Hi John,There is indeed some inconsistency here. The "_on" suffix means different things in "refl_on" and "total_on", which is the source of your complaint. I'll have to figure out if the difference is intentional (I suspect it was) and if it can be abolished. For "strict_linear_order_on" it is probably not hard, because that one is hardly used, but "total_on" is a diffent matter...
This may take a little while. Tobias On 30/06/2015 18:03, John Wickerson wrote:
Dear Isabelle, I think the definition of "strict_linear_order_on" in HOL/Order_Relation.thy may be wrong, or at least, sub-optimal. Specifically, the definition does not state that the relation r must be contained within A*A. In contrast, the non-strict version, "linear_order_on", *does* make this restriction. In which case, can I propose that the definition be updated to (something like):definition "NEW_strict_linear_order_on A r â strict_linear_order_on A r â (r â A Ã A)"On another, slightly-related, matter -- I was hoping to find a lemma capturing the "order-extension principle". Something a bit like: "if p is a strict partial order on A, then there exists a strict linear order on A that is a superset of p". [I haven't formulated this very precisely, but can do.] This doesn't immediately appear to be in the HOL library, but seems an obvious candidate for inclusion. Did I miss it somewhere under a different name? Cheers, John
Description: S/MIME Cryptographic Signature