----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: http://git.reviewboard.kde.org/r/107276/#review21816 -----------------------------------------------------------
ksmserver/server.cpp <http://git.reviewboard.kde.org/r/107276/#comment16844> why not check for both program and filename? that should then catch 1, 2 and 4, no? excludeApps could also be pre-processed to include both long paths and filenames which would then allow catching all 4 variations. probably the reason this was written to only catch 1 and 4, however, was in case there were binaries of the same name in different paths that should be treated differently (allowing differentiation by full path).. Lubos will certainly have more insight on this, however. - Aaron J. Seigo On Nov. 10, 2012, 12:45 p.m., Jekyll Wu wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > http://git.reviewboard.kde.org/r/107276/ > ----------------------------------------------------------- > > (Updated Nov. 10, 2012, 12:45 p.m.) > > > Review request for kdelibs, Plasma and Luboš Luňák. > > > Description > ------- > > It is easy to understand why the existing code (usually) fails: > > * Users are most likely to just specify short names, like > "dolphin,gwenview,okular,rekonq", instead of > "/usr/bin/konsole,/usr/bin/gwenview,/usr/bin/okular,/usr/bin/rekonq" > > * When ksmserver saves the session, it usually gets the full names, like > "/usr/bin/dolphin", unless you have started that dolphin instance by typing > "dolphin" exactly in a shell. > > > So there are four possible combinations : > > 1). config uses short name, runtime gets short name (this guy starts > everything from konsole, never using kio/krun) > 2). config uses short name, runtime gets long name (I think this is the > most common one) > 3). config uses long name, runtime gets short name > 4). config uses long name, runtime gets long name (I guess some users use > this combination because they find only that way works after trying various > workaround...) > > The existing code works with 1) and 4), this patch works with 1) and 2) . I > don't know whether it make senses to support all combinations > > > > > > This addresses bug 242760. > http://bugs.kde.org/show_bug.cgi?id=242760 > > > Diffs > ----- > > ksmserver/server.cpp eb3ac18 > > Diff: http://git.reviewboard.kde.org/r/107276/diff/ > > > Testing > ------- > > > Thanks, > > Jekyll Wu > >
