Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

PostScrip printing of Mathtime fonts in Acrobat 5

4 views
Skip to first unread message

Eddie

unread,
May 15, 2001, 9:49:38 AM5/15/01
to
Hello TeX user,

I'm using pdflatex (Miktex 2, Windows 2000) and y&y Mathtime fonts for my
documents.
With Adobe Acrobat 4.05 everything was fine, but when printing pdf documents
on a postscript printer using Acrobat 5, I get the following error message
"An error occured while downloading a font. This document might not priny
correctly." and the output is terribly wrong, some y&y math fonts are not
printed and others have been replaced. On screen everything looks fine
though, and printing on a non-postscript printer is also fine. Does anyone
have the same problem? Any solution? Is it a bug in Acrobat 5?

Regards, Edo Hulsebos


MicroPress, Inc.

unread,
May 15, 2001, 12:27:04 PM5/15/01
to
There are at least two separate bugs in Acrobat 5 (both the reader and
the full version) which result in the behaviour you are getting.

The explanation and solutions will be posted on our site later today
or tomorrow.

Eddie

unread,
May 15, 2001, 12:31:24 PM5/15/01
to
Aha, that would be nice!
Which website is it? www.micropress-inc.com?

Regards, Edo

"MicroPress, Inc." <m...@micropress-inc.com> wrote in message
news:3b0157fa...@news-server.nyc.rr.com...

MicroPress, Inc.

unread,
May 15, 2001, 12:54:59 PM5/15/01
to
Yes. I'll post the url here when it is up.

Donald Arseneau

unread,
May 15, 2001, 6:30:46 PM5/15/01
to
"Eddie" <h...@war.uu.nl> writes:

> With Adobe Acrobat 4.05 everything was fine, but when printing pdf documents
> on a postscript printer using Acrobat 5, I get the following error message

"Acrobat" is so named because of the contortions one must do to
get around its bugs.

Donald Arseneau as...@triumf.ca

Giuseppe Bilotta

unread,
May 16, 2001, 5:57:12 PM5/16/01
to
ROTFLASTC

Donald Arseneau wrote:
> "Acrobat" is so named because of the contortions one must do to
> get around its bugs.

--
Giuseppe Bilotta

Axiom I of the Giuseppe Bilotta
theory of IT:
Anything is better than MS

Eddie

unread,
May 17, 2001, 4:15:22 AM5/17/01
to
Dear sir, madame,

Is the url already up?
Regards, Edo Hulsebos

"MicroPress, Inc." <m...@micropress-inc.com> wrote in message

news:3b025f8d...@news-server.nyc.rr.com...

MicroPress, Inc.

unread,
May 17, 2001, 1:56:18 PM5/17/01
to
Yes, see

http://www.micropress-inc.com/acr5bugs.htm
http://www.micropress-inc.com/acr5bugs.htm#added

The second link gets you to the most relevant part.

Notice that the page is intended primarily for VTeX's users;
the bugs are "general" AR5, but the corrective actions may
be different.

If there is an interest, we would add small examples showing
the removed and added bugs later.

If you are aware of additional bugs, please let us know.
[We _know_ that there are more, but we do not have yet
"distilled" them to small examples.]

Thanks to D.A. for a great quote.

Louis Vosloo

unread,
May 25, 2001, 8:34:15 AM5/25/01
to
Giuseppe Bilotta wrote:

> ROTFLASTC

> Donald Arseneau wrote:
> > "Acrobat" is so named because of the contortions one must do to
> > get around its bugs.

There has been some discussion here of apparent problems with
Acrobat 5.0, pdfTeX, MathTime fonts, "extra LaTeX" fonts, etc.

(1) As far as we know, there are no problem with MathTime fonts.

The one error reported in this news group was from output of pdfTeX.
Looking at the file we note that pdfTeX replaced the first 8 zeros
of the trailing 512 zeros in the font RMTMI with nulls (character code 0).
This may be the result of some kind of counting error. Apparently
Acrobat Reader 4.0 is not sensitive to this error, while Reader 5.0 is.

pdfTeX also strips out unused Subrs *without* renumbering the remaining
Subrs. While this may seem safe, it is not a good idea, given that older
versions of ATM, for example, would not tolerate a mismatch between
the Subrs array allocation and the actual slots used. (Amusingly,
the problem occured when the array was *larger* than necessary, while
an array that was too small was OK!)

(2) As far as we know, there are no problem with Acrobat 4.05 Distiller,
or Acrobat 5.0 Distiller, or Acrobat 4.05 Reader, or Acrobat 5.0 Reader
when used with Y&Y TeX.

That is, you can read and print PDF files made by "printing to Distiller"
or by producing a PS file and having Distiller process it.

(3) There are no font rendering quality issues with Y&Y fonts.

The rasterizer in Acrobat 5.0 Reader is apparently different, but only
poorly hinted fonts are affected. (It's not clear exactly what is happening,
but it may be that this rasterizer has a new "anti-aliasing"/"font smoothing"
engine --- at the same time, the reappearance of bugs known from old
versions of ATM suggests that the new engine may be built on top of an
older version of ATM).

(4) There are no problems with "circle", "line" and "lasy" fonts from Y&Y

The "extra LaTeX + SliTeX" fonts we supply
http://www.yandy.com/latxfont.htm http://www.yandy.com/river/setdetail-LTX.htm
work just fine with Acrobat 4 and 5.

The older (and less complete) BSR/Y&Y/AMS et al font set at
http://www.ams.org/index/tex/type1-fonts.html
(fonts that are copyrighted AMS --- *not* in the public domain,
as mistated sometimes), have a benign typo in the encrypted section,
which creates a useless key / value pair. Of course, there is no
prohibition against having extra keys in font dictionaries, but some
software that does not follow the spec may stumble over this.

(5) Defensive programming wins again!

Over the years, assorted bugs have been found in ATM, Acrobat Reader,
PostScript printer interpreters, and on rare occassion even Acrobat Distiller.
These include: crash when the number of Subrs used does not match the
size of the Subrs array, crash when the number CharStrings does not match
the number of CharStrings used. Since there appears to be no continuity
in software development, or a data base of old test cases, one can expect
these errors to reappear periodically (*). Hence it is best to avoid such
programming practices.

Naturally, we welcome reports of any problems with any of our fonts or
our TeX System, whether related to Acroobat 5.0 or not. If possible,
please provide test files that illustrate the problem.

Regards, Louis.

(*) Similarly, periodically one finds bugs reappearing having to do with:
(i) mistakenly using null terminated strings (thus making use of character
code 0 impossible --- and "minus" e.g. is character code 0 in math symbol);
(ii) problems with "control characters" (char code 0--31 -- used, of course,
in CM fonts); and
(iii) problems with repeated encoding (when a glyphname appears
in more than one slot in the encoding vector --- a feature added only
to try and work around bug (ii)).


--
Y&Y, Inc. mailto:sup...@YandY.com http://www.YandY.com

Oscar Lazzarino

unread,
May 25, 2001, 8:45:15 AM5/25/01
to
Louis Vosloo wrote:
>
> The rasterizer in Acrobat 5.0 Reader is apparently different, but only
> poorly hinted fonts are affected. (It's not clear exactly what is happening,
> but it may be that this rasterizer has a new "anti-aliasing"/"font smoothing"
> engine

They added (as an option) the new "cooltype" rasterizer. You can enable
it in the preferences window. It makes use of the red, green, and blue
sub-pixel on your monitor or LCD display. Very small type looks better
when enabled. Not-so-small type looks like having a blue-cyan-reddish
halo around it... It can be somehow calibrated, but it doesn't get much
better. I still don't know if I like it or not.

Oscar Lazzarino

Giuseppe Bilotta

unread,
May 25, 2001, 2:15:23 PM5/25/01
to
Louis Vosloo wrote:
> Giuseppe Bilotta wrote:
>
> > ROTFLASTC
>
> > Donald Arseneau wrote:
> > > "Acrobat" is so named because of the contortions one must do to
> > > get around its bugs.
>
> There has been some discussion here of apparent problems with
> Acrobat 5.0, pdfTeX, MathTime fonts, "extra LaTeX" fonts, etc.

[snip]

I know. ROTFLASTC was an appreciation for the brillian quote by Mr.
Arseneau [I saw you put it in your Acrobat Bugs page]

ROTFLASTC means "Rolling On The Floor Laughing And Squeezing The Cat".

Louis Vosloo

unread,
May 26, 2001, 7:51:40 AM5/26/01
to
Louis Vosloo wrote:

> The one error reported in this news group was from output of pdfTeX.
> Looking at the file we note that pdfTeX replaced the first 8 zeros
> of the trailing 512 zeros in the font RMTMI with nulls (character code 0).
> This may be the result of some kind of counting error. Apparently
> Acrobat Reader 4.0 is not sensitive to this error, while Reader 5.0 is.

Addendum, pdfTeX is not to blame, we discovered that the font file was made with
a font converted using "CrossFont", which replaced 8 zeros with 8 nulls.

Suggestion: get the appropriate font file from the foundry for the platform you are
on,
or use quality tools for font conversion: http://www.yandy.com/products.htm#FMP

--
http://www.YandY.com/unique.htm


0 new messages