Why does the world need yet another kernel patchset? There are a couple of
reasons, basically. The first is that I have a bunch of free time coming up,
and I want to get back in sync with the kernel hacking process. The other is
that I want to strike a balance: a little more experimental than -ck, but
with less heavy updating and (I hope) potential instability than -mm.
I intend to release mostly for release kernels only; this one is for 2.6.7-rc3
because I'm impatient and because that is what Con is currently syncing up
with.
All that being said, at http://hobbs.dynup.net/patches/2.6.7-rc3-ar1/ you can
find the initial version of the patch, with staircase,
autoregulate-swappiness, supermount-ng, ext3 and reiser improvements, and a
free toy in every box. Passes the boot test on i386; if it breaks anywhere
that rc3 doesn't, I'd love to hear about it. If it makes your machine twice
as fast, I'd love to hear about that too.
Andrew Rodland < [email protected] >
P.S. I read via GMANE, so Cc:ing me if you want my attention couldn't hurt.
> find the initial version of the patch, with staircase,
> autoregulate-swappiness, supermount-ng, ext3 and reiser improvements, and a
I gather that supermount-ng is now quite dated and no longer maintained. Is
submount (http://submount.sourceforge.net/) not the current favourite to
provide such functionality?
Looking at the two, submount definitely seems more ready for inclusion based
on its non-invasive approach.
--
______________________________________________
Check out the latest SMS services @ http://www.linuxmail.org
This allows you to send and receive SMS through your mailbox.
Powered by Outblaze
On Sat, 12 Jun 2004 12:39, Pokey the Penguin wrote:
> > find the initial version of the patch, with staircase,
> > autoregulate-swappiness, supermount-ng, ext3 and reiser improvements, and
> > a
>
> I gather that supermount-ng is now quite dated and no longer maintained. Is
> submount (http://submount.sourceforge.net/) not the current favourite to
> provide such functionality?
>
> Looking at the two, submount definitely seems more ready for inclusion
> based on its non-invasive approach.
Because supermount-ng is what our users want it to be:
It's stable, mature and fully functional without any userspace changes (apart
from the fstab entry) and thousands of people use it daily without any
problems. Whether you (or the mainline kernel maintainers) like the patch or
not is a different matter. I've been porting it to each main release version.
Con
On Sat, 2004-06-12 at 04:39, Pokey the Penguin wrote:
> > find the initial version of the patch, with staircase,
> > autoregulate-swappiness, supermount-ng, ext3 and reiser improvements, and a
>
> I gather that supermount-ng is now quite dated and no longer maintained. Is
> submount (http://submount.sourceforge.net/) not the current favourite to
> provide such functionality?
>
> Looking at the two, submount definitely seems more ready for inclusion based
> on its non-invasive approach.
i believe that having the automounting inside the kernel is a bad idea.
i think there should be more people using HAL and D-BUS, together with a
userspace automounter. like ivman, http://ivman.sf.net
--
Regards, Redeeman
[email protected]
I don't intend to make announcements on a regular basis, but -ar2 is out for
2.6.7-rc3, with submount and some scheduler improvements. Since everyone
seems to be looking in the directory I posted earlier, I should probably
mention that http://hobbs.dynup.net/patches/ is holding new releases in the
obvious way. I'm still hunting for any other patches that improve user
experience, especially for desktop users, but I think I like where I am so
far, and that it's quite easily maintainable at this point.
Cheers
Andrew Rodland < [email protected]> via GMANE