It is already possible/convenient to write an adverb that returns an adverb.  
Often the returned adverb can be tacit.  An advantage is that the first binding 
of the adverb "compiles"/preparses the 2nd adverb application.  The full 
binding expression is a double adverb (u2 u1 a).  The disadvantage of double 
adverbs is that it is inconvenient to partially bind u2 without yet knowing u1.

If the proposal "let explicit conjunctions that mention v/n without u/m (or x 
or y) return a tacit adverb upon being bound with v" then the 
compiling/preparsing advantage exists while retaining full partial binding 
simplicity.

Examples

2 : ' "v' 1   NB. compiled tacit with single parameter.
"1

u1 (u2 2: ' "v')  NB. no tacit advantage, but simple partial binding semantics.
u2"u1


----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to