natural being the binding that would occur if modifier parameters were present.
ACV is natural. AVC is also not defined. Natural: A(VC) VCA is not horribly unnatural. (VC)A allows an escape for the more natural, but would be more cumbersome to do definition (V(]: C A)) CVA as natural (CV)A just saves much more parentheses. It is an adverb that can extend to right so: CVACA automatically becomes (CVA)CA instead of needing ((CV)A)CA CV(].A) accomplishes your recommendation without seeming that cumbersome. There is also the super flexible : VasC =: 1 : '[. u ].' then both CCA and ACC can have their middle C set as V VasC. or your request can also be accomplished as C([. V ].)A Since both AVC and CVA are undefined, we could make at least one of them the natural definition. AV~C could fill your original proposal, but I think if I/we can only have 1 of the 2 natural definitions, I'd pick AVC: Lets a series of pairable VC CV sequences get autopaired with ]:CV or ]:VC as leading "triplet". A problem with the modifier trains is that AA is not one in order to permit AAV CAA and AAA are natural. AAC could be defined naturally as uAACv CVA as an adverb (natural (CV)A) is more convenient than a "weird" conjunction, because we need more adverb forms for trainability. Forcing (AA) or AA]: to make a single adverb (when only one is allowed in modifier train slot, and so must be reduced to one) is cumbersome, and filling undefined modifier 3 trains in their natural adverb preferred form solves cumbersomeness in the more common trains. One of the reasons I hate AC so much is that 2-length modifier trains are "precious" They rule 4-length, 6-length trains too. Overly esoteric definitions should be implemented with 3-length trains. On Sunday, May 1, 2022, 11:07:59 p.m. EDT, Elijah Stone <elro...@elronnd.net> wrote: Why is CVA -> (CV)A more natural? Convention with trains is for the parser to make them as long as possible, since you can add the parens manually if you need the shorter train. CVA should be consistent with CCA and CVC. It's true there are multiple ways to express the same thing; there always are. Making AC consistent with CA is interesting, but since 2-trains are inherently asymmetric anyway (except CC), it seems less important. And changing AC breaks compatibility, whereas adding CVA does not. On Sun, 1 May 2022, 'Pascal Jasmin' via Programming wrote: > There is a more "obvious" interpretation of CVA > > (CV)A is the more natural extension of uCV A. (C(VA)) would be a possibility > that is less obvious. > > Your proposal can be accomplished with > > CV(].A) NB. also allows CV([.A) for flexibility of A binding. > > The current interpretation of AC could be accomplished with > > (AC[.) a: > > which would let AC be the "more natural consistent with CA" be the > conjunction uACv > > > > > On Sunday, May 1, 2022, 01:54:06 a.m. EDT, Elijah Stone <elro...@elronnd.net> > wrote: > > > > (Side-note: how come CCA and CVC are legal, but not CVA? I think CVA should > be uCv V vA. That seems fairly obvious and consistent, and it would simplify > the above definition of under_lock. AVC should be analogous. I think this > was proposed this a while ago but nothing came of it.) > > -E > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
