Author archives: Mike

RSS feed of Mike

A few tiny steps forward in OpenGLContext (So much more relaxing than Cinemon...)


Added code to cycle through defined viewpoints in a VRML97 file, to adopt the specified fieldOfView of the viewpoint, and to respect the "jump" flag in deciding whether to update position and rotation.

I think with that I'm basically done tinkering with the Viewpoint support. It's by no means a full VRML97 viewpoint implementation, but ...

Continue reading

Hard limits and heuristics (Or, my day in frustrating revelations...)


Have been doing all of our testing/development work with a partially-specified hierarchy provided by ACI. The change last week, which I thought of as creating an over-specified hierarchy in the demo/testing code had actually come darn close to simulating precisely a normal hierarchy. The extreme slowdown turns out to be a real effect that we ...

Continue reading

Long edit-test-debug cycles really are productivity-sapping (While I wait for the simulation to get to the point where it can be run again...)


Developing Cinemon got really slow sometime late last week, mostly because of the more aggressive scanning parameters and the larger physical map. On the real system these don't have that much of an effect, but the demo/dev system is just crawling. Instead of a 20s cycle it's become a 3 or 4 minute cycle. That ...

Continue reading

OpenGLContext grows basic Viewpoint support (Really, really basic...)


Fixed yet another really-really-really core system bug. Hard to believe such things could be in there that deep and never cause an error until now. This time it was in the "find" method that searches a scenegraph for out-of-order nodes. It was incorrectly generating it's flattened "todo" list, so was returning entirely invalid node-paths.

At ...

Continue reading

Old VRML never dies (It just gets shunted off to directories called oldstuff/zzserver/vrml)


Was hunting about for Eric and discovered an old link from one of the VRML guys lamenting the loss of my old VRML code collection from the VRML community (well, more like the community's implosion and everyone considering it dead). So I started hunting about looking for those old worlds. They were pretty rough. Still, ...

Continue reading

Or maybe I'll just blog more (And try to type quietly)


People I think about a lot who have gone missing-in-action from my life (mostly due to my being an idiot and not keeping up on correspondence):

Kim — She's off teaching at a University on the prairies these days
Alexei — He's off in China.
Diana — She's married, in Ottawa, with 4 children (3 ...

Continue reading

More thoughts on OpenGLContext (Or, I don't think I'm going to sleep, so may as well blog...)


Need to decide what I'm going to do about making an end-programmer-friendly UI system available. Do I go with something like PyUI, which is entirely within the OpenGL window? or do I try to make some cross-GUI-system minimal subset for 3D-oriented code? or do I write individual tools for use with a single GUI library ...

Continue reading

She never brings them home for *me* (Rosey & co go out clubbing, so the day ends early...)


Rosey and Erin (friend from St. Mary's back when we were in high school (and a rather stunning woman even back then (she's the one on the right))) took one of their friends out clubbing as part of another friend's wedding festivities. They're all staying over, so I guess I'll call it a night so ...

Continue reading

Napping isn't very productive... (Or, my day in OpenGLContext)


Spent what (small) fraction of the day I was awake working on OpenGLContext, and then watching a movie (Switch). The OpenGLContext work was fairly trivial stuff, just tightening things up in preparation for the 2.0.0 final release. This included rewriting the textobject.py demo to function as a chooser for the default fonts to use for ...

Continue reading

It's annoying when I can't seem to think straight (PyGame font should be so simple... and yet...)


Been working on fixing the PyGame font renderer in OpenGLContext. This was broken by a fix to the bitmap renderer a month ago. Problem is, the original code had this wonky set of transformations that look like they should effectively cancel each other out, but which worked perfectly well with the (wonky) previous rendering scheme ...

Continue reading