enable the 'experimental' dvi support in evince in rawhide through the fc5 test release process so testers can test and report back with no commitment to leave dvi enabled in fc5 final release. Then turn dvi support off again at the close of test phase if dvi is still not good enough to ship in fc5. -jef
Additionally id like to point out that tetex-xdvi is one of the few remaining packages in Core that requires openmotif. Moving forward towards being able to move openmotif out of Core has to be an important goal somewhere in the larger plan for Core. And the only way thats going to happen and keep dvi document support in Core is to get it turned on in evince and tested. -jef
Jef: This sounds like a great idea. Also, reducing the openmotif dep count by 1 package would make it easier to transition motif to Fedora Extras, and out of the Core OS.
Kristian: Any opinion on this? Curious what your thoughts are, etc. TIA
Even if evince was to be compiled with xdvi support, evince is NOT a good replacement for xdvi because it is missing a fundamental feature: support for dvi source specials. Source specials allow for bidirectional communication between a tex editor and a dvi viewer, it allows for a dvi document to be opened showing the place in the dvi document that corresponds to the place where the cursor is in the tex editor (and the other way around). This support is fundamental for editing tex documents with hundreds of pages like dissertations, books. Evince does not have this kind of features. If the problem is xdvi being linked with motif, then just compile xdvi with support for another toolkit (like xaw or xaw3d). Replacing xdvi with evince would be a major inconvenience for the people that actually use xdvi on a daily basis. Just my 2 cents.
(In reply to comment #4) > Even if evince was to be compiled with xdvi support, evince is NOT a good > replacement for xdvi because it is missing a fundamental feature: support for > dvi source specials. > Please file a different RFE on other feature requests. Whether it is going to replace xdvi or not is not under discussion in this report. That will have to be decided after we experimentally enable and test this option in a development cycle.
PING !! I believe the dvi backend is not marked as experimental anymore.
Marking this as a dupe of 187381 (even though this bug was filed first - this is a subset of 187381. *** This bug has been marked as a duplicate of 187381 ***