December 1999 - XEmacs-Beta - XEmacs Mailing List Archives

Oscar Figueiredo

Here's the third version of the patch introducing the new add/modify/delete functions to the LDAP API. This patch obsoletes the previous ones. Unlike the previous one, this one maintains compatibility with the former ldap-search function API released in 21.1 which should satisfy Kyle. For reference here's a summary of the changes introduced by this patch: The LDAP code design remains unchanged, it consists of two layers: - a low level layer that closely matches the C LDAP API. These functions that used to be named `*-internal' are now named more simply: `ldap-open', `ldap-close', `ldap-search-basic', `ldap-add', `ldap-modify'. They should be used directly for very specific purposes (such as multiple operations on a connection) only. - a higher-level layer provides a more convenient way to access LDAP directories hiding the subtleties of handling the connection, translating arguments and ensuring compliance with LDAP internationalization rules and formats (currently partly implemented only). This layer provides atomic operations for searches, modification, addition and deletion of multiple entries at once: ldap-search-entries, ldap-add-entries, ldap-delete-entries and ldap-modify-entries Ta maintain compatibility `ldap-search' is now a wrapper that calls either `ldap-search-basic' or `ldap-search-entries'. It's declared as obsolete since a direct call to one of these two functions is more efficient and should therefore be preferred. Please notify me if the patch is accepted so I can update the documentation too. Oscar 1999-11-27 Oscar Figueiredo <Oscar.Figueiredo(a)di.epfl.ch> * eldap.h (Fldap_search) Renamed from Fldap_search_internal: (Fldap_add, Fldap_modify, Fldap_delete): New functions * eldap.c (Qadd, Qreplace): New constant symbols (Fldap_open): Use the LDAP_OPT_RESTART option to avoid interruptions by XEmacs signals Remove unnecessary calls to slow_down_interrupts and speed_up_interrupts (Fldap_search_basic): Renamed from Fldap_search_internal Added new optional parameter VERBOSE that triggers the display of progress messages Remove unnecessary calls to slow_down_interrupts and speed_up_interrupts LDAP result code analysis rewritten (Fldap_add, Fldap_modify, Fldap_delete): New functions (syms_of_eldap): Define the new symbols and functions 1999-11-27 Oscar Figueiredo <Oscar.Figueiredo(a)di.epfl.ch> * ldap.el (toplevel): Test if LDAP support has been compiled in (ldap-verbose): New option (ldap-search-entries): Renamed from the previous ldap-search Conditionalize message display on ldap-verbose (ldap-add-entries, ldap-modify-entries, ldap-delete-entries): New defuns (ldap-search): Is now defined as an obsolete wrapper calling ldap-search-basic or ldap-search-entries

1 participants
0 comments

Re: Jobs assignments 26 years

William M. Perry

Gunnar Evermann <ge204(a)eng.cam.ac.uk> writes: > I think we need an 'FSF Liaison' much more urgently. This should be > someone who regularly tries to sync and coordinate changes/APIs with > FSFemacs. Preferably this person should be on the secret FSF mailing list > and have access to FSF pretests. Well, I have access to the FSF pretests and am on the mailing list, but Gerd asked a while ago that at least the pre-pre-pre-alpha version of the pretest not be distributed outside the small group of testers. This position may have changed now that the new redisplay is in the core Emacs 20.5.92 pretests. I'll have to ask him. Anybody else on both beta lists? If anyone has an ＠gnu.org account, you should be able to get the CVS distribution of Emacs. It is in :pserver:<acct>＠cvs.gnu.org:/gd/gnu/cvsroot - module name is 'emacs' of course. It would be much better to use kerberized rsh to get in there of course. -Bill P.

3 participants
2 comments

question regarding patches 26 years

Jeff Miller

a general question regarding patches. Normally, I submit patches generated with diff -ubw, to ignore whitespace changes. This normally makes the patches smaller and they just represent the actual code changes. This is the preferred method, right? Who, if anyone, should be re-indenting the code after patches are applied? Some of it gets pretty sloppy looking after awhile. Jeff

2 participants
1 comments

Re: New Athena widget configuration (was Re: Problem with `configure --with-sound') 26 years

Jan Vroonhof

"Daniel Pittman" <daniel(a)danann.net> writes: > The option `--with-athena' is available to specify the preferred Athena > widget set. This takes a range of values such as `flat' or `3d' or > `next' that specify what libraries to prefer. Perfect. But please make it clear in the docs that you must set one of the other options to athena for this to be usefull. > Without this option being specified, the user gets libXaw.so preferred, > with the flat or 3d headers according to the 3d-ness of libXaw.so. This > retains the current behavior of `--with-widgets=athena' -- no surprises > to people who don't pay attention. Actually I would prefer to have one of the sexier ones be the default <big>IF</big> we can detect them reliably. > In this vein, `--with-widgets=athena3d' would be the same as the new > `--with-athena=3d' and `--with-widgets=athena' for backward > compatibility. Is it really needed to be backwards compatible with a few betas? Jan

4 participants
4 comments

Re: Scanning FSF 20.5 for fixes.. 26 years

Andreas Jaeger

>>>>> Jan Vroonhof writes: > Rumour has it FSF 20.5 contains a few (minor) Y2K fixes. It would be > nice if somebody could so a scan of the Chanelogs what bugs they fixed > (not just the Y2K ones). > We don't have the embarrassing fix in timezone.el. A quick grep just revealed the following ChangeLog entries: 1999-09-07 Edward M. Reingold <reingold(a)emr.cs.uiuc.edu> * diary-lib.el (list-diary-entries): Make sure two-digit years have leading zeros. 1999-09-06 Richard Stallman <rms(a)gnu.org> * vc.el (vc-annotate-display): Use < 69 not < 70 to distinguish 20YY from 19YY. * timezone.el (timezone-parse-date): Use < 69 not < 70 to distinguish 20YY from 19YY. 1999-09-05 Richard Stallman <rms(a)gnu.org> * vc.el (vc-annotate-display): Treat 2-digit years under 70 as 20YY. 1998-08-29 Edward M. Reingold <reingold(a)emr.cs.uiuc.edu> * cal-china.el (chinese-year-cache): Change range of years from 1989-2000 to 1990-2010. 1999-08-24 Dave Love <fx(a)gnu.org> * timezone.el (timezone-parse-date): Assume 2-digit dates <70 are 2000+. I'm going to merge the cal-china.el and vc.el changes. I would prefer if somebody else checks this. Btw. we check for < 70 but shouldn't we check for < 69? AFAIK the rule is that 69-99 refers to the twentieth century. I've sent patches to xemacs-patches for these problems. Andreas -- Andreas Jaeger SuSE Labs aj(a)suse.de private aj(a)arthur.rhein-neckar.de

2 participants
1 comments

Re: More ESD ./configure bogosity 26 years

Jan Vroonhof

Raymond Toy <toy(a)rtp.ericsson.se> writes: > >> Why can't we reliably detect NAS? Isn't the existence of a few > >> include files and libaudio enough? I now remember another reason why this is probably not autodetected at the moment (but which is no longer valid). It used to be that NAS always was a thing you had to install yourself and thus was never in any obvious location. However on modern Free Unix'es that is no longer the case. > Why would you have to do that? Wouldn't configure just try to compile > something like You are right... Jan

2 participants
1 comments

[PATCH] New Athena configure support done. 26 years

Daniel Pittman

I have just posted my patch to enable the new Athena widget support in configure to xamecs-patchs, msg-id <87zovgc33g.fsf(a)inanna.danann.net>. If I can get a bunch of people to test this, that would be great. I tried as many different breakages of the Athena install as I could find, but only had the one machine on which to do this. Let me know if anything breaks for you, especially if the script makes the wrong guess about what your libraries are or where the right headers are. Daniel -- A man can no more diminish God's glory by refusing to worship Him than a lunatic can put out the sun by scribbling the word 'darkness' on the walls of his cell. -- C. S. Lewis, _The problem of pain_

1 participants
0 comments

More binary files fixed up in CVS 26 years

Martin Buchholz

Here are some more files that should be marked as binary in CVS. (martin＠lasker) ws/dev/nt/installer/Wise $ cvs-rw admin -kb gnu.bmp RCS file: /usr/CVSroot/XEmacs/xemacs/nt/installer/Wise/Attic/gnu.bmp,v done (martin＠lasker) ws/dev/nt/installer/Wise $ cvs-rw admin -kb xemacs-beta.bmp RCS file: /usr/CVSroot/XEmacs/xemacs/nt/installer/Wise/Attic/xemacs-beta.bmp,v done With this, all the *.ico and *.bmp files in CVS are marked as binary. You probably won't notice any difference, unless you are a developer who plans to modify these files sometime in the future. In that case, I recommend starting over with a fresh copy of the `nt' directory in your CVS checked-out workspace as follows: # Make sure you have no files modified in nt first rm -rf nt; cvs update -d Martin

1 participants
0 comments

Re: XEmacs 21.2.23 crashes when url-insert-file-contents 26 years

Hrvoje Niksic

Gunnar Evermann <ge204(a)eng.cam.ac.uk> writes: [...] > this instantly triggers the assertion. The assertion happens in report_extent_modification_mapper(), which I added, so it's sort of my responsibility. But the funny thing about the crashes is that no extents even have the modification functions hooked into them. Maybe report_extent_modification_mapper() is getting called while the extents are in an inconsistent state? I don't see how that can be, as signal_after_change() is the *last* of the functions to be called from buffer_insert_string_1 and friends. The only way to see what's going on is to step through the functions and try to wizard out what hoses the buffer so that carcount->bytind conversion fails.

1 participants
0 comments

Re: XEmacs 21.2.23 crashes when url-insert-file-contents 26 years

Gunnar Evermann

wmperry(a)aventail.com (William M. Perry) writes: > Shenghuo ZHU <zsh(a)cs.rochester.edu> writes: > > > When I evaluate > > (url-insert-file-contents "http://www.mail-archive.com/lists.html") > > > > #4 0x4040d868 in abort () at ../sysdeps/generic/abort.c:88 > > #5 0x808e7ca in assert_failed () at emacs.c:2825 > > #6 0x81331a7 in bufpos_to_bytind_func (buf=0x8852670, x=1025) at insdel.c:581 > > #7 0x8133b8a in bufpos_to_bytind (buf=0x8852670, x=1025) at buffer.h:985 > > #8 0x80e4cc6 in map_extents (from=1, to=1025, > > fn=0x80ea2f0 <report_extent_modification_mapper>, arg=0xbfffdf78, > > obj=142943856, after=0x0, flags=15360) at extents.c:2272 > > #9 0x80f023d in report_extent_modification (buffer=142943856, start=1, > > end=1025, inside=0x8269b34, afterp=1) at extents.c:4697 > > #10 0x8135b60 in signal_after_change (buf=0x8852670, start=1, orig_end=1, > > new_end=1025) at insdel.c:2298 > > Eek, looks like a bug in the after-change-hooks processing... indeed. We have seen something like this (at least) once before [1]. Have a look at David Byers' (excellent) report at: http://www.xemacs.org/list-archives/xemacs-beta/9905/msg00350.html the short recipe is: (switch-to-buffer "FOO") (insert "123456") (defun acf (start end len) (widen) (goto-char 4) (delete-char 1)) (let ((after-change-functions '(acf))) (save-restriction (narrow-to-region 4 4) (transpose-lines 1))) this instantly triggers the assertion. I think I looked at this some time ago, but can't remember the details. I'll try to investigate some more in the next couple of days, but any help is of course greatly appreciated. Gunnar Footnotes: [1] Actually I immediatedly remembered it, but to be honest it took me a couple of hours of grepping through my harddisk and searching the BTS until I finally tried searching in the xemacs-beta archive and found it :-)

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 1999