From: Sven P. <sve...@ae...> - 2008-10-09 10:07:17
|
On Tuesday 23 October 2007 09:36:17 Henrik R. Nagel wrote: > Steve Baker wrote: > > This is a very serious error - we rely on the application calling that > > function before getting started so we can initialise our internal > > state. Without glutInit - the program could easily crash and burn in > > all sorts of exciting and unexpected ways later on. We are right to > > treat it as an error and you should be more interested in fixing your > > real problem than in bitching to us! > > I strongly dislike your attitude. You must have a terrible personality. Well, as you have noticed, Steve likes to answer in a very direct way, ;-) but nevertheless, freeglut is behaving correctly here: http://opengl.org/documentation/specs/glut/spec3/node9.html The very first paragraph clearly states that calling glutStrokeLength without having called glutInit exactly once before is considered illegal. It might be the case that the original GLUT didn't check this constraint for all API entries, but this is not the point here. And performance is not an issue at all, performing this check is simply testing a single boolean value, which should cost only a few CPU cycles. > The following is a quote from FLTK's manual: > [...] > Notice the "Don't call glutInit()" statement. The author of the FLTK manual has obviously ignored the GLUT spec and relies on the (questionable) behavior of a specific implementation, so this part of the FLTK manual is bogus. > Perhaps you could talk directly to the people behind the otherwise > excellent FLTK about how you two groups can make your software > compatible with each other. I can't be the only one that uses both > software packages and I certainly won't be the last. Just out of curiosity: Why do you want to combine FLTK and GLUT? Cheers, S. P.S.: I've just seen that this is a very old message, but I'm just digging through my huge mail backlog... Any news on this topic? |