You are viewing a plain text version of this content. The canonical link for it is here.
Posted to server-dev@james.apache.org by "Noel J. Bergman" <no...@devtech.com> on 2005/04/11 16:46:30 UTC
Merging (was [VOTE] POJO pattern)
> the merging process, which took almost 2 years (until now)
Actually, not even one. And that only because I got *really* busy, and was
told that I should commit it finished rather than as a work-in-progress,
which meant that no one else was contributing to it.
> Nobody knows which is the real head branch, etc.
branch_2_1_fcs has always been the branch for developing the current code,
whereas MAIN (trunk/) was future code. We've already said that much of the
code change there would go away, except for the build environment, updates
to Avalon and Jason's work on IMAP. None of which we want to lose.
> There is nothing to lose, we don't use anything from the
> head anyway.
That is entirely wrong, as noted above. And part of the issue with the
merger is making sure I don't lose any of those important parts. Plus the
crazy schedule I've been keeping since last June.
Blame me if you want. Enough validity in that, I suppose. I recently
installed a VM on my laptop to carry around a JAMES build environment, so
that I can work when I catch time, and plan to start committing the merger
in parts, rather than as a whole, so that others can help out.
Will start tonight when I get back from a client's site. Now I have to go.
--- Noel
---------------------------------------------------------------------
To unsubscribe, e-mail: server-dev-unsubscribe@james.apache.org
For additional commands, e-mail: server-dev-help@james.apache.org
Re: Merging (was [VOTE] POJO pattern)
Posted by Danny Angus <da...@gmail.com>.
> I saw the point was more that we would not lose anything (since
> everything is in SVN) if we dumped what we have in MAIN (now trunk) and
> replaced it with branch_2_1_fcs. And, since we're agreeing on an
> evolutionary move away from our current dependencies, we could replace
> trunk with what's in branch_2_1_fcs to:
> - have code releases come off of trunk (simpler), and
> - only one branch to maintain (more efficient)
+1 the evolutionary approach seems to have support from people who
are prepared to put in the hours. So why not just swap branch_2_1_fcs
with HEAD and evolve "2"
> As tempting as it is, you have done way too much to deserve blame.
+1 Yeah way too much, the issue is should we expect you to do more or
take a pragmatic view of the current state of the project and the
ambitions of the community? I say we be realistic and don't expect
Noel to be a silver bullet.
> Now that we are building consensus and direction, I'd like to simplify
> our path to move us forward. While it is painful to just kill a branch,
> unless you suddenly become unemployed (or we massively and unfairly
> guilt-trip you into skipping work for a week), I see it as the simplest
> course of action.
+1.
Theres no shame in taking a fresh look at things and starting from
where we actually are, not where we had hoped to be.
d.
---------------------------------------------------------------------
To unsubscribe, e-mail: server-dev-unsubscribe@james.apache.org
For additional commands, e-mail: server-dev-help@james.apache.org
RE: Merging (was [VOTE] POJO pattern)
Posted by Jason Webb <jw...@inovem.com>.
> -----Original Message-----
> From: Serge Knystautas [mailto:sergek@lokitech.com]
> Sent: 11 April 2005 17:23
> To: James Developers List
> Subject: Re: Merging (was [VOTE] POJO pattern)
>
> Noel J. Bergman wrote:
> >>the merging process, which took almost 2 years (until now)
> >
> > Actually, not even one. And that only because I got *really* busy, and
> was
> > told that I should commit it finished rather than as a work-in-progress,
> > which meant that no one else was contributing to it.
>
> To be fair, Noel was doing yeoman's work for James in 2003 and 2004.
> Here we are in Sept 2003 already calling MAIN dead and giving Noel flack
> for not maintaining two branches...
> http://marc.theaimsgroup.com/?l=james-dev&m=106392166516868&w=2
>
> >>There is nothing to lose, we don't use anything from the
> >>head anyway.
> >
> > That is entirely wrong, as noted above. And part of the issue with the
> > merger is making sure I don't lose any of those important parts. Plus
> the
> > crazy schedule I've been keeping since last June.
>
> I saw the point was more that we would not lose anything (since
> everything is in SVN) if we dumped what we have in MAIN (now trunk) and
> replaced it with branch_2_1_fcs. And, since we're agreeing on an
> evolutionary move away from our current dependencies, we could replace
> trunk with what's in branch_2_1_fcs to:
> - have code releases come off of trunk (simpler), and
> - only one branch to maintain (more efficient)
As far as the IMAP work goes, the bits I've done already and that form the
basis of the IMAP implementation can be merged into any source tree. This
includes user attributes and my initial work on the new repositories. The
protocol stack will take some hack-n-slash to get it going, but once the new
way is clear I will start work again (I hope). If somebody does the POP3
server then the IMAP stuff will follow naturally as they are fairly similar
in layout terms. IMAP's just a little more complex :)
>
> > Blame me if you want. Enough validity in that, I suppose. I recently
> > installed a VM on my laptop to carry around a JAMES build environment,
> so
> > that I can work when I catch time, and plan to start committing the
> merger
> > in parts, rather than as a whole, so that others can help out.
>
> As tempting as it is, you have done way too much to deserve blame.
>
> Now that we are building consensus and direction, I'd like to simplify
> our path to move us forward. While it is painful to just kill a branch,
> unless you suddenly become unemployed (or we massively and unfairly
> guilt-trip you into skipping work for a week), I see it as the simplest
> course of action.
>
> --
> Serge Knystautas
> Lokitech >> software . strategy . design >> http://www.lokitech.com
> p. 301.656.5501
> e. sergek@lokitech.com
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: server-dev-unsubscribe@james.apache.org
> For additional commands, e-mail: server-dev-help@james.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: server-dev-unsubscribe@james.apache.org
For additional commands, e-mail: server-dev-help@james.apache.org
Re: Merging (was [VOTE] POJO pattern)
Posted by Serge Knystautas <se...@lokitech.com>.
Noel J. Bergman wrote:
>>the merging process, which took almost 2 years (until now)
>
> Actually, not even one. And that only because I got *really* busy, and was
> told that I should commit it finished rather than as a work-in-progress,
> which meant that no one else was contributing to it.
To be fair, Noel was doing yeoman's work for James in 2003 and 2004.
Here we are in Sept 2003 already calling MAIN dead and giving Noel flack
for not maintaining two branches...
http://marc.theaimsgroup.com/?l=james-dev&m=106392166516868&w=2
>>There is nothing to lose, we don't use anything from the
>>head anyway.
>
> That is entirely wrong, as noted above. And part of the issue with the
> merger is making sure I don't lose any of those important parts. Plus the
> crazy schedule I've been keeping since last June.
I saw the point was more that we would not lose anything (since
everything is in SVN) if we dumped what we have in MAIN (now trunk) and
replaced it with branch_2_1_fcs. And, since we're agreeing on an
evolutionary move away from our current dependencies, we could replace
trunk with what's in branch_2_1_fcs to:
- have code releases come off of trunk (simpler), and
- only one branch to maintain (more efficient)
> Blame me if you want. Enough validity in that, I suppose. I recently
> installed a VM on my laptop to carry around a JAMES build environment, so
> that I can work when I catch time, and plan to start committing the merger
> in parts, rather than as a whole, so that others can help out.
As tempting as it is, you have done way too much to deserve blame.
Now that we are building consensus and direction, I'd like to simplify
our path to move us forward. While it is painful to just kill a branch,
unless you suddenly become unemployed (or we massively and unfairly
guilt-trip you into skipping work for a week), I see it as the simplest
course of action.
--
Serge Knystautas
Lokitech >> software . strategy . design >> http://www.lokitech.com
p. 301.656.5501
e. sergek@lokitech.com
---------------------------------------------------------------------
To unsubscribe, e-mail: server-dev-unsubscribe@james.apache.org
For additional commands, e-mail: server-dev-help@james.apache.org
Re: Merging (was [VOTE] POJO pattern)
Posted by ap...@bago.org.
> branch_2_1_fcs has always been the branch for developing the
> current code, whereas MAIN (trunk/) was future code. We've
> already said that much of the code change there would go
> away, except for the build environment, updates to Avalon and
> Jason's work on IMAP. None of which we want to lose.
IMHO the build environment and IMAP work are the first one to be merged.
As I said before the Avalon update (component=>service) should be moved to
the specialized Avalon-aware components. It will be easiest to re-refactor
avalon stuff during the extraction from the POJOs. Noel, what do you think
about this sentence?
Will you/we be able to encapsulate avalon components/services in the
AvalonPOJO wrappers?
Stefano
---------------------------------------------------------------------
To unsubscribe, e-mail: server-dev-unsubscribe@james.apache.org
For additional commands, e-mail: server-dev-help@james.apache.org