[JEXL] Next steps towards release?

[Henrib-2]

What should be the next steps towards releasing JEXL-2.0 ?
The code seems pretty stable but I suspect there are other tasks waiting...
What should they be? Can we track them through JIRA ?
Thanks

[Rahul Akolkar]

On Mon, Sep 7, 2009 at 4:30 PM, Henrib<[hidden email]> wrote:
>
> What should be the next steps towards releasing JEXL-2.0 ?
<snip/>

We'll need someone to take it through the release process (a release
manager / RM), i.e. cut a release candidate, call a vote, iterate as
necessary. I won't have time for this in September or early October.
If you're interested see link below and you can also look through
archives for examples of recent releases:

  http://commons.apache.org/releases/index.html


> The code seems pretty stable but I suspect there are other tasks waiting...
> What should they be? Can we track them through JIRA ?
<snap/>

Are there any other tasks? (other than release process stuff). Release
process stuff can be tracked in JIRA if the RM desires but its not
necessary to do so.

-Rahul


> Thanks

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

[sebb-2-2]

On 08/09/2009, Rahul Akolkar <[hidden email]> wrote:

> On Mon, Sep 7, 2009 at 4:30 PM, Henrib<[hidden email]> wrote:
>  >
>  > What should be the next steps towards releasing JEXL-2.0 ?
>
> <snip/>
>
>  We'll need someone to take it through the release process (a release
>  manager / RM), i.e. cut a release candidate, call a vote, iterate as
>  necessary. I won't have time for this in September or early October.
>  If you're interested see link below and you can also look through
>  archives for examples of recent releases:
>
>   http://commons.apache.org/releases/index.html
>
>
>
>  > The code seems pretty stable but I suspect there are other tasks waiting...
>  > What should they be? Can we track them through JIRA ?
>
> <snap/>
>
>  Are there any other tasks? (other than release process stuff). Release
>  process stuff can be tracked in JIRA if the RM desires but its not
>  necessary to do so.

I think we need to perform a Clirr check on the API to see if any
important classes/methods have been dropped.

>  -Rahul
>
>
>  > Thanks
>
>
>  ---------------------------------------------------------------------
>  To unsubscribe, e-mail: [hidden email]
>  For additional commands, e-mail: [hidden email]
>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

[Rahul Akolkar]

On Sat, Sep 12, 2009 at 9:51 AM, sebb <[hidden email]> wrote:

> On 08/09/2009, Rahul Akolkar <[hidden email]> wrote:
>> On Mon, Sep 7, 2009 at 4:30 PM, Henrib<[hidden email]> wrote:
>>  >
>>  > What should be the next steps towards releasing JEXL-2.0 ?
>>
>> <snip/>
>>
>>  We'll need someone to take it through the release process (a release
>>  manager / RM), i.e. cut a release candidate, call a vote, iterate as
>>  necessary. I won't have time for this in September or early October.
>>  If you're interested see link below and you can also look through
>>  archives for examples of recent releases:
>>
>>   http://commons.apache.org/releases/index.html
>>
>>
>>
>>  > The code seems pretty stable but I suspect there are other tasks waiting...
>>  > What should they be? Can we track them through JIRA ?
>>
>> <snap/>
>>
>>  Are there any other tasks? (other than release process stuff). Release
>>  process stuff can be tracked in JIRA if the RM desires but its not
>>  necessary to do so.
>
> I think we need to perform a Clirr check on the API to see if any
> important classes/methods have been dropped.
>

<snip/>

Absolutely, must be done as part of the buildup to any RC (IOW, I
consider that release process stuff).

-Rahul

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

[Henrib-2]

There are probably too many things I dont know to be able to act as a proper and efficient RM; I can help though if told what to do (besides bug fixing) or, if no one else is candidate as RM, I would really appreciate some help and being 'shown the way'. Reading the procedure is one thing, doing it correctly another...
Anyone with some spare cycles and inclined to release 2.0 ?

[Rahul Akolkar]

On Mon, Oct 12, 2009 at 5:06 PM, Henrib <[hidden email]> wrote:
>
>
> There are probably too many things I dont know to be able to act as a proper
> and efficient RM; I can help though if told what to do (besides bug fixing)
> or, if no one else is candidate as RM, I would really appreciate some help
> and being 'shown the way'. Reading the procedure is one thing, doing it
> correctly another...
<snip/>

Right, its non-trivial (and sometimes multiple votes are needed and
that requires a lot of patience) but its not all that daunting either.

When I was getting started, I found it best to observe the process
when someone else more experienced was RM'ing. ATM, the ongoing
BeanUtils release provides such an opportunity.

I'd say if you are thinking about RM'ing, its time to do it ;-) I'll
try my best to answer any questions you may have.

-Rahul


> Anyone with some spare cycles and inclined to release 2.0 ?
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

[Henrib-2]

Rahul Akolkar wrote:

Right, its non-trivial (and sometimes multiple votes are needed and
that requires a lot of patience) but its not all that daunting either.

When I was getting started, I found it best to observe the process
when someone else more experienced was RM'ing. ATM, the ongoing
BeanUtils release provides such an opportunity.

I'd say if you are thinking about RM'ing, its time to do it ;-) I'll
try my best to answer any questions you may have.

-Rahul

I'm not thinking about RM'ing per se, I just would like 2.0 released. :-) But it also seems I'm probably one of the very few interested in it...
Besides, I'm still a "newbie" as a committer and my mistake/commit ratio is probably a bit high for this responsibility.
Anyway, as is, the trunk looks good - at least by Maven's Checkstyle & Cobertura reports -, the Javadoc seems to be 95% acceptable but I don't know how to produce a decent 'changes report' (and I'm even less capable of defining what an acceptable Clirr/JDiff output should be and have yet to grasp whether a new branch should be cut or not, etc...).
Just tell me what I should do, what the next action should be and I'll try to follow instructions.

[Rahul Akolkar]

On Tue, Oct 13, 2009 at 5:39 AM, Henrib <[hidden email]> wrote:

>
> Rahul Akolkar wrote:
>>
>>
>> Right, its non-trivial (and sometimes multiple votes are needed and
>> that requires a lot of patience) but its not all that daunting either.
>>
>> When I was getting started, I found it best to observe the process
>> when someone else more experienced was RM'ing. ATM, the ongoing
>> BeanUtils release provides such an opportunity.
>>
>> I'd say if you are thinking about RM'ing, its time to do it ;-) I'll
>> try my best to answer any questions you may have.
>>
>> -Rahul
>>
>>
>
> I'm not thinking about RM'ing per se, I just would like 2.0 released. :-)
> But it also seems I'm probably one of the very few interested in it...
> Besides, I'm still a "newbie" as a committer and my mistake/commit ratio is
> probably a bit high for this responsibility.
> Anyway, as is, the trunk looks good - at least by Maven's Checkstyle &
> Cobertura reports -, the Javadoc seems to be 95% acceptable but I don't know
> how to produce a decent 'changes report' (and I'm even less capable of
> defining what an acceptable Clirr/JDiff output should be and have yet to
> grasp whether a new branch should be cut or not, etc...).

<snip/>

A clirr report can be produced using the clirr maven plugin or by
running clirr on the command line. Then we stare at any errors or
warnings in the report and make sure each one of them can be justified
and changes are in line with the versioning guidelines. Related
resources for this part:

  http://mojo.codehaus.org/clirr-maven-plugin/

  http://commons.apache.org/releases/versioning.html


> Just tell me what I should do, what the next action should be and I'll try
> to follow instructions.
>
<snap/>

The objective is to produce a complete set of release artifacts like
this one (release notes, src and bin assemblies, m2 staging repo,
site):

  http://people.apache.org/builds/commons/digester/2.0/RC1/

That set was produced using the procedure outlined here:

  http://wiki.apache.org/commons/CreatingReleases

Once artifacts are ready, I usually post a link to dev@ for feedback
for a week or so, and if all is well its time to vote.

-Rahul

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

[Henrib-2]

Getting closer to produce an RC1, I've got a few questions.

1/ It seems the (new) template for the site does not include the various
reports (Cobertura, PMD, etc)? Am I doing something wrong ? (tried with
commons-exec which has the same behavior btw). Is this tied to
commons-parent:12 ?

2/ Reading the procedure, I went through the steps required to create a
code-signing key; not being  at ApacheCon this week ( unfortunately :-( )and
having no committers in the vicinity, I dont see how I can build the 'web of
trust' part. What should I do ? Is it OK to use my key anyway ?

2/ The clirr report is really hard to decipher; so much has changed between
jexl-1.0 and jexl-2.0 (from the parser to the API) that the output is either
obvious or noise. I only found one API change that might need to be (more)
documented, the removal of pre/post-resolvers from Expression. Not sure what
else I should expect from the tool considering how much has been refactored,
renamed or re-scoped...

3/ On a different note, it would be nice to move Apache Commons on the new
Nabble2
(http://n2.nabble.com/Migrating-forums-from-Nabble1-to-Nabble2-tp3791390p3791390.html)

Regards,
Henrib

--
View this message in context: http://old.nabble.com/-JEXL--Next-steps-towards-release--tp25335922p26157896.html
Sent from the Commons - Dev mailing list archive at Nabble.com.


---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

[Henrib-2]

Getting closer to produce an RC1, I've got a few questions.

1/ It seems the (new) template for the site does not include the various reports (Cobertura, PMD, etc)? Am I doing something wrong ? (tried with commons-exec which has the same behavior btw). Is this tied to commons-parent:12 ?

2/ Reading the procedure, I went through the steps required to create a code-signing key; not being  at ApacheCon this week ( unfortunately :-( )and having no committers in the vicinity, I dont see how I can build the 'web of trust' part. What should I do ? Is it OK to use my key anyway ?

2/ The clirr report is really hard to decipher; so much has changed between jexl-1.0 and jexl-2.0 (from the parser to the API) that the output is either obvious or noise. I only found one API change that might need to be (more) documented, the removal of pre/post-resolvers from Expression. Not sure what else I should expect from the tool considering how much has been refactored, renamed or re-scoped...

3/ On a different note, it would be nice to move Apache Commons on the new Nabble2 (http://n2.nabble.com/Migrating-forums-from-Nabble1-to-Nabble2-tp3791390p3791390.html)

Regards,
Henrib

[Rahul Akolkar]

On Mon, Nov 2, 2009 at 6:24 PM, Henrib <[hidden email]> wrote:
>
> Getting closer to produce an RC1,
<snip/>

Thanks.


> I've got a few questions.
>
> 1/ It seems the (new) template for the site does not include the various
> reports (Cobertura, PMD, etc)? Am I doing something wrong ? (tried with
> commons-exec which has the same behavior btw). Is this tied to
> commons-parent:12 ?
>
<snap/>

Don't know and can't try right now either I'm afraid (traveling this
week). Just so I understand, does going back to v11 give desired
results?


> 2/ Reading the procedure, I went through the steps required to create a
> code-signing key; not being  at ApacheCon this week ( unfortunately :-( )and
> having no committers in the vicinity, I dont see how I can build the 'web of
> trust' part. What should I do ? Is it OK to use my key anyway ?
>
<snip/>

Yes, its OK, but check in your public key here:

  http://svn.apache.org/repos/asf/commons/trunks-proper/KEYS


> 2/ The clirr report is really hard to decipher; so much has changed between
> jexl-1.0 and jexl-2.0 (from the parser to the API) that the output is either
> obvious or noise. I only found one API change that might need to be (more)
> documented, the removal of pre/post-resolvers from Expression. Not sure what
> else I should expect from the tool considering how much has been refactored,
> renamed or re-scoped...
>
<snap/>

Right, I think we've largely maintained compat with the EL syntax
itself, but the Java API has changed in many respects. Will beg the
question whether we should change package names (to oac.jexl2).


> 3/ On a different note, it would be nice to move Apache Commons on the new
> Nabble2
> (http://n2.nabble.com/Migrating-forums-from-Nabble1-to-Nabble2-tp3791390p3791390.html)
>
<snip/>

Not a nabble user, no comments from me.

-Rahul


> Regards,
> Henrib
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]