>
> His arguments on "typeless" languages is useless.
> You don't need a "scripting language" to
> get usable abstractions without the need
> to deal with low-level issues.
>
> button .b -text Hello! -font {Times 16} -command {puts hello}
>
> In Macintosh Common Lisp I'll write this as:
>
> (make-instance 'button-dialog-item
> :dialog-item-text "Hello"
> :view-font '("Times" 16)
> :dialog-item-action (lambda (item) (print "hello")))
>
> You can develop user interfaces as fast as with TCL/TK with MCL.
> Still you do it in one language, not in two. Without
> external *and* internal representation of data/procedures as strings.
>
I agree entirely. It seems to me that what sets Ousterhout's so-called
scripting languages apart from "system" languages is nothing to do with typing,
or internal representation of variables, but simply argument passing syntax.
The only thing preventing a concise creation of a user interface element (as
above) from being done in C++, is that in C++ actual args are substituted for
formals based on position rather than keyword, so you would have to give them
all (long and messy) and without keyword names (confusing). Fix this and the
barriers come down!
In fact, I believe Ada allows use of keywords to identify arguments (correct me
if I am wrong), so I guess this means Ada could find use as a scripting
language!
Regards
M.Beach
|