On Tue, Oct 23, 2007 at 08:11:57PM +0200, Bernhard R. Link wrote: > * Paul E Condon <[EMAIL PROTECTED]> [071023 18:23]: > > To make the .ps version, that I refer to in my bug report, I used the > > 'print to file' facility in xpdf. > > To make sure that it is a problem in gs-* and not in gv, you could also > try to convert it using pdf2ps (do not confuse with pdftops, pdf2ps > is ghostscript, while pdftops is xpdf) and look with gv at that file. > If that has the same problem like looking at the .pdf directly, I'd > guess it is gs. (Though I still have not understood the symptoms > correctly). > snip...
I converted the file with both pdf2ps and pdftops. The resulting files were different: -rw-r--r-- 1 pec pec 587490 Oct 22 15:04 thinking_about_mechanisms.pdf -rw-r--r-- 1 pec pec 18810638 Oct 22 15:06 thinking_about_mechanisms.ps #via xpdf -rw-r--r-- 1 pec pec 73962490 Oct 23 15:25 thinking_about_mechanisms2.ps #via pdf2ps -rw-r--r-- 1 pec pec 18810638 Oct 23 15:38 thinking_about_mechanisms2a.ps #via pdftops During the run, pdf2ps issued the identical warnings as I have already reported coming from gv. The output of pdf2ps also cannot be centered when printed by gv. (Also note the file is much larger.) During the run of pdftops no messages were displayed. and the resulting .ps file is identical to the .ps file from using xpdf to convert. I filed against gv because ghostscript has been replaced by gs in the repository, and gs is a helper script. I guess gs-common is its new name. And its not clear that it is in gs-common since gs-gpl and gs-esp behave differently. Anyway, I'd like you to choose. > If that file is multiple pages long, you could also try to get a single > page displayed incorrectly via pdftk (from package pdftk) with a command > like > pdftk yourfile.pdf cat 17 output new.pdf > to get page 17 (as example). > This might also repair some xref tables, so it would be intresting if > the result has still the same problem and the same error message. > There is a pending 'serious' bug in pdftk so I think anything I might do with it now will be suspect evidence on this problem. Since I do have a work-around, I think this bug, while real enough, can be handled at a lower priority. Your welcome to reassign it or whatever you think is appropriate for the maintainence of Debian's good reputation. Thanks very much for your help. --- Paul E Condon [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
