So no not related to idtech as such unless you count compilers
been working on fixing a rather bizarre bug in the TDM gcc compiler, first i found out that the dev used CreateMutex to lock the same named mutex in two places which ofc wont work as you would get the ERROR_ALLREADY_EXIST flag and the mutex would deadlock, so i rewrote that part but ran into a new bug which only seems to affect glib2 based sources like gdk-pixbuf2. turns out glib2 now also uses a mutex locking mechanism which collides with the one in TDM gcc basically deadlocking it again urgh... This should actually not be possible as far as i know as the TDM gcc code uses a named mutex with a unique identifier, so how it does this is beyond me unless they screwed up royally writing the locking code. so far all attempts at making it behave have met with failure and im at my wits end.
so if anyone wants to take a look feel free to contact me.
been working on fixing a rather bizarre bug in the TDM gcc compiler, first i found out that the dev used CreateMutex to lock the same named mutex in two places which ofc wont work as you would get the ERROR_ALLREADY_EXIST flag and the mutex would deadlock, so i rewrote that part but ran into a new bug which only seems to affect glib2 based sources like gdk-pixbuf2. turns out glib2 now also uses a mutex locking mechanism which collides with the one in TDM gcc basically deadlocking it again urgh... This should actually not be possible as far as i know as the TDM gcc code uses a named mutex with a unique identifier, so how it does this is beyond me unless they screwed up royally writing the locking code. so far all attempts at making it behave have met with failure and im at my wits end.
so if anyone wants to take a look feel free to contact me.