cloudstack-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Noah Slater <nsla...@apache.org>
Subject Re: [DISCUSS] Authorship of translations
Date Sat, 04 May 2013 13:58:00 GMT
(Copied from elsewhere...)

I've been involved in similar discussions about what to do with pull
requests on GitHub, etc. I think the general consensus was that as long as
there is a reasonable indication that the work was being contributed to the
project, then we are okay to include it. i.e. If somebody submits a PR to
the CloudStack mirror, then we can include that if we want to, without
doing any other checks. But if we spot a PR on a Citrix GitHub repository,
that we could apply to CloudStack, we need to contact the original author,
to make sure we have permission. The key being that we must establish
reasonable intent. And I think we have that in the scenario you describe.

In fact, we used to have a checkbox in JIRA that you used to have to tick
to indicate when you were uploading that indicated were giving permission
for the project to include your work. We removed that checkbox a while
ago. I believe we took that action, because there was consensus that
attaching a patch established intent.

As for authorship. From a legal/policy perspective, author information
should be kept out of source files. There are various reasons for this. But
the gist is that it can give the impression that individual people "own"
the various bits of code. And obviously, this can discourage participation.
This is why all Apache source files state copyright as "The Apache Software
Foundation", meaning "the lot of us", i.e. shared.

Now, I can appreciate that PO files might be a bit different. I took a look
at a few of them, and I don't see a problem from a policy perspective,
especially if these is standard meta-data, or helps the translation effort.
Now understanding how translations take place, I would ask: might having
the last translator name in there discourage other translators from
participating? As you mention, these files are machine generated, so my
guess is: no.

So, I think all that remains is a stylistic question. How do we want to
attribute the hard work and dedication of our translation team? I know that
with code, this is often done with an Author tag in Git, or in the comment,
or what have you. But there are other options. What about a THANKSfile?

On Apache CouchDB, our THANKS file is actually maintained in two ways. For
anything that does not come in with an Author tag in Git, we put it in the
file manually. For anything in Git, we actually automate that.

See:

https://git-wip-us.apache.org/repos/asf?p=couchdb.git;a=blob_plain;f=bootstrap;hb=HEAD

Grep for "THANKS" to see the code that updates the file. This is done when
we are preparing a release artefact. Which might not work for us on
CloudStack, as, presently, our release artefacts are pristine copies of our
Git repository. But we might consider making a script that updates THANKS,
and then checking in the changes.


On 27 April 2013 01:00, David Nalley <david@gnsa.us> wrote:

> Why don't we ask them their preference (keeping dev@ in the loop as well)
>
> --David
>
> On Fri, Apr 26, 2013 at 3:39 PM, Sebastien Goasguen <runseb@gmail.com>
> wrote:
> >
> > On Apr 26, 2013, at 11:22 AM, David Nalley <david@gnsa.us> wrote:
> >
> >>>> As such - I do not see a material difference in how the projects that
> >>>> are already using translate.a.o and how we function.
> >>>>
> >>>
> >>> Do we bring it up to legal-discuss ? I am happy to do so.
> >>>
> >>
> >> What question would we ask?
> >> I see two possible questions, let me know if that isn't the case.
> >>
> >> If the question is 'Is accepting contributions from a plethora of
> >> contributors to a project specific instance an acceptable way of doing
> >> business' I think the  answer is obvious that translate.a.o does
> >> exactly that mechanism and there seem to be no issues from a process
> >> standpoint.
> >>
> >> If the question is 'Can the Transifex Apache CloudStack l10n projects
> >> serve as an official contribution point' - I personally am comfortable
> >> saying that the message is currently clear that we treat them as
> >> official. I also don't see a problem with doing so. Is this a point of
> >> contention with anyone else? Is there a problem there that I am not
> >> seeing?
> >>
> >> Is there another question?
> >
> > I am fine with your statements and have no questions for legal-discuss.
> >
> > I was merely bringing it up in the open to make sure people knew about
> it.
> >
> > The only issue left IMHO is how we ack the authors of translations in
> git ?
> >
> > -sebastien
> >
>



-- 
NS

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message