On 6/11/07, Karl Fogel <kfogel@red-bean.com> wrote:
> [resending to list, as originally replied just to lsvkne@]
>
> lsuvkne@onemodel.org writes:
> > If there are no big surprises discovered beyond what was
> > described there, I wonder whether it might be possible to scope the
> > effort now. Surprises being the big wildcard, given the complex history
> > of this issue.
>
> I don't think the scoping is trivial; I certainly couldn't do it now.
>
> FWIW, in IRC earlier tonight one developer said he thought the $7500
> figure was off by an order of magnitude. I don't know if that's right
> or not (it probably is, as he's an expert in this problem), but
> statements like that point to the importance of not skimping on the
> scope/design stages! :-)
FWIW, I was that developer, and I should probably clarify.
When I was working on the current fs-atomic-renames branch the plans
we had to take that work and use it to implement atomic renames would
probably mean about 6 months of work before you could see anything
particularly useful out of it.
Now that doesn't preclude breaking it up into smaller chunks (although
following the plan we had would mean the initial smaller chunks
wouldn't be all that useful to an end user) or finding some other way
to attack the problem, but if a developer was to continue down the
path we were on back when I was hacking on it I can't see a whole lot
of progress being made for the amount you're talking about.
-garrett
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Tue Jun 12 01:05:48 2007