On Thu, 2007-07-19 at 12:07 +0100, MJ Ray wrote:
We also need to develop open service models which can be implemented in free software and are resistant against an attack of embrace-extend-extinguish. We have some of these, but we need more.
Indeed, particularly, being able to run a distributed service with similar levels of reliability as a centralised service.
I feel that placing restrictions on the output of software, as seen in Affero GPL, is an evolutionary dead-end and a way for FSF to defeat itself. The complications of GPLv3 are already a disturbance (it may be more lawyer-friendly, but it's not nice to even copyright-experienced hackers, let alone hackers who don't understand copyright yet) and that may be a threat to free software. I hope that Affero GPL gets delayed, at least long enough to allow developers to digest GPLv3 and to let FSF webmasters unbreak stet, or maybe forever.
Seems to me the current Affero draft is mostly a no-op anyway:
"You may make, run and propagate covered works that you do not convey, without conditions so long as your license otherwise remains in force" (sec. 2)
=> if I don't convey and have a license, I have no obligations.
"Mere interaction with a user through a computer network, with no transfer of a copy, is not conveying" (sec. 0)
=> network interaction isn't conveying.
"Notwithstanding any other provision of this License..." (start of sec. 13, Affero draft)
=> the Affero clause doesn't withstand 2 & 0, and they say network interaction doesn't imply obligation, therefore it's null?
It will be interesting to see how that actually gets resolved. I'm not sure how the Affero could be a simple extension to the GPLv3 without somehow overriding the basic promises made in sections 0 and 2, and the obvious connotations of unfreeness that brings.
Cheers,
Alex.