[Fonts] Re: After-XTT's extension of the encoding field.
Jungshik Shin [EMAIL PROTECTED] $B$5$s$O=q$-$^$7$?(B: (B (B On Sat, 2 Aug 2003, Chisato Yamauchi wrote: (B (B Have you seen CJK's *TYPICAL* fonts.dir of TrueType fonts? (B It is following: (B (B Not many people would be fond of tweaking fonts.dir/scale files (B these days :-) (B (BIt can be automatically generated. The /usr/sbin/fonts-config script (Bon SuSE Linux generates such TTCap entries automatically into the (Bfonts.dir if it detects that xtt is enabled in /etc/X11/XF86Config. (B (BCurrently it generates only entries for artifical bold and italic (Bfor fonts which do not yet have real bold and italic versions. (B (BI could extended it to support more features of TTCap, but I'm not (Bsure whether it is worth the effort. (B (BI agree that the old X fonts are broken beyond repair and we should (Bmove on to use fontconfig/Xft as much as possible. (B (BThe old font system must be kept for backwards compatibility of course (Bbut it is probably just a waste of effort to add more extensions the (BX11 core font system. (B (B-- (BMike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian $B?gL2ITB-$O;E;v$NE([EMAIL PROTECTED](B (B___ (BFonts mailing list (B[EMAIL PROTECTED] (Bhttp://XFree86.Org/mailman/listinfo/fonts
Re: Interesting job control bug
Juliusz Chroboczek [EMAIL PROTECTED] $B$5$s$O=q$-$^$7$?(B: (B (B You're running in a UTF-8 locale, and find a directory where the (B filenames are in ISO 8859-1. You do (B (B $ LC_ALL=en_US luit ls (B (B Ls puts its input into the pty's buffer, then quits. Depending on how (B lucky you are with scheduling, luit might get the SIGCLD before it (B deals with ls's output. Heisenbug. (B (BI see. I was lucky and always saw the output of ls in that case. (B (B I might very well change the default behaviour to quit if the child is (B dead *and* the input buffer is empty. (B (B (BLANG=en_US xterm -lc (B (Bin this xterm: (B (Bsleep 1000 (B (BThe input buffer is probably empty. But the child isn't dead. (BAnd without "-x" luit doesn't exit. (B (B-- (BMike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian $B?gL2ITB-$O;E;v$NE([EMAIL PROTECTED](B (B___ (BDevel mailing list (B[EMAIL PROTECTED] (Bhttp://XFree86.Org/mailman/listinfo/devel
Re: xterm and UTF8
Juliusz Chroboczek [EMAIL PROTECTED] $B$5$s$O=q$-$^$7$?(B: (B (B And, yes, of course xterm should start up in utf-8 mode if the locale (B encoding is UTF-8. (B (B MH Thanks, that was my original conclusion also. I had just (B MH wondered why it doesn't. Just an xterm bug I guess. (B (B *locale: true (B (B (Credit to Tomohiro Kubota.) (B (BThat makes xterm start luit automatically for non-UTF-8 locales. (B (Bxterm starts in UTF-8 mode when the locale encoding is UTF-8 even if (B"locale: false". XTerm did that for a long time already, see also (B (Bhttp://mail.nl.linux.org/linux-utf8/2001-05/msg00063.html (B (B-- (BMike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian $B?gL2ITB-$O;E;v$NE($@!#(B (B___ (BDevel mailing list ([EMAIL PROTECTED] (Bhttp://XFree86.Org/mailman/listinfo/devel
Re: [XFree86] Better non-antialiased TTF fonts with Xft?
Keith Packard [EMAIL PROTECTED] $B$5$s$O=q$-$^$7$?(B: (B (B Around 16 o'clock on Feb 16, Jeremy Buhler wrote: (B (B I can get fontconfig to do this for me, but the resulting non-antialiased (B fonts are rendered noticeably worse by Xft than by the core X rendering (B engine with the "freetype" module. (B (B The 4.2 version of "freetype" uses FreeType 1 which enables the TrueType (B byte code hinting engine which dramatically improves the quality of the (B resulting images. Xft (and XFree86 4.3 "freetype") use FreeType 2 which (B disables this by default over patent concerns. You can easily recompile (B Freetype2 to include the bytecode interpreter if US patents aren't a (B concern for you. That should make Xft and "freetype" generate identical (B results. (B (B In your freetype source directory, look in include/freetype/config/ftoption.h (B for TT_CONFIG_OPTION_BYTECODE_INTERPRETER (B (BDo you still think enabling the byte code interpreter is a good idea (Beven with the improved auto hinter in Freetype 2.1.3? (B (B(see also http://www.freetype.org/freetype2/2.1.3-explained.html) (B (BFreetype 2.1.3 also seems to have a bug that when the byte code (Binterpreter is enabled, Type1 fonts cannot be displayed anymore. (B (B(see also http://www.freetype.org/pipermail/devel/2003-February/008865.html) (B (BI guess this must be a bug, because the Type1 rendering probably (Bshould have nothing to do with the byte code interpreter. (B (B-- (BMike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian $B?gL2ITB-$O;E;v$NE($@!#(B (B___ (BXFree86 mailing list ([EMAIL PROTECTED] (Bhttp://XFree86.Org/mailman/listinfo/xfree86
Re: [XFree86] Better non-antialiased TTF fonts with Xft?
Keith Packard [EMAIL PROTECTED] さんは書きました: Around 0 o'clock on Feb 18, Mike FABIAN wrote: Wouldn't it be nice to be able to switch between the auto-hinter and the byte code interpreter with a run time option? If you have the bytecode interpreter, why would you ever use the auto-hinter for fonts with hints? And for fonts without hints? What if you have both fonts with and without hints? Is it conceivable that it might be useful to add rules to /etc/fonts/fonts.conf to switch on the byte code interpreter for some fonts (which have good hinting) and use the auto hinter for other fonts (which have bad or no hints)? When comparing rendering results of the Bitstream Vera Sans font with freetype 2.1.3 with and without byte code interpreter, most people looking at the screenshots claimed that it looks better without the byte code interpreter. Here are some screenshots (Freetype 2.1.3): http://www.suse.de/~mfabian/misc/keith-20030218/Bitstream-Vera-Sans-12-pixel-byte-code-interpreter.png Sharp and not much gray because the byte code interpreter snaps the stems to grid positions. But there is quite a lot of distortion of the glyphs, especially the dots in the ä (German ae-umlaut) are shifted to the left quite a lot. Looks like more than 2 pixels shifted. Why? Is this a bug? If yes, is it a bug in the font or in the byte code interpreter? And the spacing looks a bit wrong, the u and the nwere widened to fit to the grid and the first t was moved very close to the ä. Freetype 2.0.9 with byte code interpreter gives pixel for pixel the same screenshot by the way. http://www.suse.de/~mfabian/misc/keith-20030218/Bitstream-Vera-Sans-12-pixel-auto-hinter.png Most test persons said this looks better. It has a lot more gray and looks more blurry of course. But the ä is not distorted and the spacing looks better. http://www.suse.de/~mfabian/misc/keith-20030218/Bitstream-Vera-Sans-12-pixel-without-any-hinting.png Everybody agrees that this is ugly. http://www.suse.de/~mfabian/misc/keith-20030218/Bitstream-Vera-Sans-16-pixel-Linux-byte-code-interpreter.png a somewhat bigger size. Now the byte code interpreter suddenly shifts the dots on the ä to the right and makes each dot use 4 completely black pixels. Most viewers thought this looks very strange as well. Using the autohinter gives a decent result at this size as well. We also compared how Mac OSX rendered the same Bitstream Vera Sans font and found that the results were rather close to the results using the Freetype 2.1.3 with the auto hinter on Linux. Certainly the rendered glyphs on Mac OSX had more gray and were more blurry than the glyphs rendered by Freetype 2.1.3 with byte code interpreter. Looks almost as if Mac OSX isn't using a byte code interpreter, although there should be not patent problems for Apple. Supplying the byte code interpreter in binary form is viewed by some as an issue with the relevant US and UK patents. -- Mike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian 睡眠不足はいい仕事の敵だ。 ___ XFree86 mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/xfree86
Re: [XFree86] Better non-antialiased TTF fonts with Xft?
Keith Packard [EMAIL PROTECTED] $B$5$s$O=q$-$^$7$?(B: (B (Bftview 50 foo.pfa (B (B with foo.pfa any arbitrary Type1 font already displays garbage here (B if I compile Freetype 2.1.3 with the byte code interpreter enabled. (B (B Works great for me from vanilla Freetype 2.1.3 sources compiled on Debian (B testing (gcc version 2.95.4). I just tried Nimbus Sans L Regular (B from n019003l.pfb. (B (B I fear compiler issues... (B (BI had the problem with a gcc 3.3 prerelease. (B (BI'll try to check tomorrow whether I can reproduce it with gcc 2.95.4. (B (B-- (BMike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian $B?gL2ITB-$O;E;v$NE($@!#(B (B___ (BXFree86 mailing list ([EMAIL PROTECTED] (Bhttp://XFree86.Org/mailman/listinfo/xfree86
[Fonts] Re: A serious problem about freetype module
Chisato Yamauchi [EMAIL PROTECTED] $B$5$s$O=q$-$^$7$?(B: (B (B It's strange, but I can't reproduce the crash, although I your (B fix isn't yet included in the XFree86-4.2.99.901_CVS_20030205 (B I am currently running (B (Bxfd -fn -kochi-mincho-medium-r-normal--50-0-0-0-c-0-jisx0208.1983-0 (B (B doesn't crash when I use only the freetype module and not the xtt (B module. (B (B There are few glyphs of kochi-mincho.ttf leading to crash, (B so we may not reproduce the crash with xfd. I confirmed the (B crash using xfd with this XFLD: (B (B -kochi-mincho-medium-r-normal--24-0-0-0-c-0-jisx0208.1983-0 . (B (B Did you try xfontsel? (B (BYes, but it didn't crash either. (B (BNow I can't reproduce the crash anymore anyway because I am running (Byour fixed version already: (B (Bcvs log ftfuncs.c (B[...] (B (Brevision 1.26 (Bdate: 2003/02/09 02:44:42; author: dawes; state: Exp; lines: +3 -3 (B 864. Fix an X server crash with the "freetype" module when displaying (B kochi-mincho.ttf at a size of 18 pixels or higher (Chisato Yamauchi). (B (B (B-- (BMike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian $B?gL2ITB-$O;E;v$NE($@!#(B (B___ (BFonts mailing list ([EMAIL PROTECTED] (Bhttp://XFree86.Org/mailman/listinfo/fonts
[Fonts] Re: A serious problem about freetype module
Chisato Yamauchi [EMAIL PROTECTED] $B$5$s$O=q$-$^$7$?(B: (B (B There is a serious problem about "freetype" module in (B XFree86-4.2.99.901. (B (B If 18-pixels of kochi-mincho.ttf (a famous Japanese font) (B is selected using xfontsel with "jisx0208.1983", the X (B server crashes. (B (B[...] (B (B kochi-mincho.ttf -kochi-mincho-medium-r-normal--0-0-0-0-c-0-jisx0208.1983-0 (B (B (B First, please confirm this problem. Thanks. (B (BIt's strange, but I can't reproduce the crash, although I your (Bfix isn't yet included in the XFree86-4.2.99.901_CVS_20030205 (BI am currently running (B (B xfd -fn -kochi-mincho-medium-r-normal--50-0-0-0-c-0-jisx0208.1983-0 (B (Bdoesn't crash when I use only the freetype module and not the xtt (Bmodule. (B (B-- (BMike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian $B?gL2ITB-$O;E;v$NE($@!#(B (B___ (BFonts mailing list ([EMAIL PROTECTED] (Bhttp://XFree86.Org/mailman/listinfo/fonts
[Fonts] Re: where to include ~/.fonts.conf
James Richard Tyrer [EMAIL PROTECTED] $B$5$s$O=q$-$^$7$?(B: (B (B Mike FABIAN wrote: (B James Richard Tyrer [EMAIL PROTECTED] $B$5$s$O=q$-$^$7$?(B: (B (BMike FABIAN wrote: (B (BI wonder where in /etc/fonts.fonts.conf exactly is the best (Bplace to put the include statement for ~/.fonts.conf. (BIt cannot be at the very end of /etc/fonts/fonts.conf because (Bthen the following would not work: (B (BPut it at the end of the list starting with: (B (B dir/usr/X11R6/lib/X11/fonts/dir (B But if ~/.fonts.conf is included there (i.e. very early in (B /etc/fonts/fonts.conf), it is not possible to override the rules from (B /etc/fonts/fonts.conf in the personal ~/.font.conf file. (B (B I appear to need more sleep as I have given the correct answer, but it (B sappers to be for the wrong question -- not the one which you asked. (B (B Doesn't you: "fonts.conf" file already have the line: (B (B include ignore_missing="yes"~/.fonts.conf/include (B (B in it? (B (BYes, it is already there in the /etc/fonts/fonts.conf (Bfrom XFree86 CVS. But it is included rather close to (Bthe top of fonts.conf and followed by many rules. (BI experimented a bit and it appears to me, that I cannot (Boverride these rules by anything I write into ~/.fonts.conf (Bbecause the rule which comes last always wins. (B (BTo make it possible to override as much as possible of (Bthe global settings in /etc/fonts/fonts.conf with user (Bpreferences in ~/.fonts.conf, it seems to be better (Bto put the line (B (Binclude ignore_missing="yes"~/.fonts.conf/include (B (Bat the end of /etc/fonts/fonts.conf. (B (BObviously rules which depend on special settings in ~/.fonts.conf (Bmust follow below the line (B (Binclude ignore_missing="yes"~/.fonts.conf/include (B (Bfor example Keith's trick with "prefer_outlines". (B (BBut I couldn't find a reason to put anything else below this line, (Btherefore I wondered why it isn't already close to the bottom in (BXFree86 CVS and whether that should be changed. (B (BOr am I overlooking something? (B (B-- (BMike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian $B?gL2ITB-$O;E;v$NE($@!#(B (B___ (BFonts mailing list ([EMAIL PROTECTED] (Bhttp://XFree86.Org/mailman/listinfo/fonts
[Fonts] Re: where to include ~/.fonts.conf
James Richard Tyrer [EMAIL PROTECTED] $B$5$s$O=q$-$^$7$?(B: (B (B Mike FABIAN wrote: (B I wonder where in /etc/fonts.fonts.conf exactly is the best (B place to put the include statement for ~/.fonts.conf. (B It cannot be at the very end of /etc/fonts/fonts.conf because (B then the following would not work: (B (B Put it at the end of the list starting with: (B (B dir/usr/X11R6/lib/X11/fonts/dir (B (BBut if ~/.fonts.conf is included there (i.e. very early in (B/etc/fonts/fonts.conf), it is not possible to override the rules from (B/etc/fonts/fonts.conf in the personal ~/.font.conf file. (B (BThe last rule always seems to win. (B (BTherefore I wondered whether ~/.fonts.conf should be included (Bclose to the end of /etc/fonts/fonts.conf. Only a few rules which (Bdepend on settings in ~/.fonts.conf must be below that, like (BKeith's method to prefer outline fonts if (B (Bmatch target="pattern" (Bedit name="prefer_outline" (Bbooltrue/bool (B/edit (B/match (B (Bis in ~/.fonts.conf. (B (B-- (BMike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian $B?gL2ITB-$O;E;v$NE($@!#(B (B___ (BFonts mailing list ([EMAIL PROTECTED] (Bhttp://XFree86.Org/mailman/listinfo/fonts
[Fonts] Re: mkfontscale and family names which contain '-'
Pablo Saratxaga [EMAIL PROTECTED] $B$5$s$O=q$-$^$7$?(B: (B (B In the ttmkfdir version we use we handle such cases by using the postscript (B name of the font instead, when writting the fonts.dir file; I haven't had (B any problem so far (several years already doing that). (B (BSometimes the PostScript name can also have '-' characters: (B (Bmfabian@magellan:~$ ftdump /usr/X11R6/lib/X11/fonts/truetype/kochi-mincho.ttf | grep (Bpostscript (B postscript: Kochi-Mincho (Bmfabian@magellan:~$ (B (B-- (BMike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian $B?gL2ITB-$O;E;v$NE($@!#(B (B___ (BFonts mailing list ([EMAIL PROTECTED] (Bhttp://XFree86.Org/mailman/listinfo/fonts
[Fonts] Re: Unsafe chars in Mkfontscale
Juliusz Chroboczek [EMAIL PROTECTED] $B$5$s$O=q$-$^$7$?(B:
(B
(B Would you be so kind as to test the attached patch and confirm that it
(B does what you want? It's rather urgent, I'd like it to go into 4.3.
(B
(BThere is a 'i++;' too much:
(B
(B+ for(i = 0; i len; i++) {
(B+ if(unsafe(s[i]))
(B+ t[i] = ' ';
(B+ else
(B+ t[i] = s[i];
(B+ i++; - too much
(B+ }
(B
(Band the '-' character is missing in the list of unsafe chars:
(B
(B+ static int
(B+ unsafe(char c)
(B+ {
(B+ return
(B+ c 0x20 || c 0x7E ||
(B+ c == '[' || c == ']' || c == '(' || c == ')' || c == '\\';
(B+ }
(B
(B--
(BMike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian
$B?gL2ITB-$O;E;v$NE($@!#(B
(B___
(BFonts mailing list
([EMAIL PROTECTED]
(Bhttp://XFree86.Org/mailman/listinfo/fonts
[Fonts] mkfontscale and family names which contain '-'
If a font has a family name containing '-', mkfontscale generates the
(Bfamily name exactly like that into fonts.scale.
(B
(BThis occurs for example with the GohaTibebZemen.{otf,ttf} fonts which
(Bgot included into XFree86 CVS recently:
(B
(Bmfabian@magellan:/usr/X11R6/lib/X11/fonts/truetype$ grep "^Goha.*" fonts.scale
(BGohaTibebZemen.otf -misc-Goha-Tibeb Zemen-medium-r-normal--0-0-0-0-p-0-iso10646-1
(BGohaTibebZemen.otf -misc-Goha-Tibeb
(BZemen-medium-r-normal--0-0-0-0-p-0-adobe-fontspecific
(BGohaTibebZemen.ttf -misc-Goha-Tibeb Zemen-medium-r-normal--0-0-0-0-p-0-iso10646-1
(Bmfabian@magellan:/usr/X11R6/lib/X11/fonts/truetype$ ftdump GohaTibebZemen.otf
(BThere is 1 face in this file.
(B
(B- Face number: 0 -
(B
(Bfont name entries
(B family: Goha-Tibeb Zemen
(B style: Regular
(B postscript: GohaTibebZemen
(B
(Bfont type entries
(B FreeType driver: cff
(B sfnt wrapped:yes
(B type:scalable
(B direction: horizontal
(B fixed width: no
(B glyph names: yes
(B EM size: 1000
(B global BBox: (11,-189):(1223,802)
(B ascent: 820
(B descent: -180
(B text height: 1150
(B
(Bcharmaps
(B 0: platform: 0, encoding: 3
(B 1: platform: 1, encoding: 0
(B 2: platform: 3, encoding: 1
(B 3: platform: 7, encoding: 3
(Bmfabian@magellan:/usr/X11R6/lib/X11/fonts/truetype$
(B
(BThe XLFD thus generated doesn't work of course. How should this be
(Bfixed? Should the fonts be fixed?
(B
(BBut maybe it is no reason for the font maker to remove a '-' only
(Bbecause it doesn't work in a XLFD. Is a family name with a '-'
(Ballowed or not?
(B
(BMaybe one should work around that problem in mkfontscale by replacing
(Bthe '-' characters with ' ', for example like that:
(B
(BIndex: mkfontscale.c
(B===
(BRCS file: /cvs/xc/programs/mkfontscale/mkfontscale.c,v
(Bretrieving revision 1.3
(Bdiff -u -r1.3 mkfontscale.c
(B--- mkfontscale.c 2003/01/26 02:20:41 1.3
(B+++ mkfontscale.c 2003/02/04 18:50:28
(B@@ -317,7 +317,8 @@
(B *spacing, *full_name;
(B ListPtr encoding, entries = NULL;
(B int i, found, rc;
(B-
(B+char *p;
(B+
(B i = strlen(dirname_given);
(B if(i == 0)
(B dirname = strcat_reliable(".", "/");
(B@@ -477,7 +478,14 @@
(B fprintf(stderr, "Couldn't get family name for %s\n", filename);
(B family = entry-d_name;
(B }
(B-
(B+
(B+ while((p = strpbrk(family, "-"))) {
(B+ *p = ' ';
(B+ }
(B+ while((p = strpbrk(foundry, "-"))) {
(B+ *p = ' ';
(B+ }
(B+
(B if(!weight) weight = "medium";
(B if(!slant) slant = "r";
(B if(!sWidth) sWidth = "normal";
(B
(B--
(BMike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian
$B?gL2ITB-$O;E;v$NE($@!#(B
(B___
(BFonts mailing list
([EMAIL PROTECTED]
(Bhttp://XFree86.Org/mailman/listinfo/fonts
[Fonts] where to include ~/.fonts.conf (was: fontconfig alias question)
I wonder where in /etc/fonts.fonts.conf exactly is the best (Bplace to put the include statement for ~/.fonts.conf. (B (BIt cannot be at the very end of /etc/fonts/fonts.conf because (Bthen the following would not work: (B (BKeith Packard [EMAIL PROTECTED] $B$5$s$O=q$-$^$7$?(B: (B (B Here's what I put into my /etc/fonts/fonts.conf file (and which may go into (B the distribution for 4.3): (B (B match target="pattern" (B test name="prefer_outline" (B booltrue/bool (B /test (B test name="family" (B stringHelvetica/string (B /test (B edit name="family" mode="prepend" binding="strong" (B stringArial/string (B /edit (B /match (B (B[...] (B (B To enable these rules, I place: (B (B match target="pattern" (B edit name="prefer_outline" (B booltrue/bool (B /edit (B /match (B (B in my ~/.fonts.conf file to turn on the above substitution rules. (B (BI.e. one the include statement for ~/.fonts.conf must be above this (Brule which replaces Helvetica by Arial. (B (BBut maybe all rules in /etc/fonts/fonts.conf, which do not (Babsolutely have to be below the include statement for ~/.fonts.conf (Babove this include statement, to make it possible for each user (Bto override every rule he doesn't like. (B (BDoes that make sense? (B (B-- (BMike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian $B?gL2ITB-$O;E;v$NE($@!#(B (B___ (BFonts mailing list ([EMAIL PROTECTED] (Bhttp://XFree86.Org/mailman/listinfo/fonts
[Fonts] Re: Fontconfig release - fcpackage version 2.0
Keith Packard [EMAIL PROTECTED] $B$5$s$O=q$-$^$7$?(B: (B (B Around 16 o'clock on Jan 19, Mike FABIAN wrote: (B (B Given two fonts w/ the same name but different glyph coverages, how (B does one configure fonts.conf to ensure that the large font is chosen? (B (B If the fonts are versioned correctly, fontconfig will prefer the largest (B version number which should work. Otherwise, you should delete the broken (B font. Mike's kludge will also work. (B (BThe kludge to specify the file name didn't seem to work when I (Btried it. Apparently the file name was ignored. (B (BI also tried to use the filename to force the usage of a particular (Bbitmap font, but this didn't seem to work either. For example, if I (Bwant to use the 10x20 bitmap font, in xterm via Xft, how to do that? (B (Bfc-list gives me: (B (Bmfabian@magellan:~$ LC_ALL=en_GB.UTF-8 fc-list "fixed:pixelsize=20" file (B/usr/X11R6/lib/X11/fonts/japanese/10x20b.pcf.gz: (B/usr/X11R6/lib/X11/fonts/japanese/10x20ai.pcf.gz: (B/usr/X11R6/lib/X11/fonts/japanese/10x20a.pcf.gz: (B/usr/X11R6/lib/X11/fonts/misc/10x20-ISO8859-1.pcf.gz: (B/usr/X11R6/lib/X11/fonts/misc/10x20.pcf.gz: (B/usr/X11R6/lib/X11/fonts/japanese/10x20bi.pcf.gz: (Bmfabian@magellan:~$ LC_ALL=en_GB.UTF-8 fc-list (B"fixed:pixelsize=20:file=/usr/X11R6/lib/X11/fonts/misc/10x20.pcf.gz" file (B/usr/X11R6/lib/X11/fonts/misc/10x20.pcf.gz: (Bmfabian@magellan:~$ (B (BI.e. the last pattern gives me exactly 1 match from fc-list. (B (BNow I thought I could specify this as a parameter to xterm to get (Bexactly this font, but it doesn't seem to work like that: (B (Bmfabian@magellan:~$ FC_DEBUG=1 LC_ALL=en_GB.UTF-8 xterm -fa (B"fixed:pixelsize=20:file=/usr/X11R6/lib/X11/fonts/misc/10x20.pcf.gz" (BMatch Pattern 20 of 32 (Bantialias: FcTrue (Bautohint: FcFalse (Bdpi: 100.037 (Bfamily: "fixed" "Andale Mono"(w) "Courier New"(w) "Nimbus Mono (BL"(w) "Luxi Mono"(w) "Kochi Gothic"(w) "Kochi Mincho"(w) (B"HanyiSong"(w) "ZYSong18030"(w) "AR PL SungtiL GB"(w) "AR PL Mingti2L (BBig5"(w) "AR PL KaitiM GB"(w) "AR PL KaitiM Big5"(w) "Baekmuk (BGulim"(w) "Baekmuk Dotum"(w) "Baekmuk Batang"(w) "monospace"(w) (Bfile: "/usr/X11R6/lib/X11/fonts/misc/10x20.pcf.gz" (B (B[...] (B (BBest score 0 0 0 100 101 0 0 0 0 0 100 0 0 2.14748e+11Pattern 14 of 14 (Bantialias: FcFalse (Bcharset: set (Bfamily: "Fixed" (Bfile: "/usr/X11R6/lib/X11/fonts/japanese/10x20a.pcf.gz" (B (B[...] (B (Bmfabian@magellan:~$ (B (B-- (BMike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian $B?gL2ITB-$O;E;v$NE($@!#(B (B___ (BFonts mailing list ([EMAIL PROTECTED] (Bhttp://XFree86.Org/mailman/listinfo/fonts
[Fonts] Re: Question on xft font library.
Keith Packard [EMAIL PROTECTED] $B$5$s$O=q$-$^$7$?(B: (B (B I can pass glyphs through a matrix that shears the glyphs to make them (B look oblique, and FreeType will (probably) soon have code to artificially (B embolden glyphs, so it should be possible to generate synthetic varients (B in font style. (B (BThere seems to be code to artificially embolden glyphs in (Bfreetype-2 now, in freetype-2.1.3/src/base/ftsynth.c (Bthere is a function (B (B"FT_GlyphSlot_Embolden( FT_GlyphSlot slot )" (B (BCan this already be used via fontconfig/Xft? (B (B-- (BMike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian $B?gL2ITB-$O;E;v$NE($@!#(B (B___ (BFonts mailing list ([EMAIL PROTECTED] (Bhttp://XFree86.Org/mailman/listinfo/fonts
[Fonts] Re: Installation of Adobe's free Asian CID fonts
[EMAIL PROTECTED] (James A. Crippen) $B$5$s$O=q$-$^$7$?(B: (B (B JC There should be an automated tool like mkfontdir and mkcfm that knows (B JC how to read the CIDFonts and CMap directories and generates the proper (B JC fonts.scale file... (B (B Thanks for volunteering. (B (B I'll try a shell/sed/awk version first. If that doesn't end up too (B hard I might write a C version. But I have to get things working by (B hand before anything else... (^_^) (B (BI've added the attached perl script to SuSE Linux, which does (Bsomething like that. cidfont-x11-config Description: Binary data (B (B-- (BMike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian $B?gL2ITB-$O;E;v$NE($@!#(B
[Fonts]Re: some fonts apparently cannot be displayed with the newfreetype module
Juliusz Chroboczek [EMAIL PROTECTED] $B$5$s$O=q$-$^$7$?(B: (B (B MF XFree86 CVS version 20021219. The omegaserif fonts cannot be (B MF displayed using the new freetype2 based "freetype" of XFree86 (B MF anymore: (B (B Do they work with ftview of FreeType 2? (B (BYes, I have just tried that. (B (B-- (BMike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian $B?gL2ITB-$O;E;v$NE($@!#(B (B___ (BFonts mailing list ([EMAIL PROTECTED] (Bhttp://XFree86.Org/mailman/listinfo/fonts
[Fonts]Re: X-TT and OTF
Juliusz Chroboczek [EMAIL PROTECTED] $B$5$s$O=q$-$^$7$?(B: (B (BJC X-TT does not deal with OTF/CFF fonts. It does deal with TTF fonts. (BJC (BJC According to what you say, these are OTF/TTF fonts with the ``otf'' (BJC extension. According to Microsoft, OTF/TTF font files should have the (BJC ``ttf'' extension. Thus, apparently, the fonts included with XFree86 (BJC should be renamed to ttf. (BJC (BJC Before submitting a patch to do that, you should double-check that (BJC they are indeed TTF files. For example, use ttfdump (from (BJC ftp.freetype.org) to check whether they have a ``glyf'' table. (B (BI couldn't find ttfdump. Probably you mean ftdump? (B (BThe output of ftdump from ft2demos-2.1.3 is: (B (Bmfabian@magellan:~/ft2demos-2.1.3/bin/.libs$ ftdump (B/usr/X11R6/lib/X11/fonts/truetype/SyrCOMAntioch.otf (BThere is 1 face in this file. (B (B- Face number: 0 - (B (Bfont name entries (B family: Estrangelo Antioch (B style: Regular (B postscript: EstrangeloAntioch (B (Bfont type entries (B FreeType driver: truetype (B sfnt wrapped:yes (B type:scalable (B direction: horizontal (B fixed width: no (B glyph names: yes (B EM size: 2048 (B global BBox: (-660,-648):(2156,1560) (B ascent: 1560 (B descent: -648 (B text height: 2539 (B (Bcharmaps (B 0: platform: 1, encoding: 0 (B 1: platform: 3, encoding: 1 (B (BDoes 'glyph names:yes' mean that there is a 'glyf' table? (B (BThe output of ftdump from freetype-1 is: (B (Bmfabian@magellan:~$ /usr/bin/ftdump (B/usr/X11R6/lib/X11/fonts/truetype/SyrCOMAntioch.otf (Bfont name table entries (B-- (B (BEstrangelo Antioch - Version 1.20 (B (BPostScript name: EstrangeloAntioch (B (B?2001 Beth Mardutho: The Syriac Institute. For free distribution. Created by (BGeorge Kiraz, Christine Kiraz and Paul Nelson. Sponsered by Dr. Andreas Juckel. (B (B (B-- (B (Bcharacter map encodings (B-- (B (BThere are 2 encodings: (B (Bencoding 0: Apple Roman (Bencoding 1: Windows Unicode (B (B-- (B (Bftxcmap test (B-- (B (BThere are 2 encodings: (B (Bencoding 0: (Bfirst: glyph index 0, character code 0x0 (Bnext: glyph index 0, character code 0x1 (Blast: glyph index 0, character code 0xff (Bencoding 1: (Bfirst: glyph index 2, character code 0xd (Bnext: glyph index 3, character code 0x20 (Blast: glyph index 0, character code 0x (B (B-- (B (BGSUB table (B-- (B (B script `syrc' (index 0): (Bdefault language system: (B feature `ccmp' (index 1; lookup 0, 10, 11, 12) (B feature `calt' (index 0; lookup 13, 14) (B feature `fin2' (index 2; lookup 5) (B feature `fin3' (index 3; lookup 6) (B feature `init' (index 5; lookup 1) (B feature `medi' (index 8; lookup 2) (B feature `med2' (index 7; lookup 3) (B feature `rlig' (index 9; lookup 7) (B feature `liga' (index 6; lookup 8) (B feature `fina' (index 4; lookup 4) (B (BLookups: (B (B 0: type 2, flag 0x1 (B 1: type 1, flag 0x9 (B 2: type 1, flag 0x9 (B 3: type 1, flag 0x9 (B 4: type 1, flag 0x9 (B 5: type 1, flag 0x9 (B 6: type 1, flag 0x9 (B 7: type 4, flag 0x1 (B 8: type 4, flag 0x9 (B 9: type 4, flag 0x9 (B 10: type 6, flag 0x9 (B 11: type 6, flag 0x1 (B 12: type 6, flag 0x9 (B 13: type 6, flag 0x9 (B 14: type 6, flag 0x1 (B 15: type 1, flag 0x9 (B 16: type 1, flag 0x1 (B 17: type 1, flag 0x9 (B 18: type 2, flag 0x9 (B 19: type 2, flag 0x1 (B (B-- (B (BMemory footprint statistics: (B-- (B (B 11380 Bytes ( 12 kByte): face object (B 11084 Bytes ( 11 kByte): glyph object (B 3880 Bytes ( 4 kByte): instance object (B 9913 Bytes ( 10 kByte): exec. context object (B-- (B (B 36257 Bytes ( 36 kByte): total memory
[Fonts]Re: Scaled bitmap fonts
Juliusz Chroboczek [EMAIL PROTECTED] writes: KP Let's just rip the bitmap scaler out once and for all. If we do that (and I have no objection), we should encourage all distributions to provide scalable fonts for the Adobe base-14. The necessary font files are available with Ghostscript 6, and are already included in most distributions. Debian already makes them available to the X server. SuSE makes these available to the X server as well. We have also added :unscaled to all directories containing bitmap fonts. This caused a few problems with some applications who insisted on trying to scale bitmap fonts[1], but mostly it is an improvement. Footnotes: [1] e.g. Gimp. Gimp could do anti-aliasing for bitmap fonts by requesting them in an extremely big size first from the X-server and then calculating the anti-aliased glyph for the size desired by the users. This didn't work with :unscaled anymore, therefore I disabled it and made anti-aliasing work only with 'real' scalable fonts. xfig and dia also had some problems with :unscaled which I could fix by making them use the Ghostscript fonts instead of the -adobe-* bitmap fonts. -- Mike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian $B?gL2ITB-$O;E;v$NE($@!#(B ___ Fonts mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/fonts
[Fonts]Re: XmBDFEditor 4.5 released
Mark Leisher [EMAIL PROTECTED] writes: This is going out to the OpenType list because xmbdfed supports creation and editing of SBIT strikes up to 4bpp, which can be included in OT fonts using other tools. 4.5 is the last version with SBIT support. Nobody seems to be using it, so it's going. Werner Any reason for removing it? Design issues? I would rather stop Werner further development instead of removing it completely. The only reason to remove it was to reduce the code size. I think the reason nobody used this feature is that renderers are getting fairly sophisticated with regard to smaller point sizes. No further SBIT development was planned. The results of the rendering with freetype (1/2) at sizes between 10 and 20 pixel look much worse than the bitmaps contained for example in the free Japanese "Kochi Mincho" and "Kochi Gothic" TT fonts or Microsofts Japanese "MS Mincho" and "MS Gothic" TT fonts. -- Mike Fabian [EMAIL PROTECTED] http://www.suse.de/~mfabian $B?gL2ITB-$O;E;v$NE($@!#(B ___ Fonts mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/fonts
