Norman Vine wrote: > Erik Hofman writes: > >>I noticed that in SimGear/timing/timestamp.cxx for Windows the seconds >>are intitalized to zero while under Unix it initilized properly: >> > > You expect Windows and Unix to do things similarly ?
I expected FlighGear/SimGear would use the same variables similarly. > >>Isn't this a problem? >> > > Only after running for 45 days or so :-) > ie. > secs_min*mins_hour*hours_day*1000*46 = > 32 bit unsigned integer overflow > I'll let others comment on the probability of Win32 staying up that long > :-) No comment. > note that when > #ifdef WIN32 > the seconds variable is never used, I guess this could be better > documented, however I haven't heard of it causing any problems > and it has been this way for a while ! Yep, I only noticed because I was changing the code to use the SDL library (ionly for internal use for now) adn then I discovered this difference. > > In fact since october 1998 when I changed it to use > the Win32 API timeGetTime() instead of the <Unix'y> ftime() > which not so surprisingly would only allow a max of 18.3 fps ! > > FYI that's also when I got my first 3D accelerated GFX card > and I was %$#%# that I could still only got 18.3 fps max and > then I discovered the ftme() <-> bios timer connection in Windows :-) > > > Since then, thanks to FGFS I have learned alot about WIN32, > I was primarily an embedded systems programmer, > and if I was todo this again I think I would just add <skipped a lot of rumble> > Cheers > > Norman Thanks. I'll need it now ;-) Erik _______________________________________________ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
