I was disappointed to wake up to this from Tim O'Reilly, commenting on why GPLv3 was right to abandon the attempt to rein in free-riding SaaS companies on open source software:
Web-delivered applications are just too important to too many people for the horse to be taken back to the barn. It would have been a death blow for GPLv3, making it impossible to adopt.
I'm actually not sure where Tim falls on this. When I initially commented on his blog, I read the above to suggest that a) the web is different and b) SaaS companies need not abide the same rules as non-SaaS companies vis-a-vis open source.
Reading his comments again, it feels like he's actually just saying that GPLv3 would have died in the face of opposition from Google et al. had the Free Software Foundation pushed the issue. That's certainly what Eben said at OSBC and I can appreciate the quandary.
But it's absolutely not true (if Tim is suggesting this) that the web is different and would be stifled by actually having to contribute back the way everyone else on the planet must/can/should. There is no reason that Yahoo! should get a free pass while Citigroup does not. If you benefit from open source, you should contribute back, and the license should enforce this. The web is not going to go out of business if it is made to replenish the ecosystem that it gladly pilfers from today.
Tim is right, though, that the current crop of 20th Century open source licenses need a serious upgrade to be relevant for 21st Century software. I've argued this before. Of the current OSI-approved open source licenses, only Larry Rosen's Open Software License even seems to recognize that the Internet exists. This is highly unfortunate, as a great deal of software will be delivered over the web in the future.
To my mind, this is OSI's biggest role in the near future: charting the course into the web world in which we increasingly live. It is the OSI's lack of action on the web that has left us with attribution licenses and other things that are designed to plug this gaping hole in open source. We need to do better.
The web is not different. The web would not be crippled by having to live by the same open source rules as everyone else. GPLv3 would not have dealt a deathblow to SaaS or web applications. Not at all. A new license is forward-looking. Google wouldn't have to release its existing infrastructure had GPLv3 included SaaS provisions. (That said, I can see upgrade problems arising from this....)
Regardless, GPLv3 doesn't have such provisions. There's no point in complaining about what might have been. The real debate should be over what we do now: we need a new crop of licenses to help make open source relevant for the web. It's actually surprising that we have almost nothing so far....