[fpc-devel] [RFC] Modernising the FPC Release Process -- Proposal for Review

Michael Van Canneyt michael at freepascal.org
Fri Apr 17 11:23:45 CEST 2026



On Thu, 16 Apr 2026, Graeme Geldenhuys via fpc-devel wrote:

>
>
>
>> Side clarification - I take it this means FPC is open to AI contributions?
>> That may help break the logjam, but it may upset some people too.
>>
>
> It's a doubled edged sword.
>
> Yes it's great to get things done quickly. But flooding the FPC team with even more MRs at this time, seems to be a slippery slope.
>
> On the bright side, AI is great at doing what most developers hate. 
> Writing tests or letting it do TDD.  In such cases MRs are much easier to
> review.  More test cases validating logic is a good thing long term.

If I see a MR that has test cases included, I'll be more inclined to merge,
this is for sure.

Whether it was generated by AI or not is entirely irrelevant as far as I am
concerned. In the end it is my responsability anyway if I approve the MR.

That's a fundamental rule for me: 
The human doing the commit is the one responsible.

(Which is why I never let Claude commit anything)

Michael.


More information about the fpc-devel mailing list