Mardi, le 24 août 2010, à 20:02, Maciej Mrozowski a écrit: > On Tuesday 24 of August 2010 02:34:05 Friedrich W. H. Kossebau wrote: > > Mardi, le 24 août 2010, à 01:24, Yury G. Kudryashov a écrit: > > > BTW, what about adding one more parameter to m_l_f that tells whether > > > dependency is a runtime dep (default to FALSE)? Or even add another > > > macro because "required" flag MUST be FALSE for runtime dependencies? > > > Then we'll be able to say: > > > == THE FOLLOWING RUNTIME DEPENDENCIES ARE MISSING ON YOUR SYSTEM == > > > > That might be indeed most useful and nice to have, I agree (non-packager, > > but self-compiler :) ). > > Otherwise listing runtime-only dependencies in README file is just as good > (what said kajonggh does for instance). > > It's even better than CMake checks in case all runtime-only dependencies > are met (so no missing deps are reported) - as packager is not aware of > them.
True, good hint. So if packagers do read README file, than I will add one, too. Hm, should then be kdebase/runtime/README, or can I ever expect some packager to read kdebase/runtime/kioslave/network/README? Might be good if this could be standardized. I still like Yury's proposal for that, as it integrates into the log creation, so all dependencies can be seen at one place, no need to search somewhere else. I guess Yury also had the second listing in mind, == THE FOLLOWING RUNTIME DEPENDENCIES ARE AVAILABLE ON YOUR SYSTEM == :) Cheers Friedrich -- KDE Okteta - a simple hex editor - http://utils.kde.org/projects/okteta _______________________________________________ Kde-buildsystem mailing list Kde-buildsystem@kde.org https://mail.kde.org/mailman/listinfo/kde-buildsystem
