Web lists-archives.com

Re: CI system maintainability




In data giovedì 28 marzo 2019 16:04:01 CET, Boudhayan Gupta ha scritto:

> I don't get why mandatory code reviews are so unpopular.

It's not "unpopular". As far as the discussion goes, the opinions (from 
several parties) say that they're not a silver bullet, and that some projects 
benefit from them more than others. 

The ultimate solution is actually more developers (yeah, I know, easy...).

CI, OTOH, has been IMO very useful (despite the headaches Ben mentions) for 
all the projects in KDE. 

> I don't care if you lose time. I don't want the guys building my house to

You should if the review stays there for years when there's no one else to 
review it. 

> As a user, I simply do not want unreviewed crap running on my computer.

Well, reviews help but they're just part of the equation. CI helps as well 
(and IMO, it should be more visible as I mentioned earlier in the thread).
And perfectly reviewed code (as well as unreviewed code) can still be a 
problem (as an integrator, I see that often). 

> one-liners, you're probably too overconfident to be writing good code
> anyway, so I'm going to operate on the presumption that if the code hasn't
> had more than one pair of eyeballs ever looking at it, it's crap.

I would say that there's no need to be like this. There is bound to be 
disagreement (and there is) but not as much as to define quality on 
assumptions. 

To be clear: I'm neither on the side of "review all the things" nor on the 
anarchist side. I just want to make sure we don't engage in policies that can 
be (potentially, just potentially) harmful for some parts of KDE (while they 
are perfectly OK for others). 

-- 
Luca Beltrame
GPG key ID: A29D259B

Attachment: signature.asc
Description: This is a digitally signed message part.