Subject: Re: Large inodes for ffs
To: Bill Studenmund <email@example.com>
From: Perry E. Metzger <firstname.lastname@example.org>
Date: 03/23/1999 16:59:43
Hubert Feyrer <email@example.com> writes:
> * Are there some utilies (fsck?) to convert the FS to all-big inodes (if
> that's even wanted)? Or is this "mixed size" operation the usual mode,
> with only some inodes being "big" ones?
> * We're very close to the 1.4 cut, and if the code goes in now, there
> will probably not too much testing. Can we risk this?
I have serious concerns about this.
There are only a few days to the 1.4 quiet start. No one other than
your team has looked at this previously. There might be considerable
amounts of feedback, improvements, etc. -- and no time in which to
consider them before putting in a major new interface.
Furthermore, there might be considerable amounts of fallout from
this. I know this "shouldn't" cause trouble, since it has been tested
for a while, but lots of things that "shouldn't" cause trouble do.
Given that 1.5 is currently scheduled for fall, maybe we could wait to
commit this until after the branch, and gain the benefits of six
months of feedback on the design and testing before putting something
this big into a release?