Pavel Sanda wrote: > Peter Kuemmel wrote: > > > > Until nobody proves me wrong I assume QProcess is not usable > > > > synchronously. > > > > > > Well I've used these signals successfully both on windows and Linux so I > > > am inclined to think that there might be some problems in your code.Your > > > excessive use of processEvents might be the root of all the perceived > > > problem. I don't really understand why you need that but I unfortunately > > > don't have the time to do much today :-( > > > > Yes, the signals work, but not the functions to force synchronous > > control-flow. > > > > Attached a patch which hopefully solves the problems.
i just tested this patch in QProcess branch. it seems to work as far as the usual jobs executing is concerned. there are some small glitches - eg the splash window appears when loading file but not when loading is from svn etc. i think that this splash window makes more trouble than it resolves. afer all the long typeset jobs are supposed to be done in the background one day, so i would just drop the splash window. however there are more problematic things - i wonder about all those waitWhile(...). for example this following line if (!d.waitWhile(SystemcallPrivate::Running, 180000)) /3000 means than any typeseting or external job wont work correctly if it does not finish in this limited time? or start wihin 3s? do i understand correctly that this is just ugly workaround for the non-working qt signals? pavel
