Hi Naohiro, I see you have tagged the disk_cmds. Have you had any thoughts on the netgear development? As a guess, I think netgear will not divert much development effort into making the hfs+ journal feature (assuming they have done it and done it well) into the dev kernel, so it is up to you. So your tasks are slightly modified: 1) it would be good for you to find out what netgear's intention is 2) regardless of their intention, it would still be good for you to review their change, break it down into reviewable/commentable pieces, etc. I suspect it would take a week or two to make them buildable against current kernel, since 25 minor kernel releases is not a trivial amount of api change. Depending on 1/2, you might want to either spend more time on the netgear change, or on the disk_cmds instead. Communicating with hardware vendors and reviewing/porting other's unfinished(?) work isn't as glamourous as doing new exciting and ground-breaking work, but those need to be done and it is just as well for you to be doing those. Hin-Tak --- On Tue, 7/6/11, Hin-Tak Leung <hintak_leung@xxxxxxxxxxx> wrote: > --- On Tue, 7/6/11, Naohiro Aota > <naota@xxxxxxxxx> > wrote: > > > Hin-Tak Leung <hintak_leung@xxxxxxxxxxx> > > writes: > > > > > Naohiro Aota wrote: > > >> I've setup git repository on my github page > > >> https://github.com/naota/diskdev_cmds > > >> As you see it, it contains each versions' > history > > :-) > > > > > > Well-done :-). > > > > Added new branch "gentoo" > > https://github.com/naota/diskdev_cmds/tree/gentoo > > > > I've applied the Gentoo patch to the most new source > and > > I'm working to > > fix so many build errors. :D > > Good. I think it is also important to tag a version of what > it was originally, as it was. It is a non-trivial problem, > and not necessarily solvable by one person - but one hope > not to introduce new mistakes :-). Hence the need to keep > track of what's original and what's changed, and why. > > > > what do you mean by "cleaning the journal but > > writing"? > > > > If there is unfinished journal (cause by sudden > > power-cutdown or crash), > > ntfs-3g replay the journal. But it doesn't record > journal > > when it write > > to disks ... > > Oh - but the last statement cannot be correct - if it were > correct, new write's aren't journalled/consistent? > > > BTW, you see this mail? > > http://www.spinics.net/lists/linux-fsdevel/msg45627.html > > > > Netgear seems to have a patch for Linux kernel to > implement > > HFSPlus > > journal feature ... hm, I'm really suprised to read > this > > ... What should > > I do now ..? :S > > The objective of them is different from yours - they want > it to work for a particular (old) kernel version and one > that's quite substantially different from main-line. You > want your work to eventually ends up in mainline. Their work > makes your life a bit easier (and probably quite a lot > easier) but you still have work to do: > > 1) thoroughly understand and review what they are doing > 2) fix any bugs either in code or in documentation (or even > write newer/more detailed ones > 3) port the work forward to newer/current kernel. > > Netgear is a commercial entity so their objective (making > it work for one old version and one old device/hardware > combo) is somewhat different from yours (make it work well > for current kernel and all hardware combo). > -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
