4:45 a.m.
Hi Folks!
Earlier today SXEmacs 22.1.10 was released and tarballs were uploaded
to http://downloads.sxemacs.org/
Here are the links:
Gzip
----
http://downloads.sxemacs.org/releases/sxemacs-22.1.10.tar.gz
http://downloads.sxemacs.org/releases/sxemacs-22.1.10.tar.gz.asc
http://downloads.sxemacs.org/releases/sxemacs-22.1.10.tar.gz.md5
Bzip2
-----
http://downloads.sxemacs.org/releases/sxemacs-22.1.10.tar.bz2
http://downloads.sxemacs.org/releases/sxemacs-22.1.10.tar.bz2.asc
http://downloads.sxemacs.org/releases/sxemacs-22.1.10.tar.bz2.md5
LZMA
----
http://downloads.sxemacs.org/releases/sxemacs-22.1.10.tar.lzma
http://downloads.sxemacs.org/releases/sxemacs-22.1.10.tar.lzma.asc
http://downloads.sxemacs.org/releases/sxemacs-22.1.10.tar.lzma.md5
Patch from previous version
---------------------------
http://downloads.sxemacs.org/releases/sxemacs-22.1.9-22.1.10.diff.gz
http://downloads.sxemacs.org/releases/sxemacs-22.1.9-22.1.10.diff.gz.asc
http://downloads.sxemacs.org/releases/sxemacs-22.1.9-22.1.10.diff.gz.md5
http://downloads.sxemacs.org/releases/sxemacs-22.1.9-22.1.10.diff.bz2
http://downloads.sxemacs.org/releases/sxemacs-22.1.9-22.1.10.diff.bz2.asc
http://downloads.sxemacs.org/releases/sxemacs-22.1.9-22.1.10.diff.bz2.md5
http://downloads.sxemacs.org/releases/sxemacs-22.1.9-22.1.10.diff.lzma
http://downloads.sxemacs.org/releases/sxemacs-22.1.9-22.1.10.diff.lzma.asc
http://downloads.sxemacs.org/releases/sxemacs-22.1.9-22.1.10.diff.lzma.md5
The ChangeLog is available at
http://www.sxemacs.org/changes/ChangeLog-22.1.10
The major changes in this release are:
-------------------------------------
* Bugs closed in this release:
41, 65, 74, 75, 79, 80, 81, 82, 83, 84, 85, 87, 89, 90, 91, 92, 93, 94,
95, 96, 97, 99, 100, 101, 102, 104
* Cool New Sexy Stuff -- Zajcev, Ferreira, Freundt, Youngs.
Of course every release must have some cool new stuff. Here's what's
new in this release...
o User defined readers
o It's possible to display video in a glyph object in a buffer.
o SXEmacs is now a fairly decent image viewing/manipulation
app. There's an undo/redo stack, zoom in/out, rotate, chop/crop,
red-eye reduction. There is a whole swag of transformations
supported: charcoal, contrast, despeckle, edge, enhance, equalise,
gausian blur, negate, normalise, oil painting, radial blur, 3D
button effect, sharpen, solarise, and swirl.
Who needs The Gimp when you have SXEmacs!
o Added some MacOS icons.
o Support for sqlite3 via FFI
o New way for SXEmacs to handle different file types. It allows you to
have #'find-file take different actions based on a predicate. See
the new variable `find-file-magic-files-alist'.
o Enable PNG transparent backgrounds
Unfortunately this doesn't work if you use a background pixmap.
o FFI callbacks implemented.
o Dbus support via FFI (not completed) This should be considered very
much incomplete and experimental. Only rudimentary support is there
so far.
* Build chain fixes/improvements -- Burkhardt, Ferreira, Freundt, Youngs.
Quite a few build chain related changes were made in this release,
including...
o Work around a bug in GCC 4.5.0
o Powerpc/BSD cpu detection
o A Freedesktop.org compliant sxemacs.desktop file is now installed.
o Updated FFmpeg detection.
o OpenSSL 1.0.0 detection. OpenSSL is no on by default.
o Fix a link error when building with no MPQ
o Fix a minor bison related issue (YYSTYPE)
o Fix building with latest SoX
o Fix --without-x builds
o Fix many C compiler warnings and elisp byte-compiler warnings.
o Fix install for non-FFI and non-Mule builds
o Fix etags build for systems without getopt_long
o Fix (hopefully) parallel builds
o Fix BDWGC-enabled builds
* Minor updates to BDWGC code -- Freundt.
* Catch up with latest ffmpeg API changes -- Freundt.
* Fix for latest ediff -- Ferreira.
Changes emacs-internal to escape-quoted.
* Fix for subwindow resizing -- Zajcev.
* Various additions/updates to the contrib directory -- Ferreira
* FFI improvements -- Zajcev
* ENT fixes -- Freundt.
* TTY fixes -- Ferreira.
* Cleaner dump file (less pollution from the build/src dir) -- Freundt,
Youngs.
* Improve #'file-basename -- Ferreira
It can now take a 2nd arg for extension stripping.
* Misc doc fixes/updates -- Burkhardt, Ferreira, Freundt, Zajcev,
Youngs.
--
|---<Steve Youngs>---------------<GnuPG KeyID: A94B3003>---|
| SXEmacs - The only _______ you'll ever need. |
| Fill in the blank, yes, it's THAT good! |
|------------------------------------<steve(a)sxemacs.org>---|
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta
3:58 p.m.
Steve Youngs <steve(a)sxemacs.org> writes:
Hi Folks!
Hi,
o Dbus support via FFI (not completed) This should be considered
very
much incomplete and experimental. Only rudimentary support is there
so far.
GNU Emacs 23 offers D-Bus integration as well. A short look at your
sources show a different API.
Have you considered to offer the GNU Emacs D-Bus interface? Or to
convince the GNU maintainers to keep it in sync with SXEmacs?
Best regards, Michael.
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta
11:11 p.m.
On May 24, 2009, at 10:58 AM, Michael Albinus wrote:
Steve Youngs <steve(a)sxemacs.org> writes:
> Hi Folks!
Hi,
> o Dbus support via FFI (not completed) This should be considered
> very
> much incomplete and experimental. Only rudimentary support is
> there
> so far.
GNU Emacs 23 offers D-Bus integration as well. A short look at your
sources show a different API.
Indeed. GNU Emacs currently does not (at the best of my knowledge)
support
FFI.
The API with FFI, while different has the added advantage that once
SXEmacs is
compiled with FFI, the additional support for FFI is purely in lisp,
not with
a sizable C support like GNU Emacs one.
If you look closely all the type conversion stuff that GNU Emacs
does, is for the most
part what our FFI support giver us.
This allows us not to introduce an additional link (and deploy) time
dependency
with D-Bus, which at this time is not yet widely deployed outside of
Linux.
Have you considered to offer the GNU Emacs D-Bus interface?
To be honest, only today after your email I took a look at it. So the
answer, at least in SXEmacs is no.
Or to
convince the GNU maintainers to keep it in sync with SXEmacs?
Well since SXEmacs is GPL3 as GNU Emacs, and with past history in
mind, I believe
convincing GNU maintainers to keep it in sync is, at best, a
difficult endeavor.
However, in case it is deemed the interface to offer, as is done with
other Lisp code,
the code is periodically synced.
Best regards,
Nelson
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta
5:12 a.m.
Nelson Ferreira writes:
Indeed. GNU Emacs currently does not (at the best of my knowledge)
support FFI.
And won't. RMS is adamant on that point. It came up in an ongoing
thread, and he once again insisted that (with current technology)
dynamic loading of binaries into Emacs is not acceptable.
This allows us not to introduce an additional link (and deploy)
time dependency with D-Bus, which at this time is not yet widely
deployed outside of Linux.
I think anything that uses GTK now uses (or at least is capable of
using) D-Bus. I know that it's a dependency for a number of packages
I use via MacPorts for example, and I would suppose on other *BSD
distributions.
Well since SXEmacs is GPL3 as GNU Emacs, and with past history in
mind, I believe convincing GNU maintainers to keep it in sync is,
at best, a difficult endeavor.
I agree. GNU Emacs does not sync to other Emacsen. RMS explains this
as necessary so that Emacs can have Emacs-optimal interfaces. The
only way to get this done is to commit to doing it oneself, under
the supervision of RMS, Stefan, and Yidong. I won't ever try to do
that again, myself, but YMMV.
However, in case it is deemed the interface to offer, as is done
with other Lisp code, the code is periodically synced.
I don't have time to look at the Emacs dbus API right now. I suspect
I won't like it very much when I do see it (that "vast technical gulf"
:-), but compatibility is better than not. On the other hand, Steve Y
contributed a (now old) port of SXEmacs's FFI to XEmacs which I plan
to update and add to 21.5 shortly. So if you like, or at least find
acceptable, the Emacs dbus API, I would appreciate it (on behalf of
XEmacs) if you would sync to it.
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta
5:34 a.m.
"Stephen J. Turnbull" <stephen(a)xemacs.org> writes:
Nelson Ferreira writes:
> Indeed. GNU Emacs currently does not (at the best of my knowledge)
> support FFI.
And won't. RMS is adamant on that point. It came up in an ongoing
thread, and he once again insisted that (with current technology)
dynamic loading of binaries into Emacs is not acceptable.
I do not speak about using FFI in GNU Emacs.
D-Bus in GNU Emacs consists of dbusbind.c, dbus.el, and dbus.texi[1]. I
was thinking of writing a dbusbind.el for [S]XEmacs, reimplementing its
Lisp exported objects via ffi-dbus.el. dbus.el and dbus.texi could be
left untouched in the best case (in practice, it might be necessary to
change them as well, for example due to concurrent definitions in
dbus.el and ffi-dbus.el).
> This allows us not to introduce an additional link (and
deploy)
> time dependency with D-Bus, which at this time is not yet widely
> deployed outside of Linux.
I think anything that uses GTK now uses (or at least is capable of
using) D-Bus. I know that it's a dependency for a number of packages
I use via MacPorts for example, and I would suppose on other *BSD
distributions.
Right. Even for MS Windows there exists a D-Bus implementation (I've
never tried it).
> Well since SXEmacs is GPL3 as GNU Emacs, and with past history
in
> mind, I believe convincing GNU maintainers to keep it in sync is,
> at best, a difficult endeavor.
I agree. GNU Emacs does not sync to other Emacsen. RMS explains this
as necessary so that Emacs can have Emacs-optimal interfaces. The
only way to get this done is to commit to doing it oneself, under
the supervision of RMS, Stefan, and Yidong. I won't ever try to do
that again, myself, but YMMV.
See [1].
> However, in case it is deemed the interface to offer, as is
done
> with other Lisp code, the code is periodically synced.
I don't have time to look at the Emacs dbus API right now. I suspect
I won't like it very much when I do see it (that "vast technical gulf"
:-), but compatibility is better than not. On the other hand, Steve Y
contributed a (now old) port of SXEmacs's FFI to XEmacs which I plan
to update and add to 21.5 shortly. So if you like, or at least find
acceptable, the Emacs dbus API, I would appreciate it (on behalf of
XEmacs) if you would sync to it.
The interface was never discussed on emacs-devel. I would be happy to
get some feedback about. There are possibilities to change/enhance the
interface; of course backwards compatibility must be taken into account,
because there exists already packages on top of it.
Best regards, Michael.
[1]: I have written dbusbind.c, dbus.el, dbus.texi.
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta
5:50 a.m.
On May 25, 2009, at 12:34 AM, Michael Albinus wrote:
"Stephen J. Turnbull" <stephen(a)xemacs.org> writes:
> Nelson Ferreira writes:
>
>> Indeed. GNU Emacs currently does not (at the best of my knowledge)
>> support FFI.
>
> And won't. RMS is adamant on that point. It came up in an ongoing
> thread, and he once again insisted that (with current technology)
> dynamic loading of binaries into Emacs is not acceptable.
I do not speak about using FFI in GNU Emacs.
D-Bus in GNU Emacs consists of dbusbind.c, dbus.el, and dbus.texi
[1]. I
was thinking of writing a dbusbind.el for [S]XEmacs, reimplementing
its
Lisp exported objects via ffi-dbus.el. dbus.el and dbus.texi could be
left untouched in the best case (in practice, it might be necessary to
change them as well, for example due to concurrent definitions in
dbus.el and ffi-dbus.el).
Michael, I think that idea of implementing the dbusbind.c as a
dbusbind.el on
top of ffi-dbus.el (or not, since like SY said, it is very early), is
valid.
I for one would like to see it.
--Nelson
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta
2:07 a.m.
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta
3:33 a.m.
* Nelson Ferreira <nelson.ferreira(a)ieee.org> writes:
Michael, I think that idea of implementing the dbusbind.c as a
dbusbind.el on top of ffi-dbus.el (or not, since like SY said, it
is very early), is valid.
I for one would like to see it.
Yes, I support this idea. Michael, if you'd like to help us do this,
then probably the first thing you should do is have a look through our
SPPM, especially the VC and patches sections...
http://www.sxemacs.org/docs/sppm/Version-Control.html#Version-Control
http://www.sxemacs.org/docs/sppm/Patches.html#Patches
--
|---<Steve Youngs>---------------<GnuPG KeyID: A94B3003>---|
| SXEmacs - The only _______ you'll ever need. |
| Fill in the blank, yes, it's THAT good! |
|------------------------------------<steve(a)sxemacs.org>---|
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta
4:08 p.m.
Steve Youngs <steve(a)sxemacs.org> writes:
Yes, I support this idea. Michael, if you'd like to help us do
this,
then probably the first thing you should do is have a look through our
SPPM, especially the VC and patches sections...
http://www.sxemacs.org/docs/sppm/Version-Control.html#Version-Control
http://www.sxemacs.org/docs/sppm/Patches.html#Patches
OK, I'll do so. It might take some days to settle up; I haven't used
arch so far.
Best regards, Michael.
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta
10:02 a.m.
Michael Albinus <michael.albinus(a)gmx.de> writes:
"Stephen J. Turnbull" <stephen(a)xemacs.org> writes:
> Nelson Ferreira writes:
>
> > Indeed. GNU Emacs currently does not (at the best of my knowledge)
> > support FFI.
>
> And won't. RMS is adamant on that point. It came up in an ongoing
> thread, and he once again insisted that (with current technology)
> dynamic loading of binaries into Emacs is not acceptable.
I do not speak about using FFI in GNU Emacs.
D-Bus in GNU Emacs consists of dbusbind.c, dbus.el, and dbus.texi[1]. I
was thinking of writing a dbusbind.el for [S]XEmacs, reimplementing its
Lisp exported objects via ffi-dbus.el. dbus.el and dbus.texi could be
left untouched in the best case (in practice, it might be necessary to
change them as well, for example due to concurrent definitions in
dbus.el and ffi-dbus.el).
API Sync for D-Bus across Emacsen would be very very great!
I believe that D-Bus implementation for (S)XEmacs (with same API as
for GNU Emacs) could be done entirely in elisp, using FFI and glib's
dbus implementation. I might help you on this, but unfortunately my
spare time is greatly limited :(
[..]
--
lg
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta
5448
days inactive
5461
days old
9 comments
5 participants
participants (5)
-
Michael Albinus -
Nelson Ferreira -
Stephen J. Turnbull -
Steve Youngs -
Zajcev Evgeny