Mario Blättermann ha escrit: > Seems the project itself is dead.
Not at all. It's alive and well. The maintainer is alive as well (as far as I can judge), but as a human being, he can from time to time be occupied with some real life chores which (quite unfortunately, agreed) can distract him from attending his mails, for which he humbly begs your pardon. Now, to the matter: > The latest stable version has been released in 2011. Indeed. There are plenty of changes in my local tree which I'm going to release before the Christmas, but all of them are mostly bugfixes. All cpio formats are pretty well established and any global change will hurt both backwards compatibility and compatibility with another implementations of cpio. In that sense, cpio can certainly be considered a legacy project (that holds true for any implementation of CPIO, not only GNU). I often receive various propositions to extend CPIO headers to accomodate for 64-bit uids etc., but there's little if anything I can do about it: expanding existing formats is impossible due to the reasons explained above, and designing new formats is senseless, because it is already done by PAX. > The only possible way to keep cpio alive would be to fork it as a Red Hat > project, as long as our package management still needs it. That depends pretty well on what "alive" means for you. For what it's worth, so far our cooperation with RH maintainers was quite positive -- correct me if I'm wrong. As about further extending cpio, I have already explained my views. Regards, Sergey
