Bug 690548

Created attachment 5129 [details]
test.ps

PS file to test.

Firstly your PostScript file does not contain the whole of the TrueType fonts
that you are using, it contains subsets of them.

Because it only contains the glyphs which are actually used, and because all the
glyphs the fonts contain are used, there is no actual subsetting taking place.

When I set -dSubsetFonts=false then the resulting PDF file is not labelled by
Acrobat as containing subsets (though the file also does not render correclty
interestingly).

As far as I'm aware Acrobat is unable to tell if a font is truly subset or not,
but relies upon the font name. If the font name begins with XXXXX+ where X is a
random letter, then Acrobat deduces that the font is subset. 

At present TrueType fonts emitted by pdfwrite *always* have this prefix, and so
appear to be subset even when they are not.

Perhaps if you explained why this is causing a problem I could be more helpful.
If its simply the presence of the prefix then there is already an outstanding
bug tracker for this issue.

We produce very man single PDF - Files with Ghostscript. The Problem is when 
we merge them with Adobe Acrobat all different font subset apear in the new 
PDF Document (About 270 Pages). Saving a Document with about 80 Font Subsets 
Takes about 5 minutes for a 5 Megabyte File.

When we create the same pages with Acrobat Distiller we don't have any font 
subsets in the Single PDF. When we merge the single PDFs to one big (270) we 
are able to save the file within a few seconds. Would it be helpful to send 
test PDF files?

Hmm, don't really need any further examples, one is sufficient. I'm surprised
that you can get away with merging these without glyphs going missing, but I
guess Acrobat is being clever in some way.

As I said, when I set -dSubsetFonts=false with the current HEAD revision, the
resulting PDF file does not (according to Acrobat) contain any subset fonts. The
font names are likewise not prefixed with the 'XXXXX+' codes. 

So I'm not experiencing the problem you describe. I do see a different problem
(the file doesn't display correctly) but that is a different issue.

[later]
It looks like this has been partially addressed, see bug #689236 which describes
a partial patch.

I'm closing this issue as a duplicate of #689236 as the described problem is
identical. The remaining issues are described in the other report.

Ooops, pressed the wrong button. This time lets mark it as a duplicate instead
of 'fixed'.


*** This bug has been marked as a duplicate of 690236 ***

Can you please send me gsdll32.dll for Windows 32 Bit so that i can test it 
too?

Can you please upload me your converted PDF file.

Can you please write me the exact statement you executed?

Created attachment 5135 [details]
PostScriptFiles.rar

This are my PostScript Files which i want to Convert to PDF. I have converted
it once with Acrobat Distiller and once with GhostScript 8.64.

With Distiller I get Files with 2 Fully Embedded Fonts (Tahoma and Tahoma Bold)

With GhostScript I get Files with 2 Embedded Subset Fonts (Tahoma and Tahoma
Bold)

No Problem till now.

But when i combine this 63 Pages to one PDF File there is a big Problem.
Combined PDF Files made with Distiller has 2 Fully Emebedded Fonts. This file
is very Fast (I can save it in a few Seconds)
Combined PDF Files made with Ghost Script has 26 Subset Fonts included. This
file has fourth the size and is very slow. (Saving takes over a minute)
Working with the second File is not funny.

I used this Command to Create PDF from the PS with GhostScript:
-q -sDEVICE=pdfwrite -dSubsetFonts=false -dSAFER -dBATCH -dNOPAUSE
-sOutputFile=0100_Weiße Palette_METri.pdf -c -f 0100_Weiße Palette_METrips.ps

Created attachment 5136 [details]
PDF Files made with Distiller.rar

Created attachment 5137 [details]
PDF Files made with GhostScript.rar

Please consider the Text I added to Attachment PostScriptFiles.rar.

Mistake, this is a duplicate of 689236, not 690236.

*** This bug has been marked as a duplicate of 689236 ***