[emf-dev] Re: [modeling-pmc] Re: How to Mange Modeling Subprojects

[Ed Merks <merks@xxxxxxxxxx>]

Hi,

Given Bjorn's comments, I think it would
be beneficial to keep EMFT around, either as a Technology subproject as
it is today, or as a Modeling subproject.  It doesn't really seem
to matter which, so the path of least resistance is to leave it where it
is today.   In either case, it is a subproject that is permanently
incubating and its purpose is to contain any new component that we'd like
to add to EMF eventually.  Such a new component will start out in
EMFT until it finishes incubating (it passes the release/incubation-exit
review), at which point it will graduate and become a component in the
EMF subproject, i.e., it will be moved.  This way the EMF subproject
and its CVS modules will not ever contain incubating components and the
incubating components in EMFT can take advantage of the flexible new parallel
IP process.   We can keep both newsgroups and mailing lists alive
as well, because that seems to be working nicely so far.   Is everyone
okay with this approach?  I've added both the EMF and EMFT mailing
lists to the note, so that everyone has a chance to comment.


Ed Merks/Toronto/IBM@IBMCA
mailto: merks@xxxxxxxxxx
905-413-3265  (t/l 969)

Bjorn Freeman-Benson <bjorn.freeman-benson@xxxxxxxxxxx>
Sent by: modeling-pmc-bounces@xxxxxxxxxxx

02/09/2007 10:19 AM

Please respond to
PMC members mailing list <modeling-pmc@xxxxxxxxxxx>

To	modeling-pmc@xxxxxxxxxxx
cc
Subject	[modeling-pmc] Re: How to Mange Modeling Subprojects

Nick,
So, for example, if we start new components incubating
in EMFT, then 
move them later into Modeling/EMF, that's acceptable (even given the 
extra churn involved in moving projects, builds, UM sites, RSS feeds, 
etc.) ? 
Yes.
Of course what you point out so clearly is that we need is a way to separate
the physical location of a project (urls and repositories) from its incubation
status. I would be open to proposals where you start a component in EMFT
but that it's physical location starts in the EMF tree - we could float
that one by the membership and see if anyone objects - I can't imagine
that they would. If we can make that work, then the IT overhead of starting
a new component in the Incubation Phase EMFT would be much lower.
Or would we need to create brand-new tech projects for
each 
and every new incubating project, let it grow for a year, exit 
incubation via release review, and THEN move from 
/technology/<whatever> into /modeling/emf/<whatever>? 
What a pain that would be (for you; I'm fine with it).

Is EMFT going 
to exist as a place to incubate EMF technologies? Or is it being 
sunset after I move its remaining 9 projects into MDT, EMF, and M2T? 
I thought that the goal of EMFT was to start new technologies
so I'm perfectly content leaving it for that - it's been working well,
I think.
If we can start in EMFT that's not *too* painful; if we
must start 
fresh, that seems like extra pain just to allow the parallel IP 
process. I'm guessing that starting as a tech project and having to 
move might in fact be slower  (and certainly more work for me!) than
simply not being able to use the parallel IP process. (Please correct 
me if I'm way off here.) 
That's entirely up to you. I personally think the Parallel
IP process is a huge win - it can save months on the Legal due diligence
process. However, if we can do "most of the physical files in Modeling,
project 'existence'/web pages in EMFT", would that the best of both
of our worlds?

P.S. Of course you'll need the EMFT website to be conforming before you
can use the Parallel IP process - conforming means showing the green alien
on the project home page and the project download page - see instructions
[1,2].

- Bjorn_______________________________________________
modeling-pmc mailing list
modeling-pmc@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/modeling-pmc