On Wednesday, February 20, 2013 6:03:04 PM UTC+1, fhivert wrote: > > Hi Jean-Pierre, > > > > > * A parallel build (using the usual MAKE="make -jN") can speed up > the > > > > docbuilding time. Note that the speed-up factor is closer to N/2 > than to > > > > N because two passes are needed now. > > > > > > This is not quite true. I don't have precise timing. Put most of the > time > > > of > > > the first pass is spend parsing the files, generating the indexes and > > > pickling > > > the parse-tree. on the second pass the doc in not reread, instead the > > > parse > > > tree are unpickled and the html are generated. Usually on my laptop > the > > > first > > > pass if 4-5 time faster than the second one. So I would more expect > > > something > > > closer to 6/5N... > > > > > > Florent > > > One little rant: > > if I wanna run "make ptest" and did not set MAKE="make -j??" and the doc > > needs to be rebuilt, then it's built serially. > > Ok ! The makefile should be fixed. > > > It also seems everything is rebuilt unconditionally, even if nothing was > > modified, isn't it? That is much more worrying... Couldn't this be > avoided ? > > Ok ! Now the reference manual is splitted in several sections. If you only > want to recompile one section eg: combinat, just ask for > > sage -docbuild reference/combinat html > > and then only combinat is rebuild. I think that should solve you > problem, doesn't it ? Now if you ask > > sage -docbuild reference html > > Then sage will try to rebuild all sections. If only no file changed, then > the > recompilation of all the sections are indeed launched but then sphinx > realize > that no files changed and do nothing. Only launching the section for which > one > file changed is more difficult since the association .py -> section is > only > known by the inclusion in the .rst (usually index.rst) files that is > Sphinx > itself unless you want to rewrite the parsing of those files. > > Cheers, > > My problem today was really about issuing "make ptest" and having to wait for 5 or 10 minutes (when I set MAKE="make -j8") before the tests really began although no files were changed, and so there was no reason for Sphinx to do anything at all.
Or to e more precise, I modified some files, launched "make ptest", waited for the compilation of docs to be finished and the tests to begin, interrupted the tests, relaunched "make ptest" and then I saw the docs compiling again which I did not really enjoy. But I did not really pay attention to what was excatly going on, it just felt like it was slower than before. > Florent > -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
