On Mon, 7 Jul 2003, Maarten De Boer wrote: > I see. But I suppose that this will change at some point. I am sure > that the OpenAFS developers are not going to stick with 2.4 when the > rest of mankind moves to 2.6
Probably not. Linux 2.6 doesn't appear likely to be OpenAFS friendly, you can read linux-kernel (search for process authentication groups) if you want details. Regardless, at some point we will need to support it. > That's a bit harsh, isn't it? I'd say that the right choice is the > third choice: fix it. Has anybody investigated this matter into more > detail? Is Robert Love (preemptible patch maintainer) aware of the > problem? Is the problem with OpenAFS or with the preemptible patch? > How come a patch that does not affect any of the OpenAFS code breaks > OpenAFS? Interesting questions I'd say... And even more because they > will have to be answered for the > 2.4 future. Sure. But people aren't exactly falling over themselves to do the work, and the politics involved in making the client work completely (no loss in functionality) are unpleasant. _______________________________________________ OpenAFS-info mailing list [EMAIL PROTECTED] https://lists.openafs.org/mailman/listinfo/openafs-info
