Hi
Looking at
https://bugs.eclipse.org/bugs/enter_bug.cgi?classification=Modeling,
most of the renames are really obvious
EMF.* => *
EMFT.* => *
MDT.* => *
MMT.* => *
A few might be made a little more consistent
EMFT.Diffmerge => DiffMerge
EMFT.emf-store => EMFStore
EMFT.emfclient => EMFClient
and perhaps EMFT.b3 can vanish or containerize
Since this is a Modeling wide page, perhaps the PMC can do all these
in one go and so set a consistent style to be adopted when each
contained project promotes in project-specific style
e.g. TMF/Xtext => Xtext/Current, TMF/Xtext Backlog =>
Xtext/Backlog
Regards
Ed Willink
On 25/06/2013 07:31, Ed Merks wrote:
Wayne,
Comments below.
On 25/06/2013 5:51 AM, Wayne Beaton
wrote:
Hi Modeling PMC.
As you recall you initiated a Restructuring Review [1] late last
year that flattens the Modeling top-level project by removing
the mid-level "container" projects and terminated a large number
of inactive projects. I've captured the activity around this
restructuring in Bug 393862 [2].
Immediately following the review, I enumerated the required
steps and initiated the termination and archival of twenty
projects. In the months that followed, I have enumerated the
required steps and initiated some of the moves as outlined in
the the review document provided by the PMC.
As part of the Kepler reviews, I tried to get the Kepler
participants to engage in the moves initiated by the PMC. There
has been some push-back, but many of the projects have engaged
in the process. Many, however, have not.
I see.
I intend to remain very flexible in terms of the timing and the
natures of the moves. There are some corner cases, but most
projects don't need to move their downloads and source
repositories. I am in favour, for example, of phased migration
of download sites for projects that want to do that sort of
thing.
Yes, some of these things are painful to move, affecting scripts
and such.
I've been encouraging projects that don't have their own
products in Bugzilla to request them.
Yes, bugzilla in particular is used directly by the community and
needs to be more sensibly organized.
For
the most part, a lot of these "moves" are little more than
changes in the project's id (e.g. "modeling.emf.cdo" becomes
"modeling.cdo"). Still, even with just an id change, there can
be unanticipated side effects that we need to be sensitive to.
Yes, even those are things use within scripts and within query
links...
Many of the projects already have independent websites, but some
are blocked waiting for the Modeling website infrastructure to
be moved to Git/upgraded.
Yes, that's totally my fault. Reorganizing the whole website is
(will be) very time consuming...
Rather
than invest significant energy in upgrading the Modeling website
infrastructure, I recommend that projects be encouraged to
either create their own independent websites, or leverage the
PMI.
Yes, for the most part, I do expect projects to look after
themselves, with the overall modeling website infrastructure being
used as a front end for navigating and understand the overall
organization.
We
can set up as many redirects as possible to make this happen.
Reception of the PMI has been generally good and I am very
interested in extending it to cover functionality that the
modeling website infrastructure supports. I need input to make
that happen.
Even with specialized,carefully designed websites, it would still
be nice to have the kind of uniformity provided by the PMI....
For those projects that have provided move information, the
Webmaster has asked that we present the moves in batches;
apparently, there's a bit of grove to moving projects and it's
easier to do them this way. There are a few projects that are
ready and waiting. I intend to pull the trigger sometime next
week (following the Kepler release).
While I really do intend to be flexible with regard to
implementing this restructuring, it makes no sense to leave it
open-ended.
No, that's mostly my fault. I've not been setting a good
example...
Some
projects just don't care about the move and have done nothing to
assist with the process. At this point, I feel some reflection
is required.
Yes, that true.
Is moving everything in one big review was just too ambitious?
Perhaps, but in the end, everything needs to be done and it
requires a lot of work, and for many of us, it's hard to make that
the top priority with so many things calling for our attention...
How long is too long?
I suppose you decide that in the end.
Is it time to back out of the review and maybe attempt to do
this in smaller pieces?
Certainly we need to make incremental steps.
Do we really want/need to remove all of the container projects?
Yes, they're a point of confusion, and the need for controlling
the committers on them, when they don't host any source code has
proven to be a pita...
What are you going to do about the website?
I think I've mentioned to you before that one significant concern
I have is migrating this stuff to git and then ending up with all
the current mess in the history such that cloning the old big miss
will forever be a headache. You mentioned we could rewrite the
history when we're done.
If this restructuring is to continue, it needs to be driven by
the PMC. My role has to change to one of providing support.
Perhaps you and I could schedule a call to discuss some of the
initial details and we could follow that with a call of the PMC
and any interested project leads to discuss and overall strategy.
Certainly I have more time with Kepler out and the next release
far in the future to take steps now to get the ball rolling.
Thanks for your flexibility, assistance, and patience!
Thanks,
Wayne
[1]
http://wiki.eclipse.org/Modeling/project_termination_review_2012
[2]
https://bugs.eclipse.org/bugs/show_bug.cgi?id=393682
_______________________________________________
modeling-pmc mailing list
modeling-pmc@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/modeling-pmc
_______________________________________________
modeling-pmc mailing list
modeling-pmc@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/modeling-pmc
No virus
found in this message.
Checked by AVG - www.avg.com
Version: 2013.0.3345 / Virus Database: 3199/6437 - Release Date:
06/24/13
|