNJIGEFFFDAA.tom@kednos.com>s  B Got this yesteday.  Note the list of harware appended with one OS.    	 Dear Tom;t  L Welcome to the new HP. It is my pleasure to introduce you to Nigel Ball, whoD is the Vice President and General Manager of the Enterprise Solution Partners organization. Please J follow the enclosed link to hear first-hand from Nigel about his goals and
 vision for partners in the new HP.i  K As we previously communicated, Compaq Solutions Alliance will be integratedaI into the Developer and Solution Partners Portal, with a target completioni date of November 1st.   K To familiarize you with the offerings in the new HP, weve created a seriesoL of self-paced 15-minute modules. Just log into the members site and click onI the Merger banner to get to the Partnering with the New HP Webcast SeriesVD for a link to the modules which include information on AlphaServers,E Industry Standard Servers, NonStop Series, 9000 Servers and HP-UX ande	 OpenView.   K We look forward to our continued partnership with you - should you have anypG questions, please do not hesitate to contact the Program Center in yourcL location. Frequently asked questions about the HP/Compaq merger can be found; on the CSA Members site by clicking on the "Merger" banner.o   Carl Ramseys   ---o& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).? Version: 6.0.370 / Virus Database: 205 - Release Date: 6/5/2002h   ------------------------------  % Date: Tue, 18 Jun 2002 12:48:42 -0400i, From: "J. Scott Greig" <jsgreig@geminaq.com>G Subject: Re: DCL Puzzle: Can you produce this DIRECTORY command output?e/ Message-ID: <uguop1nkmem92b@corp.supernews.com>x  @ The only time I've seen the *.*; output was when I inadvertently: did a DIR on a spooled device (e.g. a spooled LAT printer)   $ SET DEF [000000] $ DIR LTA2:;   Directory LTA2:[000000]n   *.*;   Total of 1 file.  6 By DEFINEing  SPOOK: as a concealed device, we get the; desired "SPOOK:" in the directory command's output, however 8 I can't seem to suppress the directory specification, so  % $ DEF SPOOK _LTA2:[000000.]/TRAN=CONC 
 $ DIR SPOOK:;'   Directory SPOOK:[000000]   *.*;   Total of 1 file.  5 Even an invalid directory spec yields closer results:   / $ DIR SPOOK:[0];         <-- NB an invalid specd   Directory SPOOK:[0]s   *.*;   Total of 1 file.  5 I can't seem to suppress the directory specification.s   Scotte    ; "Alan E. Feldman" <SPAMSINK2001@YAHOO.COM> wrote in messager7 news:343f30ae.0206141439.2b094aa4@posting.google.com...' > Hello DCL users! >sG > Can you construct an argument for the DIRECTORY command that produces2/ > the following output?:  (Sorry, no prizes :-(n >s >c > Directory SPOOK: >: > *.*; >  > Total of 1 file. >s >-C > You can do whatever DCL you want to prepare the argument, but theiF > above output must be produced by a single DCL DIRECTORY command withE > the ODS-2 file system. Character set manipulations and the like arer > not allowed. > 8 > (Puzzle developed on VMS v6.1 and also works on v6.2.) >d >c > Disclaimer: JMHO > Alan E. Feldmanr$ > afeldman atski gfigroup dotski com   ------------------------------    Date: 18 Jun 2002 10:38:01 -07001 From: keithparris_NOSPAM@yahoo.com (Keith Parris)a= Subject: Disaster Tolerant advertising using Fortune magazine = Message-ID: <cf15391e.0206180938.49f5c72d@posting.google.com>e  E A four-page cover wrapper around the May 28 issue of Fortune magazine D advertised HPQ's Disaster Tolerant Solutions.  The advertisement was@ entitled 'Improve Your ROI While Meeting Your Disaster Tolerance? Needs', and described how OpenVMS and Tru64 Unix on AlphaServerr: systems, along with StorageWorks, can ensure that data and" applications are always available.  F This is the type of OpenVMS advertising I like to see -- in mainstreamC media.  And Disaster Tolerance is a hot topic in the IT world, post * 9/11, and an area where VMS really shines.  F Now I'd like to see more, in places like Business Week and Forbes (forC the CEOs) and in Computerworld and Information Week (for the CIOs),r
 and so forth.m: ----------------------------------------------------------: Keith Parris | parris <at> DECUServe <dot> decus <dot> org   ------------------------------  % Date: Tue, 18 Jun 2002 14:08:17 -0400i- From: JF Mezei <jfmezei.spamnot@videotron.ca>dA Subject: Re: Disaster Tolerant advertising using Fortune magazinee, Message-ID: <3D0F770D.88F88CB1@videotron.ca>   Keith Parris wrote:r > G > A four-page cover wrapper around the May 28 issue of Fortune magazinesF > advertised HPQ's Disaster Tolerant Solutions.  The advertisement wasB > entitled 'Improve Your ROI While Meeting Your Disaster ToleranceA > Needs', and described how OpenVMS and Tru64 Unix on AlphaServer < > systems, along with StorageWorks, can ensure that data and$ > applications are always available.  J That is most interesting when you consider all the speeches 2 weeks beforeK that when the "new" HP was unveiled with its famous roadmap gave a distinctu+ impression that Tru64 wouldn't be marketed.   M I wonder if HP realised the big mistake in made and how its pronouncements ofMM early may would generate a much bigger drop in sales (and customers) than hads< anticipated and HP must now do some serious damage control ?   ------------------------------    Date: 18 Jun 2002 05:06:27 -0700# From: fs63@volcanomail.com (Tim C.)cB Subject: Re: How can I tell which version of DECNet I am running ?= Message-ID: <399504c6.0206180406.3fa98ae4@posting.google.com>.  G > > :We have an Ethernet network of VAXes running DECNet and VMS 5.5-2. : > > :How can I tell which version of DECNet I am running ? > >   M Turns out we are running phase IV.   Thanks to all who answered my questions.6   Tim0 fs63@volcanomail.com   ------------------------------  % Date: Tue, 18 Jun 2002 08:45:56 -0400 2 From: Atlant Schmidt <atlantnospam@mindspring.com>H Subject: Re: Interesting ZDnet post about Tandem -- Can VMS do the same?. Message-ID: <3D0F2B84.C4DCE8D4@mindspring.com>   Atlant Schmidt wrote:n  3 > Can anyone from VMS get ZDnet to do the same withh > regard to VMS?   I guess not.   That's really too bad.   Atlant   ------------------------------  % Date: Tue, 18 Jun 2002 08:44:22 -040022 From: Atlant Schmidt <atlantnospam@mindspring.com>H Subject: Re: Interesting ZDnet post about Tandem -- Can VMS do the same?. Message-ID: <3D0F2B26.E44362B3@mindspring.com>   Rob Young wrote:  H >         Having started down Itanium path and forced to switch gears toJ >         Alpha sounds like it may have left behind some tough feelings asH >         they couldn't wait to get across the street and "pick up where >         they left off.",  / I was there at a few of the "ramming" meetings.S, I'd say your assessment is spot-on accurate.  * The Tandem folks were quite happy with the/ path they had chosen, but Houston spoke and thes- voice of God said "Though shalt use Alpha andr though shalt *LIKE IT*!"   Atlant   ------------------------------  # Date: Tue, 18 Jun 2002 15:13:30 GMTd1 From: "Terry C. Shannon" <terryshannon@attbi.com>dH Subject: Re: Interesting ZDnet post about Tandem -- Can VMS do the same?= Message-ID: <u6IP8.88668$6m5.74095@rwcrnsc51.ops.asp.att.net>u  8 "Rob Young" <young_r@encompasserve.org> wrote in message- news:cXEzBehluZyn@eisner.encompasserve.org...kG > In article <SwsP8.3388$_j6.237011@bin3.nnrp.aus1.giganews.com>, "Bill & Todd" <billtodd@metrocast.net> writes: > >y@ > > "Terry C. Shannon" <terryshannon@attbi.com> wrote in message; > > news:SasP8.79202$6m5.63505@rwcrnsc51.ops.asp.att.net...f > >g > > ...s > >mL > >> As an FYI, Pauline wanted to go with IPF four years ago, and that's the > > pathI > >> the Tandem Division took until the DEC acquisition rammed Alpha downo theirl> > >> throats. This could have something to do with the lust... > >4K > > To say that they've been lusting after the Itanic compilers 'for years'u atL > > least suggests a degree of continuity rather than a historical lust long agoe" > > discarded but then reacquired. > >-A > > Furthermore, four years ago - in particular, prior to the DEC 
 acquisition -<L > > there was a good deal more excuse for believing Intel's rosy projectionsJ > > about Itanic's schedule and performance (or at least far less dramaticG > > evidence to the contrary, though its schedule had already taken its  firstaL > > major slip by then).  And there was also a good deal less reason to haveL > > been optimistic about Alpha (because of the manifest incompetence of itsG > > owner) than there was after Compaq acquired it and Pfeiffer started  being anK > > booster.  So if the change to Alpha indeed had to be 'rammed down their-I > > throats', that suggests a significant degree of incompetence on theirD part -E > > and possibly exposes another division of Compaq that had a vestedL	 interest,sL > > in conflict with the best interests of the company as a whole, in seeing > > Alpha killed.u > >r > @ > While I certainly wouldn't embrace all that line of reasoning,   (Nor did I nor would I!)  H > I do agree that the statement indicates a long time thing.  Certainly,@ > if she had stated "for quite some time" my radar wouldn't have
 > gone up.  ( Hoisted by the semantic petard, she was!   >m@ > Having started down Itanium path and forced to switch gears toB > Alpha sounds like it may have left behind some tough feelings as@ > they couldn't wait to get across the street and "pick up where > they left off."u  C That's about the size of it. Note Pauline's background: she was Ms.hK AlphaServer at DEC before she went to Tandem. So she was obviously aware ofaK Alpha strengths and limitations when she made the initial Itanium decision.y  K The good news is that the decision was made, and work had progressed to thenL point that IPF had lockstepping built in. Intel didn't yank the lockstepping? when CPQ decided to drop IPF for Alpha on the Yosemite Project.    ------------------------------  % Date: Tue, 18 Jun 2002 13:30:37 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca>DH Subject: Re: Interesting ZDnet post about Tandem -- Can VMS do the same?, Message-ID: <3D0F6E3B.B632265E@videotron.ca>   Atlant Schmidt wrote: , > The Tandem folks were quite happy with the1 > path they had chosen, but Houston spoke and thee/ > voice of God said "Though shalt use Alpha andi > though shalt *LIKE IT*!"  G At the time Tandem made the decision to go with IA64, wasn't IA64 stilloL supposed to be something very succesful that Intel was about to develop as aJ replacement for the 8086 ? And at that time, wasn't MIPS going to die veryF fast with IA64 the only offer in the foreseable future with lockstep ?  N At the time Compaq told Tandem to switch to Alpha, weren't there signs alreadyI that IA64 wasn't going to be the replacement for the 8086 and that it wast2 going to be quite late and just for the high end ?  L At the time Tandem decided to eventually dump MIPS, did they attempt to talkM to Digital about using Alpha and getting Digital to add lockstep ?  If PalmerfE made sure that Tandem wouldn't choose Alpha, then he should be shot. -  J At the time Tandem was told to use Alpha, (I assume under Pfeiffer's helm)P were Tandem engineers unhappy about Alpha itself, or just the change in course ?  J Considering EV7 is to be ready this year, how close to a completed port toK Alpha had the Tandem engineers gotten by June 25 of 2001 ? Considering IA64lM won't be ready for another few years, Alpha would have still delivered bettery# performance for Tandem much sooner.u  J Had Compaq not killed Alpha and instead waited until 2005 or whenever IA64J actually outperforms the 8086 then completing the port to Alpha might haveD been a good idea. And the added volume for Alpha wouldn't have hurt.   ------------------------------    Date: 18 Jun 2002 12:53:34 -0600+ From: young_r@encompasserve.org (Rob Young)mH Subject: Re: Interesting ZDnet post about Tandem -- Can VMS do the same?3 Message-ID: <MwCJ39LGqhkW@eisner.encompasserve.org>   \ In article <3D0F6E3B.B632265E@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes:   > L > At the time Tandem was told to use Alpha, (I assume under Pfeiffer's helm)R > were Tandem engineers unhappy about Alpha itself, or just the change in course ? > L > Considering EV7 is to be ready this year, how close to a completed port toM > Alpha had the Tandem engineers gotten by June 25 of 2001 ? Considering IA64 O > won't be ready for another few years, Alpha would have still delivered better % > performance for Tandem much sooner.e >   F 	Back to the myth department.  Here is a prognostication.  I'm willingA 	to wager that Itanium servers have the highest price performancee8 	metrics for 64-bit servers - until Hamster comes along.  4 http://news.com.com/2100-1001-936936.html?tag=fd_top  I However, Sun's prices go up steeply. A four-processor server with 16GB offM memory costs $46,995, and a four-processor model with 32GB of memory goes for)	 $99,995. @   [snip]  N A four-processor Itanium 2 system will cost about $41,000 with 32GB of memory,O said Vaughn Mackie, enterprise platform marketing manager at Intel, speaking ata a media event Friday.f    @ 	It will only get better for IA64.  Intel is set on making it a B 	success and willing to spend tons of money in the process.  CouldG 	they have done it better/cheaper??  Maybe/probably/who knows.  PerhapsiF 	they should hop in their time machines and go back and change things.  C 	IA64 *systems* will stay tens of thousands of dollars cheaper thantA 	RISC, doing two things... ensuring market adoption and knocking gD 	margins from under non-IA64.  And Hamster will ensure IA64 systems  	stay "cheap."  E 	And yes, Alpha still could have been going strong flourishing in itsr 	niche even years from now.    				RobW   ------------------------------  % Date: Tue, 18 Jun 2002 05:43:21 -0700t# From: "Tom Linden" <tom@kednos.com>i Subject: Is list working?f9 Message-ID: <CIEJLCMNHNNDLLOOGNJIMEENFDAA.tom@kednos.com>l  > posted a message yesterday and never saw it, and the volume of( mail is signifcantly less than expected. ---a& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).? Version: 6.0.370 / Virus Database: 205 - Release Date: 6/5/2002(   ------------------------------  % Date: Mon, 17 Jun 2002 21:13:38 +0200oE From: Jan C. =?iso-8859-1?Q?Vorbr=FCggen?= <jvorbrueggen@mediasec.de>y& Subject: Re: Linus' comments about VMS+ Message-ID: <3D0E34E2.626A7915@mediasec.de>   B > But don't underestimate the difficulty of actual implementation.@ > The project leader of the initial VMS Distributed Lock ManagerE > said when the project was done that he never wanted to work outsideo > of user mode again.l  J Oh, I quite agree - and the lock manager was the and is the quintessentialN distributed, asynchronous, threaded, ... component of an operating system. AndL with a large number of components of the OS relying on it, he must have feltN the heavy hand of responsibility...I must admit my previous post was missing a4 smiley after the "small matter of implementation"...   	Jan   ------------------------------  % Date: Mon, 17 Jun 2002 21:19:11 +0200OE From: Jan C. =?iso-8859-1?Q?Vorbr=FCggen?= <jvorbrueggen@mediasec.de>.& Subject: Re: Linus' comments about VMS+ Message-ID: <3D0E362F.B2C8CC84@mediasec.de>a  D > >The point is that in the early 70's o/s design was not a science. > G > And you think it WAS a science at the time VMS was designed (the mid-l	 > 1970s)?)  J I disagree with both of you: the necessary design principles _were_ known.H Whether they were commercially viable on the iron of the time is anotherM matter. (VMS has some design and implementation trade-offs, some of which arehM quite ingrained, that are based on hardware restrictions of the time - today,cM we would do a number of things differently. However, some of these restraintsyJ have been relxed considerably under the hood while keeping compatibility.)  G > Unix at least went through a major rewrite in the late 1980s.  ModernvA > Unixes (Linux, Tru64) are microkernel-based.  Under the covers,-G > there's very little left of the original AT&T code.  In contrast, VMSeB > inside still is to a great extent warmed-over RSX-11M in design.  L I don't think that's true. QNX and NexTStep are/were microkernels, or nearlyF so. Linux and Tru64 are monolithic, even if very little is left of theN original AT&T code (and good riddance). Current versions of VMS have little to. do internally with what was in VMS V1, either.   	Jan   ------------------------------  / Date: Tue, 18 Jun 2002 09:12:25 +0200 (MET DST)a& From: Rudolf Wingert <win@fom.fgan.de>& Subject: Re: Linus' comments about VMS6 Message-ID: <200206180712.JAA21356@sinet1.fom.fgan.de>   Jello,  J IMHO VMS does have the modern design. Under Unix the normal method to lookI for  exceptions is polling. In my eyes this is a technology of the middleoH age. Under OpenVMS you will have ASTs. This is the better und modern one method. G Also kryptic commands are not modern. This are hacker like, nut not theoI state of the art. I think that the less storage and memory was the reasonyG for this type of commands. Modern Unixes like Solaris will use readable I commands instant of short letters. That OpenVMS goes the way to Unix with 5 unreadable commands is bad and the wrong way I think.PG To have four numbers for the year, is that modern? I thin also yes. May/F be you will find out more details, on which you can show, that OpenVMS is more modern as Unix.R   Regards Rudolf Wingert   ------------------------------  % Date: Tue, 18 Jun 2002 08:49:52 +0100g* From: "Richard Brodie" <R.Brodie@rl.ac.uk>& Subject: Re: Linus' comments about VMS, Message-ID: <aemomc$1imm@newton.cc.rl.ac.uk>  5 "Bill Todd" <billtodd@metrocast.net> wrote in message@< news:5DqP8.262359$%y.22673426@bin4.nnrp.aus1.giganews.com... >T? > "Paul Winalski" <prune@ZAnkh-Morpork.mv.com> wrote in messagec4 > news:3d0e0af7.748920280@proxy.news.easynews.com...  8 > >  Modern Unixes (Linux, Tru64) are microkernel-based. >eM > If Linux has a microkernel, it's so well hidden that one of the significant G > kernel books (Linux Device Drivers) doesn't seem to consider it worth  > mentioning  = Not to mention the Tanenbaum-Torvalds debate of 10 years ago:'9 http://www.oreilly.com/catalog/opensources/book/appa.html    ------------------------------   Date: 18 Jun 2002 12:45:05 GMT/ From: Thomas Dickey <dickey@saltmine.radix.net>n& Subject: Re: Linus' comments about VMS* Message-ID: <aena0h$nb0$2@news1.Radix.Net>  ' Rudolf Wingert <win@fom.fgan.de> wrote:;  I > Also kryptic commands are not modern. This are hacker like, nut not thesK > state of the art. I think that the less storage and memory was the reason   D indeed - VMS fits this classication nicely, since it was designed toD make most commands abbreviate within 4 characters - and perhaps it's1 been fixed - ignore characters past those 4 ;-)  p   -- -= Thomas E. Dickey <dickey@radix.net> <dickey@herndon4.his.com>l http://dickey.his.como ftp://dickey.his.com   ------------------------------    Date: 18 Jun 2002 08:12:28 -0600- From: koehler@encompasserve.org (Bob Koehler)-& Subject: Re: Linus' comments about VMS3 Message-ID: <C0F7vw2V7sIE@eisner.encompasserve.org>e  Z In article <aelesm$81ndd$1@ID-143435.news.dfncis.de>, "Hans Vlems" <hvlems@iae.nl> writes:  K > The point was that unix got started as a clean small kernel surrounded by  > dedicated tools.I > VMS was designed with more complixity built in, just have a look at VMSe  > Internals and Data Structures.F > Such a complex product is difficult to build even with all engineers > carefully selected and( > working with the same "state of mind".  F    Yes, but that somewhat more complex kernel can do things the simple-    UNIX kernel can't.  Been there, done that.g   ------------------------------    Date: 18 Jun 2002 09:18:30 -0600+ From: young_r@encompasserve.org (Rob Young)r& Subject: Re: Linus' comments about VMS3 Message-ID: <P1qWT6iLARuU@eisner.encompasserve.org>   Y In article <aemomc$1imm@newton.cc.rl.ac.uk>, "Richard Brodie" <R.Brodie@rl.ac.uk> writes:o >  > ? > Not to mention the Tanenbaum-Torvalds debate of 10 years ago: ; > http://www.oreilly.com/catalog/opensources/book/appa.htmlr >   < 	It is entertaining sometimes to poke fun at prognosticators< 	(me included) as they often miss the mark.  But who woulda 
 	thunk it?  9 http://www.oreilly.com/catalog/opensources/book/appa.htmls  N RISC chips happened, and some of them are running at over 100 MIPS.  Speeds ofO 200 MIPS and more are likely in the coming years. These things are not going toaO suddenly vanish.  What is going to happen is that they will gradually take overcK from the 80x86 line.  They will run old MS-DOS programs by interpreting the  80386 in software.  A 	What cracks me up in that prognostication isn't the obvious (x86eD 	going away - NOT) but that in light of NT (he mentions it elsewhere? 	and it not being out) he didn't assume that MS-DOS would be a h 	wimper in 4, 5 or 10 years.   				Robr   ------------------------------  # Date: Tue, 18 Jun 2002 16:32:31 GMTt0 From: prune@ZAnkh-Morpork.mv.com (Paul Winalski)& Subject: Re: Linus' comments about VMS8 Message-ID: <3d0f6009.836234431@proxy.news.easynews.com>  @ On Mon, 17 Jun 2002 22:03:58 +0200, "Hans Vlems" <hvlems@iae.nl> wrote:  M >VMS went through at least two entire code changes. VMS inside has nothing tom >do with RSX-11M whatsoever.  ? 11M and VMS share a lot of design concepts, both internally andaF externally.  Event flags, for example.  And the overall way things areE gone about in the executive.  NT shares a lot of similarity under thenA covers to VMS--I guess if you've seen one Dave Cutler-architectedc? OS, you've seen 'em all.  :-)  I grant you that the details aree completely different.u  L >I'm not underestimating the quality of the engineers, I'm commenting on the >problems ofJ >running a large software project. That takes a lot of expertise and it is >more than goodsK >programming (which is a prerequisite). Many people can design/write proper2 >if not outstandingmJ >algorithms. Doing the same thing with good quality over and over is not a >trivial task.# >That was the point of my argument.-  E Many of those problems can be avoided by dividing your large software F project into a number of small, manageable software projects.  This isA by and large the approach that the open software community takes.-  D And I'm not too sure I'd be quick to hold up VMS as a paradigm for a well-managed sotware project.>
 ---------- Remove 'Z' to reply by email.    ------------------------------  % Date: Tue, 18 Jun 2002 19:44:48 +0200o" From: "Hans Vlems" <hvlems@iae.nl>& Subject: Re: Linus' comments about VMS5 Message-ID: <aenr3l$8i4jq$1@ID-143435.news.dfncis.de>m  8 Bob Koehler <koehler@encompasserve.org> wrote in message- news:C0F7vw2V7sIE@eisner.encompasserve.org... D > In article <aelesm$81ndd$1@ID-143435.news.dfncis.de>, "Hans Vlems" <hvlems@iae.nl> writes:a >oJ > > The point was that unix got started as a clean small kernel surrounded by > > dedicated tools.K > > VMS was designed with more complixity built in, just have a look at VMS " > > Internals and Data Structures.H > > Such a complex product is difficult to build even with all engineers > > carefully selected and* > > working with the same "state of mind". >gH >    Yes, but that somewhat more complex kernel can do things the simple/ >    UNIX kernel can't.  Been there, done that.u > 7 Which completes the argument quite well, thank you Bob.h   ------------------------------  % Date: Tue, 18 Jun 2002 19:56:46 +02003" From: "Hans Vlems" <hvlems@iae.nl>& Subject: Re: Linus' comments about VMS5 Message-ID: <aens0a$8mjtd$1@ID-143435.news.dfncis.de>t  G The theory of OS design was still fairly new. IIRC Multics was a little  earlier thanK the initial work on unix and not a complete success either. Furthermore thetK distighuished gentlemen at Bell Labs started with a filesystem and a fairlymD simple monitor. SO yes the theory was there, but the people actually engineering.K new operating systems had to keep track of those developments. That makes ar
 difference, with being trained on that theory at school.G So it is not a question of commercial viability alone. Unix did not get. started like that.H It was started to make more efficient use of a costly computer system by letting moreG people working on it at the same time, internally at Bell (for documentt editing IIRC).D Later on it was sold, right. If commercial success would have been a prerequisite than thedD world would be a different place altogether. Unix led a quiet fairly academic life until thesK early 90's (give or take a couple of years). So it spent more than half itsc time in relative
 obscurity.K As for True64 I doubt whether it shares one line of source code with a Bello
 unix version.h   Hans  = Jan C. Vorbrggen <jvorbrueggen@mediasec.de> wrote in messaget% news:3D0E362F.B2C8CC84@mediasec.de...pF > > >The point is that in the early 70's o/s design was not a science. > > I > > And you think it WAS a science at the time VMS was designed (the mid-t > > 1970s)?  >hL > I disagree with both of you: the necessary design principles _were_ known.J > Whether they were commercially viable on the iron of the time is anotherK > matter. (VMS has some design and implementation trade-offs, some of whicha arefH > quite ingrained, that are based on hardware restrictions of the time - today,D > we would do a number of things differently. However, some of these
 restraintsL > have been relxed considerably under the hood while keeping compatibility.) >yI > > Unix at least went through a major rewrite in the late 1980s.  Modern.C > > Unixes (Linux, Tru64) are microkernel-based.  Under the covers, I > > there's very little left of the original AT&T code.  In contrast, VMSaD > > inside still is to a great extent warmed-over RSX-11M in design. > G > I don't think that's true. QNX and NexTStep are/were microkernels, orr nearlyH > so. Linux and Tru64 are monolithic, even if very little is left of theF > original AT&T code (and good riddance). Current versions of VMS have	 little to-0 > do internally with what was in VMS V1, either. >2 > Jan3 >0   ------------------------------  % Date: Tue, 18 Jun 2002 20:01:42 +0200i" From: "Hans Vlems" <hvlems@iae.nl>& Subject: Re: Linus' comments about VMS5 Message-ID: <aens39$8gevt$1@ID-143435.news.dfncis.de>a  < Paul the argument was not on concepts but about source code.= In concept VMS, RSX-11, RT-11 and TOPS-10 are clearly related  Add NT to the list as well.IB In terms of source code RSX and post VMS V3 have nothing in commonI except may be in the RSX product that provided MCR functionality for VMS.o   Hans  ; Paul Winalski <prune@ZAnkh-Morpork.mv.com> wrote in message 2 news:3d0f6009.836234431@proxy.news.easynews.com...B > On Mon, 17 Jun 2002 22:03:58 +0200, "Hans Vlems" <hvlems@iae.nl> > wrote: >wL > >VMS went through at least two entire code changes. VMS inside has nothing to > >do with RSX-11M whatsoever. >tA > 11M and VMS share a lot of design concepts, both internally and H > externally.  Event flags, for example.  And the overall way things areG > gone about in the executive.  NT shares a lot of similarity under theaC > covers to VMS--I guess if you've seen one Dave Cutler-architected A > OS, you've seen 'em all.  :-)  I grant you that the details arer > completely different.i >rJ > >I'm not underestimating the quality of the engineers, I'm commenting on thes > >problems ofL > >running a large software project. That takes a lot of expertise and it is > >more than good1F > >programming (which is a prerequisite). Many people can design/write proper > >if not outstandingeL > >algorithms. Doing the same thing with good quality over and over is not a > >trivial task.% > >That was the point of my argument.t >8G > Many of those problems can be avoided by dividing your large softwarehH > project into a number of small, manageable software projects.  This isC > by and large the approach that the open software community takes.. > F > And I'm not too sure I'd be quick to hold up VMS as a paradigm for a > well-managed sotware project.3 > ---------- > Remove 'Z' to reply by email.e   ------------------------------    Date: 18 Jun 2002 10:54:30 -07001 From: keithparris_NOSPAM@yahoo.com (Keith Parris)o& Subject: Re: Linus' comments about VMS= Message-ID: <cf15391e.0206180954.321c6b56@posting.google.com>s  e Jan C. Vorbrggen <jvorbrueggen@mediasec.de> wrote in message news:<3D0E362F.B2C8CC84@mediasec.de>...o  > Linux and Tru64 are monolithic  A Linux is indeed monolithic, but Tru64 comes via the OSF/1 lineagea0 based on Mach from CMU, so is microkernel-based.: ----------------------------------------------------------: Keith Parris | parris <at> DECUServe <dot> decus <dot> org   ------------------------------  % Date: Tue, 18 Jun 2002 20:03:47 +0200 " From: "Hans Vlems" <hvlems@iae.nl>& Subject: Re: Linus' comments about VMS5 Message-ID: <aens73$8cqt1$1@ID-143435.news.dfncis.de>p  E Polling is a big IBM favorite. The s/390 channel processor polls, for$	 instance. K Mideval technology perhaps, but when applied with enthousiasm it will still  do the job :-)  1 Rudolf Wingert <win@fom.fgan.de> wrote in messagef0 news:200206180712.JAA21356@sinet1.fom.fgan.de... > Jello, >aL > IMHO VMS does have the modern design. Under Unix the normal method to lookK > for  exceptions is polling. In my eyes this is a technology of the middlesJ > age. Under OpenVMS you will have ASTs. This is the better und modern one	 > method.hI > Also kryptic commands are not modern. This are hacker like, nut not the K > state of the art. I think that the less storage and memory was the reasonpI > for this type of commands. Modern Unixes like Solaris will use readablegK > commands instant of short letters. That OpenVMS goes the way to Unix with 7 > unreadable commands is bad and the wrong way I think.tI > To have four numbers for the year, is that modern? I thin also yes. May-H > be you will find out more details, on which you can show, that OpenVMS > is more modern as Unix.m >  > Regards Rudolf Wingert >e   ------------------------------    Date: 18 Jun 2002 07:00:56 -0600B From: clubley@remove_me.eisner.decus.org-Earth.UFP (Simon Clubley)E Subject: Location of DLM papers ?, was: Re: Linus' comments about VMS.3 Message-ID: <VJeUa2VD61ZK@eisner.encompasserve.org>t  s In article <3D0E0898.E618AC8B@mediasec.de>, Jan C. =?iso-8859-1?Q?Vorbr=FCggen?= <jvorbrueggen@mediasec.de> writes:, > K > The semantics of the lock manager are a copy, down to the compatible lockeM > diagram, of a paper from a conference I once stumbled across. I believe theEM > distributed deadlock detection algorithm was taken from another article (byiL > Lamport? IIRC, the source referenced it). All you then need is a compotentO > programmer to implement it. Oh, an a connection manager that's worth its saltu) > - another paper by Lamport implemented.  >   G I would like to read these papers. What are their formal titles and aret they available online ?     If so, does anybody have a URL ?   Thanks for any information.O   Simon.   -- hB Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP       + Microsoft: The Lada of the computing world.e   ------------------------------  # Date: Tue, 18 Jun 2002 13:27:16 GMT-? From: Jim.Johnson@software-exploration.nospam.com (Jim Johnson)oI Subject: Re: Location of DLM papers ?, was: Re: Linus' comments about VMSo0 Message-ID: <3d0f3424.25455382@news.demon.co.uk>  ? At least one of them is J.N. Gray, "Notes on Database OperatingaB Systems," in Operating Systems: An Advanced Course, R. Bayer et al) eds., pp.393-481, Springer- Verlag, 1978.    Jim.     On 18 Jun 2002 07:00:56 -0600,C clubley@remove_me.eisner.decus.org-Earth.UFP (Simon Clubley) wrote:n  t >In article <3D0E0898.E618AC8B@mediasec.de>, Jan C. =?iso-8859-1?Q?Vorbr=FCggen?= <jvorbrueggen@mediasec.de> writes: >> iL >> The semantics of the lock manager are a copy, down to the compatible lockN >> diagram, of a paper from a conference I once stumbled across. I believe theN >> distributed deadlock detection algorithm was taken from another article (byM >> Lamport? IIRC, the source referenced it). All you then need is a compotentTP >> programmer to implement it. Oh, an a connection manager that's worth its salt* >> - another paper by Lamport implemented. >> e >iH >I would like to read these papers. What are their formal titles and are >they available online ? >r! >If so, does anybody have a URL ?  >u >Thanks for any information. >i >Simon.t >  >-- C >Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP        , >Microsoft: The Lada of the computing world.   Jim Johnsont Software Exploration, Ltd.) (remove '.nospam' from the reply address)2   ------------------------------  % Date: Tue, 18 Jun 2002 16:36:16 +0530:= From: Kesav Tadimeti <Kesav.Tadimeti@signaltreesolutions.com>3 Subject: misc QnsdK Message-ID: <3BB00261FED5D41183FD00104B93C260020C005D@exdel01.del.mgsl.com>e  J This message is in MIME format. Since your mail reader does not understand< this format, some or all of this message may not be legible.  ( --------------InterScan_NT_MIME_Boundary$ Content-Type: multipart/alternative;1 	boundary="----_=_NextPart_001_01C216B8.2AD3EAC0"   ' ------_=_NextPart_001_01C216B8.2AD3EAC0a Content-Type: text/plain;a 	charset="iso-8859-1"   
 Hello all,! as usual I am sorry for the MIME.hL 1. How to see memory processes usage like in Task Manager? Is it the Command monitor /proc = topcpu?iK 2. My PC key Delete is mapped to VT REM key. SUppose I am at a command line/H as in $show symbol, & I place the the cursor under the character 'w' andH press the delete key to simulate REM, the command gets recalled on a new prompt like shown below:
 	$show symbol9
 	$show symbolm/ Does REM key not work as DELETE key in wondows?RK 3. What is the command to see the processor clock speed, ram etc. SHOW CPU? 8 Also how can I know the EV version (in case of an Alpha)K 4. What is the significance of SET GOLD KEY, SET KEYPAD EDT. SET KEYPAD WPSs in TPU?e   Thanks & regards Kesave  ' ------_=_NextPart_001_01C216B8.2AD3EAC0x Content-Type: text/html; 	charset="iso-8859-1"u+ Content-Transfer-Encoding: quoted-printablen  1 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">  <HTML> <HEAD>9 <META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =s charset=3Diso-8859-1">@ <META NAME=3D"Generator" CONTENT=3D"MS Exchange Server version =
 5.5.2653.12">w <TITLE>misc Qns</TITLE>s </HEAD>o <BODY>  # <P><FONT SIZE=3D2>Hello all,</FONT> ; <BR><FONT SIZE=3D2>as usual I am sorry for the MIME.</FONT>@F <BR><FONT SIZE=3D2>1. How to see memory processes usage like in Task =; Manager? Is it the Command monitor /proc =3D topcpu?</FONT>gI <BR><FONT SIZE=3D2>2. My PC key Delete is mapped to VT REM key. SUppose =iI I am at a command line as in $show symbol, &amp; I place the the cursor =oG under the character 'w' and press the delete key to simulate REM, the =tB command gets recalled on a new prompt like shown below:</FONT></P>  D <P>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; <FONT SIZE=3D2>$show =
 symbol</FONT>EE <BR>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; <FONT SIZE=3D2>$show =i
 symbol</FONT>i; <BR><FONT SIZE=3D2>Does REM key not work as DELETE key in =n wondows?</FONT>6F <BR><FONT SIZE=3D2>3. What is the command to see the processor clock =G speed, ram etc. SHOW CPU? Also how can I know the EV version (in case =  of an Alpha)</FONT></P>d  D <P><FONT SIZE=3D2>4. What is the significance of SET GOLD KEY, SET =) KEYPAD EDT. SET KEYPAD WPS in TPU?</FONT>e </P>  - <P><FONT SIZE=3D2>Thanks &amp; regards</FONT>M <BR><FONT SIZE=3D2>Kesav</FONT>t </P>   </BODY>h </HTML>k) ------_=_NextPart_001_01C216B8.2AD3EAC0--   * --------------InterScan_NT_MIME_Boundary--   ------------------------------   Date: 18 Jun 2002 01:29 CDTh' From: carl@gerg.tamu.edu (Carl Perkins) 5 Subject: Re: My conversation with Linus about VMS ...p- Message-ID: <18JUN200201295494@gerg.tamu.edu>t  6 Atlant Schmidt <atlantnospam@mindspring.com> writes... }Carl Perkins wrote: } 6 }> "DIRECTORY /NOHEADER /NOTRAILER" -> "DIR/NOHE/NOTR" }>6 }> So a reduction of 56 and 2/3 percent is "somewhat"? } 4 }Well, considering that 14 characters is still a bit1 }bigger than "ls -1"'s 5 characters (280% bigger, 6 }in fact), yeah. I'd say that only rates a "somewhat".  I So by how much can you abbreviate the unix command, or "unabbreviate" it? E In DCL, if you want to use a shorter form you can, but if you want toe? spell out the whole thing to make it clear you can do that too.i  ? }> By the way, how exactly is anyone supposed to determine that < }> when they want to see if something is in a file what they@ }> really want to do is "get a regular expression and print" it,' }> which is then abbreviated to "grep"?y } 6 }Probably in the same way that one learns that the act8 }of putting food in one's mouth is known by the verb "to8 }eat" (in some languages, anyway). At the deepest level,; }all of these words are pretty arbitrary. The amazing thing81 }about the human mind is that *MOST* are actuallym7 }capably of learning, even if it's just these arbitrarys. }associations. So while one person learns that: }the act of looking for a particular byte string in a file4 }is called "SEARCHing", another leanrs that it's the/ }use of "Generalized Regular Expression Parsingn }(and execution).V  B Learning SEARCH is vastly more simple for English speakers - it isD an actual word that describes what you want to do. DCL is like that:D is it simply a question of finding out which of the various possibleD words which you already know that could apply to an operation is the@ right one. This is considerably easier than learning the various> dialects of the martian language that Unix shells tend to use.  2 }And things naturally get cryptic when abbreviated* }for brevity. Your very own of example of: }  }  $ DIR/NOHE/NOTR } 1 }isn't exactly a model of English clarity, is it?l  H But you don't have to use the abbreviated form. When intending to teach,F for example, you usually use the full form. Where's the "full form" in
 Unix shells? a  0 }As I've said elsewhere, Unix shells and DCL are8 }both foreign languages (to varying sets of foreigners).  D VMS comes in various language variants. Does the DCL of each variantE use the matching language? I don't know - I've never seen one in use.   / }By the way, I can't recall any more; does thatb* }abbreviation of "DIRECTORY" to "DIR" work* }*WITHOUT* the user (or admin, or someone)/ }taking the active step to define a symbol? Cane/ }all DCL verbs be so abbreviated? (My LOGIN.COMk6 }was filled with abbreviation definitions from forever }ago.) }Atlantr  D Of course. You can abbreivate anything in DCL to it's minimum uniqueC set of characters - both the commands and the qualifiers (there are:G a small number of specific abbreviations defined that may not otherwisemA be possible due to conflicts, like LO for LOGOUT, DEF for DEFINE,eA C for CONTINUE, R and RU for RUN, and SH for SHOW - these use thesA "synonym" keyword in the CLD files from which the default commandn table is built).  C Symbols take precedence over the command tables, so if you define a = symbol "R" it does that instead of RUN when you enter just R.y  F You can also define a symbol such that it can be abbreviated when used
 as a command:o   $ f*oo := show symb foon $ fooS   F*OO = "SHOW SYMB FOO" $ fo   F*OO = "SHOW SYMB FOO" $ fe   F*OO = "SHOW SYMB FOO"  D What hoops do you have to jump through to get that to work with yourH favorite Unix shell? (My guess is that you have to define three separate aliases, "foo", "fo", and "f".)r   --- Carl   ------------------------------   Date: 18 Jun 2002 11:55:30 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon) 5 Subject: Re: My conversation with Linus about VMS ...>, Message-ID: <aen73i$1tq5$1@info.cs.uofs.edu>  ; In article <Xns923171292875Drbanksatarelcomau@61.9.128.12>,y4  Richard Banks <rbanks_@_arel_com_au.nospam> writes: |> lI |> Yes, I realise that (and I run VMS 7.1 here on uVax 3100's) but I was rN |> really only talking about current hardware.  The VAX is already EOL'ed and + |> unfortunately should be considered dead.t  @ I wouldn't bury it yet.  I still run VAXen here that are used byA faculty and students for real coursework every semester.  I don'tiB see that changing anytime soon.  Not even when I they finally move. that multiprocessor Alpha over here to my lab.   bill   -- tJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   c   ------------------------------  % Date: Tue, 18 Jun 2002 08:39:51 -0400p2 From: Atlant Schmidt <atlantnospam@mindspring.com>5 Subject: Re: My conversation with Linus about VMS ...a. Message-ID: <3D0F2A17.379E4459@mindspring.com>   Carl Perkins wrote:f  7 > > "DIRECTORY /NOHEADER /NOTRAILER" -> "DIR/NOHE/NOTR"  > }>8 > }> So a reduction of 56 and 2/3 percent is "somewhat"? > }e6 > }Well, considering that 14 characters is still a bit3 > }bigger than "ls -1"'s 5 characters (280% bigger, 8 > }in fact), yeah. I'd say that only rates a "somewhat". >tK > So by how much can you abbreviate the unix command, or "unabbreviate" it? G > In DCL, if you want to use a shorter form you can, but if you want to A > spell out the whole thing to make it clear you can do that too.a  2 Well, when the commands are only two or so letters: long to begin with, there's not much point in abbreviating them.a    H > You can also define a symbol such that it can be abbreviated when used > as a command:y >o > $ f*oo := show symb fooa > $ fooa >   F*OO = "SHOW SYMB FOO" > $ fo >   F*OO = "SHOW SYMB FOO" > $ f) >   F*OO = "SHOW SYMB FOO" >rF > What hoops do you have to jump through to get that to work with yourJ > favorite Unix shell? (My guess is that you have to define three separate! > aliases, "foo", "fo", and "f".)t  < AFAIK, you're correct. I've provided far fewer abbreviations; for my (already terse) Unix commands than I used to providex for my DCL commands.  9 By the way, I figured out why I couldn't remember whethern< or not DCL allowed "minimal abbrevu=iations". It was because: I had symbols covering most of the DCL commands (so, e.g.,. "COPY" became "COPY /LOG") and those *SYMBOLS*3 wouldn't have been automagically handled by the DCL 5 minimal-abbreviation logic. So, in fact, my LOGIN.COMv) file was full of stuff like "COP*Y /LOG".a    D > Learning SEARCH is vastly more simple for English speakers - it isF > an actual word that describes what you want to do. DCL is like that:F > is it simply a question of finding out which of the various possibleF > words which you already know that could apply to an operation is theB > right one. This is considerably easier than learning the various@ > dialects of the martian language that Unix shells tend to use.  0 You understand that this is pretty much a matter. of opinion. "Uhh, LOCATE, FIND, SEEK, oh yeah,, SEARCH!" It's not much improved over "grep".1 And most computer professionals are well-aware ofa2 what grep is and does. And if you really want your/ users to have the verb "SEARCH" available, justs6 alias it to fgrep; flipping the filename/pattern order4 is no big deal; I did it in reverse for years with a "grep" symbol in DCL.c    4 > }And things naturally get cryptic when abbreviated, > }for brevity. Your very own of example of: > }h > }  $ DIR/NOHE/NOTR > } 3 > }isn't exactly a model of English clarity, is it?o >eJ > But you don't have to use the abbreviated form. When intending to teach,H > for example, you usually use the full form. Where's the "full form" in > Unix shells?   ls, rm, mv, etc. :-)   Already nice and terse.   = Obviously, one could aliase those to "list", "remove", "move"w6 if one thought ones audience needed this mnemonic aid.    -=-=-=-=-=-=-=-  = The point I keep making (attempting to make?) is that command F languages are a matter of taste. I like the taste of the Unix tcshell;? I liked CCL, I *NEVER* liked DCL. And those of you (the generals@ you, not necessarily you Carl)  who believe that DCL is a reason: to choose VMS are unsupported by facts; you just have your opinion.  8 In fact, the COE effort is *ABSOLUTELY ESSENTIAL* if you; want to bring VMS to a wider audience because, by providinge= reals shells, it removes a major stumbling block for a lot ofn; Unix users. I would certainly be included in that category.t  5 Whether HP will take any advantage of COE, of course,g  remains to be seen. I'd bet not.   Atlant   ------------------------------    Date: 18 Jun 2002 08:03:26 -0600- From: koehler@encompasserve.org (Bob Koehler)o5 Subject: Re: My conversation with Linus about VMS ...o3 Message-ID: <RQFn$PruWenC@eisner.encompasserve.org>e   In article <aekoso$imh$1@new-usenet.uk.sun.com>, Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com> writes: > 4 > The joke if there is one is that the Alpha is fast6 > Sun/SPARC is slow so why does Sun sell so many boxes3 > thought process is one of the key reasons why Suna3 > does sell so many boxes when compared with Alpha.d  =    Sun/SPARC boxes sell well because they are inline with theo>    current popular trend in computing:  cheap junk that's good    enough for a lot of stuff.t   ------------------------------    Date: 18 Jun 2002 08:04:41 -0600- From: koehler@encompasserve.org (Bob Koehler).5 Subject: Re: My conversation with Linus about VMS ...m3 Message-ID: <DWQI2HGF2lVt@eisner.encompasserve.org>   o In article <Xns923171292875Drbanksatarelcomau@61.9.128.12>, Richard Banks <rbanks_@_arel_com_au.nospam> writes:. > H > Yes, I realise that (and I run VMS 7.1 here on uVax 3100's) but I was M > really only talking about current hardware.  The VAX is already EOL'ed and t* > unfortunately should be considered dead.  D    Out VAXen are alive and running just fine.  No reason to consider
    them dead.i   ------------------------------  # Date: Tue, 18 Jun 2002 14:31:26 GMTd# From: "John Smith" <a@nonymous.com>y5 Subject: Re: My conversation with Linus about VMS ...tJ Message-ID: <2vHP8.383535$t8_.319111@news01.bloor.is.net.cable.rogers.com>  : "Bob Koehler" <koehler@encompasserve.org> wrote in message- news:RQFn$PruWenC@eisner.encompasserve.org...- >-? >    Sun/SPARC boxes sell well because they are inline with theo@ >    current popular trend in computing:  cheap junk that's good >    enough for a lot of stuff.'  D I generally agree with you, but there's a whole world out there thatL purchases Fords to go from point A to point B when a Mercedes would also get you there too.  C The whole unix/Windows vs. VMS argument can be boiled down to a fewr fundamental things:   L a) People are resistant to change. If you start on unix/Windows, you tend toE stay. For VMS this is true too, but generally only for the people whofI actually do the work, not manangement who pays the bills and who are moreoH easily swayed by marketing, a round of golf, and Lakers tickets. And for those who are lemmings.i  D b) If your company is prepared to have higher staffing costs to make8 unix/Windows work acceptably, then VMS isn't your thing.  I c) If downtime and lost profits isn't a concern to your company, then VMSr isn't your thing.   H d) If you want a really wide range of 3rd party software, then VMS isn't your thing.r   ------------------------------  % Date: Tue, 18 Jun 2002 10:04:59 -0700 % From: Dean Woodward <deanw@rdrop.com> 5 Subject: Re: My conversation with Linus about VMS ...J) Message-ID: <3D0F683B.6B5921E5@rdrop.com>r   Atlant Schmidt wrote:e >  > Carl Perkins wrote:  > 9 > > > "DIRECTORY /NOHEADER /NOTRAILER" -> "DIR/NOHE/NOTR"  > > }>: > > }> So a reduction of 56 and 2/3 percent is "somewhat"? > > }n8 > > }Well, considering that 14 characters is still a bit5 > > }bigger than "ls -1"'s 5 characters (280% bigger, : > > }in fact), yeah. I'd say that only rates a "somewhat". > >eM > > So by how much can you abbreviate the unix command, or "unabbreviate" it?QI > > In DCL, if you want to use a shorter form you can, but if you want tooC > > spell out the whole thing to make it clear you can do that too.m > 4 > Well, when the commands are only two or so letters< > long to begin with, there's not much point in abbreviating > them.a > J > > You can also define a symbol such that it can be abbreviated when used > > as a command:  > >o > > $ f*oo := show symb fooa	 > > $ fooi > >   F*OO = "SHOW SYMB FOO" > > $ fo > >   F*OO = "SHOW SYMB FOO" > > $ f> > >   F*OO = "SHOW SYMB FOO" > > H > > What hoops do you have to jump through to get that to work with yourL > > favorite Unix shell? (My guess is that you have to define three separate# > > aliases, "foo", "fo", and "f".)b > > > AFAIK, you're correct. I've provided far fewer abbreviations= > for my (already terse) Unix commands than I used to providen > for my DCL commands. > ; > By the way, I figured out why I couldn't remember whethera> > or not DCL allowed "minimal abbrevu=iations". It was because< > I had symbols covering most of the DCL commands (so, e.g.,0 > "COPY" became "COPY /LOG") and those *SYMBOLS*5 > wouldn't have been automagically handled by the DCLt7 > minimal-abbreviation logic. So, in fact, my LOGIN.COM,+ > file was full of stuff like "COP*Y /LOG".l > F > > Learning SEARCH is vastly more simple for English speakers - it isH > > an actual word that describes what you want to do. DCL is like that:H > > is it simply a question of finding out which of the various possibleH > > words which you already know that could apply to an operation is theD > > right one. This is considerably easier than learning the variousB > > dialects of the martian language that Unix shells tend to use. > 2 > You understand that this is pretty much a matter0 > of opinion. "Uhh, LOCATE, FIND, SEEK, oh yeah,. > SEARCH!" It's not much improved over "grep".3 > And most computer professionals are well-aware ofh4 > what grep is and does. And if you really want your1 > users to have the verb "SEARCH" available, juste8 > alias it to fgrep; flipping the filename/pattern order6 > is no big deal; I did it in reverse for years with a > "grep" symbol in DCL.S > 6 > > }And things naturally get cryptic when abbreviated. > > }for brevity. Your very own of example of: > > }8 > > }  $ DIR/NOHE/NOTR > > }h5 > > }isn't exactly a model of English clarity, is it?: > >eL > > But you don't have to use the abbreviated form. When intending to teach,J > > for example, you usually use the full form. Where's the "full form" in > > Unix shells? >  > ls, rm, mv, etc. :-) >  > Already nice and terse.=  4 Presumes that for every user, terse equates to nice.  ? > Obviously, one could aliase those to "list", "remove", "move" 8 > if one thought ones audience needed this mnemonic aid.  H One could, though such crutches tend to bite users and administrators in
 the long run.r  F Frankly, except for actual programming features, I don't care all thatF much about the user interface.  Both *nix and VMS have features that I= need to do what I want- though logicals under *nix would be a  GoodThing(tm).  H Mostly what I worry about are reliability, security, reliability, system$ services, security, and reliability.   -- &: Dean Woodward   | Portland, OR- worst motorcycling weather0 deanw!rdrop,com | in the continental US, there. 4                 |  - Someone on rec.moto, circa 1994A ----------------+------------------------------------------------F@ '66 Duc 250 - '85 FJ1100 - '00 KLR650 - '01 Falco - '03 FJR1300*   (*Soon, but not soon enough!)w   ------------------------------   Date: 18 Jun 2002 16:34:56 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)T5 Subject: Re: My conversation with Linus about VMS ...O, Message-ID: <aennfg$267e$1@info.cs.uofs.edu>  - In article <18JUN200201295494@gerg.tamu.edu>, *  carl@gerg.tamu.edu (Carl Perkins) writes:9 |> Atlant Schmidt <atlantnospam@mindspring.com> writes...M |> }Carl Perkins wrote:- |> } -9 |> }> "DIRECTORY /NOHEADER /NOTRAILER" -> "DIR/NOHE/NOTR"r |> }>e9 |> }> So a reduction of 56 and 2/3 percent is "somewhat"?  |> } M7 |> }Well, considering that 14 characters is still a bit 4 |> }bigger than "ls -1"'s 5 characters (280% bigger,9 |> }in fact), yeah. I'd say that only rates a "somewhat".L |> >L |> So by how much can you abbreviate the unix command, or "unabbreviate" it?H |> In DCL, if you want to use a shorter form you can, but if you want toB |> spell out the whole thing to make it clear you can do that too.  ? The shells are infinitely customizable.  Simple aliasing can bex? used to not only lengthen (you can't make them much shorter :-)oA or redefine any command you want.  The System manager can do thiso@ for everyone or individuals can do it for themselves.  Most UnixC users don't because they are perfectly comfortable with the commando set the way it is.   |> hB |> }> By the way, how exactly is anyone supposed to determine that? |> }> when they want to see if something is in a file what theysC |> }> really want to do is "get a regular expression and print" it, * |> }> which is then abbreviated to "grep"? |> }  9 |> }Probably in the same way that one learns that the actt; |> }of putting food in one's mouth is known by the verb "toa; |> }eat" (in some languages, anyway). At the deepest level,s> |> }all of these words are pretty arbitrary. The amazing thing4 |> }about the human mind is that *MOST* are actually: |> }capably of learning, even if it's just these arbitrary1 |> }associations. So while one person learns thatn= |> }the act of looking for a particular byte string in a fileo7 |> }is called "SEARCHing", another leanrs that it's the 2 |> }use of "Generalized Regular Expression Parsing |> }(and execution). |> wE |> Learning SEARCH is vastly more simple for English speakers - it ispG |> an actual word that describes what you want to do. DCL is like that:eG |> is it simply a question of finding out which of the various possibledG |> words which you already know that could apply to an operation is theiC |> right one. This is considerably easier than learning the variouseA |> dialects of the martian language that Unix shells tend to use.   : If you don't like "grep", simple make it "SEARCH". Trivial8 in Unix.  Under any shell. (Or even without the shell!!)  : What you call various dialects others call flexability and8 user friendliness.  Unix doesn't have one user and it is8 not reasonable to assume that every user will want to do7 things exactly the same way.  Unix allows users to pickt8 from multiple UI command schemes.  Interestingly enough,( so did RSTS (and I think RSX as well.)    : The attitude/paradigm difference reminds me of the biggest: difference between my shop (I'm Unix) and the University's< datacenter (they're VMS).  I will make any need/deisred tool= (within reason) available.  I see my job as meeting the usersn< requirements.  The datacenter dictates, "This is how it will< be and the users will learn to live with it."  That is why I< have said all along that the big difference between Unix and= VMS is the mindset behind them from the very beginning.  Many = here think that somehow Unix had no engineering behind it.  If; have always said it did, but the engineers had a different n  target than the VMS engineers.     |> m5 |> }And things naturally get cryptic when abbreviatedt- |> }for brevity. Your very own of example of:" |> } o |> }  $ DIR/NOHE/NOTRw |> }  4 |> }isn't exactly a model of English clarity, is it? |>  K |> But you don't have to use the abbreviated form. When intending to teach,oI |> for example, you usually use the full form. Where's the "full form" in= |> Unix shells?   9 If the desire for it existed among the Unix community, itw: would be done.  It's trivial.  Radio Shack used to ship an9 MSDOS Shell with their Model-16 running Xenix.  It looked < just like COMMAND.COM and hadled translating options between< the MSDOS style commands and the Unix commands.  Most people: learned quickly that this was rather limiting and moved to a real Unix shell.   |> 13 |> }As I've said elsewhere, Unix shells and DCL aren; |> }both foreign languages (to varying sets of foreigners).o |> nG |> VMS comes in various language variants. Does the DCL of each variantVH |> use the matching language? I don't know - I've never seen one in use. |>  2 |> }By the way, I can't recall any more; does that- |> }abbreviation of "DIRECTORY" to "DIR" works- |> }*WITHOUT* the user (or admin, or someone)n2 |> }taking the active step to define a symbol? Can2 |> }all DCL verbs be so abbreviated? (My LOGIN.COM9 |> }was filled with abbreviation definitions from forevero	 |> }ago.) 
 |> }Atlant |> lG |> Of course. You can abbreivate anything in DCL to it's minimum uniqueiF |> set of characters - both the commands and the qualifiers (there areJ |> a small number of specific abbreviations defined that may not otherwiseD |> be possible due to conflicts, like LO for LOGOUT, DEF for DEFINE,D |> C for CONTINUE, R and RU for RUN, and SH for SHOW - these use theD |> "synonym" keyword in the CLD files from which the default command |> table is built).m |> rF |> Symbols take precedence over the command tables, so if you define a@ |> symbol "R" it does that instead of RUN when you enter just R. |> sI |> You can also define a symbol such that it can be abbreviated when used  |> as a command: |> s |> $ f*oo := show symb foo |> $ foo |>   F*OO = "SHOW SYMB FOO"r |> $ foo |>   F*OO = "SHOW SYMB FOO"h |> $ f |>   F*OO = "SHOW SYMB FOO"  |> nG |> What hoops do you have to jump through to get that to work with yourtK |> favorite Unix shell? (My guess is that you have to define three separatet" |> aliases, "foo", "fo", and "f".)  D Witht he existing shells, yes.  So what??  Takes what, 5 seconds andE only has to be done once.  Big deal.  But then, the average Unix usera- would only need the shortest one anyway.  :-)o   bill   -- hJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   y   ------------------------------  # Date: Tue, 18 Jun 2002 17:17:25 GMTF From: danco@pebble.org5 Subject: Re: My conversation with Linus about VMS ...s- Message-ID: <slrnagurmu.e21.danco@pebble.org>y  C In article <aennfg$267e$1@info.cs.uofs.edu>, Bill Gunshannon wrote:t  < > The attitude/paradigm difference reminds me of the biggest< > difference between my shop (I'm Unix) and the University's> > datacenter (they're VMS).  I will make any need/deisred tool? > (within reason) available.  I see my job as meeting the usersc> > requirements.  The datacenter dictates, "This is how it will/ > be and the users will learn to live with it."m  = That has zilch to do with operating systems and everything to ? do with their being the "datacenter."  You'll see that attitude < in many datacenters, but not all.  It has nothing to do with? which operating systems they use.  It has everything to do withn the datacenter management.   - Dane   ------------------------------    Date: 18 Jun 2002 05:28:38 -0700# From: fs63@volcanomail.com (Tim C.) + Subject: Re: Need DECNet packet informationh= Message-ID: <399504c6.0206180428.550facd8@posting.google.com>    > :Tim C. wrote: > :mG > :>I need DECNet packet information to use with our Ethernet sniffer.  J > :>The link below was posted in 1998 in this newsgroup, but is now a dead	 > :>link.f > :> > I >   Donno why gatekeeper is offline right now -- the system is apparentlyhL >   running, based on the errors I see returned.  I'll pass along a question! >   to some of the network folks.w > K >   I have copies of the documentation around (somewhere), if it takes morenL >   than a week or so for the folks to figure out why gatekeeper.dec.com is J >   returning the ftp 550 error (and resolve it).  (If this drags out too I >   long, I'll drop the copies of the documents I have onto the Freeware i+ >   website or the Ask The Wizard website.)  > D Mr. Hoffman, Have you by chance heard anything new on the gatekeeperB being offline?  I'm not sure who to email at DEC/Compaq/HP to ask.  < I am still trying to get to this documentation on DECNet IV.   Thanks for your help,t Timi fs63@volcanomail.com   ------------------------------  % Date: Tue, 18 Jun 2002 18:28:50 +0200y. From: "Jesper Naur" <jesper.naur@post.tele.dk>+ Subject: Re: Need DECNet packet informationt; Message-ID: <3d0f5f23$0$244$edfadb0f@dspool01.news.tele.dk>Y  . Tim C. <fs63@volcanomail.com> wrote in message7 news:399504c6.0206180428.550facd8@posting.google.com...a> > I am still trying to get to this documentation on DECNet IV. >   J I can mail it to you (sometime tomorrow) - I downloaded it once, at a timeH when the link was still alive. Let me know, if the above mail address is	 good/bad.-       Best regards     Jesper Naur0   ------------------------------  # Date: Tue, 18 Jun 2002 16:34:33 GMTa0 From: prune@ZAnkh-Morpork.mv.com (Paul Winalski)7 Subject: Re: One of Terry's shoes seems to have droppedi8 Message-ID: <3d0f619a.836635167@proxy.news.easynews.com>  F On Mon, 17 Jun 2002 16:17:40 GMT, "Bill Todd" <billtodd@metrocast.net> wrote:  J >Exactly.  The point being that there's absolutely nothing that would keepK >them from acting the same way in killing VMS should they get it into theirlE >so-called minds to do so, which is why I suggested that well-meaningeK >reassurances from the VMS group would not really carry much actual weight.   F All too true, alas.  HpQ don't have a track record of high credibility
 in this area.   
 ---------- Remove 'Z' to reply by email.h   ------------------------------  + Date: Tue, 18 Jun 2002 11:14:11 +0000 (UTC)u From: david20@alpha2.mdx.ac.ukP Subject: RE: OpenVMS, Volume Desktop OS (Re: Mark Gorham's Beer Bash in Reading)+ Message-ID: <aen4m3$nkv$1@aquila.mdx.ac.uk>m  ^ In article <Xns923080653170Cfalkarcabca@205.233.108.180>, Alfred Falk <falk@arc.ab.ca> writes:1 >koehler@encompasserve.org (Bob Koehler) wrote ine- >news:2Azg2hPl9lsI@eisner.encompasserve.org: h >d
 >> In articledI >> <BE56C50EA024184DAF48F0B9A47F5CF40266077C@kaoexc01.americas.cpqcorp.ner >t0 >> >, "Main, Kerry" <Kerry.Main@hp.com> writes:  >>> C >>>>>> (And to keep things accurate I should point out that Unix is 3 >>> somewhat less than 10 years older than VMS.)<<<t >>> E >>> Perhaps it's a bit less - Anyone know when the first UNIX systems G >>> (that now form the basics of all UNIX's today) became available?=20r >> @I >>    It's generally accepted that UNIX was first written in 1969.  It's eG >>    use of a two mode OS (kernel and user) and byte-stream files was uI >>    popular at that time.  Compare to what DEC shipped on the PDP-10 toy  >>    see internal similarities. >rD >I think 1969 is stretching things a bit.  Earliest versions were onI >PDP-7 which had a pretty limited architecture compared to what you mightaI >think of as useful today.  Subsequent PDP-11 versions looked pretty much G >like the current product.  First appearances outside Bell Labs were ini4 >1974.  Ran very nicely on PDP-11/45 with 2 RK05's.  >l  * No 1969 is definitely when Unix was born. 9 First used in anger by the BTL Patent department in 1970. M The first edition of the  "UNIX PROGRAMMER'S MANUAL [by] K. Thompson [and] D.m& M. Ritchie" is dated November 3, 1971.  M The fact it wasn't widely available outside of AT&T until 1974 is irrelevent.    Seee  / http://virtual.park.uga.edu/hc/unixhistory.htmlt   or a thousand other sites.    I
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------  # Date: Tue, 18 Jun 2002 16:50:22 GMT0" From: Alfred Falk <falk@arc.ab.ca>P Subject: RE: OpenVMS, Volume Desktop OS (Re: Mark Gorham's Beer Bash in Reading)9 Message-ID: <Xns92316E43499CBfalkarcabca@205.233.108.180>w  E david20@alpha2.mdx.ac.uk wrote in news:aen4m3$nkv$1@aquila.mdx.ac.uk:I  G > In article <Xns923080653170Cfalkarcabca@205.233.108.180>, Alfred Falkn > <falk@arc.ab.ca> writes:  2 >>koehler@encompasserve.org (Bob Koehler) wrote in. >>news:2Azg2hPl9lsI@eisner.encompasserve.org:  >> >>> In article >>> E <BE56C50EA024184DAF48F0B9A47F5CF40266077C@kaoexc01.americas.cpqcorp.nn >>> e  >>tt1 >>> >, "Main, Kerry" <Kerry.Main@hp.com> writes: D >>>> yD >>>    It's generally accepted that UNIX was first written in 1969. H >>>    It's use of a two mode OS (kernel and user) and byte-stream filesD >>>    was popular at that time.  Compare to what DEC shipped on the+ >>>    PDP-10 to see internal similarities.  >>E >>I think 1969 is stretching things a bit.  Earliest versions were onsD >>PDP-7 which had a pretty limited architecture compared to what youD >>might think of as useful today.  Subsequent PDP-11 versions lookedG >>pretty much like the current product.  First appearances outside BelleB >>Labs were in 1974.  Ran very nicely on PDP-11/45 with 2 RK05's.  >> > , > No 1969 is definitely when Unix was born. ; > First used in anger by the BTL Patent department in 1970. F > The first edition of the  "UNIX PROGRAMMER'S MANUAL [by] K. Thompson1 > [and] D. M. Ritchie" is dated November 3, 1971.L  C I don't dispute this.  What I meant was that the 1969 versions was tI decidedly embryonic.  I suspect that you wouldn't recognize much of unix a, in its pre-PDP-11 (/40 and up) incarnations.  C > The fact it wasn't widely available outside of AT&T until 1974 ish > irrelevent.   I Except that it establishes a birth date prior to 1974.  Incidentally, it IG was NOT widely available in 1974.  That didn't happen for several more A years.  lH In any case, it establishes that unix is several years older than VMS.  = This also pre-dates RSX-11M, the principal progenitor of VMS.q   --  @ ----------------------------------------------------------------A   A L B E R T A         Alfred Falk               falk@arc.ab.ca .@ R E S E A R C H         Information Systems Dept   (780)450-5185+   C O U N C I L         250 Karl Clark Roade1                         Edmonton, Alberta, Canaday http://www.arc.ab.ca/   T6N 1E4h  http://www.arc.ab.ca/staff/falk/   ------------------------------   Date: 18 Jun 2002 16:41:05 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)dP Subject: RE: OpenVMS, Volume Desktop OS (Re: Mark Gorham's Beer Bash in Reading), Message-ID: <aennr1$267e$2@info.cs.uofs.edu>  + In article <aen4m3$nkv$1@aquila.mdx.ac.uk>,e!  david20@alpha2.mdx.ac.uk writes:m |> n- |> No 1969 is definitely when Unix was born. i< |> First used in anger by the BTL Patent department in 1970.  @ Care to back that one up??  I have never seen anything that evenA hinted at dis-satisfaction by the typists using the original Unixs1 at Bell Labs.  Something on DR's homepage maybe??   P |> The first edition of the  "UNIX PROGRAMMER'S MANUAL [by] K. Thompson [and] D.) |> M. Ritchie" is dated November 3, 1971.l |>  P |> The fact it wasn't widely available outside of AT&T until 1974 is irrelevent.  A The fact it wasn't widely available outside of AT&T until 1974 ist@ primarily a legal thing and had nothing to do with the viability
 of the OS.   |> d |> See |> s2 |> http://virtual.park.uga.edu/hc/unixhistory.html  4 Nothing there about the Patent Dept. being angry....   bill   -- tJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------  + Date: Tue, 18 Jun 2002 06:44:00 +0000 (UTC) . From: "Sanface Software" <sanface@sanface.com> Subject: PLUG: txt2pdf 5.8H Message-ID: <88321b9db642d3afac10225031357099.93245@mygate.mailgate.org>  / We would like to announce txt2pdf 5.8 version. 0# http://www.sanface.com/txt2pdf.html E txt2pdf is shareware; it is a very flexible and powerful PERL5 scripteH that converts text files to PDF format files, so you can use it in every@ operating systems supported by PERL5, including MPE/iX, OpenVMS.@ If you prefer we also distribute executables for Windows, Linux,G Solaris, AIX, HP-UX, and Mac OS X. Inside the Windows version is VisualA txt2pdf, a VB GUI.   What's new in this version  C zoom feature: to set the open of first page with the selected zoom.RD txt2pdf.vim 2.4 inside contributed directory: VIM plugin to save and1 convert the edited text to PDF clicking a button!r   Test txt2pdf 5.8!s6 You can find it at http://www.sanface.com/txt2pdf.html     -- g8 Posted via Mailgate.ORG Server - http://www.Mailgate.ORG   ------------------------------  % Date: Tue, 18 Jun 2002 17:04:01 +0200a6 From: =?iso-8859-1?Q?Jean=2DFran=E7ois=20PI=C9RONNE?=  Subject: Port of SAMBA 2.2.4+ Message-ID: <3D0F4BE1.D3139984@laposte.net>l   Hi all,   K One of my colleague has done a port of the latest release of SAMBA (2.2.4),x including SWAT.yA Before officially release his port he is looking for beta-tester.o Any voluntaries ?e   Binaries: 10 MB  Sources: 14 MB    
 Jean-Franoisl   ------------------------------  % Date: Tue, 18 Jun 2002 12:03:31 +0100tU From: Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com>cA Subject: Re: Request for path balancing (Was: Re: Carly was here)e0 Message-ID: <aen424$c3b$1@new-usenet.uk.sun.com>   Rob Young wrote:   > In article <aekne7$i7l$1@new-usenet.uk.sun.com>, Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com> writes: >  >> >>Rob Young wrote: >> >> >>>In article <aed5ea$72h$1@new-usenet.uk.sun.com>, Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com> writes:. >>>a >>>g >>>>Rob Young wrote: >>>> >>>> >>>>t >>>>>In article <27vN8.207307$%o.18579187@bin3.nnrp.aus1.giganews.com>, "Bill Todd" <billtodd@metrocast.net> writes: >>>>>s >>>>>  >>>>>  >>>>>lN >>>>>>Unfortunately, the only corporation in any reasonable position to createO >>>>>>such a system would be VMS's owner.  Lots of luck.  And while open source}R >>>>>>systems will undoubtedly over time acquire some of VMS's strengths, the UnixP >>>>>>I/O model has always resisted full-blown support for asynchrony, so unlessR >>>>>>some non-Unix-like open-source system should appear I wouldn't hold out much >>>>>>hope in that direction.c >>>>>> >>>>>> >>>>>> >>>>>>( >>>>>	Lacking asynchronous support, yep. >>>>>nA >>>>>	Do you use AIO or multiple db_writers in an AIX environmenti> >>>>>	with Oracle?  There is a school of thought that multiple' >>>>>	db_writers works better than AIO.  >>>>>(: >>>>>	PowerPath and/or SecurePath?  Sure, for broken OSes. >>>>>rF >>>>>	You can layer features on an OS I suppose.  Fortunately, for VMSB >>>>>	path switching comes tacitly, naturally.  The only advantageA >>>>>	of a PowerPath/SecurePath is load balancing.  So maybe thatiK >>>>>	is a future SET DEVICE enhancement?  Meaning via a SET DEVICE settingaB >>>>>	we can auto-switch if I/O reaches a pre-set threshold and/or
 >>>>>	"load"?e >>>>>n >>>>>p >>>>>nB >>>>Any system that uses Veritas Volume Manager, HP-UX for example; >>>>has a facility called DMP this provides path balancing.a >>>>; >>>>With Solaris you can use DMP if you are running VxVM or   >>>>the native Solaris facility. >>>>. >>>>Bottom Line is most server OS's have this. >>>> >>>> >>>>B >>>	I have to quibble quite a bit with that.  Let's say you aren'tD >>>	using Veritas.  If  you are using fibre channel storage, EMC andG >>>	Compaq and IBM provide client software so that OSes can path switchdD >>>	the storage on the fly.  You certainly don't need it if you have >>>	a single HBA.c >>>  >>>v >>; >>You seem confused that IBM and EMC would provide software>@ >>product that you pay for when most OS's have the functionality+ >>you pay for built in them for free. Why ?  >> >  > $ > 	You are full of beans once again. > C > 	Clarify "most".  NT/2000 certainly doesn't have it.  AIX doesn'tr; > 	have it.  Netware doesn't have it.  The one I know aboutgI > 	is VMS.  All other OSes need a bolt-on product to path switch, whetheroB > 	free or otherwise.  They get very confused if they see a device > 	down separate paths.n >  >     ? Nope again incorrect, Solaris has a facility called MPXIO its ao( standard part of Solaris, not a bolt on.  ? You can choose to bolt on PowerPath or Veritas DMP if you value > their capabilites more than MPXIO but MPXIO is a standard part, of the OS, which bit didn't you understand ?      A >>Solaris has something called MPXIO this provides path balancingcE >>if you choose to use it. It is part of the OS and quite independantn
 >>of Veritas.> >> >> > B > 	You didn't mention that earlier.  Did you have to do a look-up?6 > 	Thanks for keeping cov abreast of Solaris features. >     0 No, I simply didn't tell you what it was called.  < And as to cov you are making comparisons between OpenVMS and< other OS's which in the case of Solaris are not correct, you8 would I am sure prefer that any factual mistakes in your: posting are corrected by an informed source wouldn't you ?    A >>FastPath from EMC provides similar functionality with the claim ? >>that it is better at balancing the load across multiple pathso4 >>than MPXIO or Veritas DMP if you are running VxVM. >> > @ > 	FastPath from what I see is for SCSI adapters.  Check out the9 > 	subject line of this thread, I am talking about fibre.f >     A If you want both SCSI and Fibre then use PowerPath which supportsh= both. Alternatively use whatever the vendor ships as standardt like MPXIO on Solaris.  @ http://www.emc.com/products/product_pdfs/ds/powerpath_l751-8.pdf   Regardsv Andrew Harrison    ------------------------------    Date: 18 Jun 2002 08:58:03 -0600+ From: young_r@encompasserve.org (Rob Young) A Subject: Re: Request for path balancing (Was: Re: Carly was here)l3 Message-ID: <D1DEA2itEtlc@eisner.encompasserve.org>,   In article <aen424$c3b$1@new-usenet.uk.sun.com>, Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com> writes: >  >  > Rob Young wrote: >    >>> < >>>You seem confused that IBM and EMC would provide softwareA >>>product that you pay for when most OS's have the functionality , >>>you pay for built in them for free. Why ? >>>a >> 6 >> h% >> 	You are full of beans once again.- >> -D >> 	Clarify "most".  NT/2000 certainly doesn't have it.  AIX doesn't< >> 	have it.  Netware doesn't have it.  The one I know aboutJ >> 	is VMS.  All other OSes need a bolt-on product to path switch, whetherC >> 	free or otherwise.  They get very confused if they see a devicec >> 	down separate paths. >> u >> t >  > A > Nope again incorrect, Solaris has a facility called MPXIO its an* > standard part of Solaris, not a bolt on. >     ? >product that you pay for when most OS's have the functionalityt(                           ^^^^^^^^^^^^^^(                           ^^^^^^^^^^^^^^  * >you pay for built in them for free. Why ?   	Not MOST.  Very few!O  C 	There are a dozen others that DO NOT have this facility AND MPXIO eE 	works well with StorEdge.  You don't use MPXIO with HP, EMC nor IBM  C 	storage, etc.  If you do, your documentaion does not reflect that  F 	capability.  There are two OSes that I know of that do not require a G 	helper utility to path switch (or work with) supported fibre attached t 	storage.  Tru64 and VMS.t   				Rob    ------------------------------  % Date: Tue, 18 Jun 2002 16:58:17 +0100GU From: Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com>eA Subject: Re: Request for path balancing (Was: Re: Carly was here)c0 Message-ID: <aenlap$hhm$1@new-usenet.uk.sun.com>   Rob Young wrote:   > In article <aen424$c3b$1@new-usenet.uk.sun.com>, Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com> writes: >  >> >>Rob Young wrote: >> >> > = >>>>You seem confused that IBM and EMC would provide softwareeB >>>>product that you pay for when most OS's have the functionality- >>>>you pay for built in them for free. Why ?  >>>> >>>> >>> % >>>	You are full of beans once again.e >>>lD >>>	Clarify "most".  NT/2000 certainly doesn't have it.  AIX doesn't< >>>	have it.  Netware doesn't have it.  The one I know aboutJ >>>	is VMS.  All other OSes need a bolt-on product to path switch, whetherC >>>	free or otherwise.  They get very confused if they see a device  >>>	down separate paths. >>>  >>>f >>>e >>A >>Nope again incorrect, Solaris has a facility called MPXIO its a.* >>standard part of Solaris, not a bolt on. >> >> >  > @ >>product that you pay for when most OS's have the functionality >>* >                           ^^^^^^^^^^^^^^* >                           ^^^^^^^^^^^^^^ >  > + >>you pay for built in them for free. Why ?- >> >  > 	Not MOST.  Very few!u > E > 	There are a dozen others that DO NOT have this facility AND MPXIO ?G > 	works well with StorEdge.  You don't use MPXIO with HP, EMC nor IBM nE > 	storage, etc.  If you do, your documentaion does not reflect that eH > 	capability.  There are two OSes that I know of that do not require a I > 	helper utility to path switch (or work with) supported fibre attached   > 	storage.  Tru64 and VMS.a >     * I am not sure what your point is are you ?  E Solaris MPXIO will work with Sun and EMC storage, I have never lookedaF at if it works with HP or IBM frankly we hardly ever see them attached' to Sun's as storage but it should work.   B So there are three OS's that you now know of that do not require a6 helper utility to path switch, Solaris, Tru64 and VMS.  A Have you looked to see if the Tru64 documentation says that Tru64n' path switching supports Sun storage ???:  B So now I have put your mind at rest we can all get back to our day jobs.    Regardsl Andrew Harrisonp   ------------------------------    Date: 18 Jun 2002 11:48:25 -0600+ From: young_r@encompasserve.org (Rob Young)vA Subject: Re: Request for path balancing (Was: Re: Carly was here)l3 Message-ID: <EfgfXQsDlgG+@eisner.encompasserve.org>    In article <aenlap$hhm$1@new-usenet.uk.sun.com>, Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com> writes: >  >  > Rob Young wrote: >  >> In article <aen424$c3b$1@new-usenet.uk.sun.com>, Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com> writes:a >> r >>>s >>>Rob Young wrote:n >>>i >>>e >> -> >>>>>You seem confused that IBM and EMC would provide softwareC >>>>>product that you pay for when most OS's have the functionalitys. >>>>>you pay for built in them for free. Why ? >>>>>o >>>>>p >>>>& >>>>	You are full of beans once again. >>>>E >>>>	Clarify "most".  NT/2000 certainly doesn't have it.  AIX doesn'ts= >>>>	have it.  Netware doesn't have it.  The one I know about K >>>>	is VMS.  All other OSes need a bolt-on product to path switch, whethersD >>>>	free or otherwise.  They get very confused if they see a device >>>>	down separate paths.> >>>> >>>> >>>> >>>PB >>>Nope again incorrect, Solaris has a facility called MPXIO its a+ >>>standard part of Solaris, not a bolt on.n >>>> >>>e >> a >> wA >>>product that you pay for when most OS's have the functionality> >>>>+ >>                           ^^^^^^^^^^^^^^s+ >>                           ^^^^^^^^^^^^^^  >> i >> l, >>>you pay for built in them for free. Why ? >>>i >> l >> 	Not MOST.  Very few! >>  F >> 	There are a dozen others that DO NOT have this facility AND MPXIO H >> 	works well with StorEdge.  You don't use MPXIO with HP, EMC nor IBM F >> 	storage, etc.  If you do, your documentaion does not reflect that I >> 	capability.  There are two OSes that I know of that do not require a nJ >> 	helper utility to path switch (or work with) supported fibre attached  >> 	storage.  Tru64 and VMS. >>   >  > , > I am not sure what your point is are you ? >   F 	Yes.  This statement of yours that is chronically wrong no matter how! 	you attempt to slice or dice it:o  : >You seem confused that IBM and EMC would provide software? >product that you pay for when most OS's have the functionalityt* >you pay for built in them for free. Why ?  B 	That isn't a true statement.  Talking about MPXIO is entertaining 	but hardly the point.  G > Solaris MPXIO will work with Sun and EMC storage, I have never lookedcH > at if it works with HP or IBM frankly we hardly ever see them attached) > to Sun's as storage but it should work.e   	This is informative, thanks.S  < 	I think very few people (any?) are doing MPXIO + EMC.  Why?5 	No record in google.com nor groups.google.com so the 8 	hidden Solaris + MPXIO + EMC users are very few indeed.   > D > So there are three OS's that you now know of that do not require a8 > helper utility to path switch, Solaris, Tru64 and VMS. > C > Have you looked to see if the Tru64 documentation says that Tru64d) > path switching supports Sun storage ???d >   D 	Tru64?, No.  But I do know it supports other fiber channel storage.4 	It still isn't most.  It is 3 out of a dozen or so.   				Robo   ------------------------------    Date: 18 Jun 2002 08:16:54 -0700. From: SPAMSINK2001@YAHOO.COM (Alan E. Feldman)+ Subject: Re: rooted devices and /TRANS=CONC = Message-ID: <343f30ae.0206180716.2f9fed3b@posting.google.com>?  ` "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message news:<3D0E94E4.28AAA3E5@fsi.net>... > "Alan E. Feldman" wrote: > >   > > Comments interspersed below. > > d > > "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message news:<3D0A9B55.336DEE2B@fsi.net>... > > > [snip]G > > > Hhmmm... F$PARSE() returns a null string - *BUT* that's what it'siJ > > > SUPPOSED to do! Now that "SY0" means "DJAS01@DKA300:" (a device nameI > > > instead of a potential filename, the entire expression equates to aoI > > > device/path that does not exist; hence, the function returns a null 
 > > > string!- > > I > > This has nothing to do with why you need a colon with /trans=conc but @ > > not without. None of your examples even include /TRANS=CONC. > C > Because it makes no difference. If you don't correctly format thehG > translation (read: "equivalence") string, the /TRANSLATION_ATTRIBUTESeF > (if any, other than the default) make no fundamental difference. TheF > difference you'll see, if any, will be determined by how the subject4 > (command) program processes its parameter strings.    * Yes it does. I even posted the difference.     > > >i- > > > > But it does not address the question.  > > > H > > > Well, yes it does. Not convinced? Let's carry that a step further: > > >n [snip]G > > All this does is show that you need a colon to distinguish a deviced > > from a filename. > * > *EXACTLY* !!!! NOW you're catching on...      This point was never in dispute.    o3 > > It does not show why you can sometimes get away H > > without using the colon ***when the equivalence name is also another > > logical name***. > C > Because it makes no fundamental difference. All you're doing withsD > /TRANS=CONC is activating another portion of the code stream(s) toE > illustrate the result of an improperly formatted translation (read:o0 > "equivalence") string in another circumstance.    E And activating that part of the code somehow suddenly doesn't let youi* "get away with" omitting a trailing colon.     > > > > The question ishK > > > > why you need a trailing colon for the case where you are defining at0 > > > > logical name to be another logical name. > > >,E > > > ...and that also answers the question, just in a different way.u > > >  [snip]  AAA:[FELDMAN]3 > > > > %DCL-I-INVDEF, AAA:[FELDMAN] does not exist- > > >-K > > > Hhmmm. We've seen *THAT* before, have we not? (Remember how F$PARSE()(* > > > behaved above? ...the COPY command?) > >  > > What's your point? > 0 > I think you may be catching on now. Read on...     ???     m% > > > > $ DEFINE AAA BBB: /TRANS=CONCtC > > > > %DCL-I-SUPERSEDE, previous value of AAA has been superseded1 > > > > $ sh def > > > >   AAA:[FELDMAN]  > > > > $ DIREC/TOTAL  > > > >  > > > > Directory AAA:[FELDMAN]- > > > >e > > > > Total of 74 files. > > > >oD > > > > Here we see that omitting the colon works fine until you addI > > > > /TRANS=CONC. If you put the trailing colon back, then /TRANS=CONCsD > > > > works fine. Why? That is what the original poster is asking.    D Here it is. Adding /TRANS=CONC makes all the difference in the world for iterative logical names.    L > > > ...and the other two examples also answered the question. Your exampleM > > > illustrates the same anomaly, just from yet another perspective. Six ofc! > > > one, half dozen of another.e > > - > > Nope. You have not answered the question.D > # > Yes, I have. Re-read the above...     , I re-read it. Nope, still doesn't answer it.    J > > > > Then there is the famous DEFINE HOME SYS$LOGIN bug for SET DEFAULT1 > > > > which I already discussed in this thread.J > > >aJ > > > Note, whoever, that SYS$LOGIN is not only already a logical name, itN > > > indicates both a device and a directory path. I believe you'll find that. > > > *THIS* distinction makes the difference. > >  > > Difference for what? >  > Did you read what follows?    ' Yes. And I repeat. Difference for what?      > > > Observe: > > >e > > > $ sh log sys$syloginA > > >    "SYS$SYLOGIN" = "SYS$MANAGER:SYLOGIN" (LNM$SYSTEM_TABLE)f > > > $ type sys$syloginM > > > %TYPE-W-SEARCHFAIL, error searching for SYS$COMMON:[SYSMGR]SYLOGIN.LIS;-  > > > -RMS-E-FNF, file not found > > >M3 > > > You expected that, right? Now check this out:  > > >: > > > $ type sys$sylogin:.com  > > > $!VERIFY = 'F$VERIFY(0)3+ > > > $!IF F$MODE() .EQS. "OTHER" THEN EXIT G > > > $! Remove the comments from the above commands to support the 'r'r > > > commandsN > > > $! when using all OpenVMS TCP/IP products.  Consult vendor documentation	 > > > fora > > > $! more information. > > > [snip] > > >i > > > Does THAT help?  > > 1 > > No, you still haven't addressed the question.s > & > Well, yes I have. Re-read the above.     Nope.a    I > > > ...or do you want the machine to do what you meant and not what youe > > > said?t > >  > > I'll try again.e > >  > > When you do something like > >  > >     $ DEFINE AAA BBB > >     $ DEFINE BBB DKA100: > > 7 > > certain commands work fine, like DIR AAA:[FELDMAN].S > E > ...but not all commands "work fine". There's you're first clue thataE > you're doing something "wrong", or at least maybe not quite kosher.c    E I'm not sure, but the only command that doesn't work with this is SETsF DEFAULT and that program is buggy in other ways as I already posted in other threads.     > > But, when you do > > $ > >     $ DEFINE AAA BBB /TRANS=CONC > >     $ DEFINE BBB DKA100: > > . > > nothing works for AAA. Change the above to > > % > >     $ DEFINE AAA BBB: /TRANS=CONCt > >     $ DEFINE BBB DKA100: > > H > > and everything works fine. OK? Nothing you've posted addresses that. > 6 > Well, yes it does. Let's try putting it another way:    : No it doesn't. You didn't post anything about /TRANS=CONC.     > G > You "meant" "BBB:", but what you told it was "BBB". No matter how yousG > slice it, BBB[FELDMAN]myfile.txt is not now, never has been and nevero# > will be a valid filespec string. D     $ DEFI AAA BBB $ DEFI BBB DKA100: $ SHOW LOG/FUL AAA,    "AAA" [super] = "BBB" (LNM$PROCESS_TABLE)0 1  "BBB" [super] = "DKA100:" (LNM$PROCESS_TABLE) $ DIREC/TOTAL AAA:[FELDMAN]c  a Directory DKA100:[FELDMAN]  o Total of 74 files. $s    B So it works, even though BBB[FELDMAN] isn't a valid file-spec. But< when you do exactly the same thing with /trans=conc added...   $ DEFI AAA BBB/TRAN=CONC; %DCL-I-SUPERSEDE, previous value of AAA has been supersededi $ SHOW LOG/FUL AAA8    "AAA" [super] = "BBB" [concealed] (LNM$PROCESS_TABLE)0 1  "BBB" [super] = "DKA100:" (LNM$PROCESS_TABLE) $ DIREC/TOTAL AAA:[FELDMAN]d6 %DIRECT-E-OPENIN, error opening AAA:[FELDMAN] as inputA -RMS-F-DEV, error in device name or inappropriate device type foru	 operation. $/  D suddenly it fails. So, why is VMS "smart enough" to do it "right" in the first case?e  E No, but DKA100:[FELDMAN] *is* a valid file-spec. Now, from the User'se Manual:   B When the system reads a file specification or device name in a DCLF command line, it examines the file specification or device name to seeA whether the leftmost component is a logical name. If the leftmost2D component ends with a colon, space, comma, or a line terminator (forB example, Return), the system attempts to translate it as a logicalB name. If the leftmost component ends with any other character, the: system does not attempt to translate it as a logical name.  A So, according to that, since AAA ends with a space or Return, and0@ should be translated to BBB. Then elsewhere the manual describes iterative translsation:i  F Logical name translation can be iterative: after the system translatesB a logical name, it repeats the translation process for any logical7 names it finds contained within the first logical name.r  F Now, BBB is clearly a logical name "contained within the first logical= name". So BBB is then translated to DKA100: which in front ofi3 [FELDMAN] forms a perfectly valid file-spec string.i    ) > The fact that "BBB" is itself a logicala: > name makes no fundamental difference, and neither do the > /TRANSLATION_ATTRIBUTES.    0 I've already demonstrated that that is not true.     [snip]H > > What I don't understand is why VMS is "smart enough" to do the rightH > > thing without /TRANS=CONC and without the colon for some commands atC > > least, but cannot to the right thing for any command if you use E > > /TRANS=CONC and don't include the colon, BOTH FOR CASES WHERE THE 2 > > EQUIVALENCE NAME IS ALSO ANOTHER LOGICAL NAME. > J > I hope the example above deepens your insight to the point where you can > start to grasp the concept.d > J > > I guess VMS is "smart enough" in these certain cases without the colon@ > > only by fortuitousness (or should that be "by serendipity"?) > & > I'd say it's more like the way that: > " > $ WRITE SYS$OUTPUT "ALAN FELDMAN >  > ...and...u > # > $ WRITE SYS$OUTPUT "ALAN FELDMAN"c > A > ...are functionally identical, even though the first example isaJ > syntactically incorrect (unbalanced quotes). It's just the nature of the > DCL beast.    D OK, this make sense. We're just wondering why VMS let's you get awayB without the trailing colon, but only if you don't add /trans=conc.    I > If you need a more nuts-and-bolts explanation, you might try writing tocG > Brian Schenkenberger off-line and arrange to visit him and review hishG > OpenVMS source listings CDs to find out *EXACTLY* why things work the0 > way they work.    $ I'd love to see the source listings.    F > On the other hand, if you make it a rule to *ALWAYS* properly formatJ > your translation (read: "equivalence") strings, you'll be a much happier > camper in the long run.e    6 I do "properly" format my equivalence strings. Agreed.    B > BTW, I've been emphasising "equivalence" for a reason. Grasp theH > concept, and you're home free (unless you're British, then "Bob's your
 > Uncle")!     ??? Who's "Bob"? t    J > I was looking for another example to underscore the point, but instead I > turned up more strangeness:h > ! > DJAS01::DDACHTERA$ def djd dka0f% > DJAS01::DDACHTERA$ dir djd:[000000]D# > %DIRECT-W-NOFILES, no files found 1 > DJAS01::DDACHTERA$ say f$parse( "djd:[000000]")h > DKA0:[000000]DKA0.;o* > DJAS01::DDACHTERA$ def djd sys$sysdevice= > %DCL-I-SUPERSEDE, previous value of DJD has been supersededn% > DJAS01::DDACHTERA$ dir djd:[000000]t > " > Directory SYS$SYSDEVICE:[000000] > @ > 000000.DIR;1               1/4        22-MAY-2000 14:48:08.28  [merciful snip]-@ > VOLSET.SYS;1               0/0        22-MAY-2000 14:48:08.28  > (RWED,RWED,RE,)l > ( > Total of 13 files, 39950/39968 blocks., > DJAS01::DDACHTERA$ sh log/fu sys$sysdeviceC >    "SYS$SYSDEVICE" [exec] = "DJAS01$DKA300:" [concealed,terminal]. > (LNM$SYSTEM_TABLE)+ > DJAS01::DDACHTERA$ def djd dka0/tran=conce= > %DCL-I-SUPERSEDE, previous value of DJD has been superseded.% > DJAS01::DDACHTERA$ dir djd:[000000] 7 > %DIRECT-E-OPENIN, error opening DJD:[000000] as inputuC > -RMS-F-DEV, error in device name or inappropriate device type fory > operationtC > DJAS01::DDACHTERA$ say f$parse( "djd:[000000]",,,, "syntax_only")  >  > DJAS01::DDACHTERA$ > > H > Now, don't *THAT* just beat all??!! Even when the logical name equatesF > to a(n unterminated) device name instead of another logical name, it > *STILL* doesn't work!     ? Of course that didn't work. You typed a device name without thee@ trailing colon! It was *you* who pointed this out in one of your previous posts!n    , > I guess what I mean and what I say are two > different things, eh?      Uh...c    D > In the end, I guess ya just gotta properly format your equivalence
 > strings.    / As I summarized in another post in this thread.a    I > Summa num batch, fargin' bastidge, miserable corksucker, stupid icehole  > thing, anyways...      Huh? Excuse me?l     Disclaimer: JMHO Alan E. FeldmanN afeldman gfigroup com    ------------------------------  % Date: Tue, 18 Jun 2002 10:23:39 -0400 + From: Michael Corbett <corbett@PROCESS.COM>I; Subject: Re: SSH v2 (was Re: OpenVMS, Volume Desktop OS...)o* Message-ID: <3D0F426B.4050705@PROCESS.COM>   Bob Ceculski wrote:S    G > TCPware's management interface is VMS like, as a matter of fact, oncerI > you understand the com's well enough, you don't even need an interface. I > If you do, it works just fine, and if you need gui, they have an add-onhI > interface you can purchase, and because it is the only ip stack for vmsrI > based on the vms kernel, it runs crisper, and switching your stack fromoH > ucx to tcpware to multinet to tcpware is not that hard ... been there, > done that ...t >     B 	For the record, there is no add on interface you need to purchaseJ for TCPware.  Most of the configuration is handled by a command procedure,H CNFNET, which prompts you for the various settings, or you can pass it aJ p1 of MENU to get a menu interface.  Most of the configuration informationD is stored in a command procedure, TCPWARE:TCPWARE_CONFIGURE.COM, viaH symbols.  You can edit it to change or add things once you know what youG are doing though we don't recommend it.  Unlike the other TCP/IP stackshG it is not a port of a Unix TCP/IP stack like TCP/IP Services which is asF port of the TRU64 stack and MultiNet which is a port of the BSD stack.   regardsr Mike   -- $K +-------------------------------------------------------------------------+mD Michael Corbett                           Email: Corbett@process.comB Process Software                          Phone: 800 722-7770 x369B 959 Concord St.                                  508 879-6994 x369= Framingham MA 01701-4682                  FAX:   508 879-0042W   ------------------------------  % Date: Tue, 18 Jun 2002 12:19:37 +0100>U From: Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com>t: Subject: Re: Sun is on the way out!  Only HP and IBM left!0 Message-ID: <aen50b$cb6$1@new-usenet.uk.sun.com>   Michael Austin wrote:r  * > Andrew Harrison SUNUK Consultancy wrote: > 4 >>Bob as you know the only bogus claims made on this9 >>group are when people start talking about Alpha Serversd6 >>being high performance systems. You know because you6 >>were asked to come up with examples that proved this7 >>claim and you didn't. Don't be too downhearted a long  >> >  > I > Not to get in the way of your current, ongoing battle, but, what planet>J > do you live on Andrew -- I would think that the Super Computer center inF > Pitt would qualify as a high performance "ALPHA" system as would theE > computing center out there in the desert at a "secret" governmentaltJ > agency. And IIRC they  are rated as "The fastest systems on the planet". >     > ES40/45s not GS boxes, the ES servers are competitive for some> workloads. But they are only workgroup servers. A Sun V880 has) more capacity and costs less for example..  : You also cite a specific workload that works in a parallel$ environment, most DBMS's etc do not.  ; So for the vast majority of commercial workloads you have at: choice of platforms, one has performance numbers that show9 that its a performance leader, one has performance claimsnA made for it but few numbers with the ones available contradictingfD these claims. The former is SPARC Solaris the latter is Alpha/Tru64.    I > Ever wonder why SPARC/Slowaris wasn't included in the RFP? too slow andC > definately not scalable. >     = No sorry, in the commercial server market, integer, DBMS, I/Oa? the numbers are in Sun Solaris's favour and have been since ther; 8400, this is the market that HP have to succeed in not the D HPC market, this is a market that Compaq claimed to have performance3 leadership in, that claim isn't/wasn't justifiable.w     Regards  Andrew Harrisont  ; An ignorant, complacent sales team is the competitions bestu weapon.    ------------------------------  % Date: Tue, 18 Jun 2002 12:39:20 +0100rU From: Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com>T0 Subject: Re: Sun Setting, Not Rising, says eWeek0 Message-ID: <aen65a$cn5$1@new-usenet.uk.sun.com>  - As a matter of interest do you agree with the-4 points made or are you simply posting for comments ?  8 I only ask because some of them are hardly clever if you( are an OpenVMS advocate. But more annon.   Keith Parris wrote:6  F > This isn't the first time a writer has seen parallels between Sun at( > present and Digital in the last 1980s. > : > http://www.eweek.com/print_article/0,3668,a=27737,00.asp >  > Sun Setting, Not Rising? >  > By  Stan Gibson  > G > When you watch something long enough, patterns emerge. When you watch F > something a really long time, you start to think you're watching the > sequel to a B movie. > C > Why is it that every time I hear Scott McNealy or listen to a SunaC > strategy pitch, I think it's circa 1987, and I'm listening to Ken A > Olsen and Digital Equipment execs? For some of you, this may be H > meaningless because you're not old enough to remember. Believe me, the > parallels are there. > F > Digital (now HP) had a proprietary operating system, VMS (now calledF > OpenVMS), and its own chip design, the VAX architecture (now Alpha).H > Sun has Solaris and SPARC. Digital contended its environment was open,G > just as Sun thinks its environment is open; most people think of themr > as proprietary.  > @ > For years, McNealy has been striving to get beyond this simpleG > paradigm, to no avail. If you remember (you probably don't), Sun rode.F > to success on the open system of Unix, which it tried to corner withE > the connivance of AT&T. Then Sun tried to create a clone market for G > the SPARC chip via companies such as Solbourne. Then McNealy tried togB > make Java VM into a computing platform that would usurp platform > dominance from Windows.p >     7 Tried to create and suceeded, Fujitsu build Sun clones.n  7 Incedentally the "corner the market" claims are BS thath; was never the intention and despite Sun and AT&T eventually < winning the OS war after the demise of OSF it didn't happen.  = And Java is upsurping Windows as a development and deploymentt	 platform.   ? Incedentally Java is the best hope that OpenVMS has so knockingd; its sucess and its aims is hardly a good plan if you are an  OpenVMS supporter.    F > Just as Olsen disparaged Unix while his company reluctantly sold it,A > so Sun is holding its nose while offering Linux products in itsa > acquired Cobalt line.a      ? Funny, Sun is the largest commerial supporter of Linux in termsl@ of actually donating stuff that sells Linux boxes and makes them: possible to compete with UNIX/Windows. The donation of theD JVM all be it under terms that some OpenSource people arn't entirelyA happy with is the single most significant donation in this space.   A And thats before you get to StarOffice/OpenOffice, Grid, etc etc. ( These pre-date Sun's purchase of Cobalt.  B If you call that holding your nose then its difficult to see which" planet you may actually reside on.     > C > And another thing: Many good people left Digital because they got H > bored. Gordon Bell and Dave Cutler come to mind. Sun lost Eric Schmidt& > several years ago and now Ed Zander. >     ? All companies change over time, Sun has a track record of being = able to adapt and re-invent. Many people thought that AndreasS0 leaving Sun was a huge problem, but we survived.  = In reality the only similarity between Sun and the old pre NTk< Digital is the focus on one product stack, Sun has proven to< be a very different company when it comes to execution as is only to obvious.     Regardsl   Andrew Harrisona   ------------------------------  + Date: Tue, 18 Jun 2002 11:33:47 +0000 (UTC)h From: david20@alpha2.mdx.ac.uk* Subject: Re: Sun/Slowaris future in doubt!+ Message-ID: <aen5qr$o13$1@aquila.mdx.ac.uk>o  h In article <d7791aa1.0206171051.41f6b319@posting.google.com>, bob@instantwhip.com (Bob Ceculski) writes:> >Since everyone here is so quick to speculate that Itanium/VMS4 >will fail, why are saying that Sun/Hammer will not?? >Makes no sense, esp. when pitting Alpha/VMS engineers againset 9 >Sun/AMD engineers, I would expect Itanium VMS chances ofm3 >survival to be a heck of alot better than Sun/AMD! : >We all know that Sparky chip is dead ... unless they can : >keep adding on more chips ... 512, 1024, 2048, to get the> >thing to look like it can run fast ... so Alpha is dead afterC >EV7, and sparc is dead now ... EV7 will outperform sparc by years,L= >the question is, why buy Sun/Slowaris?  Hammer One right nowR >looks like Itanic one ... >M; >Well computer geniuses, rational competent answers please!:  I Itanic -  Unproven EPIC architecture. Requires major compiler technology FD           improvements. Large performance penalty when running 32bit-           applications  => few applications. m  0 Hammer -  Extension of current 32bit technology.N           No new compiler developments required. However will almost certainlyH           benefit from any of the compiler technology improvements which           IA64 demands. @           No significant penalty for running 32bit applications.*           => Large number of applications.    N Itanic needs some major factor to offset this. Blistering performance on 64bitF applications compared to Hammer might do but that doesn't look likely.  M If SUN does port Solaris to Hammer (I may have missed something but I haven't.G actually seen any such announcement) then I'd say (considering Solaris''D Application portfolio) that Solaris/Hammer will do a lot better than HP-UX/Itanium.    
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------  % Date: Tue, 18 Jun 2002 17:09:25 +0100 U From: Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com>B* Subject: Re: Sun/Slowaris future in doubt!0 Message-ID: <aenlvm$hpc$1@new-usenet.uk.sun.com>   david20@alpha2.mdx.ac.uk wrote:B  j > In article <d7791aa1.0206171051.41f6b319@posting.google.com>, bob@instantwhip.com (Bob Ceculski) writes: > ? >>Since everyone here is so quick to speculate that Itanium/VMS 5 >>will fail, why are saying that Sun/Hammer will not? @ >>Makes no sense, esp. when pitting Alpha/VMS engineers againset: >>Sun/AMD engineers, I would expect Itanium VMS chances of4 >>survival to be a heck of alot better than Sun/AMD!; >>We all know that Sparky chip is dead ... unless they can [; >>keep adding on more chips ... 512, 1024, 2048, to get the0? >>thing to look like it can run fast ... so Alpha is dead afterrD >>EV7, and sparc is dead now ... EV7 will outperform sparc by years,> >>the question is, why buy Sun/Slowaris?  Hammer One right now >>looks like Itanic one ...a >>< >>Well computer geniuses, rational competent answers please! >> > K > Itanic -  Unproven EPIC architecture. Requires major compiler technology "F >           improvements. Large performance penalty when running 32bit/ >           applications  => few applications. - > 2 > Hammer -  Extension of current 32bit technology.P >           No new compiler developments required. However will almost certainlyJ >           benefit from any of the compiler technology improvements which >           IA64 demands.EB >           No significant penalty for running 32bit applications., >           => Large number of applications. >  > P > Itanic needs some major factor to offset this. Blistering performance on 64bitH > applications compared to Hammer might do but that doesn't look likely. > O > If SUN does port Solaris to Hammer (I may have missed something but I haven'thI > actually seen any such announcement) then I'd say (considering Solaris'nF > Application portfolio) that Solaris/Hammer will do a lot better than > HP-UX/Itanium. >     9 There hasn't been one though Hammer looks interesting ande9 AMD actually seem to be trying to build a range of x86-64 ; CPU's to meet low and high end systems requirements, a mucha7 more sensible appraoch than the rather chaotic approach ' being adopted by their main competitor.p   Regardsa Andrew Harrison    ------------------------------  % Date: Tue, 18 Jun 2002 17:14:18 +0100 U From: Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com>K* Subject: Re: Sun/Slowaris future in doubt!0 Message-ID: <aenm8q$ht0$1@new-usenet.uk.sun.com>   Bob Ceculski wrote:   ? > Since everyone here is so quick to speculate that Itanium/VMSe5 > will fail, why are saying that Sun/Hammer will not?a@ > Makes no sense, esp. when pitting Alpha/VMS engineers againset: > Sun/AMD engineers, I would expect Itanium VMS chances of4 > survival to be a heck of alot better than Sun/AMD!; > We all know that Sparky chip is dead ... unless they can a; > keep adding on more chips ... 512, 1024, 2048, to get theB? > thing to look like it can run fast ... so Alpha is dead afterGD > EV7, and sparc is dead now ... EV7 will outperform sparc by years,> > the question is, why buy Sun/Slowaris?  Hammer One right now > looks like Itanic one ...n >     8 Humm for a dead chip Sparky (I assume this is a spelling8 mistake and you mean SPARC) is doing pretty well. Medium6 to Large SPARC servers currently deliver significantly@ better total throughput and per CPU throughput on the comparable6 benchmarks published by both camps and market share is up.t  5 Incedentally, I keep asking you for the justifications4 for your claims of 512/1024/2048 and other BS claims/ you have been scattering about, where are they.'  1 I would suggest that if you cannot come up with ai3 rational competent justification by say Monday (youe6 have had plenty of time) that you desist from this BS.    < > Well computer geniuses, rational competent answers please! >     - How about a rational and competent question ?T   Regardsn Andrew Harrisona   ------------------------------   Date: 18 Jun 2002 16:15:40 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)e* Subject: RE: Sun/Slowaris future in doubt!, Message-ID: <aenmbc$25mg$1@info.cs.uofs.edu>  T In article <BE56C50EA024184DAF48F0B9A47F5CF40266079B@kaoexc01.americas.cpqcorp.net>,*  "Main, Kerry" <Kerry.Main@hp.com> writes: |>H |> However, the growing popularity of vendors adoption of Java means app  E I wouldn't put too much stock in the current Java craze.  Nothing newkI there.  We have had attempts at wide application popularity via "virtual" F machine before (can anyone say P-System??)  In a business as fickle asF this one, that could change tomorrow if someone comes along with a newF language with more glitz (I have visions of the Monorail salesman from Simpsons!!)T  K |> vendors and Customers can have the best of both worlds ie. rock solid OSsC |> platform from a security and ultra-high availability perspectiveYI |> combined with lots more applications and App developers (many Colleges0B |> and Universities are moving to Java as their preferred teaching
 |> language).0  ? While at the same time privately discussing all the reasons why3@ this is a bad idea.  Like it or not, education is a business and> sometimes decisions are made for marketing reasons rather than= technical reasons. Java is this years buzzword.  It sells thenD program to students and, yes, to parents as well.  I don't much care< for Ada, personally, but even I freely admit it was a better@ beginners teaching language than Java or C or C++.  I've watchedD us go from Pascal to Ada to Java in one decade.  I don't necessarilyC agree with the moves, but I do know why they were done and what theC. likely effect of not doing it would have been.   bill   -- hJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------   Date: 18 Jun 2002 16:44:07 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)c* Subject: Re: Sun/Slowaris future in doubt!, Message-ID: <aeno0n$267e$3@info.cs.uofs.edu>  + In article <aen5qr$o13$1@aquila.mdx.ac.uk>,d!  david20@alpha2.mdx.ac.uk writes:b |> g& |> If SUN does port Solaris to Hammer   B I thought one of the differentiators od Hammer was it's ability toE run existing x86 code??  Solaris already runs on that, so even before-7 a 64bit version comes out they will already support it.   P |>                                    (I may have missed something but I haven'tJ |> actually seen any such announcement) then I'd say (considering Solaris'G |> Application portfolio) that Solaris/Hammer will do a lot better thano |> HP-UX/Itanium.a  > And this ignores the fact that unlike Compaq/Alpha, Sun is not3 abandoning support for or development of the Sparc.y   bill   -- hJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   .   ------------------------------    Date: 18 Jun 2002 08:01:54 -0600- From: koehler@encompasserve.org (Bob Koehler)o Subject: Re: unix historyr3 Message-ID: <y3BQ41R6anvz@eisner.encompasserve.org>e  c In article <3D0DE18D.5273DD4B@mindspring.com>, Atlant Schmidt <atlantnospam@mindspring.com> writes:n > ) > Gee Bob, I thought we already clarifiedW& > the fact that I said "*APPLICATIONS*- > in my original posting. (I'll happily grant   > you that the OS got it right.)  ?    Nope.  In your original post you specifically said VMS.  Anyo8    application that doens't get it right is on it's own.   ------------------------------  % Date: Tue, 18 Jun 2002 09:17:25 -0400a2 From: Atlant Schmidt <atlantnospam@mindspring.com> Subject: Re: unix history-. Message-ID: <3D0F32E5.E5B39947@mindspring.com>   Bob Koehler wrote:  e > In article <3D0DE18D.5273DD4B@mindspring.com>, Atlant Schmidt <atlantnospam@mindspring.com> writes:  > > + > > Gee Bob, I thought we already clarifiedm( > > the fact that I said "*APPLICATIONS*/ > > in my original posting. (I'll happily grant-" > > you that the OS got it right.) >0A >    Nope.  In your original post you specifically said VMS.  Any_: >    application that doens't get it right is on it's own.   Yup. Go back and read it again.    Atlant   ------------------------------  % Date: Tue, 18 Jun 2002 09:20:47 -0400U2 From: Atlant Schmidt <atlantnospam@mindspring.com> Subject: Re: unix historys- Message-ID: <3D0F33AF.C1DA050@mindspring.com>l   Bob Koehler wrote:  e > In article <3D0DE18D.5273DD4B@mindspring.com>, Atlant Schmidt <atlantnospam@mindspring.com> writes:> > > + > > Gee Bob, I thought we already clarifiedn( > > the fact that I said "*APPLICATIONS*/ > > in my original posting. (I'll happily grant " > > you that the OS got it right.) >kA >    Nope.  In your original post you specifically said VMS.  Anyp: >    application that doens't get it right is on it's own.  G I will call to your attention the following phrase in my original post:   6 > we'll get to see how many C applications (on *MANY*,+ > *MANY* operating systems *INCLUDING* VMS)o  3 I was very careful to state that precisely; note ino7 particular my use of the words "applications" and "on".f   Atlant  $  -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-   Nic Clews wrote:  " > UNIX has only 67 years to go :-) >.  > VMS has another 290 CENTURIES.  2 Actually, much of the excrement's going to hit the< ventilator on Tuesday, January 19, 2038 at GMT 03:14:07 UTC.9 The next second, the clock spills over into bit <31>, andb4 we'll get to see how many C applications (on *MANY*,) *MANY* operating systems *INCLUDING* VMS)i. mistakenly treated time_t as a signed integer.   Atlant   ------------------------------   Date: 18 Jun 2002 15:59:24 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)t Subject: Re: unix historyr, Message-ID: <aenlcs$24rn$1@info.cs.uofs.edu>  ) In article <oMUmVMMuRfcg@elias.decus.ch>,v,  p_sture@elias.decus.ch (Paul Sture) writes: |> k |> a5 |> So which of the BSDs would you recommend, and why?a  6 Which tool wqould you recommend for building a house??  C Even among the existing BSD's there is enough to differentiate themuB to make it a serious engineering decision.  If you mean on generalD terms, assuming x86, Alpha or PPC, up until a few months ago I wouldA have said FreeBSD.  I would still recommend it, but maybe not the A latest release.  The product is superior, but the packaging seems>B to have fallen on bad times with the departure of some of the core
 team members.o  ? Of course, for the PDP-11 most people recommend BSD 2.11 but mycC preference is for Ultrix-11, although that's not really a BSD.  :-)s   bill   -- hJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   y   ------------------------------  % Date: Tue, 18 Jun 2002 09:21:27 -0700n# From: "Tom Linden" <tom@kednos.com>i Subject: RE: unix historye9 Message-ID: <CIEJLCMNHNNDLLOOGNJIOEFGFDAA.tom@kednos.com>S   >-----Original Message-----a9 >From: Bill Gunshannon [mailto:bill@triangle.cs.uofs.edu]i% >Sent: Tuesday, June 18, 2002 8:59 AMn >To: Info-VAX@Mvb.Saic.Com >Subject: Re: unix history >s >o* >In article <oMUmVMMuRfcg@elias.decus.ch>,- > p_sture@elias.decus.ch (Paul Sture) writes:w >|>  >|> 6 >|> So which of the BSDs would you recommend, and why? >n7 >Which tool wqould you recommend for building a house??  >uD >Even among the existing BSD's there is enough to differentiate themC >to make it a serious engineering decision.  If you mean on generaleE >terms, assuming x86, Alpha or PPC, up until a few months ago I wouldeB >have said FreeBSD.  I would still recommend it, but maybe not theB >latest release.  The product is superior, but the packaging seemsC >to have fallen on bad times with the departure of some of the corei >team members. >e@ >Of course, for the PDP-11 most people recommend BSD 2.11 but myD >preference is for Ultrix-11, although that's not really a BSD.  :-)  C Are you sure about that? I always thought that it was derived from  B BSD4.1, to which I had the sources.  It certainly behaved like it.    >r >billa >  >-- K >Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolves E >bill@cs.scranton.edu     |  and a sheep voting on what's for dinner.u >University of Scranton   |tB >Scranton, Pennsylvania   |         #include <std.disclaimer.h>    >  >---' >Incoming mail is certified Virus Free.m; >Checked by AVG anti-virus system (http://www.grisoft.com). @ >Version: 6.0.370 / Virus Database: 205 - Release Date: 6/5/2002 >i --- & Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).? Version: 6.0.370 / Virus Database: 205 - Release Date: 6/5/2002f   ------------------------------  % Date: Tue, 18 Jun 2002 12:28:23 -0400 * From: "rob kas" <rob@paychoice.nospam.com>> Subject: Re: US FDA ponders Slowaris over VMS ... what morons!3 Message-ID: <3d0f5fa5$0$1421$8e9e3842@news.atx.net>o             Hey Bob  > G > well Rob, let's see, you are going to port to Slowaris/sparky that is  > on a soon to be eol platform,.  @   mmmm Any chance you have something  you can back that up with?    ' soon to be slowaris/hammer, a chip that F > by recent examination has all the makings of Itanic One, an os beingC > ported, just like vms, to an unproven architecture (hammer), on aiE > company that is financial trouble ... looks like you are breaking aeE > few of your rules, isn't it?  I would place my bet on the Alpha/VMSnB > engineers (itanium/vms) over Sun/AMD engineers (slowaris/hammer) > everytime!    K  With all due respect Former Alpha engineers are working on AMD's chip, ands) have a lot less overhead to plow through.o: IA64/EPIC would seem to need a good deal more shaking out.@  Being able to run Solaris IA32 binarys and full speed is a huge   win.      7  Even IBM is in trouble, they just bought back secretlyUE > 34 million of their own shares to make their dismal latest earningsiC > look decent ... while hp has printers/high end revenue to help it<D > weather the storm ... I fail to see your logic ... mine looks much > better I think ..:  8      Well everyone can have a opinion even if it's wrong                               Rob .    ------------------------------  + Date: Tue, 18 Jun 2002 07:48:21 -0500 (CDT)n? From: "Doc.Cypher" <Use-Author-Supplied-Address-Header@[127.1]> 5 Subject: Re: What kind of morons work at the FDA now?W< Message-ID: <200206181248.g5ICmLXg085577@cryptofortress.com>  8 On Mon, 17 Jun 2002, bugs@pu.net (Mark Hittinger) wrote:$ >Are there kickbacks for buying VMS? >p
 >Later :-)   I'll bet Bob didn't get any.  ? He loves VMS so much he probably asked them to take more money.u     Doc. --  6 The bigger the humbug, the better people will like it.K ~ Phineas Taylor Barnum.                             https://vmsbox.cjb.netr   ------------------------------    Date: 18 Jun 2002 05:08:02 -0700% From: Bart.Zorn@xs4all.nl (Bart Zorn)t3 Subject: Re: Why is TCPIP$ROUTE.DAT in SYS$COMMON ?a= Message-ID: <a98cd882.0206180408.613a005b@posting.google.com>g  h Kilgallen@SpamCop.net (Larry Kilgallen) wrote in message news:<tsXAkpCWAxJ3@eisner.encompasserve.org>...] > In article <3D0C9C55.6060803@xs4all.nospam.nl>, Bart Zorn <B.Zorn@xs4all.nospam.nl> writes:i > > Hans Vlems wrote:gM > >> IMHO the last thing you'd want in a VMS cluster is that its availabilitye	 > >> getsoM > >> affected by a an external unit such as an IP router. Putting the routing 
 > >> database M > >> in SYS$COMMON at least suggests to have all cluster nodes in the same IPa
 > >> network.hL > >> There are probably many sites that still happily use 10.* and 192.168.* > >> networks on their LAN all- > >> routed by one box (per set of networks).  > > M > > Since when is the availability of an OpenVMS cluster dependant on TCP/IP?r > > F > > OK, if the only way to access the cluster from outside is through  > > TCP/IP, you may be rigth.  > E > Since when is the availability of a VMS cluster dependent on accessn > "from outside" ?    :-)     @ It depends on how you define availability. You could consider anB OpenVMS cluster with 100% uptime unavailable if there is no way to login to it!  	 Bart Zorni   ------------------------------    Date: 18 Jun 2002 07:22:22 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)a3 Subject: Re: Why is TCPIP$ROUTE.DAT in SYS$COMMON ?e3 Message-ID: <Z5OOKFh5PFNO@eisner.encompasserve.org>u  e In article <a98cd882.0206180408.613a005b@posting.google.com>, Bart.Zorn@xs4all.nl (Bart Zorn) writes:Mj > Kilgallen@SpamCop.net (Larry Kilgallen) wrote in message news:<tsXAkpCWAxJ3@eisner.encompasserve.org>...  F >> Since when is the availability of a VMS cluster dependent on access >> "from outside" ?    :-) >  > B > It depends on how you define availability. You could consider anD > OpenVMS cluster with 100% uptime unavailable if there is no way to > login to it!  G I define availability as able to be used in the normal application flowrG for that cluster.  That may involve all use being from VMS workstations I that are members of the cluster.  In other situations, it may involve alll use being from batch jobs.   ------------------------------  % Date: Tue, 18 Jun 2002 06:42:59 -0400o' From: "Main, Kerry" <Kerry.Main@hp.com>vB Subject: RE: Will AIT-2 or AIT-3 media/changers work with OpenVMS?T Message-ID: <BE56C50EA024184DAF48F0B9A47F5CF40266079C@kaoexc01.americas.cpqcorp.net>   Mike,y  H <<< I am currently looking for a high capacity Library solution, and the AIT-3 looks really good>>>  ? As a suggestion, the following references might be of interest: F http://www.compaq.com/alphaserver/ (select AlphaServer quick specs for options supported)B http://www.compaq.com/products/quickspecs/11332_div/11332_div.html (really high end tapes)nC http://www.compaq.com/storage/siteindex.html (StorageWorks main webi site)    Regardsm  
 Kerry Main Senior ConsultantP Hewlett-Packard Canada! Consulting & Integration Serviceso Voice: 613-592-4660  Fax   : 613-591-4477 Email: Kerry.Main@hp.com     -----Original Message-----< From: Michael carlson [mailto:mcarlson@m87-blackhole.org]=20 Sent: June 18, 2002 12:42 AM To: Info-VAX@Mvb.Saic.Come> Subject: Will AIT-2 or AIT-3 media/changers work with OpenVMS?     Specifically in 7.3a  D I am currently looking for a high capacity Library solution, and theG AIT-3 looks really good. I think OpenVMS can use AIT tape drives, but IoE dont know if it will support any drive beyond that. Any info would be>H great, as I would like to attain a spectralogic library since I've had a+ lot of success on various UNIX's with them.n  D Are there tape size limitations in VMS? And would a SDLT drive would instead?  & Any info would be greatly appreciated.   Mike Carlson   ------------------------------  % Date: Tue, 18 Jun 2002 09:50:45 -0700a' From: JMK <jeffrey.m.klopotic@lmco.com>s Subject: Re: __asm__ in DEC CXXr( Message-ID: <3D0F64E5.6BAF666B@lmco.com>   Craig, Thanks for the help!  A You know, I did my RTFM, and when a topic as blatant as assembler7; doesn't make the index or glossary of a major compiler, theS) documentation is in serious need of help.   G But it is great that we have this list, and I am most greatful that you  responded.     Thanks again for the help!  
 Jeff Klopotic    "Craig A. Berry" wrote:e > * > In article <3D0A659B.F6DF2485@lmco.com>,+ >  JMK <jeffrey.m.klopotic@lmco.com> wrote:p > I > > I have some C++ code that has some assembler in it, and it appears bytG > > the conspicuous lack of any reference to assembler in my DEC/Compaq2- > > manuals that ... oh, well, I'd just vent.e > >aK > > How does one invoke what would be an otherwise straightforward assembly ! > > routine in DEC/Compaq/HP CXX?  > H > 30 seconds with the docs reveals that there is a mechanism for in-lineI > assembly available via "#pragma intrinsic (asm)" though as someone elsebI > pointed out, assembly language is by definition implementation-defined,-I > so whatever assembly language you've got is unlikely to be VAX or Alphat > assembly language. >  > You might start with >  > $ help cxx language builtine > F > and scroll down to the section entitled "Intrinsic functions".  AlsoD > see Appendix C, "Built-in Functions" in the document "Using Compaq > C++".a   ------------------------------   End of INFO-VAX 2002.336 ************************