https://bugzilla.novell.com/show_bug.cgi?id=678222 https://bugzilla.novell.com/show_bug.cgi?id=678222#c0 Summary: texlive-2010 / pdflatex / pdfopt (ghostscript) / PDF Ver. 1.5 oddity Classification: openSUSE Product: openSUSE 11.4 Version: Factory Platform: x86 OS/Version: Other Status: NEW Severity: Normal Priority: P5 - None Component: Other AssignedTo: bnc-team-screening@forge.provo.novell.com ReportedBy: madworm_de.novell@spitzenpfeil.org QAContact: qa@suse.de Found By: --- Blocker: --- User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.2.13) Gecko/20101203 SUSE/3.6.13-3.1 Firefox/3.6.13 Welcome to Hell... I've just recently upgraded my laptop to 11.4. My Desktop is thankfully still running 11.3 and will stay that way it seems. I'm using Kile + texlive (pdflatex) to write a somewhat lengthy document (>200 pages with lots of graphics etc.) With openSUSE 11.3 / texlive-2009 everything worked just fine. Pdflatex per default produced PDF-1.4 documents, which is important. Now 11.4 comes with texlive-2010, which produces PDF-1.5 files per default, but can be forced to spit out 1.4 as well (by adding "\pdfminorversion 4" in the preamble btw.). Old workflow: a) compile all of the stuff as usual (bibtex, makeindex...) b) take the resulting pdf-1.4 (which works everywhere as far as I can tell) and use 'pdfopt' (part of the ghostscript package) to 'linearize' the pdf for better viewing perfromance. c) use 'pdfinfo' to show some info about the file d) test it with old viewers on winblows (acrobat pro 6.0) - works. On my updated laptop: If I take the exact same files (I use GIT, so I can be sure) and run my compile script on 11.4 I get problems as such: a) 'pdfinfo' complains that the PDF is probably corrupt (xref table...), but is able to recover. Same is true for okular and recent acroread. b) acrobat 6.0 pro (winblows) just crashes all the timie. It tells me that the pdf is irreparably corrupt. I tried to test certain passages of said document (images), but I can't pin it down to single files. Several combinations of disabling/enabling them are good, others are bad. What I've found out so far: a) if I skip the step with 'pdfopt' everything is sane and works (with PDF-1.5) b) if I force pdflatex to spit out PDF-1.4, running the optimizer does no harm as well. Most of my graphs were created with gnuplot/svg and postprocessed with inkscape and exported as pdfs. I can try to put together a minimal test case, but that will take some time. Lots of work to do. Reproducible: Always Steps to Reproduce: 1. pdfopt applied on a large (with images) PDF-1.5 = corrupt xref table in PDF 2. pdfopt applied on a large (with images) PDF-1.4 = fine -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.