avalon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Stephen McConnell <mcconn...@osm.net>
Subject Re: Service or not?
Date Sun, 07 Jul 2002 06:39:50 GMT


Peter Donald wrote:

> At 06:14 AM 7/7/2002 +0200, you wrote:
> It now looks like
>
>>> <dependency>
>>>   <role>blah</role>
>>>   <service classname="FooService"/>
>>>   <attributes>
>>>     <attribute ...>
>>>   <attributes>
>>> </dependency>
>>
>>
>>
>> Isn't this going to create an inconsitency with the <service/> 
>> element contained in the <services/> tag at the DTD level ?
>
>
>
> Right. Maybe it was <service-ref/> or something. Look in DTD for exact 
> structure.
>
>>> so attributes are associated with dependency directly rather than 
>>> via service reference.
>>>
>>> Like?
>>
>>
>>
>> I would prefer that we leave the entire <component-info> as is for a 
>> couple of months.
>
>
> I need it now and are are going to finalize the version in 
> containerkit now. Phoenix and myrmidon have already been converted 
> across to containerkit and I have been playing with fortress to get it 
> to do the same. 


This is bad move if we want to try and a build a common metamodel.  The 
containerkit model on the info side is ok - but there is a lot of room 
for improvement at the api and packaging levels.  The meta model in its 
current form is too difficult to work with for dynamic applications - 
there really isn't any need for the component  meta model to be an 
imutable class.  Removing that restriction is *really* needed as soon as 
you get into dynamic assembly.  There is also the completely unresolved 
issue of context at the meta versus internal entry levels.  Also core 
metamodel stuff should be seperate from kernel related stuff (i.e. 
everything in kernel and dependecies should be moved out) .  

Sound to me like we are nowhere near finalizatrion.

Steve.

>
>
> Cheers,
>
> Peter Donald
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> "Faced with the choice between changing one's mind,
> and proving that there is no need to do so - almost
> everyone gets busy on the proof."
>              - John Kenneth Galbraith
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
>
> -- 
> To unsubscribe, e-mail:   
> <mailto:avalon-dev-unsubscribe@jakarta.apache.org>
> For additional commands, e-mail: 
> <mailto:avalon-dev-help@jakarta.apache.org>
>

-- 

Stephen J. McConnell

OSM SARL
digital products for a global economy
mailto:mcconnell@osm.net
http://www.osm.net




--
To unsubscribe, e-mail:   <mailto:avalon-dev-unsubscribe@jakarta.apache.org>
For additional commands, e-mail: <mailto:avalon-dev-help@jakarta.apache.org>


Mime
View raw message