The error is reported, but understanding the error is tricky when what I see when I examine the name matches an older definition of the verb. It's a solvable problem, but why isn't 4 :(;:'1 : 2')) an error?
As for why I prefer the numeric notation, I sometimes like using J's trace facility, and: require'trace' verb 3 verb_jtrace_ 2 If I ever want to coinsert'jtrace' to help me debug something I do not want all of my verb definitions to start failing. (And I don't have a faster way of having jtrace find name definitions in some other locale.) Thanks, -- Raul On Mon, Feb 17, 2014 at 4:30 PM, Brian Schott <schott.br...@gmail.com>wrote: > Raul, > > I had recently come across your "The non-empty monad definition is > silently discarded." But I concluded that I had just misapplied dyad when I > meant verb. After all, a dyad is just a dyad, and the following error is > reported. > > example 4 > |domain error: example > | example 4 > > > You did not explain why you have stopped using the names like monad, dyad, > and verb. Do you have a reason? I suppose it is a little artificial to use > the words in place of the numbers if one does not also use the Primitive > names in place of the Primitive symbols. I am sort of indifferent to the > two verb defining methods. > > > > On Mon, Feb 17, 2014 at 3:52 PM, Raul Miller <rauldmil...@gmail.com> > wrote: > > > I have gotten into the habit of using tacit representation of explicit > > definitions. > > > > In other words: > > > > verbM=: 3 :0 > > 'example monad definition' > > ) > > > > verbD=: 4 :0 > > 'example dyad definition' > > ) > > > > verbV=: 3 :0 > > 'example monad definition' > > : > > 'example dyad definition' > > ) > > > > instead of: > > > > verbM=: monad define > > 'example monad definition' > > ) > > > > verbD=: dyad define > > 'example dyad definition' > > ) > > > > verbV=: verb define > > 'example monad definition' > > : > > 'example dyad definition' > > ) > > > > Notice, in particular: > > > > monad > > 3 > > verb > > 3 > > dyad > > 4 > > > > (Aside: note that the top links from > > http://www.jsoftware.com/phrases/explicit_def.htm are dead. Maybe > someone > > should spider jsoftware and compile a list of pages which have this > defect? > > (From experience: if you try to catalog all defects the problem becomes > > overwhelming and neglected.)) > > > > Anyways, I've been noticing an issue with the current implementation of > (4 > > :) which I think should be changed. > > > > example=: 4 :0 > > 'example monad definition' > > : > > 'example dyad definition' > > ) > > > > And that "works"... sort of: > > > > example > > > > 4 : ' ''example dyad definition''' > > > > > > See what happens here? The non-empty monad definition is silently > > discarded. I think that J should signal an error for this case. Otherwise > > when you promote a dyad definition to a dual valence definition you run > the > > risk of a bug that looks like you are loading a stale version of your > > script. > > > > This proposed change does not conflict with > > http://www.jsoftware.com/help/dictionary/d310n.htm > > > > What do you all think? > > > > (That said, note also that we now have an open source license on J - > maybe > > one of us should step up and start releasing community versions, with > > something new from 9!:14'' ?) > > > > Thanks, > > > > -- > > Raul > > Thanks, > > > > -- > > Raul > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > > > -- > (B=) > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
