On Tuesday, June 7, 2011 3:00:49 PM UTC+3, nicolas.deloof wrote:
>
> Right, we can wait yet another 4 month before doing a release just to have 
> a larger change log,
> or we can follow the best-practice to release often, and plan a new release 
> ASAP.
>

Sorry if I sounded too harsh, that was not my intention. I greatly 
appreciate the time, all of you, put into this and I 've also made small 
contributions myself (MGWT-180). I am all-in for release often and early but 
you've taken it to the extreme opposite (4 months for a change log! :)). 
Striking a good balance is an art in itself.
 

>
> This release has some untracked feature / enhancement, I agree we should 
> better log into Jira what happens in code : 
>
> The "compatibility" work on SDK was about checking for regression, 
> especially with integrated support for the javax.validation dependency, and 
> ready-to-go support for depedencies declared by gwt-user.pom. Running with 
> GWT 2.3.0 requires to override the plugin dependencies, that is only a 
> workaround.
>

So does that mean it's no longer necessary to declare the (missing) 
javax.validation dependencies in each projects pom, since they are pulled-in 
from the plugin? If that's true then that's great! This is exactly what I'm 
asking about when I say "what justifies this release".. I am not talking 
about a full fledged changelog, just a couple of remarks on what changes / 
improvements are made available with this proposed release.


> There was no work on adding support for new options, as MGWT-277 - but 
> you're welcome to contribute ! just ask, I'll grant you access to github 
> account. If you do so, please add IT-tests for them so that we can latter 
> track regressions.
>
> About our 2 blockers :
>
>    - MGWT-246 <http://jira.codehaus.org/browse/MGWT-246> is about the 
>    archetype that is only an helper feature (gwt sdk now has a -maven option, 
>    even it has not the same result), and I know many users that run it 
> without 
>    error. Maybe it needs more documentation, but I've no time by myself to 
>    check (I don't use Eclipse). Contributions are welcome ! 
>    - MGWT-278 <http://jira.codehaus.org/browse/MGWT-278>  is about running 
>    Maven+Eclipse+GoogleEclipsePlugin together. As for MGWT-246 I know many 
> user 
>    that make this work and others that complain about failures. I myself 
> didn't 
>    have this to work few times and had to re-create my project from scratch. 
> I 
>    can't see a fix for that, just expect better integration between M2E & GEP 
>    in future, especially Google including a M2E ProjectConfigurator... wait & 
>    see. 
>
> I'd be really, really pleased to see a fiable Eclipse+M2E+GEP configuration 
> and integrate it to the archetype. I just don't understand why GEP works so 
> bad when the project uses M2E and its dependency management :'( - or maybe 
> why M2E works so bad when the project uses GEP - or maybe why Eclipse is so 
> tricky :P
>

We don't use Eclipse either so we can't really help on these two. But I'll 
gladly take a look at MGWT-277, regardless if this version is released now 
or not.

-fotos
 

>
>
> 2011/6/7 Fotos Georgiadis <gfo...@gmail.com>
>
>> I go for 0 on this one.
>>
>> Given that you can compile GWT 2.3 projects with the 2.2.0 version of the 
>> plugin just fine, I think version 2.3.0 should either be compatible with the 
>> latest GWT version (e.g. MGWT-277) or at least solve some issues (there are 
>> 2 blockers and 24 majors). How does this version provide GWT 2.3.0 
>> compatibility? A change log would be really nice, otherwise I can't find 
>> anything that justifies this release besides a couple of frustrated newbie 
>> users (that search for a documented version not yet released).
>>
>> On a side note: gwt-m-p 2.3.0 works fine in our project.
>>
>> -fotos
>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe from this list, please visit:
>>
>>    http://xircles.codehaus.org/manage_email
>>
>>
>
---------------------------------------------------------------------
To unsubscribe from this list, please visit:

    http://xircles.codehaus.org/manage_email

Reply via email to