Archive-Date: Fri, 01 Mar 1996 03:15:10 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Resent-Date: Thu, 29 Feb 1996 16:47:50 -0800 (PST) Date: Thu, 29 Feb 1996 16:47:48 -0800 (PST) From: "Janet L. Newman" Reply-To: vms-emacs-testers@e.kth.se Subject: Re: "Janet L. Newman" : Re: Gnu Emacs build Resent-To: levitte@e.kth.se To: Richard Levitte - GNU on VMS hacker Resent-CC: vms-emacs-testers@e.kth.se CC: vms-emacs-testers@e.kth.se Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Transfer-Encoding: 7BIT I'm running TGV MultiNet V3.5 Rev A, Phase/IP V1.3, Secure/IP V2.0, DEC 3000 - M600, OpenVMS AXP V6.2 Hope that helps. +-----------------------+----------------+ | Janet Newman | | | jnewman@hmc.edu | You may have | | East 171 x72019 | just taken | +-----------------------+ yourself | | Harvey Mudd College | too | | 340 E. Foothill Blvd. | seriously. | | Claremont, CA 91711 | | +-----------------------+----------------+ ================================================================================ Archive-Date: Fri, 01 Mar 1996 05:21:04 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Message-ID: <2.2.32.19960301030853.006abee8@pop.j51.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Thu, 29 Feb 1996 22:08:53 -0500 To: vms-emacs-testers@e.kth.se From: JNC Consulting Group Reply-To: vms-emacs-testers@e.kth.se Subject: Subprocess Creation Problems I am running VMS 5.5, VAX C, no networking support or X11. The September release works fine. This weekm I installed the two Nov updates followed by the Feb release. Subprocess creation promptly broke with a varity of errors depending on whether I tried the "shell" command or the "compile" command. All else seems to work BTW. Looking at the source modules, it seemed to me that process.c, vmsproc.c, and callproc.c are the modules that deal with subprocess creation and they all seem to have been updated as a unit. So, I backed off of the Feb release for those three modules and rebuilt. Still no good. I backed off again to the Sept release for this three modukes and now all is well. So, what do I do now? Thanx! --- Jake Colman email: jcolman@jnc.com JNC Consulting Group Inc. Phone: (914) 368-2765 42 Olympia Lane FAX : (914) 368-2765 Monsey, NY 10952 ================================================================================ Archive-Date: Fri, 01 Mar 1996 10:43:51 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Date: Fri, 01 Mar 1996 08:52:50 +0100 From: Richard Levitte - GNU on VMS hacker Reply-To: vms-emacs-testers@e.kth.se To: vms-emacs-testers@e.kth.se Message-ID: <0099EAD4.EF544047.4@e.kth.se> Subject: Re: Subprocess Creation Problems >The September release works fine. This weekm I installed the two Nov >updates followed by the Feb release. Subprocess creation promptly broke >with a varity of errors depending on whether I tried the "shell" command or >the "compile" command. All else seems to work BTW. Oh-oh... >Looking at the source modules, it seemed to me that process.c, vmsproc.c, >and callproc.c are the modules that deal with subprocess creation and they That is correct. >So, what do I do now? Well, you've allready done a lot, just identifying when it stopped working. I would need you to tell me exactly what kind of error you get. Also, if you wish, you could help finding out exactlyt what is going wrong. I'll try to dig into this tomorrow. -- +--------------------------------------------------------------------+ ! Richard Levitte, GNU on VMS hacker ! tel: +46-8-26 52 47 ! ! Spannvägen 38, I ! fax: none for the moment ! ! S-161 43 Bromma ! Internet: levitte@e.kth.se ! ! SWEDEN ! ! +-GNUish VMS-+ You may not add me to a commercial mailing list or send me commercial advertising without my consent! See http://www.e.kth.se/~levitte/anti.html for further reference. ================================================================================ Archive-Date: Fri, 01 Mar 1996 10:48:10 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Date: Fri, 01 Mar 1996 08:49:22 +0100 From: Richard Levitte - GNU on VMS hacker Reply-To: vms-emacs-testers@e.kth.se To: vms-emacs-testers@e.kth.se Message-ID: <0099EAD4.7313293B.20@e.kth.se> Subject: Re: "Janet L. Newman" : Re: Gnu Emacs build >From: "Janet L. Newman" Welcome to the list Janet, >I'm running > >TGV MultiNet V3.5 Rev A, Phase/IP V1.3, Secure/IP V2.0, > DEC 3000 - M600, OpenVMS AXP V6.2 Yep. If I remember correctly, the bug you encountered (the one in the MultiNet TYPES.H in relation to DEC C) is well known. There was a quick remedy that I don't recall right now (it was some hacking in the MultiNet TYPES.H), as well as upgrading to a newer version of MultiNet. That's as far as I recall from last time this bug was encountered. Guys! If I recall things badly, please tell me. This should really be part of my little FAQ, but it seems to have slipped out... -- +--------------------------------------------------------------------+ ! Richard Levitte, GNU on VMS hacker ! tel: +46-8-26 52 47 ! ! Spannvägen 38, I ! fax: none for the moment ! ! S-161 43 Bromma ! Internet: levitte@e.kth.se ! ! SWEDEN ! ! +-GNUish VMS-+ You may not add me to a commercial mailing list or send me commercial advertising without my consent! See http://www.e.kth.se/~levitte/anti.html for further reference. ================================================================================ Archive-Date: Fri, 01 Mar 1996 21:53:28 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Date: Fri, 1 Mar 1996 12:32:12 -0800 (PST) From: RICK FLOWER Reply-To: vms-emacs-testers@e.kth.se To: vms-emacs-testers@e.kth.se Message-ID: <960301123212.2d1e@sdvax2.sp.TRW.COM> Subject: Re: Subprocess Creation Problems >>>>> "Jake" == vms-emacs-testers@e kth se writes: >> The September release works fine. This weekm I installed the two Nov >> updates followed by the Feb release. Subprocess creation promptly broke >> with a varity of errors depending on whether I tried the "shell" command or >> the "compile" command. All else seems to work BTW. > Oh-oh... >> Looking at the source modules, it seemed to me that process.c, vmsproc.c, >> and callproc.c are the modules that deal with subprocess creation and they > That is correct. >> So, what do I do now? > Well, you've allready done a lot, just identifying when it stopped working. > I would need you to tell me exactly what kind of error you get. > Also, if you wish, you could help finding out exactlyt what is going wrong. > I'll try to dig into this tomorrow. Did you install the patch that I needed for running VMS-5.xx?? It was the problem that I ran into with CALLPROC.C (if I recall) where there was an extra argument that was put into the call to lib$spawn that only exists in VMS 6.x?? I don't have the older mail in front of me right now to check it.. Are you getting errors regarding invalid parameters or similar?? -- Rick ================================================================================ Archive-Date: Fri, 01 Mar 1996 23:34:32 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Date: Fri, 01 Mar 1996 23:26:03 +0100 From: Richard Levitte - GNU on VMS hacker Reply-To: vms-emacs-testers@e.kth.se To: vms-emacs-testers@e.kth.se Message-ID: <0099EB4E.EBDCA62E.5@e.kth.se> Subject: Re: Subprocess Creation Problems >From: RICK FLOWER >Did you install the patch that I needed for running VMS-5.xx?? It was the >problem that I ran into with CALLPROC.C (if I recall) where there was an >extra argument that was put into the call to lib$spawn that only exists in >VMS 6.x?? I don't have the older mail in front of me right now to check it.. That patch is part of the latest distribution, yes. -- +--------------------------------------------------------------------+ ! Richard Levitte, GNU on VMS hacker ! tel: +46-8-26 52 47 ! ! Spannvägen 38, I ! fax: none for the moment ! ! S-161 43 Bromma ! Internet: levitte@e.kth.se ! ! SWEDEN ! ! +-GNUish VMS-+ You may not add me to a commercial mailing list or send me commercial advertising without my consent! See http://www.e.kth.se/~levitte/anti.html for further reference. ================================================================================ Archive-Date: Sat, 02 Mar 1996 01:01:49 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Resent-Date: Fri, 01 Mar 1996 15:54:36 -0800 (PST) Date: Fri, 01 Mar 1996 15:54:34 -0800 (PST) From: "Janet L. Newman" Reply-To: vms-emacs-testers@e.kth.se Subject: Re: "Janet L. Newman" : Re: Gnu Emacs build Resent-To: vms-emacs-testers@e.kth.se To: vms-emacs-testers@e.kth.se Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Transfer-Encoding: 7BIT On Fri, 1 Mar 1996, Richard Levitte - GNU on VMS hacker wrote: > >I'm running > > > >TGV MultiNet V3.5 Rev A, Phase/IP V1.3, Secure/IP V2.0, > > DEC 3000 - M600, OpenVMS AXP V6.2 > > Yep. If I remember correctly, the bug you encountered (the one in the > MultiNet TYPES.H in relation to DEC C) is well known. There was a quick > remedy that I don't recall right now (it was some hacking in the MultiNet > TYPES.H), as well as upgrading to a newer version of MultiNet. That's > as far as I recall from last time this bug was encountered. > > Guys! If I recall things badly, please tell me. This should really be > part of my little FAQ, but it seems to have slipped out... Well, Multinet is definitely the problem, since building without it works just fine. That's enough of a solution for me, since (frankly) we have few emacs users on our VMS cluster and as far as we know, none of them use its remote file-handling capabilites. Thanks! -- +-----------------------+----------------+ | Janet Newman | | | jnewman@hmc.edu | You may have | | East 171 x72019 | just taken | +-----------------------+ yourself | | Harvey Mudd College | too | | 340 E. Foothill Blvd. | seriously. | | Claremont, CA 91711 | | +-----------------------+----------------+ ================================================================================ Archive-Date: Sat, 02 Mar 1996 01:30:17 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Date: Sat, 02 Mar 1996 01:22:24 +0100 From: Richard Levitte - GNU on VMS hacker Reply-To: vms-emacs-testers@e.kth.se To: vms-emacs-testers@e.kth.se Message-ID: <0099EB5F.2D065BE1.29@e.kth.se> Subject: Re: "Janet L. Newman" : Re: Gnu Emacs build >From: "Janet L. Newman" >Well, Multinet is definitely the problem, since building without it works >just fine. That's enough of a solution for me, since (frankly) we have >few emacs users on our VMS cluster and as far as we know, none of them >use its remote file-handling capabilites. If you want to reenable network handling, one very quick solution would be to install NETLIB and SOCKETSHR, and link Emacs against those. both NETLIB and SOCKETSHR are available on the following sites: ftp.wku.edu:[.VMS.FILESERV] ftp.vms.stacken.kth.se:[.WKU.VMS.FILESERV] -- +--------------------------------------------------------------------+ ! Richard Levitte, GNU on VMS hacker ! tel: +46-8-26 52 47 ! ! Spannvägen 38, I ! fax: none for the moment ! ! S-161 43 Bromma ! Internet: levitte@e.kth.se ! ! SWEDEN ! ! +-GNUish VMS-+ You may not add me to a commercial mailing list or send me commercial advertising without my consent! See http://www.e.kth.se/~levitte/anti.html for further reference. ================================================================================ Archive-Date: Sat, 02 Mar 1996 01:48:26 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Resent-Date: Fri, 01 Mar 1996 16:40:27 -0800 (PST) Date: Fri, 01 Mar 1996 16:40:25 -0800 (PST) From: "Janet L. Newman" Reply-To: vms-emacs-testers@e.kth.se Subject: Re: "Janet L. Newman" : Re: Gnu Emacs build Resent-To: vms-emacs-testers@e.kth.se To: vms-emacs-testers@e.kth.se Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Transfer-Encoding: 7BIT On Sat, 2 Mar 1996, Richard Levitte - GNU on VMS hacker wrote: > If you want to reenable network handling, one very quick solution would > be to install NETLIB and SOCKETSHR, and link Emacs against those. > > both NETLIB and SOCKETSHR are available on the following sites: > > ftp.wku.edu:[.VMS.FILESERV] > ftp.vms.stacken.kth.se:[.WKU.VMS.FILESERV] Thanks, I'll be sure to look into it. -- +-----------------------+----------------+ | Janet Newman | | | jnewman@hmc.edu | You may have | | East 171 x72019 | just taken | +-----------------------+ yourself | | Harvey Mudd College | too | | 340 E. Foothill Blvd. | seriously. | | Claremont, CA 91711 | | +-----------------------+----------------+ ================================================================================ Archive-Date: Sun, 03 Mar 1996 16:31:25 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Message-ID: <2.2.32.19960303151706.006dd2c0@pop.j51.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Sun, 03 Mar 1996 10:17:06 -0500 To: vms-emacs-testers@e.kth.se From: JNC Consulting Group Reply-To: vms-emacs-testers@e.kth.se Subject: Re: Subprocess Creation Problems I'll get you specific error messages for the November release and the Februart release as soon as I am able. That might be tomorrow (Monday) or, if not, it won't be until Wednesday. The patch to CALLPROC.C which you say is part of the distribution, which release incorporated it? At 08:52 AM 3/1/96 +0100, you wrote: >Well, you've allready done a lot, just identifying when it stopped working. >I would need you to tell me exactly what kind of error you get. >Also, if you wish, you could help finding out exactlyt what is going wrong. > --- Jake Colman email: jcolman@jnc.com JNC Consulting Group Inc. Phone: (914) 368-2765 42 Olympia Lane FAX : (914) 368-2765 Monsey, NY 10952 ================================================================================ Archive-Date: Mon, 04 Mar 1996 08:30:52 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Date: Mon, 04 Mar 1996 08:16:55 +0100 From: Richard Levitte - GNU on VMS hacker Reply-To: vms-emacs-testers@e.kth.se To: vms-emacs-testers@e.kth.se Message-ID: <0099ED2B.69D3B087.28@e.kth.se> Subject: Re: Subprocess Creation Problems >From: JNC Consulting Group >The patch to CALLPROC.C which you say is part of the distribution, which >release incorporated it? Ooops, sorry. It's in the 19960225 release. -- +--------------------------------------------------------------------+ ! Richard Levitte, GNU on VMS hacker ! tel: +46-8-26 52 47 ! ! Spannvägen 38, I ! fax: none for the moment ! ! S-161 43 Bromma ! Internet: levitte@e.kth.se ! ! SWEDEN ! ! +-GNUish VMS-+ You may not add me to a commercial mailing list or send me commercial advertising without my consent! See http://www.e.kth.se/~levitte/anti.html for further reference. ================================================================================ Archive-Date: Wed, 06 Mar 1996 18:33:00 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Date: Wed, 06 Mar 1996 18:21:11 +0100 From: Richard Levitte - GNU on VMS hacker Reply-To: vms-emacs-testers@e.kth.se To: vms-emacs-testers@e.kth.se, gerrie-warner@pi.se CC: stellanl@e.kth.se, rnydahl@lbdata.se, pdellby@lbdata.se, martin@ctrl-c.liu.se, haba@stacken.kth.se Message-ID: <0099EF12.2903017A.19@e.kth.se> Subject: www.lp.se is down... Last night, www.lp.se (also ftp.lp.se) burned. Actually, it's not as serious as it sounds. It's only a small part that burned. Unfortunatelly, this part is very important. It's the contacts of the cable between the power case and slot case that have melted. Yes, melted, you read correctly. It's probably some dust (there's really no place without some dust) and a good deal of bad luck that made this happen, and which makes me looking at a piece of melted plastic... (BTW, did I say it's a MicroVAX II, in the small case (I can never remember if that's the BA23 or the BA123...)?) This has some impact on a bunch of quite popular WWW pages (http://www.lp.se/gerrie-warner/, belonging to a Gerrie Warner who's trying to organize a freenet in a part of Sweden, and who also does a few other things, like photography), as well as on the backup FTP server I've set up a few days ago to hold GNU emacs on VMS. I'm right now on my way to try to find a replacement for the melted piece of hardware, and will get up the lp.se machine as soon as possible. Meanwhile, for those trying to reach, I can only beg for some patience. At least, now you're informed. (If there is someone among those receiving this mail, preferadly those here in Sweden (for speed reasons :-)) who knows he has the piece I'm looking for, and who can give it away, I'd appreciate it a lot (I could almost lick your boots :-))) Thank you. -- +--------------------------------------------------------------------+ ! Richard Levitte, GNU on VMS hacker ! tel: +46-8-26 52 47 ! ! Spannvägen 38, I ! fax: none for the moment ! ! S-161 43 Bromma ! Internet: levitte@e.kth.se ! ! SWEDEN ! ! +-GNUish VMS-+ You may not add me to a commercial mailing list or send me commercial advertising without my consent! See http://www.e.kth.se/~levitte/anti.html for further reference. ================================================================================ Archive-Date: Thu, 07 Mar 1996 02:39:52 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Message-ID: <2.2.32.19960307012920.006c26cc@pop.j51.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Wed, 06 Mar 1996 20:29:20 -0500 To: vms-emacs-testers@e.kth.se From: JNC Consulting Group Reply-To: vms-emacs-testers@e.kth.se Subject: Re: Subprocess Creation Problems Richard, You asked for more detail concerning precisely how it crashes with which releases. Here is what I found: 1) September release works fine. 2) November release fails for both "shell" and "compile" with the error message "unable to spawn subprocess: invalid args". This sounds like what would be expected on a VMS 5.5 system since the Nov release did not include the necessary patch. 3) The February release fails for "shell" with the message "wrong type argument: processp,nill". The "compile" command fails with the message "could not deactivate process: xxxxxxxx". BTW, sysdep.c does not compile on a VMS 5.5 system, VAXC 3.x, if gethostname() is not defined by the system. This problem has existed since at least the September release. I edited the module, with Roland Roberts' assistance, so that it would compile cleanly. The February release touched sysdep.c so I hand-merged your mods into my version of it. My question is,how do I get you the necessary corrections? Can you handle the output of a standard VMS DIFFERENCES command? --- Jake Colman email: jcolman@jnc.com JNC Consulting Group Inc. Phone: (914) 368-2765 42 Olympia Lane FAX : (914) 368-2765 Monsey, NY 10952 ================================================================================ Archive-Date: Thu, 07 Mar 1996 02:45:14 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Date: Thu, 07 Mar 1996 02:39:18 +0100 From: Richard Levitte - GNU on VMS hacker Reply-To: vms-emacs-testers@e.kth.se To: vms-emacs-testers@e.kth.se Message-ID: <0099EF57.BF620765.7@e.kth.se> Subject: Re: Subprocess Creation Problems >From: JNC Consulting Group >1) September release works fine. OK. >2) November release fails for both "shell" and "compile" with the error >message "unable to spawn subprocess: invalid args". This sounds like what >would be expected on a VMS 5.5 system since the Nov release did not include >the necessary patch. Yep, that was apparently what was reported by others as well. >3) The February release fails for "shell" with the message "wrong type >argument: processp,nill". The "compile" command fails with the message >"could not deactivate process: xxxxxxxx". Now, THAT was a surprise. There are two places where LIB$SPAWN is called. One is in callproc.c, the other is vmsproc.c. I'm quite sure it's there things go wrong. Could you help me determine what happens? (I wish I had a 5.5 system handy right now... I don't...) >is,how do I get you the necessary corrections? Can you handle the output of >a standard VMS DIFFERENCES command? I do. -- +--------------------------------------------------------------------+ ! Richard Levitte, GNU on VMS hacker ! tel: +46-8-26 52 47 ! ! Spannvägen 38, I ! fax: none for the moment ! ! S-161 43 Bromma ! Internet: levitte@e.kth.se ! ! SWEDEN ! ! +-GNUish VMS-+ You may not add me to a commercial mailing list or send me commercial advertising without my consent! See http://www.e.kth.se/~levitte/anti.html for further reference. ================================================================================ Archive-Date: Thu, 07 Mar 1996 02:58:08 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Message-ID: <2.2.32.19960307014819.006e3944@pop.j51.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Wed, 06 Mar 1996 20:48:19 -0500 To: vms-emacs-testers@e.kth.se From: JNC Consulting Group Reply-To: vms-emacs-testers@e.kth.se Subject: Re: Subprocess Creation Problems >>3) The February release fails for "shell" with the message "wrong type >>argument: processp,nill". The "compile" command fails with the message >>"could not deactivate process: xxxxxxxx". > >Now, THAT was a surprise. > >There are two places where LIB$SPAWN is called. One is in callproc.c, >the other is vmsproc.c. I'm quite sure it's there things go wrong. >Could you help me determine what happens? > I'll try my best. What would you like me to do next? > >>is,how do I get you the necessary corrections? Can you handle the output of >>a standard VMS DIFFERENCES command? > >I do. > Any preference on how I run the VMS DIFF command? Is a simple "DIFF " sufficient using all default qialifers? --- Jake Colman email: jcolman@jnc.com JNC Consulting Group Inc. Phone: (914) 368-2765 42 Olympia Lane FAX : (914) 368-2765 Monsey, NY 10952 ================================================================================ Archive-Date: Thu, 07 Mar 1996 03:57:11 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Date: Thu, 07 Mar 1996 03:50:28 +0100 From: Richard Levitte - GNU on VMS hacker Reply-To: vms-emacs-testers@e.kth.se To: vms-emacs-testers@e.kth.se Message-ID: <0099EF61.B042C6EE.1@e.kth.se> Subject: Re: Subprocess Creation Problems >From: JNC Consulting Group >>There are two places where LIB$SPAWN is called. One is in callproc.c, >>the other is vmsproc.c. I'm quite sure it's there things go wrong. >>Could you help me determine what happens? > >I'll try my best. What would you like me to do next? Tell me what you found out. :-) >>>is,how do I get you the necessary corrections? Can you handle the output of >>>a standard VMS DIFFERENCES command? >> >>I do. But I must say my preference is GNU diff output when the switch -c is used. >Any preference on how I run the VMS DIFF command? Is a simple "DIFF file> " sufficient using all default qialifers? DIFF/SLP would be preferable. -- +--------------------------------------------------------------------+ ! Richard Levitte, GNU on VMS hacker ! tel: +46-8-26 52 47 ! ! Spannvägen 38, I ! fax: none for the moment ! ! S-161 43 Bromma ! Internet: levitte@e.kth.se ! ! SWEDEN ! ! +-GNUish VMS-+ You may not add me to a commercial mailing list or send me commercial advertising without my consent! See http://www.e.kth.se/~levitte/anti.html for further reference. ================================================================================ Archive-Date: Thu, 07 Mar 1996 04:46:06 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Message-ID: <2.2.32.19960307033624.006d0c58@pop.j51.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Wed, 06 Mar 1996 22:36:24 -0500 To: vms-emacs-testers@e.kth.se From: JNC Consulting Group Reply-To: vms-emacs-testers@e.kth.se Subject: Re: Subprocess Creation Problems At 03:50 AM 3/7/96 +0100, you wrote: >>I'll try my best. What would you like me to do next? > >Tell me what you found out. :-) If you tell me how to go about digging into this, I'd be happy to help you out! I have not hacked at emacs so I might need a wee bit of direction here. What kind if info would you like and how do I go about getting it? >But I must say my preference is GNU diff output when the switch -c is used. When running GNU diff (which I might actually have available), which file goes first? The original version or the one that was already modified? --- Jake Colman email: jcolman@jnc.com JNC Consulting Group Inc. Phone: (914) 368-2765 42 Olympia Lane FAX : (914) 368-2765 Monsey, NY 10952 ================================================================================ Archive-Date: Thu, 07 Mar 1996 07:44:03 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Date: Thu, 07 Mar 1996 07:37:29 +0100 From: Richard Levitte - GNU on VMS hacker Reply-To: vms-emacs-testers@e.kth.se To: vms-emacs-testers@e.kth.se Message-ID: <0099EF81.66E21B74.3@e.kth.se> Subject: Re: Subprocess Creation Problems >From: JNC Consulting Group >>>I'll try my best. What would you like me to do next? >> >>Tell me what you found out. :-) > >If you tell me how to go about digging into this, I'd be happy to help you >out! I have not hacked at emacs so I might need a wee bit of direction >here. What kind if info would you like and how do I go about getting it? This does require some understanding of the VMS debugger, The best thing to do is really to run this on a DECWindows machine, since Emacs does quite a lot of playing with the terminal parameters, something that DEBUG doesn't expect. What I had in mind was to have Emacs go through the debugger. It involves the following steps: $ MMS/IGN=W DEBUG ! creates [.VMS]TEMACS_D.EXE $ @[.VMS]TESTEMACS $ RUNTEMACS_D Now, you have a debugger window. Put a breakpoint on the functions calling LIB$SPAWN: SET MODU CALLPROC SET BREAK Fcall_process SET MODU VMSPROC SET BREAK create_process And now, just "Go" until you hit one of the breakpoints, then start single stepping. When you're near the LIB$SPAWN call, take note of variable values (the status is especially interesting, but others should not be forgotten)... You should find the info you need in the Debugger Manual. >>But I must say my preference is GNU diff output when the switch -c is used. > >When running GNU diff (which I might actually have available), which file >goes first? The original version or the one that was already modified? Original first. -- +--------------------------------------------------------------------+ ! Richard Levitte, GNU on VMS hacker ! tel: +46-8-26 52 47 ! ! Spannvägen 38, I ! fax: none for the moment ! ! S-161 43 Bromma ! Internet: levitte@e.kth.se ! ! SWEDEN ! ! +-GNUish VMS-+ You may not add me to a commercial mailing list or send me commercial advertising without my consent! See http://www.e.kth.se/~levitte/anti.html for further reference. ================================================================================ Archive-Date: Thu, 07 Mar 1996 16:28:52 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Date: Thu, 7 Mar 1996 09:54:05 -0500 Message-ID: <96030709540517@nscl.msu.edu> From: ibbotson@nscl.msu.edu (Rich Ibbotson) Reply-To: vms-emacs-testers@e.kth.se To: vms-emacs-testers@e.kth.se Subject: Re: Subprocess Creation Problems I just tried tracing this process-creation problem along the lines of Richard's suggestions. (I'm compiling with DEC C V5.2-003 on OpenVMS Alpha V6.2) Oh, I've also compiled the same thing under VMS 5.5 with no troubles. Unfortunately, I have almost no experience debugging Emacs... I stepped through "create_process" in the debugger (after issuing a "M-x shell"), and made it through the code without problems. I checked from another window, and the subprocess *was* created. The problem cropped up while stepping through "Fbyte_code". Does this mean the problem is in the lisp? I tried setting "debug-on-error", and the backtrace buffer looks like this: Signalling: (wrong-type-argument processp nil) process-command(nil) shell-mode() shell() * call-interactively(shell) execute-extended-command(nil) * call-interactively(execute-extended-command) I don't really know how to interpret this. I guess it looks like the problem *is* in process_command, but I don't see any errors while stepping through that routine. Can somebody explain this behavior? Rich Ibbotson@nscl.msu.edu ================================================================================ Archive-Date: Fri, 08 Mar 1996 18:36:08 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Message-ID: <2.2.32.19960308170436.006c7980@pop.j51.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Fri, 08 Mar 1996 12:04:36 -0500 To: vms-emacs-testers@e.kth.se From: JNC Consulting Group Reply-To: vms-emacs-testers@e.kth.se Subject: Re: Subprocess Creation Problems At 07:37 AM 3/7/96 +0100, you wrote: >This does require some understanding of the VMS debugger, The best thing >to do is really to run this on a DECWindows machine, since Emacs does >quite a lot of playing with the terminal parameters, something that DEBUG >doesn't expect. Using VMS Debugger is something I can do quite well. It's delving into Emac's that poses me some pause! I do not have access to a DECWindows system. Can I do what you suggest in a simple terminal session? --- Jake Colman email: jcolman@jnc.com JNC Consulting Group Inc. Phone: (914) 368-2765 42 Olympia Lane FAX : (914) 368-2765 Monsey, NY 10952 ================================================================================ Archive-Date: Fri, 08 Mar 1996 19:55:45 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Date: Fri, 08 Mar 1996 18:27:31 +0100 From: Richard Levitte - GNU on VMS hacker Reply-To: vms-emacs-testers@e.kth.se To: vms-emacs-testers@e.kth.se Message-ID: <0099F0A5.60649B23.30@e.kth.se> Subject: Re: Subprocess Creation Problems >From: JNC Consulting Group >I do not have access to a DECWindows system. Can I do what you suggest in a >simple terminal session? You can, but since Emacs echoes characters on its' own, it has set the terminal setinngs /NOECHO, ehich means you won't have any way to see what you write to the debugger... But it is possible. I've done such sessions myself... It's just not very funny not to see what you do... -- +--------------------------------------------------------------------+ ! Richard Levitte, GNU on VMS hacker ! tel: +46-8-26 52 47 ! ! Spannvägen 38, I ! fax: none for the moment ! ! S-161 43 Bromma ! Internet: levitte@e.kth.se ! ! SWEDEN ! ! +-GNUish VMS-+ You may not add me to a commercial mailing list or send me commercial advertising without my consent! See http://www.e.kth.se/~levitte/anti.html for further reference. ================================================================================ Archive-Date: Mon, 11 Mar 1996 01:05:38 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Message-ID: <2.2.32.19960310235327.006a4cd8@pop.j51.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Sun, 10 Mar 1996 18:53:27 -0500 To: vms-emacs-testers@e.kth.se From: JNC Consulting Group Reply-To: vms-emacs-testers@e.kth.se Subject: Re: Emacs build on VAX At 02:20 AM 3/9/96 +0100, you wrote: > >Please try this patch: > >*** ED:[EMACS-1928-HACKING.SRC]SYSDEP.C;33 Sun Feb 25 17:41:59 1996 >--- sysdep.c Sat Mar 9 02:18:26 1996 Richard, Is this the patch provided you by Roland Roberts to deal with VMS systems that do not have the gethostname() function? If so, that's the patch that I was preparing to post for you. --- Jake Colman email: jcolman@jnc.com JNC Consulting Group Inc. Phone: (914) 368-2765 42 Olympia Lane FAX : (914) 368-2765 Monsey, NY 10952 ================================================================================ Archive-Date: Mon, 11 Mar 1996 01:47:53 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Date: Mon, 11 Mar 1996 01:36:46 +0100 From: Richard Levitte - GNU on VMS hacker Reply-To: vms-emacs-testers@e.kth.se To: vms-emacs-testers@e.kth.se Message-ID: <0099F273.AC9B4B39.1@e.kth.se> Subject: Re: Emacs build on VAX >From: JNC Consulting Group >>Please try this patch: >> >>*** ED:[EMACS-1928-HACKING.SRC]SYSDEP.C;33 Sun Feb 25 17:41:59 1996 >>--- sysdep.c Sat Mar 9 02:18:26 1996 > > >Richard, > >Is this the patch provided you by Roland Roberts to deal with VMS systems >that do not have the gethostname() function? If so, that's the patch that I >was preparing to post for you. I made that patch when I got the complaint. If I got the patch from Roland, I seem to have lost it somehow (sorry, Roland). if it's the same, then everything is good! If that's the patch you were going to send me, then I guess you don't need to any more... Unless it contains more things to patch. -- +--------------------------------------------------------------------+ ! Richard Levitte, GNU on VMS hacker ! tel: +46-8-26 52 47 ! ! Spannvägen 38, I ! fax: none for the moment ! ! S-161 43 Bromma ! Internet: levitte@e.kth.se ! ! SWEDEN ! ! +-GNUish VMS-+ You may not add me to a commercial mailing list or send me commercial advertising without my consent! See http://www.e.kth.se/~levitte/anti.html for further reference. ================================================================================ Archive-Date: Mon, 11 Mar 1996 04:31:24 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Resent-Date: Sun, 10 Mar 1996 19:25:45 -0800 (PST) Date: Sun, 10 Mar 1996 19:25:45 -0800 (PST) From: "Janet L. Newman" Reply-To: vms-emacs-testers@e.kth.se Subject: Re: Emacs build on VAX Resent-To: vms-emacs-testers@e.kth.se To: vms-emacs-testers@e.kth.se Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Transfer-Encoding: 7BIT Just want to mention, thank you very much for the patch. I haven't gotten a chance to try it yet because I'm at home for spring break and it's very difficult to telnet to my school from the outside. (That last sentence took me about a minute and a fhalf to type.) I'll let you know how it goes when I do try it, probably this Wednesday. Again, thanks! I'm sure I couldn't have gotten even this far on my own. Janet -- +-----------------------+----------------+ | Janet Newman | | | jnewman@hmc.edu | You may have | | East 171 x72019 | just taken | +-----------------------+ yourself | | Harvey Mudd College | too | | 340 E. Foothill Blvd. | seriously. | | Claremont, CA 91711 | | +-----------------------+----------------+ ================================================================================ Archive-Date: Fri, 15 Mar 1996 02:08:14 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Date: Thu, 14 Mar 1996 19:39:10 -0500 Message-ID: <96031419391026@nscl.msu.edu> From: ibbotson@nscl.msu.edu (Rich Ibbotson) Reply-To: vms-emacs-testers@e.kth.se To: vms-emacs-testers@e.kth.se Subject: still confused on subprocesses Hi all, Has anyone ever sorted out this subprocess creation problem on VMS 6.2? I've been through this in the debugger and I'm a bit confused about exactly where the problem is occurring. I seem to make it through create_process fine, and a subprocess appears (I'm trying M-x shell). However, I get an error when get_buffer_process(#) returns nil. On a system with VMS 5.5, it seems to return properly with #. I tracked this to the point in get_buffer_process that loops through the entries in Vprocess_alist. It's never executed. I don't seem to be able to "see" what's in this list, though. I must misunderstand it. The variable "Vprocess_alist" isn't defined, even in the version that works. What am I doing wrong? Rich Ibbotson@nscl.msu.edu ================================================================================ Archive-Date: Mon, 18 Mar 1996 15:45:10 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Date: Mon, 18 Mar 1996 09:15:11 -0500 Message-ID: <96031809151177@aalix.pharmark.com> From: george@aalix.pharmark.com Reply-To: vms-emacs-testers@e.kth.se To: VMS-EMACS-TESTERS@e.kth.se Subject: logical name table is full Hi, When I'm using emacs in xterm, if I just type emacs, then an Emacs window pop out and everything goes fine. But when I using "emacs -nw", which let emacs running in the xterm instead of creating new widow, it comes the error message "%SYSTEM-F-EXLNMQUOTA, logical name table is full", is there anything I should take care of when using emacs with "-nw" option? Thanks in advance! George ================================================================================ Archive-Date: Mon, 18 Mar 1996 16:34:46 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Date: Mon, 18 Mar 1996 16:24:02 +0100 From: Richard Levitte - GNU on VMS hacker Reply-To: vms-emacs-testers@e.kth.se To: vms-emacs-testers@e.kth.se Message-ID: <0099F86F.C89A3A1C.9@e.kth.se> Subject: Re: logical name table is full >From: george@aalix.pharmark.com > When I'm using emacs in xterm, if I just type emacs, then an Emacs window >pop out and everything goes fine. But when I using "emacs -nw", which let >emacs running in the xterm instead of creating new widow, it comes the error >message "%SYSTEM-F-EXLNMQUOTA, logical name table is full", is there anything >I should take care of when using emacs with "-nw" option? Thanks in advance! You have a real tight job table quota. When Emacs is run as a kept subprocess (which is what happens when you don't run it in it's own window), the following line in KEPTEDITOR.COM is executed: definee/job/nolog EMACS_PARENT_PID 'f$getjpi("","PID")' Ask your sysadmin for an increase of 32 bytes in your job table quota (it's the JTquota parameter in AUTHORIZE)... That should be enough. (your sysadm could give you more, while he's at it, it's not taking much resources) (If you're the sysadmin, apply the above on yourself :-) -- +--------------------------------------------------------------------+ ! Richard Levitte, GNU on VMS hacker ! tel: +46-8-26 52 47 ! ! Spannvägen 38, I ! fax: none for the moment ! ! S-161 43 Bromma ! Internet: levitte@e.kth.se ! ! SWEDEN ! ! +-GNUish VMS-+ You may not add me to a commercial mailing list or send me commercial advertising without my consent! See http://www.e.kth.se/~levitte/anti.html for further reference. ================================================================================ Archive-Date: Thu, 28 Mar 1996 18:57:37 MET-1MET DST Sender: owner-vms-emacs-testers-archiver@ctrl-c.liu.se Date: Thu, 28 Mar 96 18:42:15 +0100 From: Tobias Johansson Reply-To: vms-emacs-testers@e.kth.se Message-ID: <9603281742.AA24810@Minsk.DoCS.UU.SE> To: vms-emacs-testers@e.kth.se UNSUBSCRIBE