Re: [PATCH] Fix futures, %worker-count shouldn't be zero to prevent halting

On 8/20/2025 4:02 PM, Felix Lechner wrote:

> On Wed, Aug 20 2025, Maxime Devos wrote:
>
>> Anything in particular?
> No, the whole thing.  Your responses were proficient and comprehensive
> (too comprehensive for Nala) yet it seems to me that you lost sight of
> the purpose of the transaction.
>
> Nala found a bug and, being a user, would like to see it fixed without
> excessive discussion.  It's not a contest.
>
> Nala, please chime in.  I don't mean to put words into your mouth.

There is no such thing as 'too comprehensive'.

All the text was about identifying and fixing the bug, so I don't get 
where you get 'lost sight of purpose of ...' from. Nowhere did I invoke 
some kind of developer/reviewer/user/... distinction (liking to getting 
things fixed easy is not dependent on any such categorisation).

Nowhere did I turn it into a contest. If anything, Nala tries to turn it 
into a compactness-despite-the-costs contest.  Sometimes when writing 
stuff there is simply a lot to write about. Also, please don't turn 
things into a 'minimise discussion' contest. Don't lose sight of how the 
purpose is to fix bugs and such, the purpose isn't review-codegolf.

Also, _what_ excessive discussion? Like, it goes like:

(1) Nala sends patch. Some bug is mentioned with a vague explanation, 
which is later known to be misidentified. At this time, the patch is 
_underdiscussed_.

(2) I notice issue about calling current-processor-count twice, so I 
send a short message

(3) Nala sends adjusted patch.

(4) I notice some other problems this time, and have a hunch for 
something that might be wrong on Artanis' side instead and should be 
easy to confirm (never received a message about that thing, so here is 
some underdiscussion as well) -> message.

(5) Nala ignores the Artanis thing, says something vague about 'mixing 
different situations', uses unclear terminology (fulfilment???), makes 
leaps about cause->effect things, and makes an impossible claim (if 
there is no thread support then you can't have worker threads ...).

(6) Since the reasoning is a bit rigorous and the previous message is 
unclear in places, I attempt to reverse-engineer what they might have 
meant, by identifying what potential meanings do not agree with the 
code, and by searching for the bug myself and after the fact determining 
in what ways the bug matches the description by Nala.   So, I do that, 
and along the way write down the thought process (and afterwards clean 
it up a little). Since debugging code and deobfuscating vagueness can be 
rather elaborate, naturally it follows that this involves a lot of text. 
And since I was unable to use Nala's messages as-is because of their 
vagueness, it naturally follows that I attempt to increase the clarity 
of own

(7) Now I get complaints about length and you find my behaviour 
outrageous???

Mirrors and kettles, people.