You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by Justin Erenkrantz <je...@apache.org> on 2002/08/01 19:13:05 UTC

Applying entries cache patch?

Now that I can acceptably update and build SVN on Darwin, I'm going
to start to turn my focus towards libsvn_wc and adding the entries
cache.  (I'm now thinking that the update hangs are somehow related
to Norton Antivirus on Darwin and, of course, the build speedup is
due to jlibtool.)

Since Philip has already merged the branch to the trunk, should
I just start committing the bits as I write and test them?  Or,
should I post them to the list first?  Or, should I start another
branch?  Do I have enough time (say a week+) before the beta is
planned to have it thoroughly vetted?

My plan is to integrate the targeted patch that I posted a while
back that only partially passes the baton around.  That should
speed up the diff and commit process.  Then, I plan on working
my way through libsvn_wc and adding the baton to all functions
that read the entries file.  I don't expect to break anything,
but it could happen.  =)  -- justin

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org

Re: Applying entries cache patch?

Posted by Philip Martin <ph...@codematters.co.uk>.
Justin Erenkrantz <je...@apache.org> writes:

> My plan is to integrate the targeted patch that I posted a while
> back that only partially passes the baton around.  That should
> speed up the diff and commit process.  Then, I plan on working
> my way through libsvn_wc and adding the baton to all functions
> that read the entries file.  I don't expect to break anything,
> but it could happen.  =)  -- justin

As I recall that patch introduced a cached interface parallel to the
current non-cached interface.  I still think that's tricky to get
right, how do you ensure that there is not a code path that bypasses
the cached interface and call the non-cached one?  I think the correct
approach is to put the baton in place first, then replace the
non-cached interface with the cached one.

-- 
Philip Martin

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org

Re: Applying entries cache patch?

Posted by Karl Fogel <kf...@newton.ch.collab.net>.
Justin Erenkrantz <je...@apache.org> writes:
> Since Philip has already merged the branch to the trunk, should
> I just start committing the bits as I write and test them?  Or,
> should I post them to the list first?  Or, should I start another
> branch?  Do I have enough time (say a week+) before the beta is
> planned to have it thoroughly vetted?

I think you should commit on the trunk.  If there's going to be a bug,
we'll find it best by people actually exercising the code...

Just test very, *very* carefully :-).

-K

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org