[mercury-users] Ambiguity

Michael Day mikeday at yeslogic.com
Wed Dec 10 17:03:53 AEDT 2008


Recent versions of the Mercury compiler (eg. from end of 2007 onwards) 
are giving us a lot of trouble with ambiguity. Some modules we cannot 
compile, as a particular predicate has excessive ambiguity including 
things such as ++ applying to either lists or strings, 'no' being either 
a bool or a maybe(T), and so on.

We've tried bumping up the ambiguity error limit, but once it goes 
beyond about 100000 no errors are produced, but compilation of the opt 
file still fails with an empty err file produced.

This is hugely annoying, as it requires renaming hundreds of symbols to 
try and reduce the ambiguity in modules that compiled fine in earlier 
versions of the compiler (eg. 2006-07-16).

Is there a way to tell the compiler to stop complaining about ambiguity 
and just get on with it? If it results in lengthy compile times or 
excessive memory usage we can always choose to interrupt it.

Best regards,


Print XML with Prince!
mercury-users mailing list
Post messages to:       mercury-users at csse.unimelb.edu.au
Administrative Queries: owner-mercury-users at csse.unimelb.edu.au
Subscriptions:          mercury-users-request at csse.unimelb.edu.au

More information about the users mailing list