Re: svn commit: rev 5194 - trunk/subversion/libsvn_fs

>
>
>Oh wait.
>
>I wonder if this is another retry_txn() kinda thing, where someone is
>using stale information that was invalidated by a Berkeley txn that
>needed to be retried. Hm...
>

You know,
I've often wondered how with all the creative stuff going on in the FS
code. There are still *0* calls to svn_fs__record_undo in trail.c.
It's a feature, I thought *I* might take advantage of. There is still a
small problem where undo_on_success functions are called before the commit.
If it hasn't been committed it isn't successful yet. Is it? I made
changes to this in my vtable backed version.

I should be submitting a new patch before first light. Got side tracked
over the weekend:-)

gat

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Tue Mar 4 06:22:39 2003