2:32 p.m.
I want to start discussing a plan to integrate the master and metamodel
branches together.
One point of discussion is whether we want to pull master over on to
metamodel on a regular basis or whether we want to just wait until the
"very end" and do a single painful integration. I did some searches and
found quite a few recommendations for ongoing integration; however, they
all specifically mentioned merges over rebasing.
Another point of discussion is merging test and matrix back together. I
specifically asked this on StackOverflow to get other perspectives on
it:
http://stackoverflow.com/questions/10888724/merging-content-of-2-director...
Still not sure of the right answer there if we decide to go for the
one-big integration event. If we do decide to move to regular
integrations from master to metamodel then I think we should merge test
and matrix back together immediately and do an integration.
--
steve(a)hibernate.org
http://hibernate.org
3:30 p.m.
On Jun 5, 2012, at 2:32 PM, Steve Ebersole wrote:
I want to start discussing a plan to integrate the master and
metamodel
branches together.
One point of discussion is whether we want to pull master over on to
metamodel on a regular basis or whether we want to just wait until the
"very end" and do a single painful integration.
+1 for regular integration
I did some searches and
found quite a few recommendations for ongoing integration; however, they
all specifically mentioned merges over rebasing.
I think now we definitely would need a merge (basically what the guys on
stack overflow recommend)
After the initial merge I think we could to rebases for the regular integrations
if we do them often enough. It will require some discipline from everyone, since
we will need to coordinate these rebases. One a rebased metamodel is pushed
to the main repo everyone will need to make sure not to continue working of this.
I guess what I am saying is that there is a little more effort for everyone involved
here.
I think it would be worth it though.
+1 for getting test + matrix together asap.
--Hardy
3:56 p.m.
Would we merge the directories on each branch and then merge branches?
From what little I know of git I think this would be best.
On
Jun 5, 2012 8:30 AM, "Hardy Ferentschik" <hardy(a)hibernate.org> wrote:
On Jun 5, 2012, at 2:32 PM, Steve Ebersole wrote:
> I want to start discussing a plan to integrate the master and metamodel
> branches together.
>
> One point of discussion is whether we want to pull master over on to
> metamodel on a regular basis or whether we want to just wait until the
> "very end" and do a single painful integration.
+1 for regular integration
> I did some searches and
> found quite a few recommendations for ongoing integration; however, they
> all specifically mentioned merges over rebasing.
I think now we definitely would need a merge (basically what the guys on
stack overflow recommend)
After the initial merge I think we could to rebases for the regular
integrations
if we do them often enough. It will require some discipline from everyone,
since
we will need to coordinate these rebases. One a rebased metamodel is pushed
to the main repo everyone will need to make sure not to continue working
of this.
I guess what I am saying is that there is a little more effort for
everyone involved here.
I think it would be worth it though.
+1 for getting test + matrix together asap.
--Hardy
4:11 p.m.
On Jun 5, 2012, at 3:56 PM, Steve Ebersole wrote:
Would we merge the directories on each branch and then merge
branches? From what little I know of git I think this would be best.
What's about doing first a merge keeping test and matrix and then combining
test+matrix on master and do another merge/rebase?
4:17 p.m.
+1 for early merges and Hardy's suggestion to merge first w/o combining test &
matrix, then again after combining.
On Jun 5, 2012, at 9:11 AM, Hardy Ferentschik wrote:
On Jun 5, 2012, at 3:56 PM, Steve Ebersole wrote:
> Would we merge the directories on each branch and then merge branches? From what
little I know of git I think this would be best.
What's about doing first a merge keeping test and matrix and then combining
test+matrix on master and do another merge/rebase?
_______________________________________________
hibernate-dev mailing list
hibernate-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/hibernate-dev
JPAV
5:27 p.m.
If we merge (not rebase) then we should be fine to do it without
everyone having to push first. So lets set a time for this to happen.
I really want to hold off branching for JPA 2.1 implementation work
until this happens, so the sooner the better. Tomorrow (6/5)?
Out of curiosity, why 2 merges?
That brings up whether we use rebase to integrate metamodel back into
master whenever we are ready. In my experience (based on pull requests
with merges of master) the rebase is really smart about weeding out
merge commits. Or do we just live with the merge commits on the main
code line?
Also, we should discuss branching 4.1 and moving metamodel to master.
I think that is really the situation we have now (only bug fixes on
master, feature dev on metamodel).
On Tue 05 Jun 2012 09:17:06 AM CDT, John Verhaeg wrote:
+1 for early merges and Hardy's suggestion to merge first w/o
combining test& matrix, then again after combining.
On Jun 5, 2012, at 9:11 AM, Hardy Ferentschik wrote:
>
> On Jun 5, 2012, at 3:56 PM, Steve Ebersole wrote:
>
>> Would we merge the directories on each branch and then merge branches? From what
little I know of git I think this would be best.
>
> What's about doing first a merge keeping test and matrix and then combining
test+matrix on master and do another merge/rebase?
>
>
>
>
>
> _______________________________________________
> hibernate-dev mailing list
> hibernate-dev(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/hibernate-dev
JPAV
--
steve(a)hibernate.org
http://hibernate.org
5:58 p.m.
On Jun 5, 2012, at 5:51 PM, John Verhaeg wrote:
On Jun 5, 2012, at 10:27 AM, Steve Ebersole wrote:
> Out of curiosity, why 2 merges?
I was just thinking it might help reduce the number of potential conflicts during the
first merge. If not, then one is fine in my mind.
Same here. Since you mentioned that the test directories have already partly diverged I
thought it would be easier to do things step by step.
Personally I think this is the more natural approach anyways. If we sort out a merge now
the collapsing of the test directories should be straight
forward then.
I guess one could run some quick experiments to see how bad things are. Personally I would
start my attempts w/ a merge first :-)
--Hardy
4336
days inactive
4336
days old
7 comments
3 participants
participants (3)
-
Hardy Ferentschik -
John Verhaeg -
Steve Ebersole