C. Michael Pilato wrote:
> Lieven Govaerts wrote:
>> Justin Erenkrantz wrote:
>>> On Dec 5, 2007 12:02 PM, Lieven Govaerts <svnlgo@mobsol.be> wrote:
>>>> the stress test script is meant to find out Mike's "double free" error
>>>> during commit. Since you're already solving the switch issue, I'll have
>>>> a look at a stress test script.
>>> I think the patch below should fix it.
>>>
>>> Anyone care to review before I commit? (I'll also write up a log msg too.)
>> Patch looks good.
>>
>> One question though, why do we only invalidate version-url for the root
>> directory? Can we assume that 'all-wc-props' is overwritten for all
>> child directories?
>
> The command-line client always does a recursive invalidation based on this
> one call. Perhaps we should just codify that that's what
> invalidate_wc_props() means. After all, invalidating non-recursively is as
> simple as calling the set_wc_prop() callback a few times with a NULL value,
> right? Alternatively, we could add a recu^H^H^H^Hdepth flag to the callback.
>
Ah, I already changed the documentation in r28289 to add that expected
recursive behavior, since we expect it to behave like that in the two
ra_dav layers.
Lieven
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Wed Dec 5 22:46:29 2007