Let me first say: This is not a proposal to make symbol names order-dependent in general. Only variants where order is needed for disambiguation would be order-dependent.
Currently, ⪋ and ⪌ are lt.gt.equiv and gt.lt.equiv. However, the symbols ⪑ and ⪒ also exist, but have not yet been named. In order to name these in our current scheme, one would need to invent yet more unnatural modifiers.
With this proposal, we would be able to distinguish between
lt.equiv.gt ⪋
lt.gt.equiv ⪑
gt.equiv.lt ⪌
gt.lt.equiv ⪒
It would also enable #75 without abusing submodules, assuming we also get #94.
(Not to mention there are examples of current naming that could potentially be improved with this proposal, such as for arrows.)
Let me first say: This is not a proposal to make symbol names order-dependent in general. Only variants where order is needed for disambiguation would be order-dependent.
Currently, ⪋ and ⪌ are
lt.gt.equivandgt.lt.equiv. However, the symbols⪑and⪒also exist, but have not yet been named. In order to name these in our current scheme, one would need to invent yet more unnatural modifiers.With this proposal, we would be able to distinguish between
lt.equiv.gt⪋lt.gt.equiv⪑gt.equiv.lt⪌gt.lt.equiv⪒It would also enable #75 without abusing submodules, assuming we also get #94.
(Not to mention there are examples of current naming that could potentially be improved with this proposal, such as for arrows.)