[aosd-discuss] AOP myths and realities article published

['Rick Bradley']

* Bill Burke (bill at jboss.org) [060217 15:42]:
> Its not about letting go of some preconceived notion.  Please tell me
> how you can refactor effectively in a dynamic untyped language with an
> unfamiliar code base.

Refactoring worked quite well when Kent Beck was doing it in Smalltalk,
and continues to work in dynamic [0] languages to date, in the same way
it works in Java:  through application of unit testing.  While compiler
advice may aid tool support [1] few would make the claim that code which
merely compiles is correct.  Without unit tests, refactoring is a Cowboy
Coding operation.

[0] I continue to decline to discuss who's (un)typed, or how-typed, or
    to what degree.
[1] Though there's never yet been a refactoring tool in other languages
    as powerful as the Smalltalk browser.

Rick
-- 
 http://www.rickbradley.com    MUPRN: 593
                       |  window manager
   random email haiku  |  config and everything
                       |  is back to normal.