[openstack-dev] Please stop reviewing code while asking questions

Julien Danjou julien at danjou.info
Fri Apr 24 15:22:07 UTC 2015


On Fri, Apr 24 2015, Doug Hellmann wrote:

> I will often ask questions like, "what is going to happen in X
> situation if we change this default" or "how does this change in
> behavior affect the case where Y happens, which isn't well tested
> in our unit tests."

Well I didn't say you weren't allowed to ask questions. I said you
cannot ask a question because you don't know something _and_ set a
score.

The cases you describe matches  my first paragraph which says:

> > _Sometimes_ there are good reasons to set -1 even when asking a
> > question. For example, when the question is a hint sent to the patch
> > author so that (s)he improves is commit message, a code comment or a
> > piece of code.

And in the case you describe I think it's even better to gently ask for
a test that proves that the case Y is not broken if the test does not
exist yet.

> If those details aren't made clear by the commit message and comments
> in the code, I consider that a good reason to include a -1 with a
> request for the author to provide more detail.

What? Why do you say this now? If I knew years ago then I would have
never have written a line of unit test! I would have replaced all of
those tests with a simple "Hey, this code totally works!" in each commit
message! ;-)

Seriously, I think this is too kind/dangerous and that's it's just even
better to ask for tests, full stop.

> Often these are cases I'm not intimately familiar with, so I ask a
> question rather than saying outright that I think something is broken
> because I expect to learn from the answer but I still have doubts that
> I want to indicate with the -1.
>
> Most of the time the author has thought about the issues and worked
> out a reason they are not a problem, but they haven't explained
> that anywhere. On the other hand, it is frequently the case that
> someone *hasn't* understood why a change might be bad and the
> question ends up leading to more research and discussion.

Yeah I don't think what you describe matches the *bad* behavior I
described in my original email.

Sorry. I can understand you wanted to appear as the bad guy I was
yelling at, but that was not you after all! ;)

-- 
Julien Danjou
-- Free Software hacker
-- http://julien.danjou.info
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 800 bytes
Desc: not available
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20150424/f154e3c3/attachment.pgp>


More information about the OpenStack-dev mailing list