[00:19:08] --- Russ has become available [01:22:25] --- haba has become available [01:55:16] --- Russ has left: Disconnected [02:42:07] --- Simon Wilkinson has become available [02:59:33] --- pod has become available [03:32:24] --- pod has left [04:24:45] --- haba has left [05:39:38] --- pod has become available [05:57:38] --- jaltman/FrogsLeap has left: Disconnected [05:57:47] --- jaltman/FrogsLeap has become available [07:27:18] --- deason has become available [07:45:51] --- reuteras has left [08:44:14] --- jaltman/FrogsLeap has left: Disconnected [10:58:51] --- jaltman/FrogsLeap has become available [11:38:23] --- jaltman/FrogsLeap has left: Disconnected [14:02:14] --- jaltman/FrogsLeap has become available [14:22:13] --- mfelliott has left [15:41:06] --- meffie has left [16:32:19] --- deason has left [17:48:27] --- rra has left: Disconnected [18:06:37] --- Russ has become available [20:40:58] myid.net is no longer an acceptable OpenID provider for Gerrit? [20:41:36] why? [20:46:56] No idea, but Gerrit won't let me log in any more. [20:47:13] "Provider is not allowed." [20:49:35] I suppose I should mail Simon. [20:50:51] Anyway, the issue I was going to comment on was 3772. I don't think there's a problem with the new implementation being under a different license there. PC clone manufacturers did clean-room reimplementations of complex and secretive code like whole computer BIOSes, where the interface layer was quite complicated and the code was sophisticated in what it had to do. This is a dead-simple interface and the code is just not that conceptually complex. I have a hard time believing there's sufficient intellectual property in the original implementation to attach to someone just by reading it. Nothing here is going to rise beyond "normal competency in the field," to borrow language from patents. [20:51:09] Also, the clean-room thing was done in anticipation of lawsuits, since the clone makers were going to get sued. I have a hard time imagining IBM is ever going to think it has grounds to sue someone for changing free software licenses for a reimplementation of this simple of an interface. [20:52:03] The risk seems really low here. I've changed licenses during that level of a reimplementation before, and I don't think full clean-room discipline is required. That's the kind of thing you do when you're going to get sued no matter what you do and you need to make sure that you don't hand your legal opponent any extra ammunition, however slight, and that's not the situation we're in. [20:53:16] Clean room implementations were required when the interfaces were not public. That isn't the case here. [20:53:58] And I've also seen other software relicensed with that level of reimplementation. Angband is a significant example, where they removed the non-commercial usage clause by rewriting all the code from people who they couldn't contact, and they didn't do clean-room for that. [20:54:08] Or at least I'm fairly sure they didn't. [20:54:39] openafs has done the same thing in order to remove the apple licensing [21:40:30] clean room implementations are fun [21:40:38] (working in an actual physical clean room) [21:49:15] --- jaltman/FrogsLeap has left: Disconnected [22:11:06] --- reuteras has become available [23:12:13] --- Russ has left: Disconnected