Re: Last call: range locking

I read the whole thing. I didn't take BYTEHEAD seriously (perhaps
it was your smiley) partly because the combination of HEAD and range
retrieval didn't seem to be adequate to discover the ways in 
which a resource could logically be decomposed, much less the 
proper way to find the names of the parts; your subsequent proposal
for LOCKSBYTE, being a kludge on top of a non-functional subproposal
didn't seem to be too serious either.

If you do mean for these to be taken seriously, perhaps you'd
want to change "BYTE" to "RANGE" and deal with the issue of
disjoint ranges, the mechanism by which AcceptRanges might
be discovered, whether AcceptRanges applies to accepting locks
on ranges, etc.

otherwise, it just bytes.

Larry
 
-- 
http://www.parc.xerox.com/masinter

Received on Wednesday, 5 March 1997 01:58:06 UTC