[neomutt-devel] #224

[Guyzmo]

Hello David, Kevin and list,

I have participated in a bunch of commits on neomutt and helped Rich get
stuff organised around here.

I've been a mutt user for over 15+ years, and I've followed at a
distance all the politics and the trolls that happened over the years.
And always have been very careful to stay at a distance of it.

I switched about 5 years ago to notmuch, and thus, started to use
mutt-kz. I had many issues using it, reported some issues, saw it get
better and more stable over time, and now I cannot live without notmuch,
so as you see karel as a guy who did an offensive fork, I thank him for
enabling a tool I love work nice with another tool I love.

I wanted to implement a feature, which is to create search windows for
notmuch searches in mutt-kz. Because the active development of kzak's
code is now happening in neomutt, I implemented the patch in neomutt,
and it made it into a release recently (for the best as for the worst).

Then, I saw a bunch of easy features in the bug tracker, so as I dug
into the sources, I just wrote them. I've even added a lua interpreter
as WIP patch.

All that to say, in this process, I never thought about reaching out to
Mutt once, because as much as you do value your time, I'd rather do
stuff I'm passionate about and good at (coding, project management), and
not lose time doing politics.

And from my understanding (please prove me wrong), the mutt project is
not interested in merging a fat feature like notmuch, and even if it
did, then you'd get a whole bunch of old school users — and I know and
have respect for some of them — that would complain about how mutt is
becoming a useless huge bloat, instead of staying the same MUA they love
since 1995.

On Thu, Feb 09, 2017 at 03:42:26PM -0800, David Champion wrote:
> 1) what does a new community bring that the established one does not

I don't know and I can only speak from my perspective.

All I know, is that if it wasn't for Karel's fork followed by Rich's, I
might have stopped using mutt, or I'd have worked on another mutt-like
MUA that has a first class notmuch integration.

Though I do appreciate you guys reaching out in a rather nice way, so we
can discuss the issue and decide how we can gather efforts, or whether
both projects have too divergent goals.

> 2) is it advantageous to create a new community rather than making any
> effort whateoever to introduce those advantages (I don't doubt that
> there are some) to the established community?

Well, it all depends on whether we'll lose more time talking politics,
or if we actually care about code.

I'm not against contributing to mutt directly, and would actually be
honoured to, if the features I'm actually interested in (mainly:
notmuch, but other troll-bait features like composing mime/alternative
mails) can be considered by the "established" community.

That being said, and after having dug a bit in (neo)mutt's sources, I do
believe that some modernisation of the codebase is really needed
(homogeneous coding style and indentation, no more globals, no more huge
functions…), the most important thing being introducing regression tests
and CI. And I'd be happy to discuss these topics in length and write
patches for that.

Politics? gah, not really.

Cheers,

-- 
Guyzmo