Some colleagues have advised me to lend my commentary regarding Jeff Atwood’s much-discussed piece entitled “Is Money Useless to Open Source Projects?” (or, as the Linux Hater put it, “You can’t even pay them to not suck“). I think I was supposed to argue the position that FFmpeg provides an efficient mechanism to convert raw currency into high quality multimedia source code. But I’m not sure I make that case or refute Jeff’s piece; in fact, I largely agree with his post. However, I can corroborate that FFmpeg does sometimes benefit from raw currency.

Let’s talk about the most visible example of the cash->code thesis as it relates to FFmpeg: Google’s Summer of Code sponsorship of FFmpeg. In 3 years, FFmpeg has earned 22 combined project slots; each slot represents a transfer for US$5000 from Google -> people involved in the project. Wow, this is the first time I have realized that Google has, to date, committed over 100 G’s to our silly little multimedia project. What has it gotten for that investment? Well, it got the foundation that made YouTube possible. What has FFmpeg been able to do with all that green? Out of 13 projects in 2006 and 2007, only 4 have currently reached acceptable status and are part of the mainline code tree.

Fortunately, this year, we have imposed much more stringent standards for the projects; students must complete their features to a level that they can be incorporated into the main tree by the end of the project. This week marks the end of GSoC 2008, so we will see how that turns out. If you have been following the FFmpeg mailing lists, you no doubt noticed a flurry of activity as students race to get all parts of their various projects approved for inclusion into the mainline tree.

Apart from GSoC, there are tales of different parts of FFmpeg development being actively funded by various organizations. The file dnxhdenc.c carries the following header:

 * VC3/DNxHD encoder
 * Copyright (c) 2007 Baptiste Coudurier 
 *
 * VC-3 encoder funded by the British Broadcasting Corporation

I hear rumors of assorted other pieces receiving formal funding but many details seem to be kept hush.

Jeff’s post has a lot of other interesting money-use cases. However, many of them seem to have the same root requirement: A central FFmpeg organization. We don’t have one. It would be nice to have one, and believe me, I’ve done research on the matter. There are basically 2 options:

1. Set up our own 501(c)(3) non-profit organization
2. Join up with another non-profit organization that specializes in being an umbrella organization, thus bestowing individual open source software projects with the benefits of 501(c)(3) status without all the administrative overhead

It was last year about this time that I was very seriously looking into starting a nonprofit organization for the FFmpeg project. First, I purchased Nonprofit Kit For Dummies. This is a very informative and thorough book. At the same time, I attended the Craigslist Foundation Nonprofit Boot Camp. That was pretty useless all around. Both the book and the boot camp made me recognize that I was not up to the task of establishing my own nonprofit organization. Essentially, setting up a 501(c) organization is the single most complex thing you can possibly endeavor to do under the U.S. tax code. This makes sense when you think about it — if you’re going to get away with not paying taxes on income, the government is going to make you toil for the privilege.

So the next option is fiscal sponsorship (the umbrella organization thing). I have tried to engage several such organizations about sponsorship. Short story: I can’t get any of them to even talk to us. I think we might have a reputation. Something to do with our cavalier attitude towards intellectual property matters.

Another problem is that a 501(c)(3) organization needs something called a “board” which is a group of people in charge of the racket. While I might be able to do the legwork for setting up the org (and I am probably the only person who would take the initiative on the matter), I fear that it might be seen as a conflict of interest, due to my day job, for me to be directly involved in its operation. That means I would need to find at least 3 people as crazy as I am who would serve on the board. For bureaucratic purposes, it would probably be easiest if they were American citizens, though that’s not a hard and fast requirement.

I’ve always thought that nonprofit status could help FFmpeg leverage money much more effectively. But for now, that remains an elusive dream.