From: Francis Cianfrocca <garbagecat10@...>
Date: 2006-06-15T00:03:45+09:00
Subject: Re: Cut-based AOP

------=_Part_337_21362007.1150297385684
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

>>>One can argue the merits (or lack there-of) of AOP all you want. But I
think you and Audtin are being overbarring to tell others they don't
need AOP. That's not your choice. <<<

Cool, I'm willing to be proved wrong. Hack it into Ruby so I can try it
(patch the language if you have to- Lord knows I've done that enough times),
write up the use cases with working examples in non-trivial programs, and
show how it's better than competing approaches.

It seems to me that Ruby has enough extensibility built in today, so that
with an IoC/DI approach you could do everything you can with AOP. But here
(and elsewhere) AOP is being presented as a new style of programming with
unique value, and that is far from convincing at the moment. I can see the
value of adding crosscuts (and DI) for system management and debugging
support, but not for extending application functionality. Until proven
otherwise, I'm going with my intuition that this is an excellent way to make
complex systems even more complex, fragile, and hard to manage.

------=_Part_337_21362007.1150297385684--