December 1998 - XEmacs-Beta - XEmacs Mailing List Archives

SL Baur

Andy Piper <andyp(a)parallax.co.uk> writes in xemacs-beta(a)xemacs.org: > At 02:27 03/12/98 -0800, SL Baur wrote: >> Even if true, it is no excuse for XEmacs package installation to puke >> its guts out. > I'm not sure I understand, but I probably agree :) Failure of gzip or tar > seems quite high in the list of things that could fail during package > installation so we should cope somehow. My point was that no matter how disk sizes have increased in recent years, they are still of finite size and may be filled close to capacity when XEmacs packages are being installed. The steady state of disks is full. -- Ken Thompson Truth be told, I'm (somewhat) less worried about filling disks than overflowing quota, but the effects on the user are identical and the steps XEmacs has to take to not puke and die are the same. ... > So you need 7Mb slack for installation - seems not unreasonable to me. I > guess the best thing to do would be to check how much space is actually > free and act accordingly. Right. > Ultimate paranoia would account for the possibility that it may not > be possible to install a newer package because it is bigger than the > previous one. Right. I like this train of thought. Excellent. :-)

2 participants
1 comments

Re: package badness 27 years

SL Baur

Andy Piper <andyp(a)parallax.co.uk> writes in xemacs-beta(a)xemacs.org: ... > Also IMHO (and this may not work on all systems). Instead of deleting the > package and then untarring the new one. The new one should be untarred to a > temp directory and then the old one moved out the way and the new one moved > in. If you're going to do this, you get into issues of quota and physical disk limits. Running out of disk/quota should fail gracefully. If this gets implemented it should probably be made optional. As a backup, I prefer keeping the previous compressed tarball around even with buildable source available.

3 participants
2 comments

Re: What happened to auto delete selection in 21.2? 27 years

Jan Vroonhof

Dan Glastonbury <Dan.Glastonbury(a)adelaide.maptek.com.au> writes: > Firstly, why is Options->Editing Options->Auto Delete Selection "greyed" > out? I find it to be a useful feature and would like to turn it back on, > but I can't (at least not from the Options menu). Have I missed > downloading a required package? Probably. Do you have pending-del.el (from the "pc" package?) > And secondly, why is 21.2 honouring the options that I set for 20.4? You mean NOT honoring the options? Because the options menu is now using custom.[1] I suggest you read the NEWS entries for 21.0 and 21.2. > From first glance the new customise face buffer isn't as simple or > easy to use as the old edit faces in 20.4. And if you mean your font settings not getting saved, you have just discovered why 'options-save-faces' was 'nil' in 20.4. You have as they say, been warned. Nobody has written an auto conversion routine for the options yet nor ported M-x edit-face RET to custom. The latter would be really appreciated but will probably be very hard. Jan Footnotes: [1] Technically it is because the options loading blurb in your .emacs checks for XEmacs major version == 20 or 19. -- Jan Vroonhof http://www.math.ethz.ch/~vroonhof/ Mathematik, vroonhof ＠ math.ethz.ch HG E16, ETH-Zentrum, Tel: +41-1-6325456/25154 Raemistrasse 101, CH-8092 Zuerich. Fax: +41-1-6321085

1 participants
0 comments

Re: fix FTP address for UK sunsite in package-get.el 27 years

Jan Vroonhof

Gunnar Evermann <ge204(a)eng.cam.ac.uk> writes: > BTW they also mirror the complete beta tree, so that's definitely an > alternative for beta testers in the UK. Since the beta dir was made open without a password, probably all mirrors mirror the beta tree. Jan

1 participants
0 comments

Re: "Pyrenean-pre8" problems on WinNT 27 years

Jan Vroonhof

Charles G Waldman <cgw(a)pgt.com> writes: > (setenv "HOME" "/home/cgw") > > (insert (getenv "HOME")) /home/cgw > (insert (expand-file-name "test" "~/test/")) c:\home\cgw\test\test It would have been more convenient if your tried this with HOME != CWD, I think it is not adding your home directory here but just the CWD. > (setenv "HOME" "\\\\reactor\\cgwhome") > > (insert (getenv "HOME")) \\reactor\cgwhome > (insert (expand-file-name "test" "~/test/")) c:\home\cgw\test\test See, no 'reactor' at all. I noticed something else > > if (IS_DIRECTORY_SEP (nm[1]) > > || nm[1] == 0) /* ~ by itself */ > > { > > if (!(newdir = (Bufbyte *) get_home_directory())) > > newdir = (Bufbyte *) ""; > > nm++; > > #ifdef WINDOWSNT > > collapse_newdir = 0; > > #endif > > } If I read correctly if nm == "~/fubar" this sets newdir = HOME and "nm = /fubar". However, later on it does /* Effectively, let newdir be (expand-file-name newdir cwd). Because of the admonition against calling expand-file-name when we have pointers into lisp strings, we accomplish this indirectly by prepending newdir to nm if necessary, and using cwd (or the wd of newdir's drive) as the new newdir. */ [..] if (!IS_DIRECTORY_SEP (nm[0])) { char * tmp = alloca (strlen (newdir) + strlen (nm) + 2); file_name_as_directory (tmp, newdir); strcat (tmp, nm); nm = tmp; } newdir = alloca (MAXPATHLEN + 1); So it only copies the newdir in front if nm[0] != '/' (because you do want to have the default-directory in front of absolute names). After that it throws newdir away. However in our case the "/" is just there by accident So try making it if (!IS_DIRECTORY_SEP (nm[0]) || !collapse_newdir) { char * tmp = alloca (strlen (newdir) + strlen (nm) + 2); and removing the "&& collapse_newdir" later on. However I must be overseeing something is as this means setting HOME on NT has never worked and that I really cannot believe. Jan

1 participants
0 comments

What happened to auto delete selection in 21.2? 27 years

Dan Glastonbury

G'Day, I've just built xemacs 21.2 b3 "Aglaia" and have a few questions. Firstly, why is Options->Editing Options->Auto Delete Selection "greyed" out? I find it to be a useful feature and would like to turn it back on, but I can't (at least not from the Options menu). Have I missed downloading a required package? And secondly, why is 21.2 honouring the options that I set for 20.4? From first glance the new customise face buffer isn't as simple or easy to use as the old edit faces in 20.4. cheers Dan -- Dan Glastonbury <dan at adelaide.maptek.com.au> `Network is unreachabl'

1 participants
0 comments

Debilan rulez -- NOT! 27 years

Hrvoje Niksic

After debugging the bloody thing for two hours at a friend's, I discovered that the source of the problem was the DUMB Debian maintainer, who set `nnmail-spool-file' to "/var/spool/mail" at a 00debian.el file. That setting didn't do what he expected, only he never bothered to CHECK. If he at least used a simple SETQ, it would have been trivial to find it. This way the only clue that anything was wrong was provided by `symbol-plist', after I remembered to check it. Sheesh. What a bunch of Penguinoid losers. :-(

1 participants
0 comments

Re: Where is this message coming from and why? 27 years

Jan Vroonhof

SL Baur <steve(a)xemacs.org> writes: > I'm trying to build XEmacs packages from a bare-ass buck-naked binary > with no installed packages besides mule-base. What is happening with > this error message? For the moment, I'm ignoring the cause of the > core dump. Issuing a prompt when running -batch is a programming > error. > -eval '(unless (featurep (quote mule)) (kill-emacs 0))' \ > -l texinfmt -f texinfo-format-buffer \ > -eval '(setq file-coding-system (quote iso-2022-7))' \ > -eval '(save-buffer 0)' > No mailcrypt; can't verify package-get DB signature, continue? (yes or no) Presumably this code in 'set-auto-mode' gets hit (I presume for texinfo-mode. (if mode (if (not (fboundp mode)) (let ((name (package-get-package-provider mode))) (if name (message "Mode %s is not installed. Download package %s" mode name) (message "Mode %s either doesn't exist or is not a known package" mode)) (sit-for 2) (error "%s" mode)) When XEmacs is trying to load a file for which there is a mode defined but you have not installed a package providing that mode it tries to suggest downloading a package. In order to do this it needs to get the package database. Before Greg's and my work this meant that XEmacs would open an ftp session to xemacs.org. [1]Now it gets a local copy. However XEmacs is currently paranoid[2] and wants to verify the signature. However you are so barebones you do not have mailcrypt so the verification fails. I think the best in this case would be to assume a "no" answer when operating non-interactively. However that would still leave the fatal error that texinfo-mode is undefined, so you need to add a "-l tex-info" too. Jan Footnotes: [1] This is what hurt Hrvoje. [2] This was added to defend against the "cache = local" type attacks that hurt the webbrowsers a lot. I think it is currently being a bit overzealous but erring on the safe site is good. -- Jan Vroonhof http://www.math.ethz.ch/~vroonhof/ Mathematik, vroonhof ＠ math.ethz.ch HG E16, ETH-Zentrum, Tel: +41-1-6325456/25154 Raemistrasse 101, CH-8092 Zuerich. Fax: +41-1-6321085

2 participants
1 comments

Re: "Pyrenean-pre8" problems on WinNT 27 years

Adrian Aichner

>>>>> "Charles" == Charles G Waldman <cgw(a)pgt.com> writes: Charles> DQoNCkphbiBWcm9vbmhvZiB3cml0ZXM6DQoNCiA+IEl0IGNvdWxkIGJlIGEgYnVnIGluIHRoZSBO <64 lines deleted by Adrian> Charles> dG9yeTogSW52YWxpZCBhcmd1bWVudCwgYzpcaG9tZVxjZ3dcBQFwP4IvLmF1dG9zYXZlDQoNCg== Is this what you call MIME? ;-> Anyways, I was able to decrypt it: CW> Jan Vroonhof writes: CW> CW> > It could be a bug in the NT expand-file-name routines (what does CW> > (expand-file-name "test" "~/test/") give under your machines). CW> CW> (setenv "HOME" "/home/cgw") CW> CW> (insert (getenv "HOME")) /home/cgw CW> (insert (expand-file-name "test" "~/test/")) c:\home\cgw\test\test CW> CW> (setenv "HOME" "C:\\home\\cgw") CW> CW> (insert (getenv "HOME")) C:\home\cgw CW> (insert (expand-file-name "test" "~/test/")) c:\home\cgw\test\test CW> CW> (setenv "HOME" "\\\\reactor\\cgwhome") CW> CW> (insert (getenv "HOME")) \\reactor\cgwhome CW> (insert (expand-file-name "test" "~/test/")) c:\home\cgw\test\test CW> CW> (setenv "HOME" "//reactor/cgwhome") CW> CW> (insert (getenv "HOME")) //reactor/cgwhome CW> (insert (expand-file-name "test" "~/test/")) c:\home\cgw\test\test Charles, you will have to change the environment variable HOME in the Control Panel and then run a new XEmacs from Start->Run... Then you will get different results for (expand-file-name ...) above. I do, at least. Even the UNC path works for me. Regards, Adrian CW> CW> > I do not know the NT port but shouldn't HOME be a directory including CW> > driveletter since otherwise it is not uniquely specified? CW> CW> I tried it with HOME=/home/cgw and HOME=C:\home\cgw, get the same CW> behaviour (No such file or directory, c:\home\cgw\home\cgw\/.autosave) CW> both ways. CW> CW> > The funny thing is that XEmacs tries to do its best to copy with funny CW> > HOME values but on NT it seems to be switched of explicitly: CW> > CW> > if (IS_DIRECTORY_SEP (nm[1]) CW> > || nm[1] == 0) /* ~ by itself */ CW> > { CW> > if (!(newdir = (Bufbyte *) get_home_directory())) CW> > newdir = (Bufbyte *) ""; CW> > nm++; CW> > #ifdef WINDOWSNT CW> > collapse_newdir = 0; CW> > #endif CW> > } CW> > CW> > The collapse_newdir explicitly tells XEmacs to not consider the fact CW> your CW> > HOME prefix starts with a / and just prepend the current dir which CW> > happened to be c:\home\cgw in your case. It seems bogus to me. CW> > CW> > Could you try removing that code and then test what happends to expand CW> > file name in all case. i.e. with CW> > CW> > 1. HOME includes driveletter CW> > 2. HOME does not include drive letter. CW> > 3. HOME is a network path CW> CW> Sorry, I was not completely clear on what code you wanted me to CW> remove. The 3 lines beginning with #ifdef WINDOWSNT, or the entire CW> block you quoted above? I tried it both ways. CW> CW> With the "#ifdef WINDOWSNT" changed to "#ifdef XXXWINDOWSNTXXX" CW> I get no warnings from make-auto-save-file-name, but clearly wrong CW> behaviour from expand-file-name: CW> CW> (setenv "HOME" "/home/cgw") CW> CW> (insert (getenv "HOME")) /home/cgw CW> (insert (expand-file-name "test" "~/test/")) c:\test\test CW> CW> (setenv "HOME" "C:\\home\\cgw") CW> CW> (insert (getenv "HOME")) C:\home\cgw CW> (insert (expand-file-name "test" "~/test/")) c:\test\test CW> CW> (setenv "HOME" "\\\\reactor\\cgwhome") CW> CW> (insert (getenv "HOME")) \\reactor\cgwhome CW> (insert (expand-file-name "test" "~/test/")) c:\test\test CW> CW> (setenv "HOME" "//reactor/cgwhome") CW> CW> (insert (getenv "HOME")) //reactor/cgwhome CW> (insert (expand-file-name "test" "~/test/")) c:\test\test CW> CW> CW> CW> With the entire block removed I get this result: CW> CW> CW> (setenv "HOME" "/home/cgw") CW> CW> (insert (getenv "HOME")) /home/cgw CW> (insert (expand-file-name "test" "~/test/")) c:\home\cgw\test\test CW> CW> (setenv "HOME" "C:\\home\\cgw") CW> CW> (insert (getenv "HOME")) C:\home\cgw CW> (insert (expand-file-name "test" "~/test/")) c:\home\cgw\test\test CW> CW> (setenv "HOME" "\\\\reactor\\cgwhome") CW> CW> (insert (getenv "HOME")) \\reactor\cgwhome CW> (insert (expand-file-name "test" "~/test/")) c:\home\cgw\test\test CW> CW> (setenv "HOME" "//reactor/cgwhome") CW> CW> (insert (getenv "HOME")) //reactor/cgwhome CW> (insert (expand-file-name "test" "~/test/")) c:\home\cgw\test\test CW> CW> CW> (1) (warning/warning) Error caught in `make-auto-save-file-name': CW> Creating directory: Invalid argument, c:\home\cgw\??/.autosave CW> CW> (2) (warning/warning) Error caught in `make-auto-save-file-name': CW> Creating directory: Invalid argument, c:\home\cgw\p?/.autosave -- Adrian Aichner European Design Center Teradyne GmbH Semiconductor Test Group Telephone +49/89/41861(0)-208 Dingolfinger Strasse 2 Fax +49/89/41861-217 (What is a Fax?) D-81673 MUENCHEN E-mail adrian.aichner(a)teradyne.com

1 participants
0 comments

Re: the splash buffer 27 years

Didier Verna

Hrvoje Niksic writes: Hrvoje> The only remark is that it won't work at 80x24 displays. Perhaps you Hrvoje> should squeeze it somehow in that case? Hmmm, I arranged it in order to make everything visible even under 80x24, appart indeed from the copyrights. Hrvoje> (And `useful' is spelled with one `l'!) oups, flyspell should be, like, dumped :-) -- / / _ _ Didier Verna http://www.inf.enst.fr/~verna/ - / / - / / /_/ / E.N.S.T. INF C201.1 mailto:verna＠inf.enst.fr /_/ / /_/ / /__ / 46 rue Barrault Tel. (33) 01 45 81 73 46 75634 Paris cedex 13 Fax. (33) 01 45 81 31 19

1 participants
0 comments

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

XEmacs-Beta December 1998