Re: [git patches] libata updates, GPG signed (but see admin notes)

[david]

On Mon, 31 Oct 2011, Linus Torvalds wrote:

> On Mon, Oct 31, 2011 at 7:17 PM,  <david@xxxxxxx> wrote:
> > what IMAP servers have you used? I find that with a good IMAP server I can
> > get away with _very_ little processing power in the client and get good
> > performance (I still am using pine, but with a cyrus IMAP server)
>
> I've used pine (and then alpine) to cyrus servers too. That was what
> LF used to have on the server side. It absolutely sucked.
>
> That combination is supposed to do server-side searching etc, but you
> couldn't tell from the performance. It was disgusting.

interesting, that's the setup I have and the server side searching works 
well for me (searching for random text works much better if I have the 
server update it's squatter indexes frequently). I did find that different 
filesystems made a HUGE difference in system performance (with ext2/3 
being the worst). I use XFS, but I see a lot of people reporting good 
results with ext4, btrfs, and ZFS.

the biggest folder I have has >250K messages in it and takes >20G. random 
text searches on that take 5-10 seconds (much better than I could do on 
my local drives)

> I'm sure it works much better with a fast local network, but quite
> frankly, that obviates the need for IMAP in the first place. If you
> have your mail locally, there are better models than IMAP for handling
> it.

I agree that it's not strictly needed if you store everything local, but I 
don't want to have all that mail on my non-mirrored, small laptop drive 
(even if it is a SSD). I've suffered too many disk failures over the years 
to not have my mail store on redundant drives :-)

David Lang

> So the only situation I've found IMAP reasonable has been at corporate
> settings where you're not talking DSL or cable modem speeds, but use
> IMAP as a way to avoid NFS-mounting the mail spool, which is even
> worse. But actually working over slowish internet connections? No
> thank you.
>
>                         Linus
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/