7:11 p.m.
Please test the release candidate.
Use "cvs update -r r21-4-9rc1" to get the RC sources. I hope to get
tarballs up tomorrow or Monday my time. For now CVS is up-to-date and
so I'm announcing for those testers who use CVS.
Use "cvs update -r candidate-21-4" to get the latest and greatest.
Don't even expect checkouts on this tag to build; I commit what I have
when I feel like it. Mostly it should work, but if it doesn't, take
two aspirin and cvs update in the morning. ;-)
"Release candidate" is something of a misnomer, there are a number of
patches in the queue that I probably will apply, and I really hope
that the syntax.c problems (aka "font-lock slows") will get fixed for
21.4.9. We'll see .... Anyway, expect at least one more candidate
before 21.4.9.
TESTING PROCEDURE
1. Get the sources.
2. Configure as usual, and if you're feeling especially chipper, try
several variants that are new and different.
3. Use M-x build-report to report each build and any problems. If you
have problems, please CC xemacs-beta. Successful builds can just
go to xemacs-buildreports.
4. Test and send reports to xemacs-beta, positive or negative.
WHAT TO TEST
1. I've included the work done by Andy and Nix to support vertical
positioning of glyphs and some other redisplay hacking needed to
support preview-latex. This is most definitely a feature, and
therefore I really should not include it in 21.4. Please try
preview-latex (http://preview-latex.sourceforge.net/ IIRC). The more
reports and the more detail I get, the more likely I am to include
it. Maybe not 21.4.9, maybe 21.4.10.
Please beat hard on this one. preview-latex is really cool and I
would like to support it in 21.4.
2. I've included Andy's revert-buffer optimization. The idea is to
compare the buffer text to the file contents, and not redo expensive
buffer analysis (eg, font-lock and semantic bovination) if the text is
the same.
There are reports on 21.5 that this code trashes markers and extents
when you revert. This is probably fixed already, but please be on the
watch for such problems, and any anomolies when you revert. Again,
this code may not make it in to 21.4.9.
3. Builds on Linux PPC, OpenBSD, and any 64-bit platforms would be
nice, I've added some configure tests and s&m pervection to support
those. MacOS/X Darwin is already in 21.4.8, but build reports on that
platform would be nice, too.
The OpenBSD changes include detection of a OpenBSD weirdness where
they rename libcurses as libtermcap or something like that. I think
it's well-guarded by an "if [ opsys = openbsd ]", but be on the
lookout for configure failures and runtime weirdness for TTY stuff.
4. XEmacs on Windows is becoming less broken. Once again I've merged
all of the changes from Andy's branch, but with the large number of
changes going in to the Windows code still, reports of the demise of
the Windows branch are most likely severely exaggerated. :-( The one
new fix not in XEmacs 21.4.7 (Windows [1]) is the format-time-string
fix from Nick Pakoulin (crash on negative times).
So Windows testing isn't terribly important; most of the code in
the RC is already in 21.4.7 Windows [1]. However, build reports
(especially Cygwin) would be nice, just to make sure the configure
hacking hasn't messed anything up.
5. Miscellanea:
-- C-x r p -- new key binding for replace-rectangle
-- Completion lists should work better now (confusion between
minibuffer and completion buffer removed)
CHANGELOG DETAILS
**** ChangeLog
2002-05-17 Stephen J. Turnbull <stephen(a)xemacs.org>
* PROBLEMS: Describe "Hannibal Lecter" crash on Linux. Reorder
Linux runtime issues section.
* INSTALL: Improve description of mail locking.
2002-05-10 Stephen J. Turnbull <stephen(a)xemacs.org>
* etc/OXYMORONS (21.4.15): New. Not a reference to Vin.
* build-msw-release.sh: Synch to Andy.
2002-03-18 Gregory Steuck <greg-xemacs-patch(a)nest.cx>
* configure.in: Improve OpenBSD support: correct check for
socklen_t, check for libtermcap is libncurses.
2002-02-28 Malcolm Purvis <malcolmpurvis(a)optushome.com.au>
* configure.in: Protect nocomboreloc linker arguments to avoid
problem on the ppc.
**** lisp/ChangeLog
2002-05-19 Stephen J. Turnbull <stephen(a)xemacs.org>
* files.el (revert-buffer):
(revert-buffer-internal):
(revert-buffer-insert-file-contents-function):
Improve docstrings and comments.
2002-05-08 Andy Piper <andy(a)xemacs.org>
* files.el (revert-buffer): observe noconfirm behavior.
2002-04-09 Andy Piper <andy(a)xemacs.org>
* files.el (revert-buffer): use revert-buffer-internal if it looks
like doing so will not result in any user-visible changes.
(revert-buffer-internal): new function. Do the actual process of
reversion and then see whether the result is any different to what
we have already. If it is not then do nothing.
2001-12-11 Andy Piper <andy(a)xemacs.org>
* menubar.el (get-popup-menu-response): re-order so that it works
on more sane/facist window systems.
2001-12-03 Andy Piper <andy(a)xemacs.org>
* faces.el (frob-face-property): don't infloop in face frobbing
from Jan Vroonhof <jan(a)xemacs.org>.
2001-11-30 Jan Vroonhof <jan(a)xemacs.org>
* faces.el (frob-face-property): Follow face fall-back hierarchy
properly for face properties without an instance. Only do manual
copy form 'default in last resort. This handles in particular
the case where 'default itself has only a fall-back (which is
the case by default on windows).
2001-11-24 Andy Piper <andy(a)xemacs.org>
* printer.el (generic-print-region): set default-frame-plist to
nil while creating the printer frame so that sizes reflect the
printed page.
* faces.el (face-complain-about-font): Don't complain on printers.
2002-05-10 Stephen J. Turnbull <stephen(a)xemacs.org>
* package-net.el: Andy synch.
2001-12-16 Andy Piper <andy(a)xemacs.org>
* package-get.el (package-get-update-all): Make sure installed.db
gets updated after updating packages.
2002-03-28 Ben Wing <ben(a)xemacs.org>
* msw-glyphs.el: Set glyph-contrib-p to nil on hscroll,
continuation, truncation so that we don't get line jumpiness from
them being larger than the line height.
* msw-font-menu.el (mswindows-font-menu-font-data): Fix bug.
2001-11-30 Jan Vroonhof <jan(a)xemacs.org>
* font.el (font-window-system-mappings): Add mapping for Gtk
(assume identical to X)
2002-02-04 Andy Piper <andy(a)xemacs.org>
* files.el (convert-standard-filename): Fix for short filename
Peter Arius <pas(a)methodpark.de>
2001-12-10 Christoph Wedler <Christoph.Wedler(a)sap.com>
* list-mode.el (default-choose-completion): completion-base-size
should refer to local value in the completion list buffer.
2001-12-03 Didier Verna <didier(a)xemacs.org>
* rect.el: autoload `replace-rectangle'.
* keydefs.el (global-map): bind it to `C-x r p'.
**** src/ChangeLog
2002-04-25 Andy Piper <andy(a)xemacs.org>
* redisplay.c (create_text_block): Remove extra comment trailer.
2002-04-24 Andy Piper <andy(a)xemacs.org>
* redisplay.c (create_text_block): Don't actually add propagation
data if the line ends after we have added a glyph.
2002-04-22 Andy Piper <andy(a)xemacs.org>
* extents.c (extent_fragment_update): check for glyphs we have
previously displayed.
* extents.c (print_extent_1): warning removal.
* extents.h: change prototype.
* redisplay-output.c (redisplay_normalize_glyph_area): calculate
widths correctly for wide glyphs.
* redisplay.c (position_redisplay_data_type): add end_glyph_width.
* redisplay.c (prop_type): add PROP_GLYPH.
* redisplay.c (struct prop_block): add glyph type
* redisplay.c (add_glyph_rune): when adding part of a glyph add it
to the propagation data.
* redisplay.c (create_text_block): if there is a glyph in the
propagation data use it to salt extent_fragment_update.
* redisplay.c (create_string_text_block): ditto.
2002-04-13 Nix <nix(a)esperi.demon.co.uk>
* redisplay.h (struct rune): Add ascent, descent, and yoffset fields.
* redisplay-output.c (compare_runes): Compare them.
* redisplay.c: Update copyright date.
* redisplay.c (pos_data): Add need_baseline_computation field.
* redisplay.c (add_glyph_rune): Update ascent, descent, and
need_baseline_computation; zero yoffset. Set max_pixmap_height
for all pixmaps, not just automatically positioned ones.
* redisplay.c (calculate_yoffset): New, compute yoffset values.
* redisplay.c (calculate_baseline): New, compute textual baseline.
* redisplay.c (add_glyph_rune): Call them.
* redisplay.c (create_text_block): Likewise.
* redisplay.c (create_overlay_glyph_block): Likewise.
* redisplay.c (add_margin_runes): Likewise.
* redisplay.c (create_string_text_block): Likewise. Fix tabdamage.
* redisplay.h: (redisplay_calculate_display_boxes): Change prototype.
* redisplay-output.c (redisplay_calculate_display_boxes): Use yoffset.
* redisplay-msw.c (mswindows_output_blank): Pass 0 as yoffset.
* redisplay-msw.c (mswindows_output_string): Likewise.
* redisplay-msw.c (mswindows_output_display_block): Pass yoffset.
* redisplay-gtk.c (gtk_output_display_block): Likewise.
* redisplay-x.c (x_output_display_block): Likewise.
2002-03-28 Ben Wing <ben(a)xemacs.org>
* redisplay.c: Fixed bug in redisplay
w.r.t. hscroll/truncation/continuation glyphs causing jumping up
and down of the lines, since they're bigger than the line
size. (It was seen most obviously when there's a horizontal scroll
bar, e.g. do C-h a glyph or something like that.) The problem was
that the glyph-contrib-p setting on glyphs was ignored even if it
was set properly, which it wasn't until now.
2002-04-24 Andy Piper <andy(a)xemacs.org>
* lisp.h (Dynarr_end): Fix definition.
2002-02-06 Adrian Aichner <adrian(a)xemacs.org>
* redisplay.c (mark_redisplay): Remove call to
update_frame_window_mirror.
2001-11-15 Andy Piper <andy(a)xemacs.org>
* win32.c (Fmswindows_shell_execute): fix handling of URL's under
cygwin (again).
2002-02-13 Andy Piper <andy(a)xemacs.org>
* event-msw.c (mswindows_wnd_proc): only mark the frame visible if
we did in fact enqueue the XM_MAPFRAME event.
2002-01-15 Adrian Aichner <adrian(a)xemacs.org>
* event-msw.c (mswindows_wnd_proc): Add handling of WM_SHOWWINDOW
to fix problem switching between virtual desktops under virtuawin
virtual window manager.
2001-07-30 Adrian Aichner <adrian(a)xemacs.org>
* event-msw.c: Typo fix.
* event-msw.c (mswindows_wnd_proc): Set FRAME_VISIBLE_P after
magic XM_MAPFRAME event has been sent.
2001-11-23 Andy Piper <andy(a)xemacs.org>
* event-msw.c (mswindows_wnd_proc): Don't pump mousewheel events.
2001-11-21 Andy Piper <andy(a)xemacs.org>
* scrollbar-msw.c (mswindows_handle_mousewheel_event): cope with
mouse events outside the frame.
2002-03-20 Andy Piper <andy(a)xemacs.org>
* menubar-msw.c (mswindows_popup_menu): warning removal.
* dialog-msw.c (dialog_popped_down): ditto.
2001-12-11 Andy Piper <andy(a)xemacs.org>
* dialog-msw.c (dialog_popped_down): new function. unset popup_up_p.
* dialog-msw.c (mswindows_make_dialog_box_internal): set
popup_up_p.
* menubar-msw.c (unsafe_handle_wm_initmenupopup_1): ditto.
* menubar-msw.c (mswindows_handle_wm_command): ditto.
* menubar-msw.c (mswindows_popup_menu): ditto.
2001-10-29 Andy Piper <andy(a)xemacs.org>
* dialog-msw.c (handle_directory_dialog_box): quit if the user
cancels.
2002-01-03 Andy Piper <andy(a)xemacs.org>
* realpath.c (ABS_LENGTH): dtrt for cygwin systems using drive
letters.
(xrealpath): ditto.
2002-03-29 Jonathan Harris <jonathan(a)xemacs.org>
* device-msw.c (plist_get_margin): Add arg specifying mm or inches
* device-msw.c (plist_set_margin): Fix multiplicand used for mm
* device-msw.c (mswindows_handle_page_setup_dialog_box):
Detect and handle case where machine is set up for metric units
2002-04-02 Andy Piper <andy(a)xemacs.org>
* dired-msw.c (mswindows_get_files):
* nt.c (mswindows_stat): SetErrorMode() so that file errors are
completely handled by XEmacs. Suggested by Thomas Vogler
<mail(a)thomas-vogler.de>.
2001-11-14 Andy Piper <andy(a)xemacs.org>
* nt.c (REG_ROOT): change registry key to XEmacs.
2002-04-27 Andy Piper <andy(a)xemacs.org>
* glyphs-msw.c (mswindows_widget_instantiate): remove dead-code.
2002-04-26 Andy Piper <andy(a)xemacs.org>
* glyphs-msw.c (mswindows_map_subwindow): observe :initial-focus
behavior.
2002-03-14 Mike Alexander <mta(a)arbortext.com>
* event-msw.c (mswindows_unwait_process): New, remove process from
wait list
* process-nt.c (nt_finalize_process_data): Call
mswindows_unwait_process
* console-msw.h: Declare mswindows_unwait_process
2002-05-14 Stephen J. Turnbull <stephen(a)xemacs.org>
* editfns.c (Fdecode_time):
(Fformat_time_string):
Check for invalid time. Thanks to Nick Pakoulin <npak(a)ispras.ru>.
(make_time): Warning elimination (change arg name).
2001-11-24 Andy Piper <andy(a)xemacs.org>
* window.c (Fsplit_window): Doc return type.
2002-04-01 Andy Piper <andy(a)xemacs.org>
* emacs.c (Fkill_emacs): Only output message box in interactive
mode.
2002-04-26 Andy Piper <andy(a)xemacs.org>
* config.h.in: pull in 21.5.x change to stop alloca warnings under
cygwin.
2002-03-18 Gregory Steuck <greg-xemacs-patch(a)nest.cx>
* unexelf.c: Use ELFSIZE or define from _LP64 to indicate a 64
bit platform.
**** man/ChangeLog
2002-05-11 Adrian Aichner <adrian(a)xemacs.org>
* xemacs-faq.texi (Top): Add Q2.0.15 under "Installation and
Trouble Shooting".
* xemacs-faq.texi (Installation): Add Q2.0.15 to Installation
menu.
* xemacs-faq.texi (Q2.0.14): Fix unnumberedsubsec argument.
* xemacs-faq.texi (Q2.0.15): New.
**** lib-src/ChangeLog
2002-05-10 Stephen J. Turnbull <stephen(a)xemacs.org>
* Makefile.in.in: Add winclient targets (Andy synch).
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
My nostalgia for Icon makes me forget about any of the bad things. I don't
have much nostalgia for Perl, so its faults I remember. Scott Gilbert c.l.py
2:01 p.m.
On 19 May 2002, Stephen J. Turnbull said:
1. I've included the work done by Andy and Nix to support
vertical
positioning of glyphs and some other redisplay hacking needed to
support preview-latex. This is most definitely a feature, and
I'd call it a bug fix. `set-glyph-baseline' didn't work as documented
(or do anything *useful* at all); certainly whatever property it set
wasn't the line's baseline. Now it works as documented.
It's not a small bug fix, but compared to e.g. the syntax table changes
(also bugfixes) it's a tiddler.
(Note that this bug also causes problems in w3m and probably in w3 too
--- I don't know, I don't use w3. It potentially causes problems
anywhere where multiple glyphs of different heights are displayed on the
same display line.)
preview-latex (http://preview-latex.sourceforge.net/ IIRC). The
more
reports and the more detail I get, the more likely I am to include
it. Maybe not 21.4.9, maybe 21.4.10.
CVS updating now; I'll build and test it as soon as I've finished
upgrading GCC (so you'll get a success/failure report for compilation of
XEmacs with GCC 3.1/binutils-2.11.2, into the bargain.)
Please beat hard on this one. preview-latex is really cool and I
would like to support it in 21.4.
For what it's worth I've been running XEmacsen with this patch on
several sites and torturing them quite hard, and I haven't seen any
problems with the patch so far. (I saw problems with earlier versions,
but they're all fixed by the version you applied.)
I can't be sure there'll be no regressions, but I haven't seen any on
i586-pc-linux-gnu, hppa2.0-hp-hpux10.20, sparc-sun-solaris2.8,
sparc-unknown-linux-gnu, or alphaev56-dec-osf4.0d. :)
2. I've included Andy's revert-buffer optimization. The idea
is to
compare the buffer text to the file contents, and not redo expensive
buffer analysis (eg, font-lock and semantic bovination) if the text is
the same.
As a heavy user of auto-revert-buffer with huge files on slow machines
with data coming from *dog* slow NFS servers, I applaud.
3. Builds on Linux PPC, OpenBSD, and any 64-bit platforms would be
nice, I've added some configure tests and s&m pervection to support
those. MacOS/X Darwin is already in 21.4.8, but build reports on that
platform would be nice, too.
I'm building a sparc64-unknown-linux-gnu environment at the moment: I'll
try it there. (I don't expect it to work, but I'll be happy if it does.)
It might be a while before I report back on this one though: I'll need
to build 64-bit versions of all *sorts* of things before I can get as
far as building XEmacs. (I haven't even got a 64-bit libc yet ;} )
--
`There are not words enough to describe how fucked up imake is.'
--- Peter da Silva
3:38 p.m.
>>>> "nix" == nix
<nix(a)esperi.demon.co.uk> writes:
nix> It's not a small bug fix, but compared to e.g. the syntax
nix> table changes (also bugfixes) it's a tiddler.
Ben's changes won't go in to 21.4 on my watch. I know I have no hope
of understanding that monster. What we're testing here is whether it
makes sense for me to try to understand your patch.
nix> As a heavy user of auto-revert-buffer with huge files on slow
nix> machines with data coming from *dog* slow NFS servers, I
nix> applaud.
Yeah, well, there's already a bug report in, against 21.4.7 Windows
[1] which is the same code I put in. :-(
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
My nostalgia for Icon makes me forget about any of the bad things. I don't
have much nostalgia for Perl, so its faults I remember. Scott Gilbert c.l.py
4:46 p.m.
On 19 May 2002, Stephen J. Turnbull said:
>>>>> "nix" == nix
<nix(a)esperi.demon.co.uk> writes:
nix> It's not a small bug fix, but compared to e.g. the syntax
nix> table changes (also bugfixes) it's a tiddler.
Ben's changes won't go in to 21.4 on my watch. I know I have no hope
of understanding that monster. What we're testing here is whether it
makes sense for me to try to understand your patch.
It isn't complex; the one bit that might be called slightly complex has
a huge comment above it to explain WTF it's doing.
It just arranges to track the baseline on a per-rune basis (which you
need to do if each rune is to have its own baseline, as we promised) and
keeps the baseline of the whole line up to date from the per-rune
baseline values.
I'll delineate it insanely verbosely for you on the off-chance that
helps (love the sound of my own voice? Never!):
Tracking the baseline is the most complex bit. We keep `ascent',
`descent' and `yoffset' in the rune to independently track all these
parameters. (We only need at most two of them at any one time, but the
inefficiency won't kill us.) We also remember if anyone explicitly set
the baseline on each rune: if nobody bothered, we don't do anything very
special and all this code reduces to almost nothing.
add_glyph_rune() is modified to remember the ascent and descent of each
rune, and keep track of the maximum height, maximum ascent and maximum
descent of any glyph on the line: this is because if the line is tall, a
short glyph won't affect the baseline of the line as a whole (the
line'll just move up or down on it).
create_text_block() and all the other functions where we used to
compute the baseline get that repeated code torn out, and instead
we call two functions.
The first one (calculate_baseline()) is the one with the big comment
above it, and is the most complicated change: it works out what the
baseline for the line as a whole should be given the linewide maximum
ascent and descent, the baseline of the default face and other
things. The comment explains it more clearly than I can here. :) Most of
this code is generalizations of code that was being run by one or
another of the routines elsewhere in redisplay (and has been removed
from them and moved here).
When this function returns, the line's new_ascent/new_descent are
known. The caller then uses these to work out the line's actual ascent
and descent in just the same way they always did (it differs for
different callers), and then they call calculate_yoffset().
calculate_yoffset() takes the ascent of the line and the ascent and
descent of each glyph rune on the line, and sets up that rune's yoffset
field: that being the amount by which the rune must be shifted down from
the topmost point the display_line is plotted at (`upper baseline'?) to
appear at the right position, given its ascent and descent, and the
ascent of the line (which the caller has turned the line's baseline
into). This is what XEmacs used to use as a `baseline' (even though it's
almost exactly the opposite). Non-glyph runes always have a zero
yoffset, which means `use the pre-existing line-wide ascent as usual'.
The net effect of all this is that the glyphs or bits of text that stick
up or down furthest on the line control the baseline of the line, while
the glyphs themselves are correctly positioned with respect to this
baseline.
The changes to redisplay-{output,msw,gtk,x}.c are all parameter changes
to pass the yoffset to the routines that do the display, and a tiny
calculation change in redisplay-output.c:redisplay_calculate_display_boxes()
to actually use the yoffset value, instead of the value it used to use;
these changes are mostly mechanical and boring.
(There's also a small change in
redisplay-output.c:redisplay_normalize_glyph_area(), which through
oversight hasn't been added to the changelog: it arranges that if a
partial-line scroll is in place and a glyph has been ascented up so it's
now completely invisible on a line that is still otherwise visible, the
glyph's height stays at something sane, rather than wrapping to a huge
positive value and mushing up the screen.)
A suitable changelog line for this would be
* redisplay-output.c (redisplay_normalize_glyph_area): Cater for
that glyphs that are pushed above the top of the visible area
on still-visible lines.
(Oops. Everything else is described in the changelog, I just checked.)
Hm. That's still not very clear, but I hope this rephrasing makes it a
bit clearer than it was. I tried to explain it in the original patch
postings, too: if you read all of these together then some sense might
emerge:)
Thanks for considering this at this late stage.
nix> As a heavy user of auto-revert-buffer with huge files on
slow
nix> machines with data coming from *dog* slow NFS servers, I
nix> applaud.
Yeah, well, there's already a bug report in, against 21.4.7 Windows
[1] which is the same code I put in. :-(
Damn:( still, until you've seen an NFS server with an effective file
transmission rate of 10K/s you haven't seen slow. Anything that
reduces file access from a box like that is good. :)
--
`There are not words enough to describe how fucked up imake is.'
--- Peter da Silva
4:54 a.m.
Well, for Cygwin there seems to be a file called lib-src/winclient.c missing
from CVS. I'm not that conversant with CVS, though, so I could have done
something wrong. I ran
cvs co -r r21-4-9rc1 -d xemacs-21.4.9 xemacs
to get the source, followed by configure, which seemed to run OK. Make,
however, failed with
make[1]: *** No rule to make target
`/pub/src/xemacs-21.4.9/lib-src/winclient.c', needed by `winclient'. Stop.
make[1]: Leaving directory `/pub/src/xemacs-21.4.9/lib-src'
make: *** [lib-src] Error 2
--Rick
--- "Stephen J. Turnbull" <stephen(a)xemacs.org> wrote:
Please test the release candidate.
Use "cvs update -r r21-4-9rc1" to get the RC sources. I hope to get
tarballs up tomorrow or Monday my time. For now CVS is up-to-date and
so I'm announcing for those testers who use CVS.
Use "cvs update -r candidate-21-4" to get the latest and greatest.
Don't even expect checkouts on this tag to build; I commit what I have
when I feel like it. Mostly it should work, but if it doesn't, take
two aspirin and cvs update in the morning. ;-)
"Release candidate" is something of a misnomer, there are a number of
patches in the queue that I probably will apply, and I really hope
that the syntax.c problems (aka "font-lock slows") will get fixed for
21.4.9. We'll see .... Anyway, expect at least one more candidate
before 21.4.9.
TESTING PROCEDURE
1. Get the sources.
2. Configure as usual, and if you're feeling especially chipper, try
several variants that are new and different.
3. Use M-x build-report to report each build and any problems. If you
have problems, please CC xemacs-beta. Successful builds can just
go to xemacs-buildreports.
4. Test and send reports to xemacs-beta, positive or negative.
WHAT TO TEST
1. I've included the work done by Andy and Nix to support vertical
positioning of glyphs and some other redisplay hacking needed to
support preview-latex. This is most definitely a feature, and
therefore I really should not include it in 21.4. Please try
preview-latex (http://preview-latex.sourceforge.net/ IIRC). The more
reports and the more detail I get, the more likely I am to include
it. Maybe not 21.4.9, maybe 21.4.10.
Please beat hard on this one. preview-latex is really cool and I
would like to support it in 21.4.
2. I've included Andy's revert-buffer optimization. The idea is to
compare the buffer text to the file contents, and not redo expensive
buffer analysis (eg, font-lock and semantic bovination) if the text is
the same.
There are reports on 21.5 that this code trashes markers and extents
when you revert. This is probably fixed already, but please be on the
watch for such problems, and any anomolies when you revert. Again,
this code may not make it in to 21.4.9.
3. Builds on Linux PPC, OpenBSD, and any 64-bit platforms would be
nice, I've added some configure tests and s&m pervection to support
those. MacOS/X Darwin is already in 21.4.8, but build reports on that
platform would be nice, too.
The OpenBSD changes include detection of a OpenBSD weirdness where
they rename libcurses as libtermcap or something like that. I think
it's well-guarded by an "if [ opsys = openbsd ]", but be on the
lookout for configure failures and runtime weirdness for TTY stuff.
4. XEmacs on Windows is becoming less broken. Once again I've merged
all of the changes from Andy's branch, but with the large number of
changes going in to the Windows code still, reports of the demise of
the Windows branch are most likely severely exaggerated. :-( The one
new fix not in XEmacs 21.4.7 (Windows [1]) is the format-time-string
fix from Nick Pakoulin (crash on negative times).
So Windows testing isn't terribly important; most of the code in
the RC is already in 21.4.7 Windows [1]. However, build reports
(especially Cygwin) would be nice, just to make sure the configure
hacking hasn't messed anything up.
5. Miscellanea:
-- C-x r p -- new key binding for replace-rectangle
-- Completion lists should work better now (confusion between
minibuffer and completion buffer removed)
CHANGELOG DETAILS
**** ChangeLog
2002-05-17 Stephen J. Turnbull <stephen(a)xemacs.org>
* PROBLEMS: Describe "Hannibal Lecter" crash on Linux. Reorder
Linux runtime issues section.
* INSTALL: Improve description of mail locking.
2002-05-10 Stephen J. Turnbull <stephen(a)xemacs.org>
* etc/OXYMORONS (21.4.15): New. Not a reference to Vin.
* build-msw-release.sh: Synch to Andy.
2002-03-18 Gregory Steuck <greg-xemacs-patch(a)nest.cx>
* configure.in: Improve OpenBSD support: correct check for
socklen_t, check for libtermcap is libncurses.
2002-02-28 Malcolm Purvis <malcolmpurvis(a)optushome.com.au>
* configure.in: Protect nocomboreloc linker arguments to avoid
problem on the ppc.
**** lisp/ChangeLog
2002-05-19 Stephen J. Turnbull <stephen(a)xemacs.org>
* files.el (revert-buffer):
(revert-buffer-internal):
(revert-buffer-insert-file-contents-function):
Improve docstrings and comments.
2002-05-08 Andy Piper <andy(a)xemacs.org>
* files.el (revert-buffer): observe noconfirm behavior.
2002-04-09 Andy Piper <andy(a)xemacs.org>
* files.el (revert-buffer): use revert-buffer-internal if it looks
like doing so will not result in any user-visible changes.
(revert-buffer-internal): new function. Do the actual process of
reversion and then see whether the result is any different to what
we have already. If it is not then do nothing.
2001-12-11 Andy Piper <andy(a)xemacs.org>
* menubar.el (get-popup-menu-response): re-order so that it works
on more sane/facist window systems.
2001-12-03 Andy Piper <andy(a)xemacs.org>
* faces.el (frob-face-property): don't infloop in face frobbing
from Jan Vroonhof <jan(a)xemacs.org>.
2001-11-30 Jan Vroonhof <jan(a)xemacs.org>
* faces.el (frob-face-property): Follow face fall-back hierarchy
properly for face properties without an instance. Only do manual
copy form 'default in last resort. This handles in particular
the case where 'default itself has only a fall-back (which is
the case by default on windows).
2001-11-24 Andy Piper <andy(a)xemacs.org>
* printer.el (generic-print-region): set default-frame-plist to
nil while creating the printer frame so that sizes reflect the
printed page.
* faces.el (face-complain-about-font): Don't complain on printers.
2002-05-10 Stephen J. Turnbull <stephen(a)xemacs.org>
* package-net.el: Andy synch.
2001-12-16 Andy Piper <andy(a)xemacs.org>
* package-get.el (package-get-update-all): Make sure installed.db
gets updated after updating packages.
2002-03-28 Ben Wing <ben(a)xemacs.org>
* msw-glyphs.el: Set glyph-contrib-p to nil on hscroll,
continuation, truncation so that we don't get line jumpiness from
them being larger than the line height.
* msw-font-menu.el (mswindows-font-menu-font-data): Fix bug.
2001-11-30 Jan Vroonhof <jan(a)xemacs.org>
* font.el (font-window-system-mappings): Add mapping for Gtk
(assume identical to X)
2002-02-04 Andy Piper <andy(a)xemacs.org>
* files.el (convert-standard-filename): Fix for short filename
Peter Arius <pas(a)methodpark.de>
2001-12-10 Christoph Wedler <Christoph.Wedler(a)sap.com>
* list-mode.el (default-choose-completion): completion-base-size
should refer to local value in the completion list buffer.
2001-12-03 Didier Verna <didier(a)xemacs.org>
* rect.el: autoload `replace-rectangle'.
* keydefs.el (global-map): bind it to `C-x r p'.
**** src/ChangeLog
2002-04-25 Andy Piper <andy(a)xemacs.org>
* redisplay.c (create_text_block): Remove extra comment trailer.
2002-04-24 Andy Piper <andy(a)xemacs.org>
* redisplay.c (create_text_block): Don't actually add propagation
data if the line ends after we have added a glyph.
2002-04-22 Andy Piper <andy(a)xemacs.org>
* extents.c (extent_fragment_update): check for glyphs we have
previously displayed.
* extents.c (print_extent_1): warning removal.
* extents.h: change prototype.
* redisplay-output.c (redisplay_normalize_glyph_area): calculate
widths correctly for wide glyphs.
* redisplay.c (position_redisplay_data_type): add end_glyph_width.
* redisplay.c (prop_type): add PROP_GLYPH.
* redisplay.c (struct prop_block): add glyph type
* redisplay.c (add_glyph_rune): when adding part of a glyph add it
to the propagation data.
* redisplay.c (create_text_block): if there is a glyph in the
propagation data use it to salt extent_fragment_update.
* redisplay.c (create_string_text_block): ditto.
2002-04-13 Nix <nix(a)esperi.demon.co.uk>
* redisplay.h (struct rune): Add ascent, descent, and yoffset fields.
* redisplay-output.c (compare_runes): Compare them.
* redisplay.c: Update copyright date.
* redisplay.c (pos_data): Add need_baseline_computation field.
* redisplay.c (add_glyph_rune): Update ascent, descent, and
need_baseline_computation; zero yoffset. Set max_pixmap_height
for all pixmaps, not just automatically positioned ones.
* redisplay.c (calculate_yoffset): New, compute yoffset values.
* redisplay.c (calculate_baseline): New, compute textual baseline.
* redisplay.c (add_glyph_rune): Call them.
* redisplay.c (create_text_block): Likewise.
* redisplay.c (create_overlay_glyph_block): Likewise.
* redisplay.c (add_margin_runes): Likewise.
* redisplay.c (create_string_text_block): Likewise. Fix tabdamage.
* redisplay.h: (redisplay_calculate_display_boxes): Change prototype.
* redisplay-output.c (redisplay_calculate_display_boxes): Use yoffset.
* redisplay-msw.c (mswindows_output_blank): Pass 0 as yoffset.
* redisplay-msw.c (mswindows_output_string): Likewise.
* redisplay-msw.c (mswindows_output_display_block): Pass yoffset.
* redisplay-gtk.c (gtk_output_display_block): Likewise.
* redisplay-x.c (x_output_display_block): Likewise.
2002-03-28 Ben Wing <ben(a)xemacs.org>
* redisplay.c: Fixed bug in redisplay
w.r.t. hscroll/truncation/continuation glyphs causing jumping up
and down of the lines, since they're bigger than the line
size. (It was seen most obviously when there's a horizontal scroll
bar, e.g. do C-h a glyph or something like that.) The problem was
that the glyph-contrib-p setting on glyphs was ignored even if it
was set properly, which it wasn't until now.
2002-04-24 Andy Piper <andy(a)xemacs.org>
* lisp.h (Dynarr_end): Fix definition.
2002-02-06 Adrian Aichner <adrian(a)xemacs.org>
* redisplay.c (mark_redisplay): Remove call to
update_frame_window_mirror.
2001-11-15 Andy Piper <andy(a)xemacs.org>
* win32.c (Fmswindows_shell_execute): fix handling of URL's under
cygwin (again).
2002-02-13 Andy Piper <andy(a)xemacs.org>
* event-msw.c (mswindows_wnd_proc): only mark the frame visible if
we did in fact enqueue the XM_MAPFRAME event.
2002-01-15 Adrian Aichner <adrian(a)xemacs.org>
* event-msw.c (mswindows_wnd_proc): Add handling of WM_SHOWWINDOW
to fix problem switching between virtual desktops under virtuawin
virtual window manager.
2001-07-30 Adrian Aichner <adrian(a)xemacs.org>
* event-msw.c: Typo fix.
* event-msw.c (mswindows_wnd_proc): Set FRAME_VISIBLE_P after
magic XM_MAPFRAME event has been sent.
2001-11-23 Andy Piper <andy(a)xemacs.org>
* event-msw.c (mswindows_wnd_proc): Don't pump mousewheel events.
2001-11-21 Andy Piper <andy(a)xemacs.org>
* scrollbar-msw.c (mswindows_handle_mousewheel_event): cope with
mouse events outside the frame.
2002-03-20 Andy Piper <andy(a)xemacs.org>
* menubar-msw.c (mswindows_popup_menu): warning removal.
* dialog-msw.c (dialog_popped_down): ditto.
2001-12-11 Andy Piper <andy(a)xemacs.org>
* dialog-msw.c (dialog_popped_down): new function. unset popup_up_p.
* dialog-msw.c (mswindows_make_dialog_box_internal): set
popup_up_p.
* menubar-msw.c (unsafe_handle_wm_initmenupopup_1): ditto.
* menubar-msw.c (mswindows_handle_wm_command): ditto.
* menubar-msw.c (mswindows_popup_menu): ditto.
2001-10-29 Andy Piper <andy(a)xemacs.org>
* dialog-msw.c (handle_directory_dialog_box): quit if the user
cancels.
2002-01-03 Andy Piper <andy(a)xemacs.org>
* realpath.c (ABS_LENGTH): dtrt for cygwin systems using drive
letters.
(xrealpath): ditto.
2002-03-29 Jonathan Harris <jonathan(a)xemacs.org>
* device-msw.c (plist_get_margin): Add arg specifying mm or inches
* device-msw.c (plist_set_margin): Fix multiplicand used for mm
* device-msw.c (mswindows_handle_page_setup_dialog_box):
Detect and handle case where machine is set up for metric units
2002-04-02 Andy Piper <andy(a)xemacs.org>
* dired-msw.c (mswindows_get_files):
* nt.c (mswindows_stat): SetErrorMode() so that file errors are
completely handled by XEmacs. Suggested by Thomas Vogler
<mail(a)thomas-vogler.de>.
2001-11-14 Andy Piper <andy(a)xemacs.org>
* nt.c (REG_ROOT): change registry key to XEmacs.
2002-04-27 Andy Piper <andy(a)xemacs.org>
* glyphs-msw.c (mswindows_widget_instantiate): remove dead-code.
2002-04-26 Andy Piper <andy(a)xemacs.org>
* glyphs-msw.c (mswindows_map_subwindow): observe :initial-focus
behavior.
2002-03-14 Mike Alexander <mta(a)arbortext.com>
* event-msw.c (mswindows_unwait_process): New, remove process from
wait list
* process-nt.c (nt_finalize_process_data): Call
mswindows_unwait_process
* console-msw.h: Declare mswindows_unwait_process
2002-05-14 Stephen J. Turnbull <stephen(a)xemacs.org>
* editfns.c (Fdecode_time):
(Fformat_time_string):
Check for invalid time. Thanks to Nick Pakoulin <npak(a)ispras.ru>.
(make_time): Warning elimination (change arg name).
2001-11-24 Andy Piper <andy(a)xemacs.org>
* window.c (Fsplit_window): Doc return type.
2002-04-01 Andy Piper <andy(a)xemacs.org>
* emacs.c (Fkill_emacs): Only output message box in interactive
mode.
2002-04-26 Andy Piper <andy(a)xemacs.org>
* config.h.in: pull in 21.5.x change to stop alloca warnings under
cygwin.
2002-03-18 Gregory Steuck <greg-xemacs-patch(a)nest.cx>
* unexelf.c: Use ELFSIZE or define from _LP64 to indicate a 64
bit platform.
**** man/ChangeLog
2002-05-11 Adrian Aichner <adrian(a)xemacs.org>
* xemacs-faq.texi (Top): Add Q2.0.15 under "Installation and
Trouble Shooting".
* xemacs-faq.texi (Installation): Add Q2.0.15 to Installation
menu.
* xemacs-faq.texi (Q2.0.14): Fix unnumberedsubsec argument.
* xemacs-faq.texi (Q2.0.15): New.
**** lib-src/ChangeLog
2002-05-10 Stephen J. Turnbull <stephen(a)xemacs.org>
* Makefile.in.in: Add winclient targets (Andy synch).
--
Institute of Policy and Planning Sciences
http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573
JAPAN
My nostalgia for Icon makes me forget about any of the bad things. I don't
have much nostalgia for Perl, so its faults I remember. Scott Gilbert c.l.py
__________________________________________________
Do You Yahoo!?
LAUNCH - Your Yahoo! Music Experience
http://launch.yahoo.com
5:06 a.m.
>>>> "nix" == nix
<nix(a)esperi.demon.co.uk> writes:
nix> Hm. That's still not very clear, but I hope this rephrasing
nix> makes it a bit clearer than it was. I tried to explain it in
nix> the original patch postings, too: if you read all of these
nix> together then some sense might emerge :)
I plan to do so.
nix> Thanks for considering this at this late stage.
The other thing you might do is see if you can get the w3m people and
Bill P to sign off on "yes, this really does make things better" with
examples, or better yet, screen shots. Sounds tedious even to me, and
it won't guarantee a decision either way. But the more evidence that
there is visible (pun intended) benefit the easier it is to say yes.
> Yeah, well, there's already a bug report in, against 21.4.7
> Windows [1] which is the same code I put in. :-(
nix> Damn :( still, until you've seen an NFS server with an
nix> effective file transmission rate of 10K/s you haven't seen
nix> slow. Anything that reduces file access from a box like that
nix> is good. :)
Doesn't mean it won't go in. It is likely that I'll rewrite it as an
option, and possibly default it to off.
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
My nostalgia for Icon makes me forget about any of the bad things. I don't
have much nostalgia for Perl, so its faults I remember. Scott Gilbert c.l.py
6:17 a.m.
Sorry about the noise on -design, follow-ups set to xemacs-beta.
>>>> "Rick" == Rick Rankin
<rick_rankin(a)yahoo.com> writes:
Rick> Well, for Cygwin there seems to be a file called
Rick> lib-src/winclient.c missing from CVS.
Yup. Thanks for the report!
There's no ChangeLog for it, so I missed adding it to my branch.
Rick> cvs co -r r21-4-9rc1 -d xemacs-21.4.9 xemacs
You're right, I was wrong. Try it again, now.
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
My nostalgia for Icon makes me forget about any of the bad things. I don't
have much nostalgia for Perl, so its faults I remember. Scott Gilbert c.l.py
8:25 a.m.
On 20 May 2002, Stephen J. Turnbull stipulated:
The other thing you might do is see if you can get the w3m people
and
Bill P to sign off on "yes, this really does make things better" with
examples, or better yet, screen shots. Sounds tedious even to me, and
it won't guarantee a decision either way. But the more evidence that
there is visible (pun intended) benefit the easier it is to say yes.
I'll try to cook something up.
I can easily generate screen shots showing visible improvement in random
bits of Lisp or, well, just about any LaTeX document in preview-latex:
does that count?
(I doubt it: you already know about these cases ;} )
--
`There are not words enough to describe how fucked up imake is.'
--- Peter da Silva
11:26 a.m.
>>>> "nix" == nix
<nix(a)esperi.demon.co.uk> writes:
nix> I can easily generate screen shots showing visible
nix> improvement in random bits of Lisp or, well, just about any
nix> LaTeX document in preview-latex: does that count?
Sure it counts; I haven't had time to get preview-latex installed and
working yet.
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
My nostalgia for Icon makes me forget about any of the bad things. I don't
have much nostalgia for Perl, so its faults I remember. Scott Gilbert c.l.py
2:55 p.m.
>>>> "Stephen" == Stephen J Turnbull
<stephen(a)xemacs.org> writes:
Stephen> Please test the release candidate.
Stephen> Use "cvs update -r r21-4-9rc1" to get the RC sources. I hope to
get
Stephen> tarballs up tomorrow or Monday my time. For now CVS is up-to-date and
Stephen> so I'm announcing for those testers who use CVS.
tried updating to this and i got this error
[rendhalver＠ulthwe xemacs-21.4]$ cvs -z3 update -r r21-4-9-rc1
cvswrapper [server aborted]: no such tag r21-4-9-rc1
have i missed something ??
Stephen> Use "cvs update -r candidate-21-4" to get the latest and
greatest.
Stephen> Don't even expect checkouts on this tag to build; I commit what I
have
Stephen> when I feel like it. Mostly it should work, but if it doesn't, take
Stephen> two aspirin and cvs update in the morning. ;-)
updating to candidate-21-4 works fine tho
Stephen> "Release candidate" is something of a misnomer, there are a
number of
--
XEmacs Advocate | Do not try the patience of Wizards,
FreeBSD Devote | for they are subtle and quick to anger.
Perl Hacker | - Elric (Technomage) , Babylon 5.
Apache God | <mailto:rendhalver＠xemacs.org>
3 p.m.
>>>> "PB" == rendhalver
<rendhalver(a)xemacs.org> writes:
>>>> "Stephen" == Stephen J Turnbull
<stephen(a)xemacs.org> writes:
Stephen> Please test the release
candidate.
Stephen> Use "cvs update -r r21-4-9rc1" to get the RC sources. I hope to
get
Stephen> tarballs up tomorrow or Monday my time. For now CVS is up-to-date and
Stephen> so I'm announcing for those testers who use CVS.
PB> tried updating to this and i got this error
PB> [rendhalver＠ulthwe xemacs-21.4]$ cvs -z3 update -r r21-4-9-rc1
PB> cvswrapper [server aborted]: no such tag r21-4-9-rc1
PB> have i missed something ??
oops sorry
where's that rubber mallet ?
lack of sleep (and typing ability) again
just ignore me
--
XEmacs Advocate | Do not try the patience of Wizards,
FreeBSD Devote | for they are subtle and quick to anger.
Perl Hacker | - Elric (Technomage) , Babylon 5.
Apache God | <mailto:rendhalver＠xemacs.org>
4:57 p.m.
I've just tried to build the XEmacs 21.4.9 release candidate (r21-4-9rc1) on
the latest release of Cygwin (1.3.10-1). It fails building device-msw.c with
the following error:
gcc -c -g -O3 -Wall -Wno-switch -Winline -Wmissing-prototypes -Wsign-compare
-Wshadow -Wpointer-arith -Demacs -I. -DHAVE_CONFIG_H -Wno-sign-compare -fn
o-caller-saves -I/usr/X11R6/include device-msw.c
device-msw.c: In function `mswindows_handle_page_setup_dialog_box':
device-msw.c:872: `LOCALE_RETURN_NUMBER' undeclared (first use in this
function)
device-msw.c:872: (Each undeclared identifier is reported only once
device-msw.c:872: for each function it appears in.)
make[1]: *** [device-msw.o] Error 1
I've attached the configure log to the message.
I also checked the 'LOCALE_RETURN_NUMBER' flag on MSDN, where it says that
it is only in Windows 98/Me, Windows NT 4.0 and later, so it may not be the
right flag to be using in a generic win32 version anyway?
I can't find any discussion of this flag on the Cygwin mailing lists. The
only mention on the XEmacs lists seems to be the patch that added its use
(http://list-archive.xemacs.org/xemacs-beta/200203/msg00676.html and
http://list-archive.xemacs.org/xemacs-beta/200203/msg00679.html and
http://list-archive.xemacs.org/xemacs-beta/200203/msg00697.html)
Hope this helps.
// Conrad
5:27 a.m.
>>>> "Conrad" == Conrad Scott
<Conrad.Scott(a)dsl.pipex.com> writes:
Conrad> I also checked the 'LOCALE_RETURN_NUMBER' flag on MSDN,
Conrad> where it says that it is only in Windows 98/Me, Windows NT
Conrad> 4.0 and later, so it may not be the right flag to be using
Conrad> in a generic win32 version anyway?
OK, we'll flag this question for Andy et al.
The reason it's not in the release candidate is that it's not
ChangeLog'd and there's no comment to explain what it's for. I
assumed it was cruft. This patch (all that's left of the diff against
the windows branch that I merged) should get you a compile, at least,
and I'll add it for the next RC.
? candidate-21-4->windows-21-4.diff
? windows-21-4->current.diff
Index: src/syswindows.h
===================================================================
RCS file: /pack/xemacscvs/XEmacs/xemacs/src/syswindows.h,v
retrieving revision 1.2.2.1.4.2
retrieving revision 1.2.2.2
diff -u -r1.2.2.1.4.2 -r1.2.2.2
--- src/syswindows.h 17 Dec 2001 05:49:27 -0000 1.2.2.2
+++ src/syswindows.h 30 Mar 2002 21:46:09 -0000 1.2.2.1.4.2
＠＠ -180,6 +180,9 ＠＠
#ifndef MAC_CHARSET
#define MAC_CHARSET 77
#endif
+#ifndef LOCALE_RETURN_NUMBER
+#define LOCALE_RETURN_NUMBER 0x20000000
+#endif
/***************************************************************/
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
My nostalgia for Icon makes me forget about any of the bad things. I don't
have much nostalgia for Perl, so its faults I remember. Scott Gilbert c.l.py
6:16 a.m.
OK, next cygwin problem. Make fails with:
device-msw.c: In function `mswindows_handle_page_setup_dialog_box':
device-msw.c:872: `LOCALE_RETURN_NUMBER' undeclared (first use in this
function)
device-msw.c:872: (Each undeclared identifier is reported only once
device-msw.c:872: for each function it appears in.)
make[1]: *** [device-msw.o] Error 1
make[1]: Leaving directory `/pub/src/xemacs-21.4.9/src'
make: *** [src] Error 2
I didn't have much time to look at it, but it's related to NLS support. For
LOCALE_RETURN_NUMBER to be #defined in winnls.h (included from windows.h),
WINVER has to be #defined to a value >= 0x0400, meaning you have to be running
NT 4.0 or greater. I'm trying to compile on Windows 2000, so that's not an
issue.
I hope to have more time to look at it tomorrow. If anyone already knows the
answer though...
Rick
--- "Stephen J. Turnbull" <stephen(a)xemacs.org> wrote:
Sorry about the noise on -design, follow-ups set to xemacs-beta.
>>>>> "Rick" == Rick Rankin <rick_rankin(a)yahoo.com>
writes:
Rick> Well, for Cygwin there seems to be a file called
Rick> lib-src/winclient.c missing from CVS.
Yup. Thanks for the report!
There's no ChangeLog for it, so I missed adding it to my branch.
Rick> cvs co -r r21-4-9rc1 -d xemacs-21.4.9 xemacs
You're right, I was wrong. Try it again, now.
--
Institute of Policy and Planning Sciences
http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573
JAPAN
My nostalgia for Icon makes me forget about any of the bad things. I don't
have much nostalgia for Perl, so its faults I remember. Scott Gilbert c.l.py
__________________________________________________
Do You Yahoo!?
LAUNCH - Your Yahoo! Music Experience
http://launch.yahoo.com
8:07 a.m.
New subject: [AC 21.4] #define LOCALE_RETURN_NUMBER. Why, I don't know.
APPROVE COMMIT 21.4
Andy, Jonathan, what's going on here?
Followup to xemacs-beta (Reply-To set).
Updated tags for r21-4-9rc1.
2002-05-21 Stephen J. Turnbull <stephen(a)xemacs.org>
* syswindows.h: #define LOCALE_RETURN_NUMBER. Why, I don't know.
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
My nostalgia for Icon makes me forget about any of the bad things. I don't
have much nostalgia for Perl, so its faults I remember. Scott Gilbert c.l.py
1:17 p.m.
"Stephen J. Turnbull" <stephen(a)xemacs.org> wrote:
>>>>> "Conrad" == Conrad Scott
<Conrad.Scott(a)dsl.pipex.com> writes:
Conrad> I also checked the 'LOCALE_RETURN_NUMBER' flag on MSDN,
Conrad> where it says that it is only in Windows 98/Me, Windows NT
Conrad> 4.0 and later, so it may not be the right flag to be using
Conrad> in a generic win32 version anyway?
OK, we'll flag this question for Andy et al.
The reason it's not in the release candidate is that it's not
ChangeLog'd and there's no comment to explain what it's for. I
assumed it was cruft. This patch (all that's left of the diff against
the windows branch that I merged) should get you a compile, at least,
and I'll add it for the next RC.
Thanks, I picked up the change from CVS and that fixed the build problem.
I've now got a clean working build on cygwin (and it builds CVS HEAD of the
packages too).
I'll be submitting a build-report just as soon as I fix something else in my
setup that I broke last night . . .
// Conrad
4:55 p.m.
At 11:38 PM 5/19/02 +0900, Stephen J. Turnbull wrote:
nix> As a heavy user of auto-revert-buffer with huge files on
slow
nix> machines with data coming from *dog* slow NFS servers, I
nix> applaud.
Yeah, well, there's already a bug report in, against 21.4.7 Windows
[1] which is the same code I put in. :-(
Not true. The installer version doesn't contain the last bugfix I did for
Ben. The changes you pulled in do.
andy
5 p.m.
I think this was required by one of Jonathan's patches. Cygwin doesn't
provide 100% compatible headers.
Sorry about the lack of ChangeLog.
andy
At 01:27 PM 5/21/02 +0900, Stephen J. Turnbull wrote:
>>>>> "Conrad" == Conrad Scott
<Conrad.Scott(a)dsl.pipex.com> writes:
Conrad> I also checked the 'LOCALE_RETURN_NUMBER' flag on MSDN,
Conrad> where it says that it is only in Windows 98/Me, Windows NT
Conrad> 4.0 and later, so it may not be the right flag to be using
Conrad> in a generic win32 version anyway?
OK, we'll flag this question for Andy et al.
The reason it's not in the release candidate is that it's not
ChangeLog'd and there's no comment to explain what it's for. I
assumed it was cruft. This patch (all that's left of the diff against
the windows branch that I merged) should get you a compile, at least,
and I'll add it for the next RC.
? candidate-21-4->windows-21-4.diff
? windows-21-4->current.diff
Index: src/syswindows.h
===================================================================
RCS file: /pack/xemacscvs/XEmacs/xemacs/src/syswindows.h,v
retrieving revision 1.2.2.1.4.2
retrieving revision 1.2.2.2
diff -u -r1.2.2.1.4.2 -r1.2.2.2
--- src/syswindows.h 17 Dec 2001 05:49:27 -0000 1.2.2.2
+++ src/syswindows.h 30 Mar 2002 21:46:09 -0000 1.2.2.1.4.2
＠＠ -180,6 +180,9 ＠＠
#ifndef MAC_CHARSET
#define MAC_CHARSET 77
#endif
+#ifndef LOCALE_RETURN_NUMBER
+#define LOCALE_RETURN_NUMBER 0x20000000
+#endif
/***************************************************************/
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
My nostalgia for Icon makes me forget about any of the bad things. I don't
have much nostalgia for Perl, so its faults I remember. Scott Gilbert c.l.py
11:23 p.m.
New subject: [21.4] Re: #define LOCALE_RETURN_NUMBER. Why, I don't know.
Stephen,
Andy, Jonathan, what's going on here?
My screwup - I didn't notice that it wouldn't work on Win95. I don't
think anyone's testing that platform but I think they would have seen
the same old broken behaviour that the patch was intended to fix.
The patch below should work on all versions of MS Windows, but I only
have Win98se and WinXP available to test. Rick's patch to #define
LOCALE_RETURN_NUMBER to fix cygwin builds could now be safely reverted,
but is harmless. I'll apply a similar patch to 21.5 myself.
Jonathan.
ChangeLog for src:
2002-05-21 Jonathan Harris <jonathan(a)xemacs.org>
* device-msw.c (mswindows_handle_page_setup_dialog_box):
Fix detection of metric units to work on Win95
Index: src/device-msw.c
===================================================================
RCS file: /pack/xemacscvs/XEmacs/xemacs/src/device-msw.c,v
retrieving revision 1.24.2.3.2.1
diff -u -r1.24.2.3.2.1 device-msw.c
--- src/device-msw.c 2002/03/30 21:45:55 1.24.2.3.2.1
+++ src/device-msw.c 2002/05/21 22:05:47
＠＠ -866,11 +866,12 ＠＠
{
Lisp_Devmode *ldm = decode_devmode (device);
PAGESETUPDLG pd;
- DWORD data;
+ TCHAR measure[2];
+ int data;
- GetLocaleInfo (LOCALE_USER_DEFAULT,
-
LOCALE_IMEASURE|LOCALE_RETURN_NUMBER,
-
(LPTSTR) &data, sizeof(data));
+ GetLocaleInfo (LOCALE_USER_DEFAULT, LOCALE_IMEASURE,
+
measure, sizeof(measure));
+ data = (strcmp (measure, "0"));
memset (&pd, 0, sizeof (pd));
pd.lStructSize = sizeof (pd);
--
Jonathan Harris | jhar(a)tardis.ed.ac.uk
London, England | Jonathan.Harris(a)symbian.com
10:21 a.m.
[Second attempt: get the address right this time . . .]
"Stephen J. Turnbull" <stephen(a)xemacs.org> wrote:
>>>>> "Conrad" == Conrad Scott
<Conrad.Scott(a)dsl.pipex.com> writes:
Conrad> I also checked the 'LOCALE_RETURN_NUMBER' flag on MSDN,
Conrad> where it says that it is only in Windows 98/Me, Windows NT
Conrad> 4.0 and later, so it may not be the right flag to be using
Conrad> in a generic win32 version anyway?
OK, we'll flag this question for Andy et al.
The reason it's not in the release candidate is that it's not
ChangeLog'd and there's no comment to explain what it's for. I
assumed it was cruft. This patch (all that's left of the diff against
the windows branch that I merged) should get you a compile, at least,
and I'll add it for the next RC.
Thanks, I picked up the change from CVS and that fixed the build problem.
I've now got a clean working build on cygwin (and it builds CVS HEAD of the
packages too).
// Conrad
9:56 a.m.
>>>> "Stephen" == Stephen J Turnbull
<stephen(a)xemacs.org> writes:
Stephen> Please test the release candidate.
Stephen> Use "cvs update -r r21-4-9rc1" to get the RC sources. I hope to
get
Stephen> tarballs up tomorrow or Monday my time. For now CVS is up-to-date and
Stephen> so I'm announcing for those testers who use CVS.
Stephen> Use "cvs update -r candidate-21-4" to get the latest and
greatest.
Stephen> Don't even expect checkouts on this tag to build; I commit what I
have
Stephen> when I feel like it. Mostly it should work, but if it doesn't, take
Stephen> two aspirin and cvs update in the morning. ;-)
Stephen> "Release candidate" is something of a misnomer, there are a
number of
Stephen> patches in the queue that I probably will apply, and I really hope
Stephen> that the syntax.c problems (aka "font-lock slows") will get
fixed for
Stephen> 21.4.9. We'll see .... Anyway, expect at least one more candidate
Stephen> before 21.4.9.
Stephen> TESTING PROCEDURE
Stephen> 1. Get the sources.
Stephen> 2. Configure as usual, and if you're feeling especially chipper, try
Stephen> several variants that are new and different.
Stephen> 3. Use M-x build-report to report each build and any problems. If you
Stephen> have problems, please CC xemacs-beta. Successful builds can just
Stephen> go to xemacs-buildreports.
Stephen> 4. Test and send reports to xemacs-beta, positive or negative.
hey stephen
just thought i would let you know i have been using r21-4-9rc1 since
you sent this email and i havent had any problems yet :)
havent noticed anything weird either
Stephen> WHAT TO TEST
Stephen> 1. I've included the work done by Andy and Nix to support vertical
Stephen> positioning of glyphs and some other redisplay hacking needed to
Stephen> support preview-latex. This is most definitely a feature, and
Stephen> therefore I really should not include it in 21.4. Please try
Stephen> preview-latex (http://preview-latex.sourceforge.net/ IIRC). The more
Stephen> reports and the more detail I get, the more likely I am to include
Stephen> it. Maybe not 21.4.9, maybe 21.4.10.
Stephen> Please beat hard on this one. preview-latex is really cool and I
Stephen> would like to support it in 21.4.
just downloading preview-latex, will let you know how it goes
Stephen> 2. I've included Andy's revert-buffer optimization. The idea is
to
Stephen> compare the buffer text to the file contents, and not redo expensive
Stephen> buffer analysis (eg, font-lock and semantic bovination) if the text is
Stephen> the same.
Stephen> There are reports on 21.5 that this code trashes markers and extents
Stephen> when you revert. This is probably fixed already, but please be on the
Stephen> watch for such problems, and any anomolies when you revert. Again,
Stephen> this code may not make it in to 21.4.9.
havent noticed any strangeness here either
Stephen> 3. Builds on Linux PPC, OpenBSD, and any 64-bit platforms would be
Stephen> nice, I've added some configure tests and s&m pervection to
support
Stephen> those. MacOS/X Darwin is already in 21.4.8, but build reports on that
Stephen> platform would be nice, too.
i want a G4 Powerbook now :(
Stephen> The OpenBSD changes include detection of a OpenBSD weirdness where
Stephen> they rename libcurses as libtermcap or something like that. I think
Stephen> it's well-guarded by an "if [ opsys = openbsd ]", but be on
the
Stephen> lookout for configure failures and runtime weirdness for TTY stuff.
hmm maybe i can install OpenBSD on my server :) and test XEmacs
need that cable connection :(
--
XEmacs Advocate | Do not try the patience of Wizards,
FreeBSD Devote | for they are subtle and quick to anger.
Perl Hacker | - Elric (Technomage) , Babylon 5.
Apache God | <mailto:rendhalver＠xemacs.org>
5:20 a.m.
New subject: [21.4] Re: #define LOCALE_RETURN_NUMBER. Why, I don't know.
APPROVE COMMIT 21.4-windows
-----Original Message-----
From: Jonathan Harris [mailto:jhar＠tardis.ed.ac.uk]
Sent: Tuesday, May 21, 2002 3:23 PM
To: xemacs-beta(a)xemacs.org
Cc: xemacs-patches(a)xemacs.org; andy(a)xemacs.org
Subject: [21.4] Re: #define LOCALE_RETURN_NUMBER. Why, I don't know.
Stephen,
> Andy, Jonathan, what's going on here?
My screwup - I didn't notice that it wouldn't work on Win95. I don't
think anyone's testing that platform but I think they would have seen
the same old broken behaviour that the patch was intended to fix.
The patch below should work on all versions of MS Windows, but I only
have Win98se and WinXP available to test. Rick's patch to #define
LOCALE_RETURN_NUMBER to fix cygwin builds could now be safely reverted,
but is harmless. I'll apply a similar patch to 21.5 myself.
Jonathan.
ChangeLog for src:
2002-05-21 Jonathan Harris <jonathan(a)xemacs.org>
* device-msw.c (mswindows_handle_page_setup_dialog_box):
Fix detection of metric units to work on Win95
Index: src/device-msw.c
===================================================================
RCS file: /pack/xemacscvs/XEmacs/xemacs/src/device-msw.c,v
retrieving revision 1.24.2.3.2.1
diff -u -r1.24.2.3.2.1 device-msw.c
--- src/device-msw.c 2002/03/30 21:45:55 1.24.2.3.2.1
+++ src/device-msw.c 2002/05/21 22:05:47
＠＠ -866,11 +866,12 ＠＠
{
Lisp_Devmode *ldm = decode_devmode (device);
PAGESETUPDLG pd;
- DWORD data;
+ TCHAR measure[2];
+ int data;
- GetLocaleInfo (LOCALE_USER_DEFAULT,
-
LOCALE_IMEASURE|LOCALE_RETURN_NUMBER,
-
(LPTSTR) &data, sizeof(data));
+ GetLocaleInfo (LOCALE_USER_DEFAULT, LOCALE_IMEASURE,
+
measure, sizeof(measure));
+ data = (strcmp (measure, "0"));
memset (&pd, 0, sizeof (pd));
pd.lStructSize = sizeof (pd);
--
Jonathan Harris | jhar(a)tardis.ed.ac.uk
London, England | Jonathan.Harris(a)symbian.com
7994
days inactive
8007
days old
21 comments
8 participants
participants (8)
-
Andy Piper -
Andy Piper -
Conrad Scott -
Jonathan Harris -
Nix -
Rendhalver [Peter Brown] -
Rick Rankin -
Stephen J. Turnbull