On Wed, Nov 14, 2012 at 4:34 AM, Brian Knox <[email protected]> wrote:
> I fail to see the "failure" here. A new feature that people wanted
> was added to a release candidate. Problems were found, in the release
> candidate. The feature, based on community feedback, was reworked and
> fixed.
>
> To me, this looks like a success!
Indeed.
I'm sure some people enjoy to rant. Good show. Others prefer to make
patches. We've seen this lots of times. Unless someone can point to an
actual systemic problem with the process, or code base, there's no
reason to take such venting too seriously IMO.
Nicholas, if you still want me to reply, I'll try again. But I fear my
sarcastic "So, you can spend a day writing this post but not an hour
making a pull request?" comment would be unhelpful. :-)
The thing about C4, we don't check that a patch doesn't break
regression tests before merging them. This is the contributor's job.
The rule about "must build and pass tests" allows anyone who disagrees
with the patch to revert it. Mistakes are welcome, that's how we
learn.
-Pieter
Ps. I still get a couple of test failures in the libzmq master. Anyone
else seeing these?
lt-test_connect_delay: test_connect_delay.cpp:153: int main():
Assertion `rc == 0' failed.
/bin/bash: line 5: 29787 Aborted (core dumped) ${dir}$tst
lt-test_raw_sock: test_raw_sock.cpp:217: int main(): Assertion `rc == 0' failed.
/bin/bash: line 5: 29910 Aborted (core dumped) ${dir}$tst
_______________________________________________
zeromq-dev mailing list
[email protected]
http://lists.zeromq.org/mailman/listinfo/zeromq-dev
