From: Werner LEMBERG <[EMAIL PROTECTED]>
> > If there are some remarks, please let me know. I've only access to
> > a teTeX distribution on Debian GNU/Linux.
>
> Sorry, I'm away from good internet access for the next three weeks.
No problem. I've also got to get those HBF fonts working, so that'll
keep me busy.
> > I currently need four hlatex2agl.pl commands, one for each .sfd
> > file: UKS-HLaTeX.sfd, UKS-HLaTeX-Hangul.sfd,
> > UKS-HLaTeX-Hangul_Hanja.sfd and UKS-HLaTeX-Hangul_Symbol.sfd. So it
> > would be nice if hlatex2agl.pl could also have an extra argument to
> > choose which .sfd is to be used.
>
> Patch, please :-)
Patch coming up soon!
> > The u??.pfb files created by hlatex2agl have exactly the same name
> > as the original HLaTeX counterparts. Do I have to overwrite the
> > original fonts?
>
> Yes. They are improved versions with corrected glyph names; nothing
> else has changed.
So I can use HLaTeX with these new fonts as well?
Hmmm. Processing hlguide.tex doesn't seem to break. A lot of
warnings, but nothing serious (probably .fd files that are not
complete on my side).
> > I have problems with dinaru and pilgia when I run dvips or pdflatex.
> > Well, warnings actually. With dvips I get:
> >
> > dvips: Checksum mismatch in udn10
> > [...]
>
> This is not related to the PFB fonts but to TFM (and VF) files. PFBs
> don't contain a checksum at all. Apparently the DVI file you are
> using for testing purposes has been created with a different set of
> metric files (probably the old ones, whatever they are; maybe you have
> still something from earlier experiments).
I didn't remove the original HLaTeX fonts, because I wasn't sure
whether they could be overwritten or not.
I removed the afm and type1 fonts that have the same name, yet the
checksum problem persists with dvips/dvipdf.
Besides, why would only dinaru and pilgia have this problem, and not
the others?
Perhaps it's because the *00.pfb files don't get modified (ut infra)?
> > I noticed that hlatex2agl didn't put a comment in the dinaru and
> > pilgia .pfb fonts. I think the reason is that these two fonts were
> > the only two not created by Fontographer 4.1 but by FontForge 1.0.
> > Perhaps hlatex2agl botches something up?
>
> The FontForge version used to create those PFBs doesn't add a
> `Generated' comment line which hlatex2agl uses as an anchor to insert
> its additional comment. Perhaps it would be useful to apply the
> following patch.
>
>
> --- hlatex2agl.pl 2006-04-25 17:11:50.000000000 +0200
> +++ hlatex2agl.pl.new 2006-08-09 10:31:35.000000000 +0200
> @@ -455,7 +455,7 @@
> print(OUTFILE $_);
>
> # Add comment.
> - if (/^% Generated/) {
> + if (/^% CreationDate/) {
> print(OUTFILE "% AGL compliant glyph names added "
> . "by script $prog 2005-Jul-27.\n");
> }
>
Thanks.
> A similar problem occurs with udn00.pfb: The very first line doesn't
> contain `001.000', so the script is not able to correctly update the
> version number. You might extend the regular expression to handle
> this; I'm too lazy for it.
Something like
s/Altsys\ Fontographer\ 4\.1\ 1\/10\/95$/001.001/;
perhaps?
I think the reason that udn00.pfb doesn't get processed is because
udn00 doesn't appear in wdn*.vf. BTW, udn00.pfb is made with Altsys
Fontographer, udn1?.pfb are made with FontForge.
For example, here's how wdn00.vf looks like:
(MAPFONT D 0
(FONTNAME udn10)
(FONTCHECKSUM O 20132433534)
(FONTAT R 1.0)
(FONTDSIZE R 10.0)
)
(MAPFONT D 1
(FONTNAME udn11)
(FONTCHECKSUM O 22026147241)
(FONTAT R 1.0)
(FONTDSIZE R 10.0)
)
(MAPFONT D 2
(FONTNAME udn12)
(FONTCHECKSUM O 33161023114)
(FONTAT R 1.0)
(FONTDSIZE R 10.0)
)
BTW, is it normal that udn12.pfb and upga12.pfb don't get uni names
instead of ks ones?
> > I get the following warnings:
> >
> > gothic/log:Use of uninitialized value in hash element at
> > hlatex2agl.pl line 359.
> > [...]
>
> Interesting. Please tell me which fonts are involved. I suspect this
> is due to problems with Koaung-Hi's VF files.
gothic/log:
Creating virtual property list file `wgtb29.vf.vpl'...
Use of uninitialized value in hash element at hlatex2agl.pl line 359.
Reading `wgtb29.vf.vpl'...
graphic/log:
Creating virtual property list file `wgrb07.vf.vpl'...
Use of uninitialized value in hash element at hlatex2agl-hangul_symbol.pl line
359.
Reading `wgrb07.vf.vpl'...
jamo/log:
Creating virtual property list file `wjsr9.vf.vpl'...
Bad VF file: Character 228 in font 0 will be ignored
...
Bad VF file: Character 254 in font 0 will be ignored
Use of uninitialized value in array element at hlatex2agl-hangul.pl line 341.
Use of uninitialized value in sprintf at hlatex2agl-hangul.pl line 348.
Use of uninitialized value in array element at hlatex2agl-hangul.pl line 349.
Reading `wjsr9.vf.vpl'...
> > Currently each font is its own "italics" and its own "typewriter"
> > font. Only the glue would be different. Would you prefer to have
> > for example every \texttt{} use Tada, or should the user himself be
> > smart enough to use \CJKfamily{tt} and to reconfigure \texttt and \tt
> > if s/he wishes so? We might have to rethink how the .fd files are
> > configured.
>
> *This* you should discuss with Koaung-Hi :-) I have no idea how Korean
> fonts are used in the real world.
OK. Despite what my surname suggests, I'm not Korean but Chinese, so
I have no idea either. I'll ask him later. =)
PS: I'm going to change ISP soon. I don't know how long I'll be
offline and what my email address will be. So do keep CC'ing the
messages to the mailing list. ;)
Cheers!
Danai SAE-HAN
韓達耐
--
題目:《霰》
作者:楊萬里（１１２７－１２０６）
雪花遣霰作前鋒，勢破張皇欲暗空。
篩瓦巧尋疏處漏，跳階誤到暖邊融。
寒聲帶雨山難白，冷氣侵人火失紅。
方訝一冬暄較甚，今宵敢嘆臥如弓。
_______________________________________________
Cjk maillist - [email protected]
http://lists.ffii.org/mailman/listinfo/cjk
