On Jun 3, 2006, at 12:13 AM, David Jencks wrote:
On Jun 2, 2006, at 8:42 PM, John Sisson wrote:
Kevan Miller wrote:
It seems that there are now +1's from 3 committers for this
change. Although I also support the change (once copyrights are
corected), I cannot offer a +1. Apologies for being a pedant, but
to my knowledge the current RTC "rules" we are living under are:
'I have applied this patch and tested it and found it good'
I'm pretty sure that there's only one person who has done this
and his name is Sachin. Although it's likely that I could apply
the patch and build devtools, I wouldn't know how to integrate
and test the change.
I think many of us feel that the requirement to apply and test a
patch is too restrictive and cumbersome -- especially in smaller
subprojects such as devtools and daytrader. Rather than ignoring
this directive, let's get the process changed or at least start a
discussion... I'll start a thread.
I will retract my +1, I mistakenly thought that the testing
requirement was relaxed, but after reviewing the mailing lists I
didn't see a response to Greg's comments on this topic:
http://marc.theaimsgroup.com/?l=geronimo-dev&m=114850042818334&w=2
Agree that we need to discuss getting the process relaxed
officially with Ken's approval.
Looks like I have to retract my +1 also. I merely spent an hour or
so carefully studying sachin's patch, but did not apply it or test
it. Based on my careful review I think that the patch should be
applied.
I support changing the process to require careful review rather
than applying and testing.
I've posted my proposal for an interpretation of an RTC +1 vote. Your
votes were consistent, IMO, with this interpretation. Here's my +1
under my interpretation of RTC.
Sachin committed his change with 3 +1's in effect. There have not
been any -1's. So, I don't see any reason for Sachin to revert his
commit. I don't think holding Sachin hostage does us any good as we
noodle over the finer points of RTC.
--kevan
thanks
david jencks
John
--kevan
On Jun 1, 2006, at 9:42 AM, Sachin Patel wrote:
The following big change is part 1 of 2 to correctly handle
loading of the deployment plan editors. To briefly summarize,
the change introduces a new extension point which requires an
implementation of IGeronimoFormContentLoader. Previously the
editor extension point loaded the editor for 1.0 plans. Now the
editor must handle multiple versions of the deployment plans so
now there is a single editor extension def which loads a single
editor impl, SharedDeploymentEditor which loads all the
IGeronimoFormContent loaders (one impl for 1.0 and another for
1.1) and delegates to the appropriate emf models to generate the
UI for the plans. With this commit, the 1.0 plans load
correctly and there is some work still need to be done for 1.1
to be done. There is currently some duplicate code which will
need to be refactored, part 2 will address this, but since the
patch is about 3000 lines I wanted to go ahead and commit.
Please vote...
Thanks.
<patch2.txt>
-sachin
