I can take it on since this impacts our getting WildFly certified for
EE10. What GitHub org would I use for these forked repos?
On Mar 3, 2022 at 10:25:39 AM, James Perkins <jperkins(a)redhat.com> wrote:
I do think this would be nice if someone is willing to manage it. I
can
say with RESTEasy for the Jakarta REST dependencies only I'm using the
Jakarta EE staging repository which is not ideal. It essentially pollutes
anyones local maven repository that who builds RESTEasy on it's main
branch. I also have to do some ugly hacks to get the TCK to download to use
it as a dependency.
As for changes, I can say there is likely at least one TCK change coming
to the Jakarta REST TCK, but it's just a test change.
On Wed, Mar 2, 2022 at 4:21 PM Scott Stark <sstark(a)redhat.com> wrote:
> None of the specs is under development, so at most what I would expect to
> see is there might need to be a respin of the staged content to update the
> javadocs. That has happened, but no one with a staged final API artifact
> has made a change to it. It might happen for example with JSTL or faces
> where significant restructuring of the project and building of the tlds has
> changed and might show a problem as the TCK is finalized.
>
> On Wed, Mar 2, 2022 at 12:26 PM Darran Lofthouse <
> darran.lofthouse(a)jboss.com> wrote:
>
>> On Wed, Mar 2, 2022 at 6:22 PM Brian Stansberry <
>> brian.stansberry(a)redhat.com> wrote:
>>
>>>
>>>
>>> On Wed, Mar 2, 2022 at 12:05 PM Darran Lofthouse <
>>> darran.lofthouse(a)jboss.com> wrote:
>>>
>>>> On Wed, Mar 2, 2022 at 5:08 PM Brian Stansberry <
>>>> brian.stansberry(a)redhat.com> wrote:
>>>>
>>>>>
>>>>>
>>>>> On Wed, Mar 2, 2022 at 10:50 AM Brian Stansberry <
>>>>> brian.stansberry(a)redhat.com> wrote:
>>>>>
>>>>>> As of last week there were no milestone releases for the EE 10
specs
>>>>>> listed at [1] available in maven central. It's not clear when
public
>>>>>> releases of some of these specs might be possible. I've
directly addressed
>>>>>> the WFLY component leads for the JIRA components that integrate
those
>>>>>> because I'd like to get answers to 3 questions:
>>>>>>
>>>>>> 1) Would near term (say next couple weeks) production of a
-jbossorg
>>>>>> variant of a milestone of a spec be useful to you? E.g. unblock
other
>>>>>> work. Probably this could be done by forking the official spec
branch into
>>>>>> our existing JBoss fork repos and publishing a -jbossorg variant
of
>>>>>> an existing tag.
>>>>>>
>>>>>
>>>> It might be worth thinking about how this will look and the flows in
>>>> git for this.
>>>>
>>>> We probably need to keep the option open that the module we are
>>>> forking is not final so subject to change so we may need subsequent
>>>> jbossorg releases.
>>>>
>>>> Where we fork a maven project it sounds like it will be two steps to
>>>> tag, the first being the migration to add jboss-parent as the parent so
we
>>>> inherit the configuration to publish to nexus. Secondly would be the
>>>> version change to add the jbossorg suffix.
>>>>
>>>> Where we need to sync from upstream maybe it would be sufficient to
>>>> use "git merge" to pull in the upstream changes, add another
commit for our
>>>> next jbossorg version and tag again?
>>>>
>>>> Or we could create a new branch each time we need to tag, fork from
>>>> upstream, apply the two changes, tag.
>>>>
>>>
>>> Sounds right, although my belief is that we're largely talking about
>>> specs that won't be changing much. I *believe* this is mostly about
specs
>>> where the API is done and available in staging repos, but won't be
>>> published in maven central until the spec TCK is finished and the vote
>>> happens. If specs are still actively developing their API hopefully they
>>> would publish a public milestone.
>>>
>>
>> +1 it sounds like assume we will do it once, but just be prepared in
>> case a subsequent release is needed. Early naming conventions may be
>> influenced by the preferred approach.
>>
>>
>>>
>>>>
>>>>>
>>>>>> 2) If not useful near term, an info on when the lack of a public
API
>>>>>> milestone would be a problem for you EE 10 work would be useful.
>>>>>>
>>>>>> 3) If useful, could you help?
>>>>>>
>>>>>> I'll create a google sheet to record input.
>>>>>>
>>>>>
>>>>>
>>>>>
https://docs.google.com/spreadsheets/d/1ql5hpsqs5ZjawErhyVQX8idigFfm0uZZx...
>>>>>
>>>>> Anyone can read that; if you want to write to it please send me your
>>>>> google account privately. Or you can give input via this thread and
I'll
>>>>> add it.
>>>>>
>>>>>>
>>>>>> Note that I'm not talking about using these releases for the
long
>>>>>> term; it's just about avoiding needing to use staging repos
for WF 27
>>>>>> development while we wait for releases in maven central.
Obviously there is
>>>>>> a cost/benefit tradeoff here, hence the request for input.
>>>>>>
>>>>>> [1]
>>>>>>
>>>>>> Authentication
>>>>>>
>>>>>> Concurrency
>>>>>>
>>>>>> Connectors
>>>>>>
>>>>>> Enterprise Java Beans (no biggie; just a micro)
>>>>>>
>>>>>> JSON Binding
>>>>>>
>>>>>> JSON Processing
>>>>>>
>>>>>> Messaging
>>>>>>
>>>>>> RESTFul Webservices
>>>>>>
>>>>>> Server Pages
>>>>>>
>>>>>> Servlet
>>>>>>
>>>>>> Standard Tag Library
>>>>>>
>>>>>> Websocket
>>>>>>
>>>>>> Best regards,
>>>>>> Brian
>>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> Brian Stansberry
>>>>> Principal Architect, Red Hat JBoss EAP
>>>>> He/Him/His
>>>>> _______________________________________________
>>>>> wildfly-dev mailing list -- wildfly-dev(a)lists.jboss.org
>>>>> To unsubscribe send an email to wildfly-dev-leave(a)lists.jboss.org
>>>>> %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s
>>>>
>>>>
>>>
>>> --
>>> Brian Stansberry
>>> Principal Architect, Red Hat JBoss EAP
>>> He/Him/His
>>>
>> _______________________________________________
>> wildfly-dev mailing list -- wildfly-dev(a)lists.jboss.org
>> To unsubscribe send an email to wildfly-dev-leave(a)lists.jboss.org
>> %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s
>
> _______________________________________________
> wildfly-dev mailing list -- wildfly-dev(a)lists.jboss.org
> To unsubscribe send an email to wildfly-dev-leave(a)lists.jboss.org
> %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s
--
James R. Perkins
JBoss by Red Hat