Quoth Alexandre Oliva: > Taking Free sources and applying patches is a better way to go anyway, > why would you want to steer people towards the non-Free Software in > upstream just to (hopefully) throw it all away afterwards? :-) > > Most of the time, patches should apply just fine on top of Linux-libre, > but it would be nice to check that they don't add back non-Free stuff, > running deblob-check on them, like we do on Freed-ora. These runs of > deblob-check are far less demanding of resources, for they're only > heuristic, without bringing in all the knowledge (and regexps) about the > false positives and known blobs in Linux.
OK, I've updated the gentoo ebuilds so that it's now trivial to add any patchsets we want. So there's now a linux-libre version of the vanilla, default gentoo patchset, and hardened gentoo patchset kernels available (all attached to the aforementioned bug). > > Is my experience of the deblob script being very slow and heavy > > common? Is there much we could do about it? > > Unfortunately, it is normal. There's a long post of mine from a few > days ago in which I delve into some experiments I've been doing to try > to address this very problem. Look for “flex” and “perl” in > http://www.fsfla.org/pipermail/linux-libre/2009-May/000610.html That looks like great work. I look forward to seeing it progress. Take care, Nick White -- GPG : 0x04E4653F / 9732 D7C7 A441 D79E FDF0 94F6 1F48 5674 04E4 653F IM /OTR : [email protected] / 7E3C82CC D6AB2CEA E8000300 E429A122 D984111B SIP/ZRTP: [email protected] WWW : http://www.njw.me.uk
pgp7ubAuRbXKY.pgp
Description: PGP signature
_______________________________________________ linux-libre mailing list [email protected] http://www.fsfla.org/cgi-bin/mailman/listinfo/linux-libre
