I’ve finally had time to get back to the XSI exporter over the last few days, and it’s shaping up nicely. I refactored it all today since now I’ve learned a lot about how XSI ‘thinks’ I could see much better ways to structure the code so I could add the features I wanted. The current version now exports meshes fully, dealing with multiple object / tree selections, automatically index-organises the vertex data for efficiency (ie grouping vertex data by triangle usage), merges XSI mesh objects into single SubMesh instances where the material is the same (you can turn this off if you want your sub-objects to remain separate for whatever reason, but it allows you to keep modelling using separate objects but have a more efficient runtime mesh if you use it), and I also now deal with polygon clusters, which is XSI’s way of allowing you to change materials in the middle of a mesh.
You heard right. We crossed the million ‘hits’ mark in less than 3 days. Yowsers. In terms of unique IPs we’ve had about 11,000 so far and it’s still going up. The mind boggles. I’m presently trying to clear a few patches and digging back into XSI. I spent a little time just getting familiar with it again, and fixed a small transform bug. Lots more to do though, but it’s been long overdue.
Wow, now we have full access to server logs (which we didn’t on Sourceforge), we finally know just how popular the site is, and it’s terrifying. The new server was overloaded really fast at first through sheer weight of demand, luckily we switched from apache to thttpd which lowered the memory requirements significantly, to the extent that we can actually run the site now (and it’s quite nippy). The scary thing is that our bandwidth is averaging a projected 90Gb/month at the moment, which is well above what we expected (and more than double what we’re paying for right now).
We have a new server from www.linode.com set up to run www.ogre3d.org now, and we’re just waiting for the DNS to update for it to be available. It appears I made a typing error in my first update, shame on me, resulting in the DNS records being wrong at the moment - please be patient while they update. Normal service will resume imminently.
Just as I sat down to do something useful today, I find someone has hacked the OGRE forums and deleted several thousand posts from the General Discussion and Recruitment forums. Words cannot adequately describe what I want to do to these people, but it involves inadequately sharpened implements and soft tissues.
Rant warning! Here’s the proof that democracy doesn’t work, and corporate greed is all that makes a difference in the world: Groklaw report on today’s EC decision I think most people knew that the EC, like any large organisation, is rife by bureaucrats and corrupution; but given the amount of high-level opposition to the CIID I thought this might be different. We even had the Legal Affairs Committee on our side.
That’s it. Stick a fork in it, because OGRE 1.0.0 is done and in the hands of the masses (washed or unwashed we don’t know, that’s the beauty of the Internet). Looking back, I’m quite amazed we got here in one piece. I certainly never imagined OGRE be anywhere near as successful as it’s been; I still remember bashing away at it’s early classes on a crappy old Pentium laptop with little hardware acceleration to speak of - I had to wait until I got to my GeForce1 desktop before I could test anything much 😀At the time I remember reading lots of code from other projects and examples (legally, I might add - I just won’t name any names here), and bitching about how difficult they were to read, and how unnecessary most of the illegibility and obsfucation was.
Well, a surprising amount of little things have come up which has meant the 1.0.0 final isn’t quite done yet, but it’s really, really close now. The SDK is working for both VC6 abd VC7.0 now, I just need to ask nfz to build one for VC7.1 for me, now the scripts are all done & tested. I have onr last patch to evaluate before we put the seal on it.
I’ve been trying to take it a little easier this last week since the RC1 release since I needed a bit of a break. I’ve largely succeeded, although there have been some bugs found in wider user testing of RC1 which have needed addressing; which was the whole point of doing a release candidate of course. As of right now all the serious ones have been fixed and things are looking pretty peachy for the final.