SIGGRAPH and Open Source

#1

dunno if cross-posting between forums is bad ettiquete here on cgtalk :shrug:, i posted this article in cg news and the moment i pressed submit i remembered this forum. i think this is pertinent here:

http://it.slashdot.org/it/04/08/22/1721254.shtml?tid=152&tid=117&tid=9

Quote:
The knee-jerk reaction that may be some executives’ first objection: our code is a strategic advantage, giving it away would be throwing away money. If we can do hair and our competitors can’t, we’ll make better films then they can (and, if it’s a visual effects studio, we’ll win contracts based on that unique ability).

Bull honkus. If your competitors need hair, they’ll write hair software, no problem. Another quote from the Pixar RenderMan user’s group, this one by a RenderMan developer (paraphrased): “this is based on the subsurface scattering papers from a couple years ago. Everybody does this, based on those papers.” Nope, I don’t see strategic advantage there: I see waste.

#2

great post. I completely agree with this idea, especially being an advocate for open source software.
This industry especially would benefit from even initial open source projects.
Long has many have thought of producing an animated short using entirely open source software. I think this would help prove what can be achieved when everybody works together.

my 2 cents.

-M@

#3

requiring published papers to come with at least somewhat complete pseudo code would probably improve the quality of the papers that overall has been dwindling rather badly in Siggraph for the last 5 years or so.

having said that : everyone writing similar proprietary software is far from waste. first of : everyone has a different approach to the problem, and second, it -is- a strategic advantage if you have superior tools. that everyone “does it based on the same paper” doesn’t mean everyone does the same thing or has even remotely the same needs for a given system.

#4

I concur. Many communication protocols are freely available, but the implementations of a given protocol can vastly differ. Look at the TCP/IP stack first developed by Microsoft for NT 3.51 compared to the TCP/IP stack in BSD. They both achieve the same end, but at what cost?

Sure, a company could write their own hair plugin, but how sloppy is their code? How ineficient are their data structures? IMHO software development is a trade, and a superior developer’s product will have an advantage in efficiency/speed or ease of use.

#5

The main downside I can see to this kind of thing (which, in principle, I really agree with) is that a lot of programmers will lose their jobs…

Why should a company employ X programmers to develop a piece of software or plugin when they can get it from another company…

#6

Well, I’m a full-time paid programmer who works exclusively on GNU’d code. The thing about Open Source code is that it works great for the explicit purpose of the developer. If your company wants 90% of the GNU code, then they’ll need a programmer who can understand other people’s code (usually without any documentation, just comments in the header files if lucky) and make it work how the boss sees fit.

Now, to bring the thread a little back on course, there are many late nights when I’m reading mathematical documents and am just praying for some nice pseudo code to make it make sense.

#7

From a programmer’s point of view, I totally agree with you - I’ve had the same experience with reading docs and really wanting some code that surely someone has already written.

#8

i can only see this open source approach working for applications that don’t give you a technology advantage you could sell in a pitch to a client - so it looks pretty bad for all the inhouse dynamics, fluid, hair rendering, water simulation, you-get-what-i’m-talking-about stuff
but I don’t see any reason why it could’nt work with management und utility software: file renamer, sequence viewer, chaching systems, file formats, converter etc. - every vfx house has loads of this stuff developed inhouse… sharing this kind of code would only bring benefits to your company.

#9

even something as basic as an image viewer has very specific requirements depending on who’s using it. file name padding scripts on the other hand aren’t considered “tools” : they are something every user who pretends to be computer literate should be capable of writing :wink:

#10

Open source works very well for certain scenarios but I’m not convinced it works well for the types of software that are developed in-house by larger animation studios. I don’t feel too strongly about it, by the way, but here are some arguments against it (all from the point of view of the studio that would be considering releasing their software) that I think have merit:

  • First, the argument offered by the original poster (“this is based on the subsurface scattering papers from a couple years ago. Everybody does this, based on those papers”) misses the point that the difficult part of large-scale animation is not the low-level algorithm but the high-level workflow and organization of work. To the extent that technology offers a competitive advantage to one company over another, it’s the latter that confers that advantage, and usually in-house software is inextricably tied with that, and could give away secrets, not technical, but business process. (Obviously this information drifts around as people change companies, but that’s a slower process than people often seem to think it is.)

  • Second, even if the competitive advantage of differences in workflow were illusory, that in-house software is so deeply tied with both workflow and the balance of technical vs. artistic skills that already exist in a company makes it questionable whether that software would be very useful to parties outside. Years ago I worked in a group that maintained Disney’s 2D ink-and-paint system called CAPS, and the joke there was that instead of keeping it secret we should give it away to our competitors because the work necessary to integrate it into their facilities would drive them out of business.

  • Third, releasing in-house software as open source would introduce a difficult issue of whose bug fixes get priority for in-house developer time. For example – a severe bug gets identified that all the outside users hit but that isn’t an issue due to the workflow differences of the studio that released the software. This would put that studio into the position of either having a PR problem in the animation world or sacrificing the priorities of their own productions.

  • Fourth, if you look at all the successful open-source projects out there, they’re all built to stable requirements. The requirements of in-house software are not nearly so stable, it’s often rewritten, and leaving behind multiple old versions of the software in the hands of outsiders would do nothing to help the releasing studio do their work.

  • Fifth: To contradict the original poster, there really aren’t that many people who can come up with a fully production-usable hair implementation, for example, and even at the big-name companies these people are notoriously hard to identify and hire. This suggests that such technologies might be a useful competitive advantage even if all the theory about how to do the work is public.

Note: I am not involved in these kinds of strategy decisions for my employer, so I’m just laying out the arguments as I see them, not stating why any particular company hasn’t done any particular thing.

– Mark

#11

>CAPS … drive them out of business.

i liked the CHIP launch party though…

#12

I confess to not having followed some of this stuff in the years since… what’s CHIP?

– Mark

#13

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.