[TYPO3-v4] Verified / Code Review

[Ernesto Baschny [cron IT]]

Hi Krystian,

Krystian Szymukowicz schrieb am 07.03.2011 18:58:

> Sorry - its me again about the gerrit :)

No need to be sorry, better ask many questions, because others will
probably have similar ones!

> Is there some rules explaining the "pass" for RFC? I mean all the +1 /
> +2 etc. I can see some incosistency.

Yes, we kind of "decided" on some rules, but it's not written in stone
yet because we wanted to try out the workflow first. The issue
discussing this is:

http://forge.typo3.org/issues/13468

Please add your comments if something is unclear.

> Example 1:
> 
> https://review.typo3.org/#change,1066
> It was always a fact that person who brings the patch can not vote for it.
> In the bug above the same person "verifies" and do "code review".
> Is it right?

That's right but it doesn't count for our policy. It's fine that the
patch submitter reviews his own code, but sometimes someone brings in
code that is not from himself, which makes it better to always at least
state (with reviews) that the code was reviewed and verified.

To get submitted however, this one needs one more review by testing
("Verified"), because only the author and one core developer (Steffen
Kamper) has done so.

> Example 2:
> 
> https://review.typo3.org/#change,1062
> Core developer have done "Code Review". The system is saying that only
> simple non-core "Need Verified +1 (Verified)" is needed to pass the RFC?
> Do not we need here?
> a) one more +1 from core/non-core for "Code Review"
> b) one more +1 from core for "Verified"

Steffen decided to click on "+2" right away because he knows this code
(he wrote it) and he probably identified this bug by reading. But he
probably hasn't got the time to test it.

The policy still remains: Still needing one more code review (by
anyone), but two people (one of them a core developer) must still test
this issue before a core developer decides to submit.

Cheers,
Ernesto