proposal for cleaning up 2.x series releases / trees

[BRIAN FOX-5]

I'm -0 on the 2.0.11 release.

On Tue, Jun 30, 2009 at 11:10 PM, Brett Porter<brett@...> wrote:

>
> On 01/07/2009, at 6:01 AM, Brian Fox wrote:
>
>> On Tue, Jun 30, 2009 at 11:58 AM, Brett Porter<brett@...> wrote:
>>>
>>>
>>> On 01/07/2009, at 1:47 AM, nicolas de loof wrote:
>>>
>>>> I'm also fine with this, just would like to avoid some "EOL" tag on 2.0
>>>> that
>>>> may be considered as lack of support by some corporate users using (old)
>>>> maven releases
>>>
>>> Sure, we can use a different name. All I meant EOL to mean here was that
>>> we
>>> don't plan to make any more releases (unless something is found to be
>>> really, horribly, wrong). EOD (end of development) is probably more
>>> appropriate.
>>>
>>
>> Exactly, and IMO, we're at that point today with 2.0.10
>
> Ok, but are you leaning towards a -0 or a -1 on a 2.0.11 release?
>
> I'm happy to burn the small amount of my time on it and clean up the release
> process along the way (given the issues we had with 2.2.0).
>
> I'm not looking to add any more changes, just release the 37 already merged
> in there so we have a proper end point. It should be a short cycle since
> it's stuff already in 2.1.0+, but there were a couple of critical ones (eg,
> POM plugin ordering regression) that are worth having IMO.
>
> Cheers,
> Brett
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@...
> For additional commands, e-mail: dev-help@...
>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Jörg Schaible-2]

Brian Fox wrote:

> Yeah get rid of it. Is there really demand for the fixed in 2.0.11? I
> feel like it's EOL now.

The point is, in 6 months nobody knows axaclty anymore what is in
2.0.11-SNAPSHOT. That will actually stop any bugfix release ever.

- Jörg


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Paul Benedict-2]

My preference is to release 2.0.11 as it is now (37 issues fixed). The
remaining issues should move to 2.2.1. If critical bugs remain in
2.0.x, then build build a 2.0.12 issue list as people require it.

- Paul

On Wed, Jul 1, 2009 at 10:33 AM, Jörg Schaible<joerg.schaible@...> wrote:

> Brian Fox wrote:
>
>> Yeah get rid of it. Is there really demand for the fixed in 2.0.11? I
>> feel like it's EOL now.
>
> The point is, in 6 months nobody knows axaclty anymore what is in
> 2.0.11-SNAPSHOT. That will actually stop any bugfix release ever.
>
> - Jörg
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@...
> For additional commands, e-mail: dev-help@...
>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[John Casey-5]

+1

Paul Benedict wrote:

> My preference is to release 2.0.11 as it is now (37 issues fixed). The
> remaining issues should move to 2.2.1. If critical bugs remain in
> 2.0.x, then build build a 2.0.12 issue list as people require it.
>
> - Paul
>
> On Wed, Jul 1, 2009 at 10:33 AM, Jörg Schaible<joerg.schaible@...> wrote:
>> Brian Fox wrote:
>>
>>> Yeah get rid of it. Is there really demand for the fixed in 2.0.11? I
>>> feel like it's EOL now.
>> The point is, in 6 months nobody knows axaclty anymore what is in
>> 2.0.11-SNAPSHOT. That will actually stop any bugfix release ever.
>>
>> - Jörg
>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: dev-unsubscribe@...
>> For additional commands, e-mail: dev-help@...
>>
>>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@...
> For additional commands, e-mail: dev-help@...
>

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Christian Schulte]

Paul Benedict schrieb:
> My preference is to release 2.0.11 as it is now (37 issues fixed). The
> remaining issues should move to 2.2.1. If critical bugs remain in
> 2.0.x, then build build a 2.0.12 issue list as people require it.
>
> - Paul
>

+1

2.0.x is the last JDK 1.4 release. Users of the GPG plugin simply cannot
use 2.1.x.

--
Christian


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[brettporter]

Ok, for starters I've moved all the open issues from 2.0.11 to 2.2.1  
and am now going through them to cull them down where possible.

I've also confirmed that the ITs pass for 2.0.11-SNAPSHOT as it is.

Once I get the 2.1.x bits cleaned up (per original mail that everyone  
seems in favour of), I'll spin an RC and see what everyone thinks.

- Brett

On 02/07/2009, at 1:46 AM, John Casey wrote:

> +1
>
> Paul Benedict wrote:
>> My preference is to release 2.0.11 as it is now (37 issues fixed).  
>> The
>> remaining issues should move to 2.2.1. If critical bugs remain in
>> 2.0.x, then build build a 2.0.12 issue list as people require it.
>> - Paul
>> On Wed, Jul 1, 2009 at 10:33 AM, Jörg  
>> Schaible<joerg.schaible@...> wrote:
>>> Brian Fox wrote:
>>>
>>>> Yeah get rid of it. Is there really demand for the fixed in  
>>>> 2.0.11? I
>>>> feel like it's EOL now.
>>> The point is, in 6 months nobody knows axaclty anymore what is in
>>> 2.0.11-SNAPSHOT. That will actually stop any bugfix release ever.
>>>
>>> - Jörg
>>>
>>>
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail: dev-unsubscribe@...
>>> For additional commands, e-mail: dev-help@...
>>>
>>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: dev-unsubscribe@...
>> For additional commands, e-mail: dev-help@...
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@...
> For additional commands, e-mail: dev-help@...
>

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Jason van Zyl-5]

On 1-Jul-09, at 9:47 AM, Brett Porter wrote:

> Ok, for starters I've moved all the open issues from 2.0.11 to 2.2.1  
> and am now going through them to cull them down where possible.
>

You need to leave the bugs raised against 2.0.x because there is no  
way around the fact that 2.0.x is going to be the dominant version  
used for quite some time. We can't just stop bug fixing the 2.0.x  
line. If you moved them all how are you going to know what applies?

> I've also confirmed that the ITs pass for 2.0.11-SNAPSHOT as it is.
>
> Once I get the 2.1.x bits cleaned up (per original mail that  
> everyone seems in favour of), I'll spin an RC and see what everyone  
> thinks.
>
> - Brett
>
> On 02/07/2009, at 1:46 AM, John Casey wrote:
>
>> +1
>>
>> Paul Benedict wrote:
>>> My preference is to release 2.0.11 as it is now (37 issues fixed).  
>>> The
>>> remaining issues should move to 2.2.1. If critical bugs remain in
>>> 2.0.x, then build build a 2.0.12 issue list as people require it.
>>> - Paul
>>> On Wed, Jul 1, 2009 at 10:33 AM, Jörg  
>>> Schaible<joerg.schaible@...> wrote:
>>>> Brian Fox wrote:
>>>>
>>>>> Yeah get rid of it. Is there really demand for the fixed in  
>>>>> 2.0.11? I
>>>>> feel like it's EOL now.
>>>> The point is, in 6 months nobody knows axaclty anymore what is in
>>>> 2.0.11-SNAPSHOT. That will actually stop any bugfix release ever.
>>>>
>>>> - Jörg
>>>>
>>>>
>>>> ---------------------------------------------------------------------
>>>> To unsubscribe, e-mail: dev-unsubscribe@...
>>>> For additional commands, e-mail: dev-help@...
>>>>
>>>>
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail: dev-unsubscribe@...
>>> For additional commands, e-mail: dev-help@...
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: dev-unsubscribe@...
>> For additional commands, e-mail: dev-help@...
>>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@...
> For additional commands, e-mail: dev-help@...
>

Thanks,

Jason

----------------------------------------------------------
Jason van Zyl
Founder,  Apache Maven
http://twitter.com/jvanzyl
http://twitter.com/SonatypeNexus
http://twitter.com/SonatypeM2E
----------------------------------------------------------

First, the taking in of scattered particulars under one Idea,
so that everyone understands what is being talked about ... Second,
the separation of the Idea into parts, by dividing it at the joints,
as nature directs, not breaking any limb in half as a bad carver might.

   -- Plato, Phaedrus (Notes on the Synthesis of Form by C. Alexander)


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[brettporter]

On 02/07/2009, at 3:38 AM, Jason van Zyl wrote:

>
> On 1-Jul-09, at 9:47 AM, Brett Porter wrote:
>
>> Ok, for starters I've moved all the open issues from 2.0.11 to  
>> 2.2.1 and am now going through them to cull them down where possible.
>>
>
> You need to leave the bugs raised against 2.0.x because there is no  
> way around the fact that 2.0.x is going to be the dominant version  
> used for quite some time. We can't just stop bug fixing the 2.0.x  
> line. If you moved them all how are you going to know what applies?

There weren't any changes to the "affects" version.

- Brett


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Paul Benedict-2]

It's logical to believe that 2.1 and 2.2 contain almost all the
unresolved bugs of 2.0.x. Since 2.0.x is no longer being supported,
there's no good reason to keep them attached to that version. You only
want to backport the issues that will get fixing -- not potential
fixes UNLESS the issue is exclusively a 2.0.x issue.

-- Paul

On Wed, Jul 1, 2009 at 12:38 PM, Jason van Zyl<jvanzyl@...> wrote:

> You need to leave the bugs raised against 2.0.x because there is no way
> around the fact that 2.0.x is going to be the dominant version used for
> quite some time. We can't just stop bug fixing the 2.0.x line. If you moved
> them all how are you going to know what applies?

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Jörg Schaible-2]

Christian Schulte wrote:

> Paul Benedict schrieb:
>> My preference is to release 2.0.11 as it is now (37 issues fixed). The
>> remaining issues should move to 2.2.1. If critical bugs remain in
>> 2.0.x, then build build a 2.0.12 issue list as people require it.
>>
>> - Paul
>>
>
> +1
>
> 2.0.x is the last JDK 1.4 release. Users of the GPG plugin simply cannot
> use 2.1.x.

No. 2.1.x is also JDK 1.4. That was the whole point for starting 2.2.x.

- Jörg



---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Jason van Zyl-5]

On 1-Jul-09, at 10:52 AM, Paul Benedict wrote:

> It's logical to believe that 2.1 and 2.2 contain almost all the
> unresolved bugs of 2.0.x. Since 2.0.x is no longer being supported,
> there's no good reason to keep them attached to that version. You only
> want to backport the issues that will get fixing -- not potential
> fixes UNLESS the issue is exclusively a 2.0.x issue.
>

Unfortunately this may not be the case because the code bases are now  
pretty different. My only concern is that the 2.0.x line becomes the  
ugly stepchild meanwhile this is where the vast majority of our users  
live.

> -- Paul
>
> On Wed, Jul 1, 2009 at 12:38 PM, Jason van Zyl<jvanzyl@...>  
> wrote:
>
>> You need to leave the bugs raised against 2.0.x because there is no  
>> way
>> around the fact that 2.0.x is going to be the dominant version used  
>> for
>> quite some time. We can't just stop bug fixing the 2.0.x line. If  
>> you moved
>> them all how are you going to know what applies?
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@...
> For additional commands, e-mail: dev-help@...
>

Thanks,

Jason

----------------------------------------------------------
Jason van Zyl
Founder,  Apache Maven
http://twitter.com/jvanzyl
http://twitter.com/SonatypeNexus
http://twitter.com/SonatypeM2E
----------------------------------------------------------



---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[John Casey-5]

FYI, you can still build 1.4 projects safely in Maven 2.2.0:

http://maven.apache.org/guides/mini/guide-building-jdk14-on-jdk15.html

-john

Christian Schulte wrote:

> Paul Benedict schrieb:
>> My preference is to release 2.0.11 as it is now (37 issues fixed). The
>> remaining issues should move to 2.2.1. If critical bugs remain in
>> 2.0.x, then build build a 2.0.12 issue list as people require it.
>>
>> - Paul
>>
>
> +1
>
> 2.0.x is the last JDK 1.4 release. Users of the GPG plugin simply cannot
> use 2.1.x.
>

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[brettporter]

On 02/07/2009, at 4:06 AM, Jason van Zyl wrote:

> On 1-Jul-09, at 10:52 AM, Paul Benedict wrote:
>
>> It's logical to believe that 2.1 and 2.2 contain almost all the
>> unresolved bugs of 2.0.x. Since 2.0.x is no longer being supported,
>> there's no good reason to keep them attached to that version. You  
>> only
>> want to backport the issues that will get fixing -- not potential
>> fixes UNLESS the issue is exclusively a 2.0.x issue.
>>
>
> Unfortunately this may not be the case because the code bases are  
> now pretty different. My only concern is that the 2.0.x line becomes  
> the ugly stepchild meanwhile this is where the vast majority of our  
> users live.

Ok, even so - I think there was some agreement that we wouldn't  
explicitly plan for a 2.0.12+ release, which was the motivation for  
the change I made. If, in the process of fixing an issue, the  
committer decides it really should be backported to 2.0.x that's still  
a possibility (or if someone else comes along and requests it).

But I get the feeling that those sticking to 2.0.x are "happy" - in  
that they've got things working the way they want and probably won't  
jump up to further 2.0.x releases, let along 2.2.x. If we put out a  
2.0.11 release and say "this is the last, barring critical issues -  
start looking at 2.2", we'll fairly soon hear about it if that's not  
what users want.

At the same time, if we do start pushing fixes into 2.2.x, that gives  
more people incentive to try it, and help us identify if there are  
further barriers to moving across, in addition to continuing to build  
out more integration test cases that benefit us across the board.

- Brett


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[John Casey-5]

Brett Porter wrote:

>
> But I get the feeling that those sticking to 2.0.x are "happy" - in that
> they've got things working the way they want and probably won't jump up
> to further 2.0.x releases, let along 2.2.x. If we put out a 2.0.11
> release and say "this is the last, barring critical issues - start
> looking at 2.2", we'll fairly soon hear about it if that's not what
> users want.
>
> At the same time, if we do start pushing fixes into 2.2.x, that gives
> more people incentive to try it, and help us identify if there are
> further barriers to moving across, in addition to continuing to build
> out more integration test cases that benefit us across the board.
>
> - Brett

Personally, I think this makes a lot of sense. I think we shouldn't go
out of our way to freak out our user base, but at the same time we
shouldn't spend too much time pushing the envelope with 2.0.x now that
we've decided to move on. If we announce that we're doing critical fixes
only on 2.0.x - and not spending time "cleaning up" -  then people who
have a problem with this should become visible. It's a good way to
engage with our community to figure out why people won't make the jump, IMO.

If it's just about an arbitrary version number, I'm not sure how to
reassure those people without making a largely symbolic 2.2.1 release.

-john

>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@...
> For additional commands, e-mail: dev-help@...
>

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Paul Benedict-2]

Jason, I apologize for misspeaking. I meant what Brian said: the
"affected" version should stay the same. It's okay to remove the "Fix"
for version which was altered to 2.2.1

On Wed, Jul 1, 2009 at 1:29 PM, Brett Porter<brett@...> wrote:

>
>
> On 02/07/2009, at 4:06 AM, Jason van Zyl wrote:
>
>> On 1-Jul-09, at 10:52 AM, Paul Benedict wrote:
>>
>>> It's logical to believe that 2.1 and 2.2 contain almost all the
>>> unresolved bugs of 2.0.x. Since 2.0.x is no longer being supported,
>>> there's no good reason to keep them attached to that version. You only
>>> want to backport the issues that will get fixing -- not potential
>>> fixes UNLESS the issue is exclusively a 2.0.x issue.
>>>
>>
>> Unfortunately this may not be the case because the code bases are now
>> pretty different. My only concern is that the 2.0.x line becomes the ugly
>> stepchild meanwhile this is where the vast majority of our users live.
>
> Ok, even so - I think there was some agreement that we wouldn't explicitly
> plan for a 2.0.12+ release, which was the motivation for the change I made.
> If, in the process of fixing an issue, the committer decides it really
> should be backported to 2.0.x that's still a possibility (or if someone else
> comes along and requests it).
>
> But I get the feeling that those sticking to 2.0.x are "happy" - in that
> they've got things working the way they want and probably won't jump up to
> further 2.0.x releases, let along 2.2.x. If we put out a 2.0.11 release and
> say "this is the last, barring critical issues - start looking at 2.2",
> we'll fairly soon hear about it if that's not what users want.
>
> At the same time, if we do start pushing fixes into 2.2.x, that gives more
> people incentive to try it, and help us identify if there are further
> barriers to moving across, in addition to continuing to build out more
> integration test cases that benefit us across the board.
>
> - Brett
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@...
> For additional commands, e-mail: dev-help@...
>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Christian Edward Gruber-2]

As a user... +1

On Jul 1, 2009, at 3:41 PM, John Casey wrote:

> Brett Porter wrote:
>> But I get the feeling that those sticking to 2.0.x are "happy" - in  
>> that they've got things working the way they want and probably  
>> won't jump up to further 2.0.x releases, let along 2.2.x. If we put  
>> out a 2.0.11 release and say "this is the last, barring critical  
>> issues - start looking at 2.2", we'll fairly soon hear about it if  
>> that's not what users want.
>> At the same time, if we do start pushing fixes into 2.2.x, that  
>> gives more people incentive to try it, and help us identify if  
>> there are further barriers to moving across, in addition to  
>> continuing to build out more integration test cases that benefit us  
>> across the board.
>> - Brett
>
> Personally, I think this makes a lot of sense. I think we shouldn't  
> go out of our way to freak out our user base, but at the same time  
> we shouldn't spend too much time pushing the envelope with 2.0.x now  
> that we've decided to move on. If we announce that we're doing  
> critical fixes only on 2.0.x - and not spending time "cleaning up"  
> -  then people who have a problem with this should become visible.  
> It's a good way to engage with our community to figure out why  
> people won't make the jump, IMO.
>
> If it's just about an arbitrary version number, I'm not sure how to  
> reassure those people without making a largely symbolic 2.2.1 release.
>
> -john
>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: dev-unsubscribe@...
>> For additional commands, e-mail: dev-help@...
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@...
> For additional commands, e-mail: dev-help@...
>

Christian Edward Gruber
christianedwardgruber@...
http://www.geekinasuit.com/


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...