Need for testing [was Re: one more time.lisp patch]

Aleksandar Bakic wrote:
> Line numbers are wrt. my older local copy, but you can see that usec was
> missing.

Thanks for the patch, Alex. This is about the 4th or or 5th fix to the
microsecond patch I accepted last year. While improvements to CLSQL
are welcome, it is suboptimal to incorporate patches that have so many
bugs.

A robust approach, I think, would be to require patches to CLSQL to
also add to the test suite. I read Beck's "Test-Driven Development" a
few years ago. His stringent approach is only write code after one has
written a test to verify the correct operation of the code.  For
CLSQL, a good adaption might be to only commit patches which also have
tests that show the incorrect previous behavior and correct current
behavior.

Kevin

case, I think a good ada




> Alex
> 
> @@ -932,9 +938,9 @@
>      (print-date time style)))
>  
>  (defun print-date (time &optional (style :daytime))
> -  (multiple-value-bind (usec second minute hour day month year dow)
> +  (multiple-value-bind (second minute hour day month year dow)
>        (decode-time time)
> -    (declare (ignore usec second))
> +    (declare (ignore second))
>      (multiple-value-bind (hours meridian)
>          (time-meridian hour)
>        (ecase style
> @@ -967,9 +973,8 @@
>           (format nil "~d/~d/~d" month day year))))))
>  
>  (defun time-element (time element)
> -  (multiple-value-bind (usec second minute hour day month year dow)
> +  (multiple-value-bind (second minute hour day month year dow)
>        (decode-time time)
> -    (declare (ignore usec))
>      (ecase element
>        (:seconds
>         second)
> 
> 
> 	
> 		
> __________________________________ 
> Yahoo! Mail - PC Magazine Editors' Choice 2005 
> http://mail.yahoo.com
> _______________________________________________
> CLSQL-Devel mailing list
> [email protected]
> http://lists.b9.com/mailman/listinfo/clsql-devel
-- 
Kevin Rosenberg
kevin-HJRc7zDS/[email protected]