I don't know if it's a regression. My logic was that we don't want to
do a release that fails our own test suite. I think it will be less
work down the line if we just fix it tonight. Cheers,
Adam
Sent from my iPhone
On Jun 29, 2009, at 5:29 PM, Damien Katz <[email protected]> wrote:
Unless this is a regression, I think we shouldn't hold up 0.9.1 for
this. We can add the fix to the branch, and if we do a 0.9.2 it will
be there.
-Damien
On Jun 29, 2009, at 3:59 PM, Paul Davis wrote:
Like I said on the ticket, unless I'm mistaken, _restart only exists
for testing purposes. This fix is more about fixing the test so that
people aren't alarmed which to me doesn't change actual behavior as
most people shouldn't be hitting _restart outside of test suites.
As for the mechanics of doing a vote, I'd just wait for those that
expressed concern about the test case failing to verify that a new
0.9.1 tarball doesn't fail and if so then move on with our lives.
On Mon, Jun 29, 2009 at 3:50 PM, Noah Slater<[email protected]>
wrote:
On Mon, Jun 29, 2009 at 02:47:25PM -0400, Adam Kocoloski wrote:
We don't necessarily need to hold up 0.9.1 while we discuss the
behavior
of /_restart, but I think it's very likely that the
implementation of
/_restart will change substantially in the near future. I would
go so
far as to recommend that people not use the /_restart feature in
0.9.1
and below. Best,
Jan, Chris, Damien, Chris, Paul, everyone else? Thoughts?
--
Noah Slater, http://tumbolia.org/nslater
