Re: Plans for Midnight Commander development

[Oswald Buddenhagen <oswald buddenhagen gmx de>]

On Sun, May 31, 2015 at 08:30:49PM +0200, Yury V. Zaytsev wrote:
> * Regarding mc^2 fork, I would like to review the code and get it
> merged. After that, we can make a major 5.0.0 release, which may take
> some time to stabilize. I would need help with code review to make this
> happen in finite time though.
you can count on at least some attention from me.

> * If you want to invest a large amount of work into something to prove
> yourself, but are unsure about it, ask on the list. If you want to
> tell me (or anyone) what and how we should do, but aren't ready to
> invest large amounts of work yourself, then please don't.
you certainly realize that the boundary of what is considered
constructive input/feedback and useless noise is highly individual and
context-sensitive. luckily, the effort to reject (or outright ignore) a
particular contribution is roughly inversely proportional to how
outlandish it is. just don't overdo it.

> * Regarding the infrastructure & the development process that we have
> set up and followed while still being active: nothing is set in stone.
>
> However, keep in mind that the best way to convince me is to actually do
> the work, e.g. if someone has been triaging Trac for some time, and can
> explain why it's been painful, and how a proper migration to service X
> that he is ready to prepare will solve all the problems and make him
> über productive, I'm likely to fall for this argument.
i have no problem whatsoever with trac as a bug tracking system
(obviously, as long as the performance doesn't break down entirely). it
isn't awesome, but it certainly does everything a project of the size of
mc needs.

however, as i have expressed multiple times over the past years, i do
have a problem with it being used as a code review tool - because it
simply isn't one. even you guys effectively demonstrated that: while you
used it as a patch tracker (with rather clumsy workflow management), you
did all the actual feedback on jabber.
so while i'm not exactly a fan of github (as indicated in my other
mail), i certainly consider it a significant improvement over trac, and
would fully support making it the official channel for patches.

one might argue that the fragmentation of tools is a disadvantage, but i
don't really see a significant problem with that.
one idea, though: it would be awesome if you could set up github
authentication for trac. https://trac-hacks.org/wiki/OAuth2Plugin

however, there *is* a significant problem with people using the github
issue tracker while trac is the official one - i'd disable the github
one instantly.