Christopher Lenz wrote:
> Am 18.04.2002 16:30:59, schrieb Christopher Lenz <[EMAIL PROTECTED]>:
> 
>>Am 18.04.2002 13:46:34, schrieb Jean-Philippe Courson <[EMAIL PROTECTED]>:
> 
> [...]
> 
>>>ps : for more informations on proposed modifications, please see
>>>  http://www.mail-archive.com/[email protected]/msg02876.html
>>
>>I've just re-read the discussion (which I didn't closely follow, originally).
>>
>>One point: IMHO ContentInterceptor should have been an interface from the 
>>start. As the mechanism obviously hasn't been used much (due to it's limited 
>>usability), why not make it an interface now if we're changing the API anyway ? 
>>And provide an abstract class AbstractContentInterceptor to ease 
>>implementation.
>>
>>Are there any good reasons ContentInterceptor isn't an interface ? Would the 
>>above be too much of a change ?
> 
> 
> ...and,
> 
> [sorry but I hadn't been dealing with the whole ContentInterceptor API before]
> 
> if there is a preStoreContent(), why shouldn't there be preRetrieveContent() and 
> preRemoveContent() ? preRetrieveContent() might not make sense, but there are 
> probably use cases for preRemoveContent(). Either way, the interface should be 
> consistent and have pre/post hooks for each event. Further, there are no hooks 
> for the fork/merge "events". Should those be added to ?
>
> We should try to get the API right if we're changing it anyway, IMHO.

I agree

> -chris
> _______________________________________________
>  /=/ cmlenz at gmx.de
> 
> 
> 
> 
> 
> --
> To unsubscribe, e-mail:   <mailto:[EMAIL PROTECTED]>
> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
> 
> 
> 




--
To unsubscribe, e-mail:   <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>

Reply via email to