05.09.2013 19:20, John Arthorne wrote:
If Christian is interested in
becoming a committer
on this effort then it is very easy. If he is a committer, and
he wrote
100% of the code himself, he can commit it. Even if he is not
a committer,
we would just need to submit a CQ to get the code reviewed. If
it is written
100% by a single author, and the author is involved in the
process, this
is quite easy (but may cause some delay).
I am interested to become a committer as I want to work actively on
this project. So there should not be any problems with that.
That part you will need to resolve
with Sharon and
the rest of the IP team. If you actually want to *include*
this code and
it has uncertain history then it might become a blocker. If
you want to
make a dependency on it, where it is one of multiple scripting
engines
that you support, that is easier to deal with. Would it be
possible at
all to separate the Jython-specific piece from the rest of the
code base
or is it a fundamental part. In another Eclipse project I know
a case where
they host separate plugins/extensions that rely on "unclean"
libraries in GitHub and host the rest of the project at
Eclipse, for example.
Currently 4 script languages are included in EScript: Rhino, Jython,
Groovy and JRuby.
The Rhino bundle comes already as part of JSDT, so we have no IP
problems with that.
For each of the other languages I currenty created a plugin
containing eg the Jython interpreter. Integration in EScript is done
by another plugin that depends on the script engine classes. For
example we have an org.jython plugin and an
...escript.jython.interpreter plugin. The former may contain code
where we need to clarify licensing issues the latter contains only
code written by myself.
So I guess your second suggestion would work as we could put the
engines plugins on github sites. On the long run it would of course
be more convenient to host all necessary stuff within eclipse.
Making committers on e4 is very easy. We
typically
start an election as soon as they introduce themselves on the
e4-dev mailing
list. The election takes one week, and then in 1-2 more weeks
it is all
done.
I'd say it would be far better to host all this in an e4 repository
than starting a migration after some months.
As long as
all the commits are from people who will be committers on the
effort, it
is easy to bring that work into Eclipse a bit later.
That seems to be the case so far. We might continue on github until
we are ready to move to e4, but we should trigger an e4 setup for
our project as soon as possible.
In the meantime it would be great to know which rules we should
apply on commits, file headers and so on to make migration later on
as smooth as possible.
Christian
|