On Oct 15, 2004, at 06:59, gabriele renzi wrote:

> this is interesting, but still requires a lot of typing from you.
> I'd prefer florian gross' approach to extract data from embedded 
> comments. No performance hit, real samples, no extraneous code.

Florian's approach definitely has advantages... the one killer 
advantage to me of the direction I'm taking is its simplicity on the 
testing framework side. Parsing Ruby code is not something I want to do 
in test/unit, so I'm going with an approach that doesn't require it. Of 
course I'll have to parse some stuff in RDoc to get it integrated with 
the documentation, but RDoc already does Ruby parsing.

Of course, I'll stop no one from using Florian's library... I think 
it's a great idea that can peacefully coexist with test/unit2.

Thanks for the feedback,


Nathaniel
Terralien, Inc.

<:((><