Date: Thu, 28 Sep 2000 13:21:14 +0000 (GMT) From: Terry Lambert <tlambert@primenet.com> To: geoffb@chuggalug.clues.com (Geoff Buckingham) Cc: freebsd-fs@FreeBSD.ORG Subject: Re: XFS Message-ID: <200009281321.GAA27231@usr05.primenet.com> In-Reply-To: <20000928130419.A2374@chuggalug.clues.com> from "Geoff Buckingham" at Sep 28, 2000 01:04:19 PM
next in thread | previous in thread | raw e-mail | index | archive | help
> Sorry if this is old hat but on reading the caveats for SGIs XFS beta release. > > http://oss.sgi.com/projects/xfs/beta_caveats.html > > It seems rather badly hindered by the linux kernely. > > Was a FreeBSD port written off because of the GPL? I have talked to SGI's chief scientist about the license, and he is rather adamant. This is not surprising, since I think it was Larry McVoy's and Jeremy Allison's GPL influence that got the ball rolling on them releasing anything at all. He also didn't seem to get that any improvements coming from the community would not be allowed to be integrated back into IRIX, unless all of IRIX were to become GPL'ed. Rather than mark him off as "clueless" (despite a "why can't FreeBSD go GPL?" comment), I rather think they don't believe that they will get improvements that they care about from the community; either this is hubris ("we're professionals, not hackers!"), a loss leader, or it's just a way to get into the press. Without a license change, XFS can never ship compiled into the FreeBSD GENERIC kernel by default, and it can therefore never really be used as a boot filesystem for anything important or useful, apart from individual's hobby machines or commercial machines which are never, ever sold to customers. I tried to convince him about the existance of serious FS researchers mostly doing their work on BSD, and that he was locking out a lot of contributions, not to mention improvements that could be rolled into IRIX, but he was pretty much just not interested. My main interest has been in the recovery characteristics following a failure, since the soft updates "come up and run the cylinder group cleanup in the background" can not be achieved without NVRAM, and then only in the case of 3 out of 5 types of failures (see the recent discussion on this topic on this list). So at this point, it's an interesting curiousity, but not worth any developement effort. Duplicating it under a different license might be worthwhile; doing so would certainly undercut the SGI folks, if the intent was to limit it to hobby use, in the first place. SGI looks pretty much ready to roll up the sidewalks, anyway, at this point; their only big market is a niche in movie special effects, and "The Matrix" and "Titanic" have shown that Open Source OSs are makinginroads there, too, so they are probably not long for this world. 8-(. Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-fs" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200009281321.GAA27231>