[Dev] Code collaboration platform GitLab acquires rival Gitorious, will shut it down on June 1
fabio at pesari.eu
Wed Mar 4 20:07:46 GMT 2015
I would prefer Ruby too but the fact that Gogs is written in Go is not a
problem, in my opinion. Go was developed by Google, true, but it's Free
Software and there is even a GCC frontend for it. Aside from maybe
patents, Google can do nothing to halt the distribution of Go's source
code. And in this particular case, the language is so simple I think a
Ruby programmer would not have any problems using it, especially if you
know a bit of C.
The broad implication of your concern is important though - software
like this should be written in a popular language. Go is a bit of a hip
language right now but I doubt it will become as popular as Python or
Ruby are - people will soon realize that there are many proper use cases
for Go but they are almost all related to concurrency, as Go's
highly-touted performance is not that impressive compared to Scala or
Haskell, which are also higher-level (but incredibly more complex).
Fortunately, as I said, Go is an extremely simple language and suited
for back-end work, so I think that in this case it makes an acceptable
choice, at least compared to other less popular languages (such as OCaml).
I agree with you that permissive licenses should not be supported by
Free Software activists, as they inevitably end up powering proprietary
software, but at the same time MIT and BSD software is still Free
Software and I'd rather have that than proprietary software. It's really
hard to convince some people to adopt the GNU licenses, especially when
there's money involved, and open source supporters have their own agenda
which is incompatible with our ideals.
But if those people want to make their software cross-platform and that
does not have any impact on GNU/Linux or BSD or Plan 9 or any other Free
OS then I say let them do it. Besides, these days it's really hard to
write GNU/Linux specific code. Mac OS is a descendant of UNIX and
Windows has Cygwin. Even if they decided to use the freest language out
there, Emacs Lisp, their code would still be cross-platform; Go's static
linking does make things easier but not much, plus who in their right
minds would run a server on Mac OS or Windows?
On 03/04/2015 04:37 PM, hellekin wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA512
> On 03/04/15 11:07, Digit wrote:
>> https://notabug.org/ fulfills many of those criteria listed by Fabio,
>> and seems the best option around (so far, given an absence of other
>> freedom respecting git webware).
> *** It's "MIT" as well, and written in Go, a language created by Google.
> I don't have anything against Go, I like and use Pond for example, but
> I can't hack it, and I can hack Ruby.
> "The goal of this project is to make the easiest, fastest, and most
> painless way to set up a self-hosted Git service. With Go, this can be
> done via an independent binary distribution across ALL platforms that Go
> supports, including Linux, Mac OS X, and Windows."
> So they chose Go to support proprietary software. I don't see that as a
> better option for software freedom that the Gitlab "open core" model. If
> Gitlab stands with the FSF, they have more chances than say, GNU
> Savannah, to be a proper replacement platform than Github.
> "What's wrong with MIT License?" is the question to tackle. When your
> whole stack is "MIT", it's a logical thing to do to keep building on the
> same license: that's what Gitlab does. Gogs.io licensed their software
> as "MIT" as well, although Go itself is licensed under the terms of "a
> BSD-style license". In both cases, the vocabulary does not support the
> FSF view.
> The Github-Reddit-Twitter-Google-Facebook-Apple-Amazon storm cloud is
> going to rain on us any day if there's no alternative. Why is it that
> free software developers do not support the FSF?
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v2
> -----END PGP SIGNATURE-----
> Dev mailing list
> Dev at lists.parabola.nu
More information about the Dev