On Tue, 2011-03-22 at 01:48 -0500, Norbert Thiebaud wrote: > I think the issue here is that if you make the top level default > target (all) mean 'build-only' you decrease the odd of the tests being > covered.
As a data point, post-merge of new build system stuff, passive component registration and so forth a lot of the new tests are disabled at the moment seeing as they need a bit of love to use the new passive reg, on turning them on one at a time I see that... a) smoketest merge went wrong, now fixed hopefully b) i18npool merge went very wrong, regression test detected the failure, now fixed hopefully c) starmath had a (new?) bug which the regression test picked up, now fixed. Which argues to me that a test not run by default isn't worth a whole pile if when I enable three tests all three tests immediately detect regressions. I've sympathy for the "linking the sw tests is *so* slow, can I skip the unit test" argument, I wish I could have faith that people would run a "make check", but the evidence so far is that few run the current "make check" target :-( C. _______________________________________________ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice