Re: [PATCH v3 1/1] scripts: Add add-maintainer.py

[Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>]

On 30/08/2023 01:16, Guru Das Srinagesh wrote:
> On Aug 28 2023 21:45, Krzysztof Kozlowski wrote:
>> On 28/08/2023 21:41, Mark Brown wrote:
>>> On Mon, Aug 28, 2023 at 07:59:54PM +0200, Krzysztof Kozlowski wrote:
>>>> On 28/08/2023 19:56, Guru Das Srinagesh wrote:
>>>
>>>>> Your function adds mailing lists also in "To:" which is not ideal, in my view.
>>>>> You've mentioned before that To or Cc doesn't matter [1] which I disagree
>>>>> with: it doesn't matter, why does Cc exist as a concept at all?
>>>
>>>> To/Cc does not matter when sending new patch, because maintainers know
>>>> they are maintainers of which parts. I know what I handle.
>>>
>>> That might be true for you (and also is for me) but I know there are
>>> people who pay attention to if they're in the To: for various reasons, I
>>> gather it's mostly about triaging their emails and is especially likely
>>> in cases where trees have overlaps in the code they cover.
>>
>> True, there can be cases where people pay attention to addresses of
>> emails. Just like there are cases where people pay attention to "To/Cc"
>> difference.
>>
>> In my short experience with a few patches sent, no one complained to me
>> that I put him/her/they in "To" field of a patch instead of "Cc" (with
>> remark to not spamming to much, so imagine I send a patch for regulator
>> and DTS). Big, multi-subsystem patchsets are different case and this
>> script does not solve it either.
> 
> Not sure what you mean by "does not solve it" - what is the problem being
> referred to here?

Exactly, no one even knows what problem you want to solve by swapping
To-Cc between patches...

> 
> In case of multi-subsystem patches in a series, the commit message of this
> patch explains exactly the actions taken.
> 
>> Anyway, if it is not ideal for Guru, I wonder how his LKML maintainer
>> filters work that it is not ideal? What is exactly not ideal in
>> maintainer workflow?
> 
> I am not a maintainer - only an individual contributor - and as such, even
> though I may get patches on files I've contributed to, I deeply appreciate the
> distinction between being Cc-ed in a patch vs To-ed in one. The distinction
> being that if I'm in "To:" I ascribe higher priority to it and lesser if I'm in
> "Cc:".

That's your feeling, quite subjective. I understand it comes from
corporate world, but again...

> 
> If this script is accepted and gains adoption, maintainers like yourself will
> only be To-ed in patches that touch files that you're a direct "Maintainer" or
> "Reviewer" of. 

It will not get traction because:
1. People should use b4, not this script.
2. Remaining people will just use get_maintainers.pl.
3. People cannot get right even basic commands, so we will never be able
to rely on To or Cc distinction. I can give you example: my email
address in get_maintainers.pl is a bit different. Does it matter? Often
not. Entire bunch of folks were Ccing me on different address. Even
though every tool told them not to...

> For all other patches in the series you'll be in "Cc:". I
> imagine that this can be very useful regardless of the specifics of your
> workflow.

Zero usefulness for me.

Best regards,
Krzysztof