From: "Marek Novotny" <mnovotny(a)redhat.com>
To: "Max Rydahl Andersen" <max.andersen(a)redhat.com>, "Rafael
Benevides" <benevides(a)redhat.com>
Cc: jbossdeveloper(a)lists.jboss.org, "Tomas Repel" <trepel(a)redhat.com>,
"Nikoleta Ziakova" <nziakova(a)redhat.com>
Sent: Friday, October 31, 2014 2:17:34 PM
Subject: Re: [jbossdeveloper] <repository /> on quickstarts pom.xml
On 31.10.2014 09:02, Max Rydahl Andersen wrote:
> On 30 Oct 2014, at 22:29, Rafael Benevides wrote:
>
>> On 9/22/14 10:54, Max Rydahl Andersen wrote:
>>> On 16 Sep 2014, at 11:56, Pete Muir wrote:
>>>
>>>> On 16 Sep 2014, at 00:27, Max Rydahl Andersen
>>>> <max.andersen(a)redhat.com> wrote:
>>>>
>>>>> On 9 Sep 2014, at 16:07, Rafael Benevides wrote:
>>>>>
>>>>>> We also need to check that with JBDS, because it's odd that
>>>>>> stacks only
>>>>>> uses "redhat-techpreview-all-repository" as additional
repository.
>>>>>
>>>>> because it was the only thing we needed originally.
>>>>>
>>>>>> Max/Fred,
>>>>>>
>>>>>> Is there any issues having
>>>>>>
https://maven.repository.redhat.com/earlyaccess/all/ as
declared
>>>>>> inside
>>>>>> quickstarts pom.xml ?
>>>>>
>>>>> sorry - but I need to grok when earlyaccess/all is even needed
>>>>> before answering that ? :)
>>>>>
>>>>> My default position would be that if you need earlyaccess you are
>>>>> expected to be able to understand why you need it.
>>>>>
>>>>> But i'm not sure if that is true anymore - hence wanting to ask
>>>>> why/when earlyaccess is ever needed ?
>>>>> is it for running quickstarts targeting a beta version of EAP ?
>>>>> (meaning no snapshots, but actual all releases in it?)
>>>>
>>>> Right, it’s where we put beta and alpha releases.
>>>>
>>>>>
>>>>> If yes, why doesn't /all get split up in /supported and
>>>>> /earlyaccess and /all will serve the combined result ?
>>>>
>>>> Propose this on the wolf list?
>>>
>>> Added as topic on tomorrows wolf call:
>>>
http://pad.engineering.redhat.com/wolf
>>>
>>> /max
>>> /max
>>>
http://about.me/maxandersen
>>
>> Max,
>>
>> What was the result of this subject in wolf call ?
>
> For now the split would stay, but we remove the /all grouping that was
> introduced by error (it was a neither or).
>
> the final grouping says on line 110 on
>
http://pad.engineering.redhat.com/wolf
>
> "Pete and David to follow up offline, fix up mojo doc, and send to
> list for signoff."
>
> I haven't seen this yet. Maybe I missed something while I was on PTO ?
>
>> Today I was helping Alexandre Porcelli from BRMS team to setup Ticket
>> Monster as a demo. But as he had issues with unresolved dependencies
>> and it seems that he have some restrictions to use the settings.xml
>> file I'm bringing this subject again to be discussed.
>
> What restrictions does he have ?
>
> And I guess the issue here is ticketmonster master needs the
> earlyaccess bits ?
>
> I think ticketmonster master could add the earlyaccess to its pom but
> then just make sure once it is releaesed it gets removed which should
> be fine since then it should be using any earlyaccess bundles, right ?
Not really, ticket-monster master branch doesn't exist. The closest to
master is 2.6.0.Final -
https://github.com/jboss-developer/ticket-monster/tree/2.6.0.Final -
which is the tight to WFK 2.6.0.GA and all GAVs should be resolvable
from
http://maven.repository.redhat.com/techpreview/all/
I think that Rafael knows that 2.7.x-develop is unstable and includes
internal product GAVs which are not in all or earlyaccess.
So if Alexandre uses 2.6.0.Final branch/tag he should get it to work
just with online redhat repository
http://maven.repository.redhat.com/techpreview/all/. If not please
advice him to send errors to Vineet or me.
I did verify 2.6.0.Final against techpreview/all not so long ago, and everything worked
fine.
I do recall an issue I ran into several weeks ago, involving some ShrinkWrap dependency,
but I can no longer reproduce it.
2.7.x-develop does require the EAP 6.3.1 and WFK 2.7.0 ER1 maven repos, so it wont work
with techpreview/all at the moment.
>
>> Due to his short time to prepare the Demo, he will use Petclinic from
>> Spring as he found that Ticket Monster is too tricky to build and
>> run. Maybe he can give us more details/feedback but it was good to
>> remember that we should improve our JBoss Developer Materials usability.
you should have the same result for jboss-wfk-quickstarts so I can't
believe it works and Timo doesn't.
>
> Agreed.
>
> /max
>
http://about.me/maxandersen
--
Marek Novotny
--
WFK and Seam Product Lead
Red Hat Czech s.r.o.
Purkynova 99
612 45 Brno
_______________________________________________
jbossdeveloper mailing list
jbossdeveloper(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/jbossdeveloper