Re: moving back to 32-bit swap_filens ?

>
> I'd like to have COSS. I also second your idea to try and maintain the
> byte boundary for data structures, but since I'm not touching them,
> I'm in the hands of those who do.
> If possible, I'd move all of my storage to it, esp. if it was available in
> async/diskd flavours.

Well, see, one of the aims of the diskio branch was to separate
the disk IO layer from the filesystem layer. That way, we save space on
duplicated code (only one 'ufs' codebase now) and people with
strange platforms can implement their own disk IO and have it appear
in all the filesystems that support it(read: windows :-)

The trouble is that I can't just use the read/write buffers that squid
supplies in the async layer, since the buffers may or may not be there
when it comes time for the IO to occur.

It _would_be a good branch to finish though.

Hrmm. Yeah. It would be actually, as I could reuse the code in other
places.

If only I could finish the http / storage manager rewrite, all the
cool network/disk code Henrik and I have been tossing about over the
last 12months could become _useful_. :-)

Adrian

-- 
Adrian Chadd			Yeah, for me its (XML) like the movie Titanic.
<adrian@creative.net.au>	  Everybody loves it.
				    I want to be different, so I hate it.
					--Duane Wessels