12:52 p.m.
There is a small error in font.el of the Xft branch leading to a
malfunctioning font-default-font-for-device.
Have a look at
http://cvs.xemacs.org/viewcvs.cgi/XEmacs/xemacs/lisp/font.el.diff?r1=1.16...
Obviously the else path has not been removed when the if statement for
font-running-xemacs was removed.
The "patch" below removes the remains.
Stephen: The patch from Daniel Pittman
http://list-archive.xemacs.org/xemacs-patches/200507/msg00058.html
gives a nice speed gain. He hasn't merged the patch himself yet. Would
you like to do that, while you're at the code? :)
Thanks for your work!
--- lisp/font.el 2005-09-07 13:41:16.000000000 +0200
+++ lisp/font.el 2005-09-07 13:41:19.000000000 +0200
＠＠ -637,11 +637,7 ＠＠
(or device (setq device (selected-device)))
(font-truename
(make-font-specifier
- (face-font-name 'default device)))
- (let ((font (cdr-safe (assq 'font (frame-parameters device)))))
- (if (and (fboundp 'fontsetp) (fontsetp font))
- (aref (get-font-info (aref (cdr (get-fontset-info font)) 0)) 2)
- font)))
+ (face-font-name 'default device))))
;;;###autoload
(defun font-default-object-for-device (&optional device)
7:52 a.m.
New subject: [AC sjt-xft] Re: Xft branch - code cut error in font.el
APPROVE COMMIT sjt-xft
Courtesy notice of private branch commit.
>>>> "Clemens" == Clemens Fruhwirth
<clemens(a)endorphin.org> writes:
Clemens> Obviously the else path has not been removed when the if
Clemens> statement for font-running-xemacs was removed.
Clemens> The "patch" below removes the remains.
Thanks, I'll apply that, with this ChangeLog:
2005-09-07 Clemens Fruhwirth <clemens(a)endorphin.org>
* font.el (font-default-font-for-device): Remove vestigial code
from incomplete removal of conditional.
Clemens> Stephen: The patch from Daniel Pittman
Clemens> http://list-archive.xemacs.org/xemacs-patches/200507/msg00058.html
Clemens> gives a nice speed gain. He hasn't merged the patch
Clemens> himself yet. Would you like to do that, while you're at
Clemens> the code? :)
Actually, that's been in for weeks, along with some refactoring of
that series of functions which may have confused the issue. Please
check and make sure that you do have the right file after a CVS
update. You should find that the only call to an Fc*Init* () function
is in the XEmacs init code at the end of the file, and the only call
to FcFontList now has a NULL FcConfig argument. If not, let me know
and we'll debug our CVS (cvs diff against the file I have claims that
that's what is in CVS).
--
School of Systems and Information Engineering http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Ask not how you can "do" free software business;
ask what your business can "do for" free software.
7:36 p.m.
New subject: [AC sjt-xft] Re: Xft branch - code cut error in font.el
Stephen J. Turnbull <stephen(a)xemacs.org> wrote:
Actually, that's been in for weeks, along with some refactoring
of
that series of functions which may have confused the issue. Please
check and make sure that you do have the right file after a CVS
update.
Ah, my local CVS tree was checked out with sjt-xft-21-5-21 and not
sjt-xft. Now everything appears a bit faster. Thanks!
Here is another small refinement. The tree won't built without
"--with-xft", because error checking is done for a variable that doesn't
existing, because it's only created when USE_XFT is define.
Best Regards, Clemens
Index: glyphs-x.c
===================================================================
RCS file: /pack/xemacscvs/XEmacs/xemacs/src/glyphs-x.c,v
retrieving revision 1.80.2.6
diff -u -r1.80.2.6 glyphs-x.c
--- glyphs-x.c 2005/03/15 05:24:21 1.80.2.6
+++ glyphs-x.c 2005/09/10 18:27:47
＠＠ -2352,11 +2352,13 ＠＠
}
+#ifdef USE_XFT
/* #### sanity check, should wrap in appropriate ERROR_CHECK macro */
if (!rf && !fs)
warn_when_safe_lispobj
(intern ("xft"), Qdebug,
Fcons (build_string ("missing font in update_widget_face"),
Fface_name (face)));
+#endif
}
wv->change = VISIBLE_CHANGE;
/* #### Megahack - but its just getting too complicated to do this
Index: ChangeLog.Xft
===================================================================
RCS file: /pack/xemacscvs/XEmacs/xemacs/src/Attic/ChangeLog.Xft,v
retrieving revision 1.1.2.7
diff -u -r1.1.2.7 ChangeLog.Xft
--- ChangeLog.Xft 2005/09/06 08:24:50 1.1.2.7
+++ ChangeLog.Xft 2005/09/10 18:28:55
＠＠ -1,3 +1,8 ＠＠
+2005-09-10 Clemens Fruhwirth <clemens(a)endorphin.org>
+
+ * glyphs-x.c (update_widget_face): Wrap error checking of rf in an
+ #ifdef USE_XFT.
+
2005-09-06 Stephen J. Turnbull <stephen(a)xemacs.org>
* console-x-impl.h (struct x_frame): Rewrite comment.
5:26 p.m.
New subject: [AC sjt-xft] fix glyphs-x.c for --with-xft=none
APPROVE COMMIT sjt-xft
>>>> "Clemens" == Clemens Fruhwirth
<clemens(a)endorphin.org> writes:
Clemens> Here is another small refinement. The tree won't built
Clemens> without "--with-xft", because error checking is done for
Clemens> a variable that doesn't existing, because it's only
Clemens> created when USE_XFT is define.
--
School of Systems and Information Engineering http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Ask not how you can "do" free software business;
ask what your business can "do for" free software.
7025
days inactive
7045
days old
3 comments
2 participants
participants (2)
-
Clemens Fruhwirth -
Stephen J. Turnbull