Clearly there were some design flaws and whatever happened that caused the huge delay (microsoft COM ??? or the whole Twister issue) you arrived late to the party with a half baked cake.
You should be super proud of what you achieved, amongst others XSI has been pivotal in the construction of companies like The Mill 3D department and the torrent of awards on the 10 year period has been a clear testimony of both talent and toolset, I can tell you that. thanks for the insight. Jordi Bares [email protected] On 8 Mar 2014, at 18:25, Luc-Eric Rousseau <[email protected]> wrote: > On Sat, Mar 8, 2014 at 12:05 PM, Eugen Sares <[email protected]> wrote: >> One cardinal mistake back then was keeping so many aspects of the >> development in house. Once a critical mass was reached, the burden became >> too big. >> An open SDK concept would have been needed - platform like. Unlocking as >> many doors as possible, let 3rd parties do the extensions. >> >> A question, Luc-Eric (no sarcasm, just interest!): >> Why was there even any abstraction layer (the SDK) introduced in the first >> place? Why not allow 3rd parties to hook in 'first-class'? > > In my personal opinion, we just didn't have the right people, and > nobody was thinking in those terms. The SDK was tacked-on and limited > in Softimage|3D as well (devkit, then saaphire, then GDK). We added > scripting in the design of XSI only in 1998, after Maya was released. > Even if we remove Maya from the picture, in 1997 it was already clear > that 3dsmax had a winner with its SDK-oriented design and third party > support. > > We had the right people when it came to keying, animation mixing, > rendering, general workflows. > > In the team's defense, though, developing Twister and then killing > that off made everyone waste a lot of time they might have used for > other things. There seemed to have been the magical belief that using > the microsoft APIs would give some extensibility "for free".
