> 
> I had some fs corruption from hfsutils....  No problems with the lkm. 
> Apparently some people did, though, but only when mounting read-write. 
> The corruption was from two particular versions, 0.8 and 0.8.1.  I don't
> think the prerelease versions had this (I'm using a pre-release version
> w/o any trouble at all, with a large amount of reading and writing for a
> couple of months), and the bug has been found and corrected in 0.8.2 (at
> least so they say). 
> 
> As I understand it, there are two porting efforts for NetBSD, though I'm
> not sure who they're by (was Bill Studenmund doing one?).  One of them was
> an attempt at hfs support that would not be GPL'ed, thus it could be
> included in the kernel, instead of requiring an LKM.  It currently only
> supports reading, and as I understand it, works quite well.  In addition,
> I _think_ someone is porting the linux lkm, which would have to stay an
> lkm for copyright reasons (GPL), which would support reading and writing,
> but, as I said, could not be included in a kernel.

I was working w/ Paul Goyette on getting macfs ported. We've got something
which kinda works. There's a buffer utilization problem which we never
found. If you're running single user, it's ok. Multi-user, and the system
crashes.

The other problem is that macfs has restrictive licensing terms which,
though not GPL, will keep it out of the kernel.

Paul and I talked about porting hfs_fs. I know Paul Hargrove personally,
and would like to get it ported. But I've been working on the serial
ports, and on trying to finish my PhD. So I've let hfs_fs slip. I'm
not sure how Paul is doing.

If someone wants to help, I'm sure Paul would be happy for the help, and
I can chip in the occasional comment too.

Take care,

Bill