A related mutation that'd be interesting (to me, at least) would be an APL that only allowed definitions in the form of Dyalog's direct definitions (d-fns). This would include all of the associated mechanisms, like lexical scoping and guards. Maybe even closures...
On Thu, 2014-05-15 at 22:28 +0200, Daniel H. Leidisch wrote: > Go for it! I thought about this, too. Give it proper lexically scoped > lambdas, namespaces, seamless access to some kind of dictionary data > type (maybe something like in K), make it extendable from within itself > like Lisp (or at least allow definition of primitives in APL, like in > NGN APL, and as planned for NARS2000), and I'll definitely take a closer > look at it, or help you out, if I can. I see no reason why you shouldn't > call it APL, if it adheres to the core principles. But, of course, > that's up to you. There were APLs before the standard, and there will be > APLs after it – unless it just dies a slow death. The extensions and > nonconformances of today might be tomorrow's standards – if there will > ever be another one.
