I'm converted the attached PostScript file into PDF with GS head and when i'm trying to open the PDF file in Acrobat Reader i get "Cannot extract the embedded font" error message. The PDF file created by Adobe Distiller opens fine in AR. PS. i got this problem with all gs8.x builds, it works fine, however, with gs7.03
Created attachment 910 [details] PostScript file (zipped with winzip)
Created attachment 911 [details] PDF file created with gs8.32
I cannot reproduce this with Acrobat Reader 5.x on Linux. What version of Reader are you using?
I tried it with Acrobat Reader 6.0 for windows on several computers and got the same error mesage. With Acrobat Reader v5.0 this document opens with no error message, however the fonts are distorted on the first page - i'm attaching the PDF file created with Adobe Distiller, so you can compare the results (see the hieroglyphs in the first line of the first page).
Created attachment 922 [details] PDF file created with Distiller
I've reproduced the error with the gs8.32-created PDF with Acrobat Reader 6.0 on MacOS X.
I've reproduced the effect with Adobe Reader 6.0 on Windows. Actually the attached file causes an indeterministic behavior of the Adobe Reader 6.0. For a better stability of the effect, run Adobe Reader with no document, then use File/Open menu with the copy of the attached file. Current HEAD generates same PDF as the attached one, except for file IDs and CreationDate. When I disable the PDF stearm compression, the effect becomes less probable, and I could not find how to get it stable with no compression.
There is a problem painting the first 2 hieroglyphs on the first page, with both Adobe Reader 6 and Acrobat reader 5.0 . When glyphs are painted, they are highly distorted.
Trying to render the generated PDF with Ghostscript, got : Error: /invalidfont in --@0x1003af90-- This should help to the analyzis.
Tracing GS with the distilled document, found that default_get_outline returns empty string for the following glyph indices : 3, 262, 263, 264, 265, 363, 407, 391, 314, 367, 392, 326, 312, 653, 370, 328, 338, 312, 425, 324,350 - 'loca' specifies zero length. Likely they are missed from the font subset while the distillation.
Patch http://www.ghostscript.com/pipermail/gs-cvs/2004-September/004894.html Note another Bug 687716 related to same document.
Ported the patch to GS_8_1X : http://www.ghostscript.com/pipermail/gs-cvs/2004-September/004895.html