Re: XFS for 2.4

[davidsen@tmr.com (bill davidsen)]

In article <20031202182037.GD17045@work.bitmover.com>,
Larry McVoy  <lm@bitmover.com> wrote:

| This is a process.  The process is supposed to screen out bad change.
| Maybe XFS got into 2.5/2.6 inspite of the process rather than because
| of it.  Maybe not.  Whatever the answer is, it's perfectly reasonable
| for the maintainer of the 2.4 tree to want someone he trusts to step
| forward and say "yeah, it's fine".  The fact that other VFS people
| aren't jumping up and down and saying this should go in is troublesome.
| If I were Marcelo the more the XFS people push without visible backing
| from someone with a clear vision of the VFS layer the more I'd push back.
| 
| Don't get me wrong, I have not looked at or used XFS in years.  I have
| no opinion about it at this point.  But I do have an opinion about process
| and what is going on here, in my opinion, violates the Linux development
| process.  Patches shouldn't go in just because you want them in, they go
| in because the maintainer chooses to bless them or someone he trusts chooses
| to bless them.

It has been my experience that a few hundred normal users actually
running code without problems is a LOT more reliable predictor of
stable operation than any one person reading the code and saying it
looks good. Users check out the exception handling paths better ;-)

-- 
bill davidsen <davidsen@tmr.com>
  CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.