maven-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jason van Zyl <ja...@tesla.io>
Subject Re: 1.5 Annotations for Mojo
Date Sat, 28 Apr 2012 13:29:17 GMT
On Apr 27, 2012, at 9:55 PM, Hervé BOUTEMY wrote:

> yes, +1 for the first step - ie annotations for plugin.xml generation

Yes, this is the wisest, most compatible way to go. It can easily be done using ASM to process
the annotations and produce the plugin.xml. Works fine. This also allows people who want to
create different development patterns for plugins remain compatible. All you require in the
end are instructions that are encoded in a plugin.xml.

> but -0 for discovery with runtime annotation instead of plugin.xml
> 

-1

This will cause massive compatibility problems. Roughly 30% of the user base has migrated
to Maven 3.x which means the vast majority of users will not be able to use plugins that require
classpath scanning and processing.

Anything that produces a plugin.xml by whatever means can be made to work in Maven 2.x and
3.x.

> and yes, I took time to document (end reference JFrog's implementation) but 
> don't have really time to code it even if the feature would be a great benefit 
> IMHO
> 
> thenks for your work on that: I'll review and help
> 
> Regards,
> 
> Hervé
> 
> Le vendredi 27 avril 2012 20:01:34 Olivier Lamy a écrit :
>> I would prefer not be core change dependent. Some folks are still using 2.x
>> .
>> And 'classloader scanning' at runtime level will have IHMO a performance
>> cost !
>> So I would prefer plugin metadata generation (the trick will be having the
>> mix of doclet and annotations : perso I'm not sure it's a good idea to have
>> both ? )
>> 
>> --
>> Olivier
>> Send from a VT100 phone console
>> 
>> Le 27 avr. 2012 18:26, "Robert Scholte" <apache@sourcegrounds.com> a écrit
:
>>> +100
>>> 
>>> I've had a talk about this with Simone this month.
>>> He'd done some investigation already.
>>> I think there are two separate steps to be taken.
>>> 1. Generate the plugin.xml based on both annotations and doclets
>>> 2. Extend Maven Core to understand Mojo Annotations as well
>>> 
>>> The first one would already be a huge improvement for all mojo-devs.
>>> 
>>> -Robert
>>> 
>>> Op Fri, 27 Apr 2012 16:36:57 +0200 schreef Olivier Lamy
>>> <olamy@apache.org
>>> 
>>> Hi,
>>> 
>>>> I'd like to work on 1.5 Annotations for Mojos.
>>>> Hervé started documentation here:
>>>> https://cwiki.apache.org/**confluence/display/MAVEN/Java+**
>>>> 5+Annotations+for+Plugins<https://cwiki.apache.org/confluence/display/
>>>> MAVEN/Java+5+Annotations+for+Plugins>
>>>> 
>>>> The Stephen's idea for named without Mojo prefix looks fine (at least
>>>> for me). But I would prefer to not implement (yet) the other idea with
>>>> 
>>>> synthetic bridging classes.
>>>> 
>>>> I can start the job next week (probably in a branch).
>>>> 
>>>> Comments ?
>>>> 
>>>> Thanks,
>>> 
>>> ------------------------------**------------------------------**--------
>>> -
>>> To unsubscribe, e-mail:
>>> dev-unsubscribe@maven.apache.**org<dev-unsubscribe@maven.apache.org>
>>> For additional commands, e-mail: dev-help@maven.apache.org
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@maven.apache.org
> For additional commands, e-mail: dev-help@maven.apache.org
> 

Thanks,

Jason

----------------------------------------------------------
Jason van Zyl
Founder,  Apache Maven
http://twitter.com/jvanzyl
---------------------------------------------------------

First, the taking in of scattered particulars under one Idea,
so that everyone understands what is being talked about ... Second,
the separation of the Idea into parts, by dividing it at the joints,
as nature directs, not breaking any limb in half as a bad carver might.

  -- Plato, Phaedrus (Notes on the Synthesis of Form by C. Alexander)





Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message