To follow-up,
Considering the feedback I got on the TTM list, my current proposal is to use
this new mapping:
{<=} | ⊆
{!<=} | ⊈
{>=} | ⊇
{!>=} | ⊉
{<} | ⊂
{!<} | ⊄
{>} | ⊃
{!>} | ⊅
The format is to take the generic order-comparison operators and delimit them
with curly braces.
There already exist precedents in various places for overloading the compare ops
to mean sub/superset etc.
I wanted to use modified versions partly because one can potentially use the
original ops on any 2 Muldis D values, including 2 relations, so we avoid
conflicts, but more important I think these are different enough that they
should look different.
Curly braces are already a mnemonic in Muldis D and elsewhere that is associated
with sets (or unordered groups in general), so seeing them here may suggest "set
operator". Also, the braces in combination with < or > doesn't conflict
syntactically with any other use of braces that I have now.
Using symbolics rather than words is an advantage because not everyone agrees on
whether the unqualified word "subset" means "proper subset" or "subset or equal
to", and so the symbolic version should be easier for people to disambiguate.
So, barring any argument to do differently, the above is what I've decided on.
Thank you. -- Darren Duncan
_______________________________________________
muldis-db-users mailing list
muldis-db-users@mm.darrenduncan.net
http://mm.darrenduncan.net/mailman/listinfo/muldis-db-users
