Autodesk Buys Softimage XSI

[anamous]

All of them (including Houdini) are at the end of their journey anyway. They are ~10 years old, and there are a number of cycles one software can go through reasonable. There was Softimage 3D| XSI... and there was Prisms | Houdini, then there will be another step of evolution. Autodesk is preparing itself for that step rather then investing in any of Softimage code.

Of course I would have a problem to prove anything I said here

While I agree with you on a general level, I think that the Houdini codebase is vastly different (conceptually and implementation-wise) from most other commercial products. AFAIK Houdini is a textbook implementation of C++ object orientation that even some of the fathers of C++ openly admired. And the internal representation of the node trees and the very clean transport protocol between nodes are - to a software engineer - things of beauty. You will find way, way more hackery and shortsightedness in the other commercial products - just take a look at their API's...

Of course at some point every codebase is "too old" and has to be renewed. I believe that Houdinis can take a few years more of abuse than its main competitors

cheers,

Abdelkareem

[digitallysane]

sry for twin posting, but can u explain this statement a little more ?

There are 2 sides of this:

First, there was, and is, continuous flak from the Houdini community towards SESI on the subject of better Houdini marketing and such, a topic I was pretty vocal many times. And while (like any responsible "internet user" ) I still have firm opinions on what SESI should or should not do, I always was of the opinion that what matters are the results. And the results are that on a marketplace like this, SESI is still in business while others are not. Which proves that there seems to be something right in their approach.

Second, I appreciate the fact that they were so long in this business as a small, privately owned company. Most or all of their competitors sold themselves to big shareholding corporations for the hope of getting access to bigger R & D and marketing resources. Well, for Alias it paid off (but they were big already) but others weren't so lucky. Softimage in particular, in the last 14 years, was basically a tool used by various huge corporations for some other purposes. First for Microsoft to force the market to switch to Windows NT for 3D animation (and indirectly CAD) and then for Avid to gain access to a next generation editing and finishing suite (Softimage DS) and, less importantly, to broaden its portfolio with 3D animation tools which could make it a more viable competitor to Discreet / Autodesk.

I liked XSI a lot some time ago. We own a license and for various jobs is very good and very fast. However, I began to dislike them quite a lot lately, for various reasons: too much (dishonest) marketing; too little R & D: Softimage brought almost nothing new to the world of 3D, except for good implementations of concepts copied from elsewhere (ICE for example, or Gator); also there were lots of "patches" which offered dubious functionality for the sake of looking good in a feature list (putting a compositing module based on outdated technology pulled from Avid's trashcan of 10 years ago); lots of dependencies from outside vendors because they were not able to develop their own: rendering, hair, dynamics; it was also pretty clear that their presence on the market was only because they were part of Avid, not because they were sustainable by themselves, and that they were doing basically *anything* to gain marketshare: you simply cannot sell a product like Foundation (500USD), do high-end R & D and still survive on your own money.

I was discussing many months ago with a colleague and I was telling him that Avid will get rid of Softimage at some moment and that I really don't understand why they aren't doing it already as Soft are for sure losing them money and are of no benefit to their core business. I was hoping that (like Quantel some years ago) it might become independent again. I still think that, considering the extremely low price for which it was sold, Softimage's management could have rised some money and buy the company themselves. Maybe if it weren't for the global crisis, that would have been possible. Microsoft bought them for 130 million, then sold them for 285 million (making a nice profit) and Avid sold them for 35 million?? As a side note, like it or not, it shows that M$ are the only ones who really know business from all those

So, to get back, I like that SESI are steadily going on, in their own somewhat strange way, so it seems again that they are doing something right by staying independent, with all those "limited" (but sure!) resources.

Dragos

[kumpa]

Thanks Dragos for sharing your thoughts, very interesting reading, i wonder how many software vendors actually exists as privately hold companies. I bet not many.

Kim Davidson, we love you and God bless you.

Edited October 24, 2008 by kumpa

[Wolfwood]

All of them (including Houdini) are at the end of their journey anyway. They are ~10 years old, and there are a number of cycles one software can go through reasonable. There was Softimage 3D| XSI... and there was Prisms | Houdini, then there will be another step of evolution. Autodesk is preparing itself for that step rather then investing in any of Softimage code.

Of course I would have a problem to prove anything I said here

I don't think rewrites are possible any more. A ground up redesign takes years, and while that product is being worked on the original one stagnates. The instant you stop working on application and go to "the next best thing" someone else will jump in and fill your place. Look at what happened when Apple took Shake and stopped working on it. The industry shrugged and started using Nuke.

These days you have to rewrite sections of your application in chunks. Ie, redo the interface but leave the geometry architecture alone. etc etc

[JDenker]

Autodesk - 3 | Others - 0 | Please insert more coins to play again!

I really dont like the way it goes. Sure Autodesk isnt bad at all, I guess they have way bigger interests in CG than Avid does but anyways,

people at SI put their heart, soul and of course really really hard work into XSI and tried to stand out as a small company with their always innovative tools (crosswalk,e.g.) against big Autodesk. & now... this is all over, who knows what the future will bring us but i really like SI|XSI and hope they dont put maya-xsi together in the far future.

lets hope the best.

Cheers

a sad Jenz!

[zasid]

the day is not far when we all will read.Auto desk buys Blender.

[ykcosmo]

maya would add the ice........maya would be node based 3d software.....-.-!

[old school]

Probably not though. Architectures too different. Houdini's architecture easily supports vops though (arguably similar to ice). Houdini has a file sop. They don't. Same for Copy sop, time sops, etc. There in lies the difference.

As for incremental enhancements of Houdini and Mantra's foundations, this happens somewhat frequently. Lest we all forget SOP unloading was a huge enhancement in h7 and that was introduced in an incremental update.

[edward]

I still think that, considering the extremely low price for which it was sold, Softimage's management could have rised some money and buy the company themselves. Maybe if it weren't for the global crisis, that would have been possible. Microsoft bought them for 130 million, then sold them for 285 million (making a nice profit) and Avid sold them for 35 million??

Interesting comment. My first thought was that 35 million seemed a bit high compared to the 57 million when Alias was bought from SGI in 2004. The fact that it was then sold a year later for 182 million to Autodesk is a different story.

[digitallysane]

Interesting comment. My first thought was that 35 million seemed a bit high compared to the 57 million when Alias was bought from SGI in 2004. The fact that it was then sold a year later for 182 million to Autodesk is a different story.

I think the situation was kinda similar. In both cases, the owner (Avid or SGI) was bleeding money and probably wanted to get rid of anything that kept them from focusing on their core business. So probably the 57 million for AW was low as well at that time.

And in both cases the buyer was someone nobody expected

Dragos

[lisux]

Probably not though. Architectures too different. Houdini's architecture easily supports vops though (arguably similar to ice). Houdini has a file sop. They don't. Same for Copy sop, time sops, etc. There in lies the difference.

As for incremental enhancements of Houdini and Mantra's foundations, this happens somewhat frequently. Lest we all forget SOP unloading was a huge enhancement in h7 and that was introduced in an incremental update.

[offtopic]

Please Jeff I think that a video explaining more about unloading will be great, is avery useful feature when scenes becomes big and many people don't know about it.

[/offtopic]

[kumpa]

hahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahaha

[Leon Amion]

hahahahahahahahahahahahahahahahahahahahahahahahahahahahaha

I deleted some of the haha