On 10-Aug-08, at 9:05 PM, Milos Kleint wrote:
Jason,
The issues I'm finding (or my userbase actually) are not with mevenide
integration. It's also not something I could test on my side. It's in
99% of cases incompatibilities with 2.0.x. And it's not a reoccuring
pattern, no trunk-to-trunk regressions. So no test could save me from
it anyway. And *if* I wrote tests, it would be in maven not mevenide.
That's where it belongs IMHO.
So please don't make it sound like it's my own fault anyway.
I'm not trying to make it sound like your fault. The specific case you
mentioned was a trunk-to-trunk regression so that's what I based my
message. Sorry, if it sounded like I was trying to pin it on you.
There is a bigger problem with digression between the trunk and what's
released and a lot of this surfaces in plugins. That is rapidly being
addressed in the work going on with testing the plugins and setting up
a set of plugin ITs we can run against trunk to find problems. I bet
when we run them we'll have a lot of information to work with.
So I'm definitely not trying to pin anything on you. We have had a
hard time automating tests in m2eclipse, and after creating
integration with Tycho to the point where we can automate them without
PDE headless build we are in good shape and it will be harder for
regressions to slip through.
So it appears there will be a mediating 2.1.x branch, and if you're
willing to take what's on trunk (or at least work off it) then I'm
happy to cut a release as raw as it is. It's really only going to be
for embedder users in the next few weeks anyway. We'll make it clear
what we're doing and we'll have to work fast to get some parity. I
think with a plugin IT harness in place all sorts of crap will fall out.
more inline
On Sun, Aug 10, 2008 at 10:18 PM, Jason van Zyl <[EMAIL PROTECTED]>
wrote:
I think having the intermediary bridge is a good idea, and I would be
comfortable finding the last stable version of trunk that works with
Mevenide and then release that and then leave that as a stable
branch for
you to work off.
One of the problems is that your code seems not to be very testable
from
your own description and there's nothing we could run to validate
changes in
trunk haven't busted you without you manually testing. You have to do
something about that because asking you to manually try things isn't
tenable. We'll make the stable branch of 3.0.x, and then we can
leave that
pretty much static except for bug fixes you want to put in there.
I need a release just as badly as you for the Eclipse IP process.
So if you
we can match what you're using and find that point in time where
you're
happy we'll roll back trunk to there, cut a 3.0.x branch and you
will have
something stable.
well, any of the 6 or so snapshots I've used in the past 6-8 months
had some (different) rough edges. So from my perspective it can be
current trunk, no need to go backwards.
I want to continue getting Mercury and Shane's new project
builder in because to me that will be a massive stabilization in the
artifact mechanism and Shane is just tearing out all sort of cruft
that's
built up in the POM builder and we're just not going to be able to
create a
spec'd process, mixins support, multi-format/version support, and a
many
other things with these two changes.
It's all cool and I want it in the released bits just as you do but it
IMHO pushes the release into more distant future. That's what I was
pointing to my previous emails.
regards
Milos
Releasing trunk as 2.1 I think would be a very bad idea, but I'm
happy to
rollback/patch to whatever point in time makes you comfortable. I
would
prefer to plough along on trunk which looks like would become 3.1.x
in this
scheme. You would probably stay away from Mercury and we are really
going to
need a harness from you we can run. The ITs will get better and be
protection at the CLI level but you seem to keep getting bitten at
the
embedder level.
Let me know what you would like to do.
On 10-Aug-08, at 10:37 AM, Milos Kleint wrote:
On Sat, Aug 9, 2008 at 11:32 AM, Mauro Talevi
<[EMAIL PROTECTED]> wrote:
Milos Kleint wrote:
please, please, let's not add anything else to trunk (2.1) and
stabilize it. I've been waiting for a stable embeddable version
for 2
years and with the number of work (complete rewrites of
everything)
in the branches, a stable maven.next looks years ahead again.
Not having an embeddable maven that works in the IDE integrations
hurts the adoption and trust of users.
Milos, isn't just a case of renaming trunk/2.1 to trunk/3 ?
well, the version number itself is of little interest to me, but I
see
a lot of people cooking new stuff at branches. I suppose the
intention
is to get this code into trunk. The question for me is wheher it
gets
into trunk before the maven.next is released or after (be it 2.1 or
3.0 ). The maven.next that's interesting to me is the version that
is
embeddable.
Also cutting an alpha or beta would enable IDE devs to work to that
without
having sleepless nights about stabilisation.
Well, if the alphas and betas get cut from a stable branch that will
ultimately become the next final release, I'm cool with it.
Milos
Cheers
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
Thanks,
Jason
----------------------------------------------------------
Jason van Zyl
Founder, Apache Maven
jason at sonatype dot com
----------------------------------------------------------
We know what we are, but know not what we may be.
-- Shakespeare
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
Thanks,
Jason
----------------------------------------------------------
Jason van Zyl
Founder, Apache Maven
jason at sonatype dot com
----------------------------------------------------------
You are never dedicated to something you have complete confidence in.
No one is fanatically shouting that the sun is going to rise tomorrow.
They know it is going to rise tomorrow. When people are fanatically
dedicated to political or religious faiths or any other kind of
dogmas or goals, it's always because these dogmas or
goals are in doubt.
-- Robert Pirzig, Zen and the Art of Motorcycle Maintenance
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
