Re: [Resteasy-developers] Some Thoughts on the Interceptor API

[Ryan J. M. <ry...@da...>]

On Dec 29, 2008, at 4:26 PM, Bill Burke wrote:

>
>
> Ryan J. McDonough wrote:
>> *Web Beans Interceptors*
>> JSR-299 also defines/enhances the notion of interceptors in Java  
>> EE. What I don't want to be doing is creating yet another micro  
>> interceptor API that is kind of the same, yet a little bit  
>> different than EJB and Web Bean interceptors. However, not  
>> everything Web Beans is offering may be totally applicable to what  
>> we are doing.
>
> Its not :(
>
> Unfortunately, we have asynchronicity.  We need to be intercepting  
> at different points in the call cycle.  Simple method interception  
> is just not gonna cut it so neither is Web Beans (although the Web  
> Beans injection and component model will help tremendously)

Yep, that's my takeaway. But, as you point out in your last email, the  
way Seam does meta-annotations is exactly what we want to do.
>
>
>> What I'm saying is that we have very specific entry points for each  
>> of these features and there's too much ambiguity around and  
>> interceptor framework right now. I think our efforts would be  
>> better spent creating a pluggable layer at these entry points.  
>> Thoughts?
>
> An interceptor model, IMO, is better than specific entry points for  
> most things.  Specifically gzip, security, caching, async jobs.  It  
> might even solve most content negotiation customization.

I;m still a bit dubious, but I have started to go through the current  
interceptor impl you have and will most likely be using that as a  
starting point. However, I think for content negotiation, I may want  
address that a bit later. Thanks for the input.

Ryan-


>
>
> -- 
> Bill Burke
> JBoss, a division of Red Hat
> http://bill.burkecentral.com