Yes, definitely. Just trying to address them ...

Werner

Jerome Lacoste wrote:
> On Fri, Apr 11, 2008 at 1:01 PM, Werner Guttmann
> <[EMAIL PROTECTED]> wrote:
>>
>> Jerome Lacoste wrote:
>>> On Thu, Apr 10, 2008 at 2:34 PM, Werner Guttmann <[EMAIL PROTECTED]>
>>> wrote:
>>>> Hi,
>>>>
>>>>  I have been looking at the existing code base of the Castor plugin for
>>>> some
>>>> days now, and it looks like I will have to break existing contracts to
>>>> get
>>>> it working again (fully, against all Castor releases).
>>>>
>>>>  Right now (SVN), the plugin will run against versions of Castor starting
>>>> from 1.0.5 to now. But I'd like to get it working against older versions
>>>> of
>>>> Castor as well, but the way it's implemented would not allow that.
>>>>
>>>>  There's an arae of code where the plugin tries to outsmart the way
>>>> Castor
>>>> normally works. And the way this has been implemented .. well, moves a
>>>> couple of configuration parameters supplied by the means of a file to the
>>>> plugin configuration section.
>>>>
>>>>  As already said, I'd like to get it working again against all versions
>>>> of
>>>> Castor, and would like to refactor the plugin. For this, I would have to
>>>> drop this 'smartness'.
>>>>
>>>>
>>>>  My question: is this actually an option with Mojo plugins. Can we (well,
>>>> not) easily drop some configuration parameters even though they have been
>>>> supported in the past. Of course we'd make this clear in the plugin
>>>> configuration (as much as we don't like this), but I cannot see a way
>>>> around.
>>> I think Trygve was very against that idea some months ago arguing that
>>> we can not break backwards compatibility.
>> Which I can perfectly understand, as per my explanation(s) above. I am just
>> arguing with myself whether the pragmatic approach as outlined by yourself
>> doesn't complicate things for users.
>>
>> Personally, I think that the benefit that has been provided by the current
>> implementation with regards to 'over-engineering' plugin configuration (and
>> thus configuration of the underlying Castor source generator) is limited, if
>> not tending towards 0. I mean, the same can be achieved by supplying a
>> source generator configuration file as part of the plugin configuration. And
>> as it's this duality that makes the code hard to maintain, I'd like to get
>> rid of it.
>>
>> Anyhow, I think you've got a point there, and we'll definitely consider it.
> 
> Hei,
> 
> What's the status with/plans for the castor plugin ? Do you plan to
> release 1.0.2 or 2.0 as next release (POM says 2.0, Jira says 1.0.2) ?
> 
> I've commented some of the issues in Jira:
> http://jira.codehaus.org/browse/MCASTOR
> 
> Cheers,
> 
> Jerome
> 
> ---------------------------------------------------------------------
> 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