Control: tags 648330 + unreproducible moreinfo Am 10.11.2011 um 16:59 teilte cfr mit:
Hi Clea, I'm going through some old bugs and check if they are still valid. https://bugs.debian.org/648330 As Atsuhito I'm failing to reproduce the issue. I've opened the TeX file in texworks, wrote some text and compiled the document. Then the viewer opens. I can now change the document and recompile, the viewer refreshes the view. I don't see any dat file around. Is the issue still reproducible for you? Hilmar > I set up a merge file as follows: > filename.dat [data to be merged] > filename.tex [tex file to be processed] > Processing filename.tex produces filename.pdf. If I now ask for the file > to be re-typeset while filename.pdf is frontmost, TeXWorks attempts to > process filename.dat rather than filename.tex. (And this obviously > doesn't work since filename.dat just contains a bunch of data.) > > I would expect that asking for the document to be reprocessed while > filename.pdf was active would cause TeXWorks to typeset filename.tex. > This might work for two reasons: first, TeXWorks might remember which > file actually produced the pdf (clearly typesetting the .dat file has > never yielded any pdf at all); second (more likely), I'd expect TeXWorks > to prefer filename.tex to filename.dat for any filename and to process > the former rather than the latter whenever possible. > > filename.tex also includes the standard TeX programme tag line at the > top which filename.dat obviously doesn't so TeXWorks could use that > information although that would probably be unnecessarily complicated. > > Because TeXWorks associates the .pdf with the .dat rather than the .tex, > this also means that opening the .dat in TeXWorks also opens the .pdf > which is rather unexpected. (Opening the .tex also opens the .pdf.) > Similarly, closing the .tex doesn't close the .pdf if the .dat is also > open. (Neither does closing the .dat if .tex is also open.) But although > I find this behaviour odd, it isn't in itself objectionable. It is the > typesetting relationship which is really problematic. > > (I realise that I could probably work around this by using different > filenames but in lots of cases I've got customised scripts which would > need changing and that's more work!) > -- sigfault #206401 http://counter.li.org
signature.asc
Description: OpenPGP digital signature
