Buildroot Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Yann E. MORIN <yann.morin.1998@free.fr>
To: buildroot@busybox.net
Subject: [Buildroot] Buildroot, github and nopullrequest.com
Date: Sat, 16 Jul 2016 18:46:04 +0200	[thread overview]
Message-ID: <20160716164604.GA3882@free.fr> (raw)
In-Reply-To: <CA+TH9VnO2kF00pQJB3K_f6NcRboq3afw-MmxM1sNHxXWpjdThw@mail.gmail.com>

Angelo, All,

On 2016-07-16 14:44 +0200, Angelo Compagnucci spake thusly:
> Personally, I'm really sad that the buildroot infrastructure is
> currently down from a few days, buildroot is such a wonderful project
> that not merits such a shame.

Yes, DDoS are painful... :-(

> Obviously, the infrastructure behind buildroot is not up to date to
> the current standards, I think we should change.

That's something that had been floating for the past year or so...

> So, I'm drafting a tentative approach, and I'm sharing it with you.
> Obviously I'm not entitled in acting in any way, but hey, can I dream
> for a better Buildroot?!

Your input is much welcome! :-)

We all believe self-hosting is a waste of time: it's too difficult, we
don't really need it, and it would not be much better than what we have
today. We all believe a forge of some sort is what we want to go to.

> 1) Website: we have a github account, we have github pages, we should
> switch there! The only downside here is that github pages doesn't
> suppo SSI, so we should write a simple bash script to assemble the
> website before publication. Github pages resides on their own
> repository, so we can have the website as always in the main
> repository and the bash script will assemble the website in an outside
> directory where the other repository it's cloned.

There is also Gitlab. I would favour Gitlab over github, if at least
because it is an open source project, not a closed, walled garden like
Github is. All things being equal, Gitlab is on par with github
regarding the features we need, plus a killer one: we can disable PRs
alltogether with Gitlab, which we can't with Github.

And if we were ever to re-self-host, we could take the data as-is with
our own instance of Gitlab. And even if moving to something else than
Gitlab, taking the data out of Gitlab is easier than out of Github.

> 2) We have a github account, we should use it as the main repository.
> No other words here.

Or Gitlab! ;-)

> 3) Githb pull requests. Obviously, we won't github pull requests, and
> here enters the game https://nopullrequests.com.

Or Gitlab, which allows to disable PRs altogether.

> It's a simple github
> webhook that closes a PR with a message. The source is publicly
> available on github. We can use that service as is, or run an instance
> ourselves on GAE.

As Thomas said, I'm not too comfortable with allowing a third party
automatic access to our repo. Especially since it looks like it wants a
Google account.

As for running an instance ourselves, it would mean we self-host again,
which wee don;t want.

> I had a look at the source code and it's pretty
> secure, once it's authorized on github, it receives a message on each
> pull requests and replies immediately without keeping anything
> (caching or disk). Hosting our instance on GAE will have the benefit
> on personalizing the commit message with something more specific than
> the pretty aseptic default message.
> 
> 4) Issues: well, everything is better than bugzilla!

Issues in Github are a pain IMHO: except for their "template", there is
no possibility to have arbitrary fields. Git lab is not much better in
this respect either.

And there's still the emails for each commit: neither Github nor Gitlab
can do that; they can only send an email per push, not per commit.

However, with Gitlab, it looks like we could implement it and do a PR.
With Github, there's no way we can do it.


So, lemme recap:

                            | Github            | Gitlab            |
----------------------------+-------------------+-------------------+
License                     | Proprietary       | Open Source       |
Network effect              | Definitely        | Somewhat          |
git tree                    | Yes               | Yes               |
Web pages                   | Static            | Static            |
web hooks                   | Yes               | Yes               |
C.I.                        | Yes, many [0]     | Yes, many [1]     |
emails                      | on push           | on push [2]       |
IRC                         | No?               | irker             |
Markup language             | Markdown (MD)     | Markdown (MD)     |
README                      | Yes, MD           | Yes, MD           |
Issue tracker               | Yes, minimal, MD  | Yes, minimal, MD  |
Snippets                    | Yes               | Yes               |
Export project data         | No?               | Yes               |


So there is no clear winner. The only things that would tip the scale is
the license and the possibility to have per-commit emails.

My favourite is obviously Gitlab. ;-)


[0] of which, Travis
[1] of which, Jenkins
[2] The source code is available, we could do the modifications and
    submit back to Gitlab; for now, I opened an issue there;
    https://gitlab.com/gitlab-org/gitlab-ce/issues/19901

Regards,
Yann E. MORIN.

-- 
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| +33 223 225 172 `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
'------------------------------^-------^------------------^--------------------'

  parent reply	other threads:[~2016-07-16 16:46 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-07-16 12:44 [Buildroot] Buildroot, github and nopullrequest.com Angelo Compagnucci
2016-07-16 13:09 ` Thomas Petazzoni
2016-07-16 13:58   ` Angelo Compagnucci
2016-07-16 14:08     ` Thomas Petazzoni
2016-07-16 14:27       ` Angelo Compagnucci
2016-07-16 16:46 ` Yann E. MORIN [this message]
2016-07-16 16:52   ` Angelo Compagnucci
2016-07-16 16:57     ` Angelo Compagnucci

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20160716164604.GA3882@free.fr \
    --to=yann.morin.1998@free.fr \
    --cc=buildroot@busybox.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox