INFO-VAX Tue, 17 Jul 2007 Volume 2007 : Issue 388 Contents: Re: Debugging shareable images weirdness Re: Debugging shareable images weirdness Re: Debugging shareable images weirdness Re: Debugging shareable images weirdness Re: Debugging shareable images weirdness Re: Debugging shareable images weirdness Re: Debugging shareable images weirdness Re: Debugging shareable images weirdness RE: FTSO/FASTCopy replacement Re: Linux (was Re: How many people here use Itanium w VMS) Re: Linux (was Re: How many people here use Itanium w VMS) OT: Their shamelessness knows no bounds. Re: OT: Their shamelessness knows no bounds. PC EDT keypad (was EDT Replacement) Re: PC EDT keypad (was EDT Replacement) Re: symlink to another device What does GEM mean? Re: What does GEM mean? Re: What does GEM mean? Re: What happen to the Deathrow cluster Re: What OS version to run on a DEC 3000-M600 Re: What OS version to run on a DEC 3000-M600 ---------------------------------------------------------------------- Date: Tue, 17 Jul 2007 00:01:07 -0700 From: dooleys@snowy.net.au Subject: Re: Debugging shareable images weirdness Message-ID: <1184655667.455244.29330@x35g2000prf.googlegroups.com> On Jul 17, 7:35 am, Neil Lowden wrote: > Hi VAXman > > Thanks for that, what you say makes perfect sense. Do you know if this > is documented anywhere? > > Regards > > -N The debugger manual suggests that you don't install it "You do not have to install a shareable image to debug it. Instead, you can debug your own private copy by assigning a logical name to it." and "Define a logical name to point to the local copy of the shareable image. You must specify the device and directory as well as the image name." Phil ------------------------------ Date: Tue, 17 Jul 2007 11:04:17 +0200 From: "P. Sture" Subject: Re: Debugging shareable images weirdness Message-ID: In article <1184655667.455244.29330@x35g2000prf.googlegroups.com>, dooleys@snowy.net.au wrote: > On Jul 17, 7:35 am, Neil Lowden wrote: > > Hi VAXman > > > > Thanks for that, what you say makes perfect sense. Do you know if this > > is documented anywhere? > > > > Regards > > > > -N > The debugger manual suggests that you don't install it > > "You do not have to install a shareable image to debug it. > Instead, you can debug your own private copy by assigning a logical > name to it." > > and > > "Define a logical name to point to the local copy of the shareable > image. > You must specify the device and directory as well as the image name." > In case it's not obvious from that, define it as a process logical name, not a system wide one, so that it's completely private. -- Paul Sture ------------------------------ Date: Tue, 17 Jul 2007 09:25:48 -0400 From: John Reagan Subject: Re: Debugging shareable images weirdness Message-ID: VAXman- @SendSpamHere.ORG wrote: > * The Alpha uses a PALcode to implement the BPT. I don't have the IA64 > instruction manual before me but it too has a similar mechanism. > The Itanium has an instruction called "break" that takes a literal argument. One of those literal values has been reserved for breakpoint. To set a breakpoint, the debugger rewrites the appropriate bundle with a break instruction in one of the slots. I'll ask Jeff about a documentation pointer. -- John Reagan OpenVMS Pascal/Macro-32/COBOL Project Leader Hewlett-Packard Company ------------------------------ Date: Tue, 17 Jul 2007 13:41:44 GMT From: VAXman- @SendSpamHere.ORG Subject: Re: Debugging shareable images weirdness Message-ID: In article , John Reagan writes: > > >VAXman- @SendSpamHere.ORG wrote: > >> * The Alpha uses a PALcode to implement the BPT. I don't have the IA64 >> instruction manual before me but it too has a similar mechanism. >> > >The Itanium has an instruction called "break" that takes a literal >argument. One of those literal values has been reserved for breakpoint. > To set a breakpoint, the debugger rewrites the appropriate bundle with >a break instruction in one of the slots. I knew there was an instruction but I did not recall the specifics as to whether it was a whole bundle supplanted or just an instruction in the bundle. Generally, it needs to modify the instruction stream as with the Alpha and the VAX. -- VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)COM "Well my son, life is like a beanstalk, isn't it?" ------------------------------ Date: Tue, 17 Jul 2007 06:48:18 -0700 From: "Tom Linden" Subject: Re: Debugging shareable images weirdness Message-ID: On Tue, 17 Jul 2007 02:04:17 -0700, P. Sture wrote: > In article <1184655667.455244.29330@x35g2000prf.googlegroups.com>, > dooleys@snowy.net.au wrote: > >> On Jul 17, 7:35 am, Neil Lowden wrote: >> > Hi VAXman >> > >> > Thanks for that, what you say makes perfect sense. Do you know if this >> > is documented anywhere? >> > >> > Regards >> > >> > -N >> The debugger manual suggests that you don't install it >> >> "You do not have to install a shareable image to debug it. >> Instead, you can debug your own private copy by assigning a logical >> name to it." >> >> and >> >> "Define a logical name to point to the local copy of the shareable >> image. >> You must specify the device and directory as well as the image name." >> > > In case it's not obvious from that, define it as a process logical name, > not a system wide one, so that it's completely private. > define/user ? -- PL/I for OpenVMS www.kednos.com ------------------------------ Date: Tue, 17 Jul 2007 14:46:18 GMT From: VAXman- @SendSpamHere.ORG Subject: Re: Debugging shareable images weirdness Message-ID: <_C4ni.94$ip4.69@newsfe12.lga> In article , "Tom Linden" writes: > > >On Tue, 17 Jul 2007 02:04:17 -0700, P. Sture > wrote: > >> In article <1184655667.455244.29330@x35g2000prf.googlegroups.com>, >> dooleys@snowy.net.au wrote: >> >>> On Jul 17, 7:35 am, Neil Lowden wrote: >>> > Hi VAXman >>> > >>> > Thanks for that, what you say makes perfect sense. Do you know if this >>> > is documented anywhere? >>> > >>> > Regards >>> > >>> > -N >>> The debugger manual suggests that you don't install it >>> >>> "You do not have to install a shareable image to debug it. >>> Instead, you can debug your own private copy by assigning a logical >>> name to it." >>> >>> and >>> >>> "Define a logical name to point to the local copy of the shareable >>> image. >>> You must specify the device and directory as well as the image name." >>> >> >> In case it's not obvious from that, define it as a process logical name, >> not a system wide one, so that it's completely private. >> >define/user ? That would work if you execute the image that will invoke it before the definition is no longer valid. A simple $ DEFINE will deifne the logical in the process address space such that it can be subsequently debugged. I, for one, would not INSTALL anything on a system unless it was exhaust- ively debugged first. I don't know why the PO was INSTALLing in the first place but localizing the debugging process with a process logical is a far better method. -- VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)COM "Well my son, life is like a beanstalk, isn't it?" ------------------------------ Date: Tue, 17 Jul 2007 07:55:55 -0700 From: "Tom Linden" Subject: Re: Debugging shareable images weirdness Message-ID: On Tue, 17 Jul 2007 07:46:18 -0700, VAXman- <@SendSpamHere.ORG> wrote: > In article , "Tom Linden" > writes: >> >> >> On Tue, 17 Jul 2007 02:04:17 -0700, P. Sture >> wrote: >> >>> In article <1184655667.455244.29330@x35g2000prf.googlegroups.com>, >>> dooleys@snowy.net.au wrote: >>> >>>> On Jul 17, 7:35 am, Neil Lowden wrote: >>>> > Hi VAXman >>>> > >>>> > Thanks for that, what you say makes perfect sense. Do you know if >>>> this >>>> > is documented anywhere? >>>> > >>>> > Regards >>>> > >>>> > -N >>>> The debugger manual suggests that you don't install it >>>> >>>> "You do not have to install a shareable image to debug it. >>>> Instead, you can debug your own private copy by assigning a logical >>>> name to it." >>>> >>>> and >>>> >>>> "Define a logical name to point to the local copy of the shareable >>>> image. >>>> You must specify the device and directory as well as the image name." >>>> >>> >>> In case it's not obvious from that, define it as a process logical >>> name, >>> not a system wide one, so that it's completely private. >>> >> define/user ? > > That would work if you execute the image that will invoke it before the > definition is no longer valid. A simple $ DEFINE will deifne the logical > in the process address space such that it can be subsequently debugged. Right. I use it, for example, when debugging a new PLIRTL, since there is already the older one installed, so it won't interefer with other users. > > I, for one, would not INSTALL anything on a system unless it was exhaust- > ively debugged first. I don't know why the PO was INSTALLing in the > first > place but localizing the debugging process with a process logical is a > far > better method. > -- PL/I for OpenVMS www.kednos.com ------------------------------ Date: Tue, 17 Jul 2007 17:21:34 +0000 (UTC) From: moroney@world.std.spaamtrap.com (Michael Moroney) Subject: Re: Debugging shareable images weirdness Message-ID: JF Mezei writes: >re: defining a process logical to point to your own copy of a shareable >image. >Some applications require that such logical be defined with /EXEC. Not >sure how it is implemented (perhaps they get the shareable image in via >lib$find_image_symbol and before that, they do their own restrictted >logical name translation to ensure a non privileged user cannot >subsititute the real image with his own evil image. Being a process logical and being defined /EXEC are not mutually exclusive. A logical can be defined processwide, jobwide, groupwide or systemwide with increasing scopes. A logical can be defined with trust levels as well, /USER (logical deleted at image rundown), the default /SUPERVISOR, and /EXEC. Images installed with privileges will only look at /EXEC definitions as part of VMS security. Otherwise it would be fairly simple to create a boobytrap LIBRTL or something to take advantage of a priviledged image. ------------------------------ Date: Tue, 17 Jul 2007 13:17:28 +0000 From: "Main, Kerry" Subject: RE: FTSO/FASTCopy replacement Message-ID: > -----Original Message----- > From: Pierre [mailto:pierre.bru@gmail.com] > Sent: July 13, 2007 5:59 PM > To: Info-VAX@Mvb.Saic.Com > Subject: Re: FTSO/FASTCopy replacement > > > > yes. we are currently using FTSO but we exprient some problems > with > > > it > > > (for ex. sometimes copy jobq vanish) and would like some more > > > robust and supported product. > > > > Are you sure you are not running into some process quota issues ? > > as each job is a separate entry in FTSO batch queue, to what kind > of > process quota issue are you thinking of ? > > Pierre. Not sure, but easy way to monitor if you have VMS V8.3 is with Monitor Proc= ess (click Q and it displays quota of process being monitored. Other watch = process dcl utilities are around for older versions of OpenVMS. Also, Avail= ability Manager (AM) is free on the HP web site (my preferred utility for m= onitoring). AM runs on OpenVMS or Windows clients. http://h71000.www7.hp.com/openvms/products/availman/index.html Regards Kerry Main Senior Consultant HP Services Canada Voice: 613-592-4660 Fax: 613-591-4477 kerryDOTmainAThpDOTcom (remove the DOT's and AT) OpenVMS - the secure, multi-site OS that just works. ------------------------------ Date: Tue, 17 Jul 2007 13:52:08 GMT From: "Jeffrey H. Coffield" Subject: Re: Linux (was Re: How many people here use Itanium w VMS) Message-ID: JF Mezei wrote: > Ron Johnson wrote: >> Just about *every* open source app that runs in 32-bit Linux also >> runs in the PPC64, Alpha, x86-64 & IA64 Linux. > > What about stuff like the Adobe reader ? Isn't Oracle now available on > Linux ? Does Oracle provide binaries for all CPU platforms ? The Linux Itanium servers I sold to my customer only run Apache/Tomcat in a DMZ. All the database are on OpenVMS on the internal network. The customer is migrating all their desktops from MS to Linux so they have a great deal of experience with Linux and since their previous web servers ran Linux, they considered that safer than running Apache/Tomcat on OpenVMS. Jeff Coffield ------------------------------ Date: Tue, 17 Jul 2007 07:18:21 -0700 From: "Tom Linden" Subject: Re: Linux (was Re: How many people here use Itanium w VMS) Message-ID: On Tue, 17 Jul 2007 06:52:08 -0700, Jeffrey H. Coffield wrote: > JF Mezei wrote: >> Ron Johnson wrote: >>> Just about *every* open source app that runs in 32-bit Linux also >>> runs in the PPC64, Alpha, x86-64 & IA64 Linux. >> What about stuff like the Adobe reader ? Isn't Oracle now available on >> Linux ? Does Oracle provide binaries for all CPU platforms ? > > The Linux Itanium servers I sold to my customer only run Apache/Tomcat > in a DMZ. All the database are on OpenVMS on the internal network. The > customer is migrating all their desktops from MS to Linux so they have a > great deal of experience with Linux and since their previous web servers > ran Linux, they considered that safer than running Apache/Tomcat on > OpenVMS. Why not upgrade to WASD? > > Jeff Coffield -- PL/I for OpenVMS www.kednos.com ------------------------------ Date: Tue, 17 Jul 2007 14:47:59 GMT From: VAXman- @SendSpamHere.ORG Subject: OT: Their shamelessness knows no bounds. Message-ID: Of personal interest... http://tmesis.com/seirea.html Please help. -- VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)COM "Well my son, life is like a beanstalk, isn't it?" ------------------------------ Date: Tue, 17 Jul 2007 12:08:16 -0400 From: JF Mezei Subject: Re: OT: Their shamelessness knows no bounds. Message-ID: <85add$469ce9b6$cef8887a$1522@TEKSAVVY.COM> VAXman- @SendSpamHere.ORG wrote: > Of personal interest... > > http://tmesis.com/seirea.html Listen to radio stations hosted outside the USA. Or if you operate your own, move your server over to canada or whatever other country that is nearest to you. Note that many USA radio stations now have their web sites setup to refuse connections from IPs originating outside the USA, claiming the RIAA/whatever has now prevented them from broadcasting outside of the USA due to those royalties things. ------------------------------ Date: Tue, 17 Jul 2007 08:13:33 -0800 From: "C.W.Holeman II" Subject: PC EDT keypad (was EDT Replacement) Message-ID: <139pn3si2c2ic49@corp.supernews.com> Bob Koehler wrote: > emacs has had EDT keybindings for a long, long time. But not all > keyboards have those keys. The keyboard keypad that I am using has a double height "+" key. Are there keyboards that work on a PC that have the EDT keypad? -- C.W.Holeman II | cwhii@Julian5Locals.com-5 http://JulianLocals.com/cwhii To only a fraction of the human race does God give the privilege of earning one's bread doing what one would have gladly pursued free, for passion. I am very thankful. The Mythical Man-Month Epilogue/F.P.Brooks ------------------------------ Date: 17 Jul 2007 12:14:13 -0500 From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) Subject: Re: PC EDT keypad (was EDT Replacement) Message-ID: <3ET1haRZU2Lx@eisner.encompasserve.org> In article <139pn3si2c2ic49@corp.supernews.com>, "C.W.Holeman II" writes: > Bob Koehler wrote: >> emacs has had EDT keybindings for a long, long time. But not all >> keyboards have those keys. > > The keyboard keypad that I am using has a double height "+" key. Are > there keyboards that work on a PC that have the EDT keypad? I'm using a USB keypad from Targus that actually has more keys than a DEC VT style keypad. I'm quite happy with it. Picked it up at Target fairly cheap. ------------------------------ Date: Tue, 17 Jul 2007 13:43:26 GMT From: "John E. Malmberg" Subject: Re: symlink to another device Message-ID: <1I3ni.30997$Fc.8230@attbi_s21> Pierre wrote: > On Jul 16, 3:19 pm, s...@antinode.org (Steven M. Schweda) wrote: > >> >> As the documentation says, >> >>http://h71000.www7.hp.com/doc/83final/5763/5763pro_021.html#symlink_chap >> >>the C RTL is expecting symlink text to be a POSIX pathname, which yours >>is not. I don't have a convenient test system at the moment, but you >>might try using more slashes. For example, "/$1$dka100/test/foo.com". See section 12.3.1 of the above link. The C RTL has nothing to do with the interpretation of symbolic links, it is all done in RMS. Until logical name support as having those names be implied mount points is in RMS, the POSIX compliant pathname support in the CRTL will not be of much use to most programmers. Think of it as a preview. The DECC$POSIX_COMPLIANT_PATHNAME settings other than 0 disable the use of logical names in the resolution of "UNIX" pathnames in the CRTL, for all libraries/shared images that are written in C. Most of those libraries absolutely depend on logical names working in UNIX pathnames in order to function. Some of them expect search lists to work, so that just following the UNIX trick of adding additional mount points or symbolic links to compensate will not work. Symbolic links are usable now as long as the POSIX root is set up, and do not depend on any DECC feature setting, as it is RMS that is doing the interpretation of the link. I do not know when the RMS support for logical names in Posix pathnames will be released. Also, like on UNIX, the contents of a symbolic link do not have to be a pathname, but can be just random text to be retrieved by a readlink() routine. Any program that was doing UNIX to VMS filename translations on its own will probably break if it gets a "^UP^" filename. -John wb8tyw@qsl.network Personal Opinion Only ------------------------------ Date: Tue, 17 Jul 2007 08:35:49 -0700 From: Neil Rieck Subject: What does GEM mean? Message-ID: <1184686549.609123.60580@g37g2000prf.googlegroups.com> MIPS, in particular, became known as much for its compiler technology as for its processor architectures. MIPS pursued an approach to compiler systems involving language-specific "front ends" that convert programs into one common intermediate encoded form. A common "back end" then analyzes and optimizes the intermediate expression of the program and generates actual machine instructions. A compiler system composed of such front and back ends can be modified easily as languages change, as another language must be supported, and as new hardware implementations require different optimizations. Digital Equipment Corporation developed the well-respected GEM compiler technology at a time when its line of VAX systems (CISC) was complemented by a line of MIPS-based systems (RISC), before the Alpha architecture (RISC) arrived. This GEM technology made it possible to offer compatible language compilers for both VAX and Alpha systems, thus facilitating a migration of customer applications from 32- to 64- bit systems, especially those for the OpenVMS programming environment. I've already killed an hour today and haven't been able to find out what the letters in GEM mean? Anyone out there know? Neil Rieck Kitchener/Waterloo/Cambridge, Ontario, Canada. http://www3.sympatico.ca/n.rieck/ ------------------------------ Date: Tue, 17 Jul 2007 13:06:47 -0400 From: "FredK" Subject: Re: What does GEM mean? Message-ID: "Steve Davidson" wrote in message news:576ni.36614$G23.1262@newsreading01.news.tds.net... > GEM didn't stand for anything. It came into being as a result of other > names in use at the time. OPAL and PRISM come to mind... > That was my guess also. ------------------------------ Date: Tue, 17 Jul 2007 10:50:53 -0700 From: "Tom Linden" Subject: Re: What does GEM mean? Message-ID: On Tue, 17 Jul 2007 08:35:49 -0700, Neil Rieck wrote: > > MIPS, in particular, became known as much for its compiler technology > as for its processor architectures. MIPS pursued an approach to > compiler systems involving language-specific "front ends" that convert > programs into one common intermediate encoded form. A common "back > end" then analyzes and optimizes the intermediate expression of the > program and generates actual machine instructions. A compiler system > composed of such front and back ends can be modified easily as > languages change, as another language must be supported, and as new > hardware implementations require different optimizations. Mips designed their HW in conjunction with their compilers and the instruction set reflected the requirements of their intermediate (quads) form. We were doing that in 1978, with PL/I Cobol fortan Pascal. VCG was Digital's effort to do essentially the same and they did have at one time PL/I, Pascal, SCAN, C, Pearl and Coral66. Owing to lack of strong management at Digital this never achieved the same success as it had at Prime, DG and Wang. From a theoretical point of View, I don't believe that GEM advances the state of the art substantially beyond that achieved with VCG. It certainly has added capability, e.g., OOE, but that is an implementation issue that can be easily integrated once the program is reduced to an acyclic flow graph. In my view, VPO (Very Portable Optimizer, by Jack Davidson) an earlier version of which is the genesis of the gnu backend(s) is more interesting employing a grammar for hardware definition and using Register Transfer Lists. > > Digital Equipment Corporation developed the well-respected GEM > compiler technology at a time when its line of VAX systems (CISC) was > complemented by a line of MIPS-based systems (RISC), before the Alpha > architecture (RISC) arrived. This GEM technology made it possible to > offer compatible language compilers for both VAX and Alpha systems, > thus facilitating a migration of customer applications from 32- to 64- > bit systems, especially those for the OpenVMS programming environment. > As I said before, we had essentially the same compilers on, for example, DG and Prime, using a retargetable backend, so this was nothing new. > > I've already killed an hour today and haven't been able to find out > what the letters in GEM mean? Anyone out there know? > > Neil Rieck > Kitchener/Waterloo/Cambridge, > Ontario, Canada. > http://www3.sympatico.ca/n.rieck/ > -- PL/I for OpenVMS www.kednos.com ------------------------------ Date: Tue, 17 Jul 2007 19:34:53 +0800 From: "Richard Maher" Subject: Re: What happen to the Deathrow cluster Message-ID: They're baaaack! Cheers Richard Maher "Richard Maher" wrote in message news:f6i0ek$4ab$1@news-01.bur.connect.com.au... > Hi Phil > > Last I heard it was a CPU fan, but it does seem to be taking some time. > (Having said I personally am yet to donate money, hardware, or time to the > cause so I'm certainly not gonna start whinging just yet :-) > > Cheers Richard Maher > > wrote in message > news:1183608700.159244.49960@w5g2000hsg.googlegroups.com... > > I tried to connect, but I get a page not available or no route to > > host? So if the Deathrow cluster is no more, that brings up a question > > or two> > > 1. What are the pro/cons or offering free VMS accounts? > > 2. Can you mix hobby license/with "real VMS licenses" ie WPS/ > > OfficeServer/and yes even PL/1 > > 3. Is there even a need for free VMS accounts with so many VAX/Alpha > > flooding the market (ebay, come get the Alpha/VAX or its trash etc) > > tks > > phil > > > > ------------------------------ Date: Tue, 17 Jul 2007 09:42:13 +0200 From: Michael Kraemer Subject: Re: What OS version to run on a DEC 3000-M600 Message-ID: Slor schrieb: > > Also, if increasing the RAM would help it be more suited for a specific OS > version, I can consider looking for more memory as well. > there's this guy, Ben Myers (ben_myers@charter.net), who posts on comp.sys.dec every now and then. A couple of months ago he had a pile of old DEC memory on sale, including MS15-DA IIRC. About $2 per 8MB or so. ------------------------------ Date: Tue, 17 Jul 2007 00:59:30 -0700 From: urbancamo Subject: Re: What OS version to run on a DEC 3000-M600 Message-ID: <1184659170.266058.235280@o11g2000prd.googlegroups.com> I have OpenVMS 8.3 running on a DEC 3000-M600 with 64MB and a PMGD-BA (a 24 bit 3D graphics card). It was very easy to install. The box was running tru64 V5.0. If you need help obtaining media for OpenVMS 8.3 let me know, I may be able to help. Regards, Mark ------------------------------ End of INFO-VAX 2007.388 ************************