Re: [Sbcl-devel] Likely regression between d299ab6..2ea1111

[Jan M. <jmo...@te...>]

On Tue, 2014-04-22 at 02:11 +0200, Jan Moringen wrote:
> On Mon, 2014-04-21 at 22:49 +0200, Jan Moringen wrote:
>
> However, a different problem can be illustrated with
> 
>         (defun a (q)
>           (coerce q '(complex rational)))
> 
> This causes NUMERIC-TYPE-FORMAT to return NIL for which the COERCE
> transform is not prepared, generating

A sketch of a fix and a corresponding test is attached. What do you
think?

Kind regards,
Jan