[imp-hackers] ansi test issues

[Raymond Toy]

>>>>> "Sam" == Sam Steingold <sds at gnu.org> writes:

    >> However, before adding different return values, wouldn't the changes
    >> need to be discussed?

    Sam> of course any functional change should be discussed and documented!

    Sam> E.g., consider print-level.[89]
    Sam> the text and examples in
    Sam>  http://www.lispworks.com/documentation/HyperSpec/Body/v_pr_lev.htm
    Sam> seem to indicated that
    Sam> (write-to-string #S(foo) :print-level 0)
    Sam> should return "#" and not "#S(foo)" as the tests mandate.

While I'm grateful that someone is looking into these issues,
shouldn't this discussion be on the ansi-test mailing list?

Having said that, I'm glad to see the discussion here.  One less
mailing list I need to subscribe to. :-)

And I do hope someone will fix the subtype tests which assume that
simple-base-string and simple-string are different (and base-char and
character).  This is not true in cmucl for either the 8-bit or Unicode
version; simple-base-string and simple-string are the same types
because base-char and character are the same.

    Sam> In fact, I argue that this is a case where the test is _wrong_ as
    Sam> opposed to two possible return values being both acceptable.

I suppose that discussions on the interpretation of a test against the
spec would be appropriate here, though, with the intent of getting
implementations to agree, perhaps.

Ray