[merge][rfc] remove python -O from make check

Martin Pool mbp at sourcefrog.net
Sun Aug 19 05:24:30 BST 2007


On 8/17/07, John Arbash Meinel <john at arbash-meinel.com> wrote:
> I wonder if we wouldn't rather always run with -O. Since we shouldn't be
> relying on assert foo to do any actual error checking. (It should only
> be used to assert that developers are using the apis properly.)

That's an even better idea.

> Other than that, I don't mind pulling out one of the selftests. I think
> our bigger lag is just because the LP tests are running at the same
> time. When an LP submission isn't pending, I'm usually shocked at how
> fast I get a response.

The other people who use that machine are in the process of moving to
a separate new machine.  I've asked for our pqm to be upgraded to a
newer/faster machine, but that may take a while.

-- 
Martin



More information about the bazaar mailing list