Re: Moving pkgsrc-wip away from SourceForge

[Aleksej Saushev <asau%inbox.ru@localhost>]

Benny Siegert <bsiegert%gmail.com@localhost> writes:

> On Sat, Jul 11, 2015 at 4:22 PM, Aleksej Saushev <asau%inbox.ru@localhost> wrote:
>> First, I find it annoying that some minority of people decide for others
>> without giving any chance to speak.
>
> The whole point of this mail was to get a discussion going, a
> discussion in which you are taking place. You have been given a chance
> to speak, and you have spoken. Is that not enough?

If you read your original mail again, you'll see that it is written in a
form that suggests that consensus has been reached. It is written in a
form as to notify others rather than in a form to start a discussion.

>> Only because of that "so let's do this"
>> must be rejected. Even if this is a consensus of people in the room,
>> it may not be a consensus in the whole community.
>
> Stop talking in hypothetical terms. Are you against this change? Which
> part? VCS change or hoster change?

I don't see rationale neither for VCS change nor for hoster change,
and I'd like to see it.

I strongly object to converting to git  for following reasons:
1. git doesn't support partial checkouts and other partial operations.
2. git doesn't support interoperating with pkgsrc the way CVS does.
(In particular, "cvs up" updates pkgsrc and wip transparently along the trunk.
git doesn't do that.)
3. Merge abilities of git are total mess. Worse than CVS. (Yes, I've heard
all those lies many times. I cannot count merging head.1 with pom.6 as
"better merge support.")

1 and 2 apply to other similar version control systems. Changeset-only VCS
is very horrible fit for pkgsrc and pkgsrc-wip.

>> Second, when moving away from sourceforge was discussed? Has it ever been
>> discussed at all?
>
> It has been discussed several times. The most recent one happened over
> on tech-repository but the writing has been on the wall for longer.

I have checked the list archives, and what I see that this is the first time
or nearly the first time this is discussed. Whatever happened on tech-repository
(if happened at all) is irrelevant here. We're not talking about NetBSD right now.

>> Third, how is moving away from sourceforge (to be discussed still) related
>> to changing VCS? Moving CVS repository is a matter of copying files,
>> moving to another VCS is not.
>
> True. But coalescing both changes would save work and reduce the
> number of disruptions.
>
>> Fourth, your "b" is not evident. In fact, this is not even a fact yet.
>
> This was a quote from someone at the table.

Then you shouldn't have presented it as an evident fact.
It would be a lot better, if such "facts" were either given with some
supporting evidence or left for elaboration by those who presented them.

>> Where and when was it presented in public?
>
> At pkgsrccon 2015.

This is not public by any means. There's a number of people actively
using pkgsrc-wip who have never attended pkgsrcCon. The number of people
who didn't attend pkgsrcCon 2015 is even larger. Because of this
discussion happening at pkgsrcCon is essentially the same as behind
closed doors.

>> How did you determine that
>> the number of those contributions is not a noise?
>
> I did not, so far, but one look on the page would tell you that they are not.

"One look" doesn't demonstrate anything. You do understand that letting
opponent to hunt your "evidence" down is at least not ethical, don't you?
So far your claim is to be considered false until you demonstrate it in
some sensible way.

>> Fifth, I don't see how Joyent is fundamentally different from myself in
>> receiving and processing change requests for pkgsrc-wip. I receive them
>> by mail and on IRC (sometimes via paste service, sometimes directly),
>> Joyent may receive them by any other means.
>
> This is true, but I don't see how that applies to the discussion
> above. The thing I was trying to say is:
>
> The github workflow (fork, commit, pull request) is in fact something
> that people _are_ using to try to contribute changes to wip. So,
> considering that we want these changes and that we want the barrier to
> entry to be as low as can be, we should be offering this way of
> accepting contributions. Does this make more sense?

Yes, and I don't see how this prevents Joyent from passing changes to
pkgsrc-wip right now. Pull requests are not commits, so they have to act
as a gateway anyway. It is the same as me accepting changes in personal
mail or pastebin service. Yes, some people do use mail and pastebin service
to contribute to wip.


-- 
HE CE3OH...