Re: 2.1 Release -- Banging the drum

2008-01-30 Thread Jason Dillon

So, what is the new timeframe to TCK and release this puppy?

--jason


On Jan 29, 2008, at 2:15 PM, Joe Bohn wrote:


Kevan Miller wrote:

All,
This note is a bit overdue (it's been a distracting start to the  
New Year for me). Time, IMO, for us to get focused on our 2.1  
release.
As David Jencks has pointed out. We need to start cleaning out the  
2.1 Jiras. It looks like I've got several open that have been  
fixed, either by additional development activities or redundant  
jira's. First step is to take a look at Jira's that you've created  
and make sure they are still valid and if you think it's important  
that they be fixed for 2.1.
We also need to be taking a close look at our current  
functionality. Make sure things are working the way we want them  
to... Especially need to cast a critical eye on our the usability  
aspects of the new 2.1. Along the way, will be great if we can  
start pulling docs together.
I started running tests last night. Right away, I'm noticing little  
things like warning messages being sent to STDOUT, etc. I'll start  
registering problem areas that I'm seeing.
I'd like to set a target of 2 weeks for reviewing and fixing  
problems. After that would start the branching, final tck, and  
packaging work. If we feel this might negatively impact post-2.1  
development activities. We can consider creating a 2.1 branch  
sooner...

Thoughts?
--kevan


I've recently upgraded our image to utilize the newly released xbean  
3.3, myfaces 1.2.2, geronimo-transaction 2.1, and geronimo-connector  
2.1.  Earlier I had updated the Tomcat build image and created a  
private build for Pluto.


I think the only snapshots left in our image are some of the specs  
(JACC  JASPI ... and it looks like Guillaume is working those along  
with upgrading others) and OpenEjb.


Joe




Re: 2.1 Release -- Banging the drum

2008-01-30 Thread David Jencks
BTW when we branch for 2.1 I plan to remove the jaspi dependent code  
and remove the jaspi spec dependency.


meanwhile someone :-) should make sure the in-vote specs pass the tck.

thanks
david jencks
On Jan 30, 2008, at 1:01 PM, Jason Dillon wrote:


So, what is the new timeframe to TCK and release this puppy?

--jason


On Jan 29, 2008, at 2:15 PM, Joe Bohn wrote:


Kevan Miller wrote:

All,
This note is a bit overdue (it's been a distracting start to the  
New Year for me). Time, IMO, for us to get focused on our 2.1  
release.
As David Jencks has pointed out. We need to start cleaning out  
the 2.1 Jiras. It looks like I've got several open that have been  
fixed, either by additional development activities or redundant  
jira's. First step is to take a look at Jira's that you've  
created and make sure they are still valid and if you think it's  
important that they be fixed for 2.1.
We also need to be taking a close look at our current  
functionality. Make sure things are working the way we want them  
to... Especially need to cast a critical eye on our the usability  
aspects of the new 2.1. Along the way, will be great if we can  
start pulling docs together.
I started running tests last night. Right away, I'm noticing  
little things like warning messages being sent to STDOUT, etc.  
I'll start registering problem areas that I'm seeing.
I'd like to set a target of 2 weeks for reviewing and fixing  
problems. After that would start the branching, final tck, and  
packaging work. If we feel this might negatively impact post-2.1  
development activities. We can consider creating a 2.1 branch  
sooner...

Thoughts?
--kevan


I've recently upgraded our image to utilize the newly released  
xbean 3.3, myfaces 1.2.2, geronimo-transaction 2.1, and geronimo- 
connector 2.1.  Earlier I had updated the Tomcat build image and  
created a private build for Pluto.


I think the only snapshots left in our image are some of the specs  
(JACC  JASPI ... and it looks like Guillaume is working those  
along with upgrading others) and OpenEjb.


Joe






Re: 2.1 Release -- Banging the drum

2008-01-30 Thread Joe Bohn
I guess that Kevan needs to chime in on that ... but I think Friday is 
still reasonable to branch and start the release work if we can get the 
critical issues resolved by then.


I've been running TCK continuously and fixing and/or pointing things out 
when I notice problems.  We're generally at 100% with problems creeping 
up every few days and getting resolved.


Joe


Jason Dillon wrote:

So, what is the new timeframe to TCK and release this puppy?

--jason


On Jan 29, 2008, at 2:15 PM, Joe Bohn wrote:


Kevan Miller wrote:

All,
This note is a bit overdue (it's been a distracting start to the New 
Year for me). Time, IMO, for us to get focused on our 2.1 release.
As David Jencks has pointed out. We need to start cleaning out the 
2.1 Jiras. It looks like I've got several open that have been fixed, 
either by additional development activities or redundant jira's. 
First step is to take a look at Jira's that you've created and make 
sure they are still valid and if you think it's important that they 
be fixed for 2.1.
We also need to be taking a close look at our current functionality. 
Make sure things are working the way we want them to... Especially 
need to cast a critical eye on our the usability aspects of the new 
2.1. Along the way, will be great if we can start pulling docs together.
I started running tests last night. Right away, I'm noticing little 
things like warning messages being sent to STDOUT, etc. I'll start 
registering problem areas that I'm seeing.
I'd like to set a target of 2 weeks for reviewing and fixing 
problems. After that would start the branching, final tck, and 
packaging work. If we feel this might negatively impact post-2.1 
development activities. We can consider creating a 2.1 branch sooner...

Thoughts?
--kevan


I've recently upgraded our image to utilize the newly released xbean 
3.3, myfaces 1.2.2, geronimo-transaction 2.1, and geronimo-connector 
2.1.  Earlier I had updated the Tomcat build image and created a 
private build for Pluto.


I think the only snapshots left in our image are some of the specs 
(JACC  JASPI ... and it looks like Guillaume is working those along 
with upgrading others) and OpenEjb.


Joe







Re: 2.1 Release -- Banging the drum

2008-01-30 Thread Kevan Miller


On Jan 30, 2008, at 4:01 PM, Jason Dillon wrote:


So, what is the new timeframe to TCK and release this puppy?


Well, my two week period for reviewing and fixing problems ends today.  
Things are looking pretty good, IMO.


I think we branch on Friday and start turning the screws.

--kevan



Re: 2.1 Release -- Banging the drum

2008-01-30 Thread Jason Dillon

Aighty, sounds good to me :-)

--jason


On Jan 30, 2008, at 1:17 PM, Kevan Miller wrote:



On Jan 30, 2008, at 4:01 PM, Jason Dillon wrote:


So, what is the new timeframe to TCK and release this puppy?


Well, my two week period for reviewing and fixing problems ends  
today. Things are looking pretty good, IMO.


I think we branch on Friday and start turning the screws.

--kevan





Re: 2.1 Release -- Banging the drum

2008-01-29 Thread Joe Bohn

Kevan Miller wrote:

All,
This note is a bit overdue (it's been a distracting start to the New 
Year for me). Time, IMO, for us to get focused on our 2.1 release.


As David Jencks has pointed out. We need to start cleaning out the 2.1 
Jiras. It looks like I've got several open that have been fixed, either 
by additional development activities or redundant jira's. First step is 
to take a look at Jira's that you've created and make sure they are 
still valid and if you think it's important that they be fixed for 2.1.


We also need to be taking a close look at our current functionality. 
Make sure things are working the way we want them to... Especially need 
to cast a critical eye on our the usability aspects of the new 2.1. 
Along the way, will be great if we can start pulling docs together.


I started running tests last night. Right away, I'm noticing little 
things like warning messages being sent to STDOUT, etc. I'll start 
registering problem areas that I'm seeing.


I'd like to set a target of 2 weeks for reviewing and fixing problems. 
After that would start the branching, final tck, and packaging work. If 
we feel this might negatively impact post-2.1 development activities. We 
can consider creating a 2.1 branch sooner...


Thoughts?

--kevan




I've recently upgraded our image to utilize the newly released xbean 
3.3, myfaces 1.2.2, geronimo-transaction 2.1, and geronimo-connector 
2.1.  Earlier I had updated the Tomcat build image and created a private 
build for Pluto.


I think the only snapshots left in our image are some of the specs (JACC 
 JASPI ... and it looks like Guillaume is working those along with 
upgrading others) and OpenEjb.


Joe


Re: 2.1 Release -- Banging the drum

2008-01-22 Thread Jacek Laskowski
On Jan 22, 2008 12:44 AM, Jay D. McHugh [EMAIL PROTECTED] wrote:
 +1 - Two weeks sounds good to me.

It's 6 days ago when Kevan put it to discussion so it's really a week
away from that day ;-)

Jacek

-- 
Jacek Laskowski
http://www.JacekLaskowski.pl


Re: 2.1 Release -- Banging the drum

2008-01-21 Thread Kevan Miller


On Jan 20, 2008, at 1:42 PM, Jason Dillon wrote:

I'm going to start working on this... looks like there are more  
problems that I thought, though not hard to fix... just a PITA.


--jason


On Jan 20, 2008, at 8:13 AM, Kevan Miller wrote:



On Jan 19, 2008, at 1:18 PM, Jacek Laskowski wrote:

On Jan 18, 2008 3:15 AM, Kevan Miller [EMAIL PROTECTED]  
wrote:


I agree with Jason. We shouldn't be carrying forward the current  
structure.
And, I think we have enough time to fix this problem while we are  
fixing

other issues with the release.


Even though I tend to agree I understand the pain of our end users  
who

suffer from working with 2.0.2 when we keep telling them use the
unofficial 2.1 release and I wish we could release G 2.1 as soon as
possible. No issues should be counted any more. Just go ahead,  
release

and keep working on 2.2 release.


Jacek,
Let's level-set for a second. From my original note on this subject:

On Jan 16, 2008, at 10:27 AM, Kevan Miller wrote:



I'd like to set a target of 2 weeks for reviewing and fixing  
problems. After that would start the branching, final tck, and  
packaging work. If we feel this might negatively impact post-2.1  
development activities. We can consider creating a 2.1 branch  
sooner...


Are you ok with the 2 week target for reviewing the current trunk  
codebase and resolving issues?


The structure of our pom's are one of the issues that I think have  
been identified in our current codebase. Seems like we can resolve  
the problem within our 2 week timeframe. So, I'm all for fixing the  
poms...


--kevan






Re: 2.1 Release -- Banging the drum

2008-01-21 Thread Jay D. McHugh

+1 - Two weeks sounds good to me.

The big feature that I had been waiting for was the Dojo upgrade and 
that is done.


I can start looking at the 'STDOUT' messages.

I would assume that sending messages to the console during tests would 
be fine (yes?) and that the problem would really be when messages from a 
running server get sent there as well.


Or should even the test messages be getting sent to a log?

Jay

Kevan Miller wrote:

All,
This note is a bit overdue (it's been a distracting start to the New 
Year for me). Time, IMO, for us to get focused on our 2.1 release.


As David Jencks has pointed out. We need to start cleaning out the 2.1 
Jiras. It looks like I've got several open that have been fixed, either 
by additional development activities or redundant jira's. First step is 
to take a look at Jira's that you've created and make sure they are 
still valid and if you think it's important that they be fixed for 2.1.


We also need to be taking a close look at our current functionality. 
Make sure things are working the way we want them to... Especially need 
to cast a critical eye on our the usability aspects of the new 2.1. 
Along the way, will be great if we can start pulling docs together.


I started running tests last night. Right away, I'm noticing little 
things like warning messages being sent to STDOUT, etc. I'll start 
registering problem areas that I'm seeing.


I'd like to set a target of 2 weeks for reviewing and fixing problems. 
After that would start the branching, final tck, and packaging work. If 
we feel this might negatively impact post-2.1 development activities. We 
can consider creating a 2.1 branch sooner...


Thoughts?

--kevan








Re: 2.1 Release -- Banging the drum

2008-01-21 Thread Kevan Miller


On Jan 21, 2008, at 6:44 PM, Jay D. McHugh wrote:


+1 - Two weeks sounds good to me.

The big feature that I had been waiting for was the Dojo upgrade and  
that is done.


I can start looking at the 'STDOUT' messages.

I would assume that sending messages to the console during tests  
would be fine (yes?) and that the problem would really be when  
messages from a running server get sent there as well.


Or should even the test messages be getting sent to a log?


Well, I was thinking of warning/error messages that aren't really  
warning/error messages. For example:


19:25:17,715 WARN  [AbstractGBeanReference] GBean references are not  
using proxies


and the Tomcat restricted listeners error message...

--kevan



Re: 2.1 Release -- Banging the drum

2008-01-20 Thread Kevan Miller


On Jan 19, 2008, at 1:18 PM, Jacek Laskowski wrote:


On Jan 18, 2008 3:15 AM, Kevan Miller [EMAIL PROTECTED] wrote:

I agree with Jason. We shouldn't be carrying forward the current  
structure.
And, I think we have enough time to fix this problem while we are  
fixing

other issues with the release.


Even though I tend to agree I understand the pain of our end users who
suffer from working with 2.0.2 when we keep telling them use the
unofficial 2.1 release and I wish we could release G 2.1 as soon as
possible. No issues should be counted any more. Just go ahead, release
and keep working on 2.2 release.


Jacek,
Let's level-set for a second. From my original note on this subject:

On Jan 16, 2008, at 10:27 AM, Kevan Miller wrote:



I'd like to set a target of 2 weeks for reviewing and fixing  
problems. After that would start the branching, final tck, and  
packaging work. If we feel this might negatively impact post-2.1  
development activities. We can consider creating a 2.1 branch  
sooner...


Are you ok with the 2 week target for reviewing the current trunk  
codebase and resolving issues?


The structure of our pom's are one of the issues that I think have  
been identified in our current codebase. Seems like we can resolve the  
problem within our 2 week timeframe. So, I'm all for fixing the poms...


--kevan

Re: 2.1 Release -- Banging the drum

2008-01-20 Thread Jacek Laskowski
On Jan 20, 2008 5:13 PM, Kevan Miller [EMAIL PROTECTED] wrote:

 Are you ok with the 2 week target for reviewing the current trunk codebase
 and resolving issues?

I'm ok as long as it won't take us longer than 2 weeks to release
Geronimo 2.1 as is with all identified issues described in
RELEASE_NOTES or such.

Jacek

-- 
Jacek Laskowski
http://www.JacekLaskowski.pl


Re: 2.1 Release -- Banging the drum

2008-01-20 Thread Jason Dillon
I'm going to start working on this... looks like there are more  
problems that I thought, though not hard to fix... just a PITA.


--jason


On Jan 20, 2008, at 8:13 AM, Kevan Miller wrote:



On Jan 19, 2008, at 1:18 PM, Jacek Laskowski wrote:


On Jan 18, 2008 3:15 AM, Kevan Miller [EMAIL PROTECTED] wrote:

I agree with Jason. We shouldn't be carrying forward the current  
structure.
And, I think we have enough time to fix this problem while we are  
fixing

other issues with the release.


Even though I tend to agree I understand the pain of our end users  
who

suffer from working with 2.0.2 when we keep telling them use the
unofficial 2.1 release and I wish we could release G 2.1 as soon as
possible. No issues should be counted any more. Just go ahead,  
release

and keep working on 2.2 release.


Jacek,
Let's level-set for a second. From my original note on this subject:

On Jan 16, 2008, at 10:27 AM, Kevan Miller wrote:



I'd like to set a target of 2 weeks for reviewing and fixing  
problems. After that would start the branching, final tck, and  
packaging work. If we feel this might negatively impact post-2.1  
development activities. We can consider creating a 2.1 branch  
sooner...


Are you ok with the 2 week target for reviewing the current trunk  
codebase and resolving issues?


The structure of our pom's are one of the issues that I think have  
been identified in our current codebase. Seems like we can resolve  
the problem within our 2 week timeframe. So, I'm all for fixing the  
poms...


--kevan




Re: 2.1 Release -- Banging the drum

2008-01-19 Thread Jacek Laskowski
On Jan 18, 2008 3:15 AM, Kevan Miller [EMAIL PROTECTED] wrote:

 I agree with Jason. We shouldn't be carrying forward the current structure.
 And, I think we have enough time to fix this problem while we are fixing
 other issues with the release.

Even though I tend to agree I understand the pain of our end users who
suffer from working with 2.0.2 when we keep telling them use the
unofficial 2.1 release and I wish we could release G 2.1 as soon as
possible. No issues should be counted any more. Just go ahead, release
and keep working on 2.2 release.

Jacek

-- 
Jacek Laskowski
http://www.JacekLaskowski.pl


Re: 2.1 Release -- Banging the drum

2008-01-17 Thread Jason Dillon
I think we need to fix the pom parentage post reorganization before we  
can branch for a 2.1 release.  IMO the reorg is only half done... and  
really needs to be finished.


--jason


On Jan 16, 2008, at 7:27 AM, Kevan Miller wrote:


All,
This note is a bit overdue (it's been a distracting start to the New  
Year for me). Time, IMO, for us to get focused on our 2.1 release.


As David Jencks has pointed out. We need to start cleaning out the  
2.1 Jiras. It looks like I've got several open that have been fixed,  
either by additional development activities or redundant jira's.  
First step is to take a look at Jira's that you've created and make  
sure they are still valid and if you think it's important that they  
be fixed for 2.1.


We also need to be taking a close look at our current functionality.  
Make sure things are working the way we want them to... Especially  
need to cast a critical eye on our the usability aspects of the new  
2.1. Along the way, will be great if we can start pulling docs  
together.


I started running tests last night. Right away, I'm noticing little  
things like warning messages being sent to STDOUT, etc. I'll start  
registering problem areas that I'm seeing.


I'd like to set a target of 2 weeks for reviewing and fixing  
problems. After that would start the branching, final tck, and  
packaging work. If we feel this might negatively impact post-2.1  
development activities. We can consider creating a 2.1 branch  
sooner...


Thoughts?

--kevan





Re: 2.1 Release -- Banging the drum

2008-01-17 Thread Jacek Laskowski
On Jan 17, 2008 6:39 PM, Jason Dillon [EMAIL PROTECTED] wrote:
 I think we need to fix the pom parentage post reorganization before we
 can branch for a 2.1 release.  IMO the reorg is only half done... and
 really needs to be finished.

I disagree. We've been living with it for a while and am sure we can
live with it a bit longer. I'm against any issues/tasks that would
make the 2.1 release delayed. I don't want to tell our users that 2.1
is the solution to their problems as it doesn't exist yet. I've seen a
lot of answers with 2.1 as the solution so if we've suggested using
2.1 it's ready. There's no need to wait any longer and it should be
released now provided it passes TCK. Any other issues can be fixed in
2.2.

Jacek

-- 
Jacek Laskowski
http://www.JacekLaskowski.pl


Re: 2.1 Release -- Banging the drum

2008-01-17 Thread Jason Dillon
It should take a day or two to fix, nothing significant.  It should  
have been done when the modules were reorganized... and I have no idea  
why it was not.  The reorg task should be completed before we  
release.  I don't understand why folks tend to discount build related  
issues.  Maybe we should consult the resident m2 expert?  :-P


--jason


On Jan 17, 2008, at 2:22 PM, Jacek Laskowski wrote:


On Jan 17, 2008 6:39 PM, Jason Dillon [EMAIL PROTECTED] wrote:
I think we need to fix the pom parentage post reorganization before  
we

can branch for a 2.1 release.  IMO the reorg is only half done... and
really needs to be finished.


I disagree. We've been living with it for a while and am sure we can
live with it a bit longer. I'm against any issues/tasks that would
make the 2.1 release delayed. I don't want to tell our users that 2.1
is the solution to their problems as it doesn't exist yet. I've seen a
lot of answers with 2.1 as the solution so if we've suggested using
2.1 it's ready. There's no need to wait any longer and it should be
released now provided it passes TCK. Any other issues can be fixed in
2.2.

Jacek

--
Jacek Laskowski
http://www.JacekLaskowski.pl




Re: 2.1 Release -- Banging the drum

2008-01-17 Thread Kevan Miller


On Jan 17, 2008, at 7:46 PM, Jason Dillon wrote:

It should take a day or two to fix, nothing significant.  It should  
have been done when the modules were reorganized... and I have no  
idea why it was not.  The reorg task should be completed before we  
release.  I don't understand why folks tend to discount build  
related issues.  Maybe we should consult the resident m2 expert?  :-P


I agree with Jason. We shouldn't be carrying forward the current  
structure. And, I think we have enough time to fix this problem while  
we are fixing other issues with the release.


--kevan


--jason


On Jan 17, 2008, at 2:22 PM, Jacek Laskowski wrote:


On Jan 17, 2008 6:39 PM, Jason Dillon [EMAIL PROTECTED] wrote:
I think we need to fix the pom parentage post reorganization  
before we
can branch for a 2.1 release.  IMO the reorg is only half done...  
and

really needs to be finished.


I disagree. We've been living with it for a while and am sure we can
live with it a bit longer. I'm against any issues/tasks that would
make the 2.1 release delayed. I don't want to tell our users that 2.1
is the solution to their problems as it doesn't exist yet. I've  
seen a

lot of answers with 2.1 as the solution so if we've suggested using
2.1 it's ready. There's no need to wait any longer and it should be
released now provided it passes TCK. Any other issues can be fixed in
2.2.

Jacek

--
Jacek Laskowski
http://www.JacekLaskowski.pl