Am 16.01.11 19:54, schrieb Felix Meschberger:
> Hi,
> 
> Am Sonntag, den 16.01.2011, 18:06 +0100 schrieb Carsten Ziegeler: 
>> Felix Meschberger  wrote
>>> Hi,
>>>
>>> Am Samstag, den 15.01.2011, 11:16 +0100 schrieb Carsten Ziegeler: 
>>>> Yes, but in these cases all components implementing the service would be
>>>> active though only one is used.
>>>
>>> No. As long as you only have the ServiceReference in hand delayed
>>> components are not (fully) activated, i.e. the classes not instantiated
>>> (and potentially not even loaded).
>>>
>> Yepp exactly - I realized this directly after sending my mail above :)
>>
>> But still, using this technique your component has a 1..n reference to a
>> service though actually it's a 1..1 (or 0..n vs 0..1).
> 
> Yes, but so what ? ;-)
> 
It feels like what it is - a workaround where the spec could be slightly
changed to support this use case :)

Carsten


-- 
Carsten Ziegeler
[email protected]
_______________________________________________
OSGi Developer Mail List
[email protected]
https://mail.osgi.org/mailman/listinfo/osgi-dev

Reply via email to