Re: "mc is over!?" - post by Ilia Maslakov on Russian-speaking IT site

[Volodymyr Buell <vbuell gmail com>]

On Sat, May 30, 2015 at 8:47 AM, Yury V. Zaytsev <yury shurup com> wrote:

On Sat, 2015-05-30 at 08:23 -0400, Volodymyr Buell wrote:
>
> That's the reason to decommission existing infrastructure asap - you
> pay for the things that work against your productivity.

I've heard this before, and you still haven't explained how it works
against *my* productivity, or the productivity of Andrew.

I was referring to productivity of the team as a whole.

And if you want to hear an explanation, here is this:

* the process of code reviewing is non transparent enough for the community - attaching the patches to the tickets and reviewing these patches is not the same as pull requests
* people tend to use the tools if they are good. If there are few options - hack the project on github and hack another project on trac - I'd say majority will choose the github

I personally couldn't do much else in 1 hour per week that I'm spending
on it anyways. Andrew likes it and it does make him productive: check
the git log if you need statistics. Do you think that if the tracker is
migrated to Github, I will magically be able to review 500 tickets in
this 1 hour per week or what?

You did get me wrong. I'm not saying about your personal productivity. Sorry for miscommunication. There are lot of people saying that dvc is bad because they used to share their work by sending *.patch files and they don't need anything else. Does that mean that it works well for other team members? I'd say no.
What I meant is that it's much easier to work with PRs instead of patch files. It's easier for community to help to review these requests. It's easy to newcomers to get to it...

 

A valid reason for moving in my opinion would be to reduce reliance on
privately owned stuff, and I have been slowly working in this direction,
and hope to take it further in the future, but other than that, I see no
other reasons currently to do so.

On Sat, 2015-05-30 at 08:23 -0400, Volodymyr Buell wrote:
> I see that you not so interested in migration as you didn't answer my
> question in private.

It's not just a matter of interest; realistically, I can scrap up to 5
hours per week for mc, which means process the mailing list ~2 times per
week; processing huge emails full of very questionable content by some
posters takes hours, so there we are. I saw your mails among others, and
I'll try to reply tomorrow.

Now in what concerns the interest, yes, it is low. For once I
wholeheartedly agree with Oswald. There need to be some very important
advantage in the migration, and if we go for it, it should be done
properly.

One advantage could be that person X steps up and shows enough
commitment to prepare a migration like Slava did, and which was later
completed by Oswald. He also declares it as a pre-requisite for him
taking over and investing serious time in the project.

Under these circumstances, I can stick my own (very negative) opinion of
Github issue tracker somewhere deep down, and accept that the tools are
chosen by those people who do the real work. If they like Github issues
and they make them productive, so be it.

But I don't buy unsubstantiated arguments about magical community of
productive and qualified members appearing out of nowhere, and doing
quality code review over large spans of time. Instead, what will happen
is that Github issue tracker will become just as dead swamp of issues
and patches, as Trac has become now. I've been part of too many
projects, and I know how successful open source projects work: there is
a lot happening behind the scenes.

--
Sincerely yours,
Yury V. Zaytsev

--

Thanks,
Volodymyr