Hi Jason,

Jason Thorpe wrote:
>  Guenther Grau <Guenther.Grau@marconicomms.com> wrote:
>  > LFS has been in the tree only for a few month now.
> 
> Nonsense!  LFS has been in the source tree for a *very* long time (since
> at least the 4.4BSD-Lite code came in).

:-) Well, yes, (of course :-) you're right. Sorry about this! 
I should have said it's been in the tree in a useable form for a
few month, thanx to/kudos to Konrad Schroder (and others lately).

[...]
> Well, while there are still bugs, it's been shaken out quite a bit
> in the last few months by Konrad Schroder, and works pretty well,

Yes, especially after the bug fixes that came in a couple of
days ago, which seemed to fix just about all known bugs.
The only problem that might still remain is LFS vs. a
full filesystem, or has that been dealt with as well?

> for the most part.  However, LFS isn't good for all types of workloads,
> so this is one of the reasons it's not more widely used.

Hm, I can't comment on this. Did anybody do some testruns
with an LFS from -current and do some testruns under
different types of workloads? What are the cases it
doesn't handle very well? (Maybe I should go reading
papers about it first :-)

  Guenther