cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Stefano Mazzocchi <>
Subject Re: FYI: Using Java to define Cocoon Flow
Date Mon, 12 May 2003 23:15:17 GMT
on 5/12/03 11:53 AM Bertrand Delacretaz wrote:

> Le Lundi, 12 mai 2003, à 18:42 Europe/Zurich, Reinhard Pötz a écrit :
>>Ovidio points in one of his latest blogs to a Java implementation
>>of the Cocoon control flow:
> (That's Ovidiu by the way ;-)
> Just got a nice comment from Alex Krut at 
>, he says 
> "Kudos to people who architected cocoon and avalon - awesome work.". 
> There's no way I can count me in that group (yet ;-) but you know who 
> you are...
> -Bertrand

I'll tell you what: if I had a java implementation of flow in cocoon
*today*, I would still use the javascript version for these reasons:

 1) ease of use: scripting is the perfect language for gluing components
 2) try/fail cycle: instantaneous in javascript, less so in java (how
would a java-enabling continuation engine work with dynamic
compilation?) [also requires the compiler in memory!]
 3) syntax sugar: I'm getting more and more lazy so things like

   sendPage("blah", { whatever : whatever })

are IMO, much more friendly than any

   Map map = new HashMap();

and the ability to do things like


instead of

   Request request = ObjectModelHelper.get(ObjectModelHelper.REQUEST);
   String action = request.getParameter("action");

simply make life easier.

Besides, I don't think a java implementation of the flowscript would
make that difference in performance (which is the only thing I think it
would be valuable) and, also, it's not clear if those java-based
continuations would be long-term persistante as we can do by serializing
rhino continuations to disk today.

So, besides performance (but I would like to see the numbers first!), i
don't see what you gain with a java flowscript implementation (besides
pleasing those who don't get the use of weakly typed languages for
framework composition)


View raw message