It’s been a bit of a ‘bitty’ start to the week after the Easter break, I of course had a ton of email / forum posts to catch up on to begin with. One nice outcome of that was seeing what is AFAIK the first dedicated OSX-only application using Ogre, Live Interior 3D. Looks pretty darned nice too. There is also the Summer of Code which came to a head yesterday - we had a last-minute conflict come up with Mozilla which had to be sorted out. That’s all resolved now, and Google should be announcing the successful students late today PDT, which is unfortunately the middle of the night here so I won’t post about it on the Ogre site until tomorrow now.
It’s also our 5th wedding anniversary tomorrow and as a typical male, today was the day to rush about sorting something out. As well as the day itself, we’d been thinking of taking a long weekend in France at the end of next week but hadn’t organised anything yet, so that was one of my jobs today.
I had a couple of items regarding the LiSPSM shadow projection on my TODO too, one of which is really yanking my chain. The focus region just didn’t seem optimal in some cases - turns out it’s mostly right in fact due to the way you have to calculate it for directional lights, but in spending lots of time on that I discovered what I think might be a clipping bug that’s been avoided by an overly generous focus region in some cases. Whenever I try things to optimise more it blows up and although I thought I was doing something wrong earlier, I think there might be an underlying issue there. It’s really tough to visualise & trace though, and it’s eating a ton of time. Ugh - I’m going to have to keep at it even though it feels like I’m getting nowhere.
Gah no, I think everything is ok after all. I just forgot to adjust the ‘nopt’ value significantly enough for a directional light so it looked worse than it should have. Damn, what a load of wasted time.