[armedbear-devel] false-positive ansi tests

[Mark Evenson]

On 7/28/09 12:42 AM, Peter Tsenter wrote:
[…]

> (disassemble (eval '(function (lambda (x y) (cons x y)))))
>
> returns the following
>
> ; 'jad' is not recognized as an internal or external command,
> ; operable program or batch file.
> NIL

ABCL depends on the presence of ['jad'][1] in its PATH to work, using 
the external binary to disassemble the Java bytecode.  If one were 
officially keeping score for the ANSI tests, we might lose to the judges 
on whether we pass this test.  This interaction should certainly be 
better documented.

[1]: http://www.varaneckas.com/jad

[Note to ABCL developers]

Perhaps we should attempt to default to using 'javap' if Jad is not 
found?  The problem with this is that a JDK might not always be present, 
so this isn't a failproof solution either.

And 'jad' looks like it is eventually going to go away, as the author 
who never released the source, has stopped distributing binaries (the 
above link is to a "mirror").


-- 
"A screaming comes across the sky.  It has happened before, but there
is nothing to compare to it now."