On Mon, Feb 9, 2015 at 4:38 PM, Branko Čibej <br...@wandisco.com> wrote:
> On 09.02.2015 16:34, Stefan Fuhrmann wrote: > > On Mon, Feb 9, 2015 at 3:05 PM, Evgeny Kotkov > > <evgeny.kot...@visualsvn.com <mailto:evgeny.kot...@visualsvn.com>> > wrote: > > > > I committed a failing test in r1658406. As for fixing this issue, > > I think that > > we should entirely revert this changeset. > > > > > > Authz performance has always been a sore spot. > > I think making hurt less is worth some code. > > But we're doing that on the authzperf branch, where the benefits are > actually really tangible. Not on trunk while we're trying to stabilize > it for the release and without at least minimal performance comparison > on real-life data. > That branch did not exist in Mai 2013, when the bug got introduced. -- Stefan^2.
