To answer the question regarding the new type failure: it's very unlikely to be a bug in IGNORE-ERRORS. Does the evaluation of subforms in your evaluator handle multiple values correctly? (Another possibility that I seem to remember seeing fly past on IRC somewhere was that the error handling in FLOAT-RADIX is not quite right, but that is based on possibly faulty memory).
Could you update the patch to HEAD? There'll be a conflict in the change to NEWS, I'm afraid. Thanks.
To answer the question regarding the new type failure: it's very unlikely to be a bug in IGNORE-ERRORS. Does the evaluation of subforms in your evaluator handle multiple values correctly? (Another possibility that I seem to remember seeing fly past on IRC somewhere was that the error handling in FLOAT-RADIX is not quite right, but that is based on possibly faulty memory).
Could you update the patch to HEAD? There'll be a conflict in the change to NEWS, I'm afraid. Thanks.