Re: [VOTE] Apache Geronimo Specs JCDI_2.0 1.2, Annotation_1.3 1.3, AtInject_1.0 1.2, Interceptor_1.2 1.2, JSONP_1.1 1.4, JSONB_1.0 1.3

[Romain Manni-Bucau]

Le mar. 28 avr. 2020 à 18:25, Mark Struberg  a écrit :

> As explained in the other thread:
>
> What is the difference between the various specs and their previous
> versions?
> Afaict the only difference is jakartaEE, and we WILL  NOT MANAGE to do all
> via simple replacement. This is an absolute dead end.
>

Mark, you keep saying that but each time I ask for a proof of that you
never answer.
Most of the spec are 1-1 (it was the deal of jakartaee8 and some spec had
been forbidden to replace/remove anything).
So please list the diff to let us fix that point if accurate.


> There have been plenty of smallish methods removed in e.g. EJB and
> servlet. But also in other specs. So we WILL need to go full scale. And I
> also expect more changes to come for JakartaEE9.
>


Ok, think it is your method removal ;).
It is not that accurate for this vote, both jars are not in the scope of
this vote.

I agree on EE9 point and here we will just create the jar from scratch as
usual, with the right package directly, no ambiguity or discussion IMHO.


>
> EE8 is FINISHED. There is NO change!
>

Does not mean we don't need to release the jar, please just have a look to
the versions of this thread (which is <1/6  of all specs), spec were
finished when we did 1.0 but we are at 1.4 for some jars so not a point for
me.


>
> All which happens is done in JakartaEE. And we have all the work done
> since a year?
> I could do a release of those jars today.
>

If you can do all jakarta jars I'm happy to cancel this vote as
mentionned already, my goal was just to get jakarta artifacts for free (and
this is what this vote does) to enable the CDI-SE/JSON-B case which starts
to get pressure to be useable in jakarta namespace more than others.


> Also your list of specs is not final. There is quite a few missing.
>

Not sure what you mean, I released the ones I announce for CDI SE + JSON-B
stacks.


>
> So why not release from here?
> http://svn.apache.org/repos/asf/geronimo/specs/branches/jakarta/
>

Not sure what would be the point to create another branch, we can keep
specs/ still it is specs, no?

However, I'm -1 to change the artifact id to contain jakarta. Worse case we
could do geronimo--api to try to simplify the naming, avoid a
2-versions based convention and be less rude to end user + use a jpms
friendly default name (even if we put an automatic name it avoids issues in
some envs/ide).


>
> Actually I'd move this to
>
> http://svn.apache.org/repos/asf/geronimo/jakarta-specs/trunk
>

If we would move anything I would try to use gitbox but can wait after the
first release which is, IMHO, the most urgent.


>
> and then do the release.
>

>
> LieGrue,
> strub
>
>
>
> Am 28.04.2020 um 07:59 schrieb Romain Manni-Bucau :
>
> Hi everyone,
>
> Here is the vote for some of our spec with jakarta shades.
>
> Tags:
> -
> http://svn.apache.org/repos/asf/geronimo/specs/tags/geronimo-jcdi_2.0_spec-1.2/
>  (rev 1877103)
> -
> http://svn.apache.org/repos/asf/geronimo/specs/tags/geronimo-annotation_1.3_spec-1.3/
>  (rev
> 1877106)
> -
> http://svn.apache.org/repos/asf/geronimo/specs/tags/geronimo-atinject_1.0_spec-1.2/
>  (rev
> 1877109)
> -
> http://svn.apache.org/repos/asf/geronimo/specs/tags/geronimo-interceptor_1.2_spec-1.2/
>  (rev
> 1877112)
> -
> http://svn.apache.org/repos/asf/geronimo/specs/tags/geronimo-json_1.1_spec-1.4/
>  (rev
> 1877115)
> -
> http://svn.apache.org/repos/asf/geronimo/specs/tags/geronimo-jsonb_1.0_spec-1.3/
>  (rev
> 1877118)
> Dist area:
> https://repository.apache.org/content/repositories/orgapachegeronimo-1124
> Staging repo: https://dist.apache.org/repos/dist/dev/geronimo/specs/
> My key is still the same.
>
> Please vote:
>
> [ ] +1 release it
> [ ] -1 dont' release it ${cause}
>
> Vote is open for 3 days or until we get enough bindings as usual.
>
> Romain Manni-Bucau
> @rmannibucau  |  Blog
>  | Old Blog
>  | Github
>  | LinkedIn
>  | Book
> 
>
>
>

Re: [VOTE] Apache Geronimo Specs JCDI_2.0 1.2, Annotation_1.3 1.3, AtInject_1.0 1.2, Interceptor_1.2 1.2, JSONP_1.1 1.4, JSONB_1.0 1.3

[Mark Struberg]

As explained in the other thread:

What is the difference between the various specs and their previous versions?
Afaict the only difference is jakartaEE, and we WILL  NOT MANAGE to do all via 
simple replacement. This is an absolute dead end. 
There have been plenty of smallish methods removed in e.g. EJB and servlet. But 
also in other specs. So we WILL need to go full scale. And I also expect more 
changes to come for JakartaEE9.

EE8 is FINISHED. There is NO change!

All which happens is done in JakartaEE. And we have all the work done since a 
year?
I could do a release of those jars today.
Also your list of specs is not final. There is quite a few missing.

So why not release from here?
http://svn.apache.org/repos/asf/geronimo/specs/branches/jakarta/ 


Actually I'd move this to 

http://svn.apache.org/repos/asf/geronimo/jakarta-specs/trunk 


and then do the release.


LieGrue,
strub



> Am 28.04.2020 um 07:59 schrieb Romain Manni-Bucau :
> 
> Hi everyone,
> 
> Here is the vote for some of our spec with jakarta shades.
> 
> Tags:
> - 
> http://svn.apache.org/repos/asf/geronimo/specs/tags/geronimo-jcdi_2.0_spec-1.2/
>  
> 
>  (rev 1877103)
> - 
> http://svn.apache.org/repos/asf/geronimo/specs/tags/geronimo-annotation_1.3_spec-1.3/
>  
> 
>  (rev 1877106)
> - 
> http://svn.apache.org/repos/asf/geronimo/specs/tags/geronimo-atinject_1.0_spec-1.2/
>  
> 
>  (rev 1877109)
> - 
> http://svn.apache.org/repos/asf/geronimo/specs/tags/geronimo-interceptor_1.2_spec-1.2/
>  
> 
>  (rev 1877112)
> - 
> http://svn.apache.org/repos/asf/geronimo/specs/tags/geronimo-json_1.1_spec-1.4/
>  
> 
>  (rev 1877115)
> - 
> http://svn.apache.org/repos/asf/geronimo/specs/tags/geronimo-jsonb_1.0_spec-1.3/
>  
> 
>  (rev 1877118)
> Dist area: 
> https://repository.apache.org/content/repositories/orgapachegeronimo-1124 
> 
> Staging repo: https://dist.apache.org/repos/dist/dev/geronimo/specs/ 
> 
> My key is still the same.
> 
> Please vote:
> 
> [ ] +1 release it
> [ ] -1 dont' release it ${cause}
> 
> Vote is open for 3 days or until we get enough bindings as usual.
> 
> Romain Manni-Bucau
> @rmannibucau  |  Blog 
>  | Old Blog 
>  | Github  
> | LinkedIn  | Book 
> 

Re: [VOTE] Apache Geronimo Specs JCDI_2.0 1.2, Annotation_1.3 1.3, AtInject_1.0 1.2, Interceptor_1.2 1.2, JSONP_1.1 1.4, JSONB_1.0 1.3

[Raymond Auge]

+1

- Ray

On Tue, Apr 28, 2020 at 2:14 AM Francois Papon 
wrote:

> +1 (non-binding)
>
> regards,
>
> Françoisfpa...@apache.org
>
> Le 28/04/2020 à 07:59, Romain Manni-Bucau a écrit :
>
> Hi everyone,
>
> Here is the vote for some of our spec with jakarta shades.
>
> Tags:
> -
> http://svn.apache.org/repos/asf/geronimo/specs/tags/geronimo-jcdi_2.0_spec-1.2/
>  (rev 1877103)
> -
> http://svn.apache.org/repos/asf/geronimo/specs/tags/geronimo-annotation_1.3_spec-1.3/
>  (rev
> 1877106)
> -
> http://svn.apache.org/repos/asf/geronimo/specs/tags/geronimo-atinject_1.0_spec-1.2/
>  (rev
> 1877109)
> -
> http://svn.apache.org/repos/asf/geronimo/specs/tags/geronimo-interceptor_1.2_spec-1.2/
>  (rev
> 1877112)
> -
> http://svn.apache.org/repos/asf/geronimo/specs/tags/geronimo-json_1.1_spec-1.4/
>  (rev
> 1877115)
> -
> http://svn.apache.org/repos/asf/geronimo/specs/tags/geronimo-jsonb_1.0_spec-1.3/
>  (rev
> 1877118)
> Dist area:
> https://repository.apache.org/content/repositories/orgapachegeronimo-1124
> Staging repo: https://dist.apache.org/repos/dist/dev/geronimo/specs/
> My key is still the same.
>
> Please vote:
>
> [ ] +1 release it
> [ ] -1 dont' release it ${cause}
>
> Vote is open for 3 days or until we get enough bindings as usual.
>
> Romain Manni-Bucau
> @rmannibucau  |  Blog
>  | Old Blog
>  | Github
>  | LinkedIn
>  | Book
> 
>
>

-- 
*Raymond Augé* 
 (@rotty3000)
Senior Software Architect *Liferay, Inc.* 
 (@Liferay)

Re: Jakarta artifacts?

[Romain Manni-Bucau]

Le mar. 28 avr. 2020 à 11:03, Mark Struberg  a écrit :

> Why is there any pain?
> There is _nothing_ which we will change for the current specs. They are
> binary verified and frozen.
> The only thing we changed in the last 2 years has been the java9 module
> name, but that is different in jakartaee anyway plus it's no code change
> but just a maven thingy.
>

Nop ;), we also changed OSGi meta and we always think it is frozen until we
re-release them (most of the spec are not in 1.0 so I tend to think we can
need minor adjustments).

Now if you have time to duplicate all specs, I'm happy with that, I just
don't want to handle the duplicated work maintenance since it does not
bring much IMHO for us and users but if  somebody else does it I'm happy
with whatever solution gives us jakarta artifacts (side note being it
includes jira versions, downloads, potential patches if any and dist area
management to be very explicit).

Don't think it prevents the pending vote to finish though since it does not
affect what we will do, it just adds experimental classified (as mvn
classifier) artifact so no issue I guess.


>
> LieGrue,
> strub
>
>
> Am 28.04.2020 um 10:38 schrieb Romain Manni-Bucau :
>
> Clearly the shade option is easier today because it does not require to
> maintain 2 branches.
>
> Do you have pointers on the removed methods? Most specs I reviewed were
> just 1-1.
> Anyway, it is a first step for us to enable people to play with jakarta
> package, we can still adjust things and if the fork is too big we would
> have to use jakarta branch but it would be a pain for no gain IMHO.
> If it is just about dropping a few methods we can add a shade transformer
> to do it IMHO.
>
> Romain Manni-Bucau
> @rmannibucau  |  Blog
>  | Old Blog
>  | Github
>  | LinkedIn
>  | Book
> 
>
>
> Le mar. 28 avr. 2020 à 10:30, Mark Struberg  a écrit :
>
>> We also have a branch for jakartaee already.
>>
>> The question is which one is easier. There are btw slight differences. It
>> is NOT 1:1! Some methods got removed!
>> So I'd rather go the branch way.
>>
>> LieGrue,
>> strub
>>
>> Am 28.04.2020 um 07:37 schrieb Romain Manni-Bucau > >:
>>
>> Hi everyone,
>>
>> Should we release jakarta artifacts?
>> Short term I'm just expecting a shade added to the default artifact since
>> for now it is 1-1 and for jakarta9 we would do the new bundle/artifact as
>> usual probably (seems the less costly compromise to me).
>>
>> If so (we do it now) I can take CDI SE stack + JSONB stack at my charge.
>> Happy to get help if anyone is interested.
>>
>> Romain Manni-Bucau
>> @rmannibucau  |  Blog
>>  | Old Blog
>>  | Github
>>  | LinkedIn
>>  | Book
>> 
>>
>>
>>
>

Re: Jakarta artifacts?

[Mark Struberg]

Why is there any pain?
There is _nothing_ which we will change for the current specs. They are binary 
verified and frozen.
The only thing we changed in the last 2 years has been the java9 module name, 
but that is different in jakartaee anyway plus it's no code change but just a 
maven thingy.

LieGrue,
strub


> Am 28.04.2020 um 10:38 schrieb Romain Manni-Bucau :
> 
> Clearly the shade option is easier today because it does not require to 
> maintain 2 branches.
> 
> Do you have pointers on the removed methods? Most specs I reviewed were just 
> 1-1.
> Anyway, it is a first step for us to enable people to play with jakarta 
> package, we can still adjust things and if the fork is too big we would have 
> to use jakarta branch but it would be a pain for no gain IMHO.
> If it is just about dropping a few methods we can add a shade transformer to 
> do it IMHO.
> 
> Romain Manni-Bucau
> @rmannibucau  |  Blog 
>  | Old Blog 
>  | Github  
> | LinkedIn  | Book 
> 
> 
> Le mar. 28 avr. 2020 à 10:30, Mark Struberg  > a écrit :
> We also have a branch for jakartaee already.
> 
> The question is which one is easier. There are btw slight differences. It is 
> NOT 1:1! Some methods got removed!
> So I'd rather go the branch way.
> 
> LieGrue,
> strub
> 
>> Am 28.04.2020 um 07:37 schrieb Romain Manni-Bucau > >:
>> 
>> Hi everyone,
>> 
>> Should we release jakarta artifacts?
>> Short term I'm just expecting a shade added to the default artifact since 
>> for now it is 1-1 and for jakarta9 we would do the new bundle/artifact as 
>> usual probably (seems the less costly compromise to me).
>> 
>> If so (we do it now) I can take CDI SE stack + JSONB stack at my charge.
>> Happy to get help if anyone is interested.
>> 
>> Romain Manni-Bucau
>> @rmannibucau  |  Blog 
>>  | Old Blog 
>>  | Github 
>>  | LinkedIn 
>>  | Book 
>> 

Re: Jakarta artifacts?

[Romain Manni-Bucau]

Clearly the shade option is easier today because it does not require to
maintain 2 branches.

Do you have pointers on the removed methods? Most specs I reviewed were
just 1-1.
Anyway, it is a first step for us to enable people to play with jakarta
package, we can still adjust things and if the fork is too big we would
have to use jakarta branch but it would be a pain for no gain IMHO.
If it is just about dropping a few methods we can add a shade transformer
to do it IMHO.

Romain Manni-Bucau
@rmannibucau  |  Blog
 | Old Blog
 | Github  |
LinkedIn  | Book



Le mar. 28 avr. 2020 à 10:30, Mark Struberg  a écrit :

> We also have a branch for jakartaee already.
>
> The question is which one is easier. There are btw slight differences. It
> is NOT 1:1! Some methods got removed!
> So I'd rather go the branch way.
>
> LieGrue,
> strub
>
> Am 28.04.2020 um 07:37 schrieb Romain Manni-Bucau :
>
> Hi everyone,
>
> Should we release jakarta artifacts?
> Short term I'm just expecting a shade added to the default artifact since
> for now it is 1-1 and for jakarta9 we would do the new bundle/artifact as
> usual probably (seems the less costly compromise to me).
>
> If so (we do it now) I can take CDI SE stack + JSONB stack at my charge.
> Happy to get help if anyone is interested.
>
> Romain Manni-Bucau
> @rmannibucau  |  Blog
>  | Old Blog
>  | Github
>  | LinkedIn
>  | Book
> 
>
>
>

Re: Jakarta artifacts?

[Mark Struberg]

We also have a branch for jakartaee already.

The question is which one is easier. There are btw slight differences. It is 
NOT 1:1! Some methods got removed!
So I'd rather go the branch way.

LieGrue,
strub

> Am 28.04.2020 um 07:37 schrieb Romain Manni-Bucau :
> 
> Hi everyone,
> 
> Should we release jakarta artifacts?
> Short term I'm just expecting a shade added to the default artifact since for 
> now it is 1-1 and for jakarta9 we would do the new bundle/artifact as usual 
> probably (seems the less costly compromise to me).
> 
> If so (we do it now) I can take CDI SE stack + JSONB stack at my charge.
> Happy to get help if anyone is interested.
> 
> Romain Manni-Bucau
> @rmannibucau  |  Blog 
>  | Old Blog 
>  | Github  
> | LinkedIn  | Book 
> 

[jira] [Created] (GERONIMO-6771) Reduce memory usage of circuit breaker by using BitSet

[Francois Papon (Jira)]

Francois Papon created GERONIMO-6771:


 Summary: Reduce memory usage of circuit breaker by using BitSet
 Key: GERONIMO-6771
 URL: https://issues.apache.org/jira/browse/GERONIMO-6771
 Project: Geronimo
  Issue Type: Improvement
  Security Level: public (Regular issues)
  Components: Safeguard
Reporter: Francois Papon
 Fix For: Safeguard_1.2.2


As discuss in the mailing list, we should replace Boolean[] with BitSet to 
reduce memory usage.

[https://lists.apache.org/thread.html/r72096fc8daed456f1e60e0a534d7cd51e9f2b75c49bf2478d289ab20%40%3Cdev.geronimo.apache.org%3E]

 



--
This message was sent by Atlassian Jira
(v8.3.4#803005)