Re: Load User module
On 8 Nov 2002, at 7:09, kbochert at copper.net wrote:
>
> Hi Derek Parnell, you wrote on 11/8/02 5:26:38 AM:
>
> >
> >Robert,
> >is this the sort of thing you are suggesting for us ? ...
> >
> >include linux/ui.e as lui
> >include win32/ui.e as wui
> >include dos/ui.e as dui
> >.
> >.
> >.
>
> How about a few more keywords---
>
> Linux_include linux/ui.e
> Win32_include win/ui.e
> Dos32_include dos/ui.e
>
> ...
>
> There could even be a User-defined platform:
>
> if ... then
> User = TRUE
> endif
>
> User_include conditional/code.e
>
> To get really extreme, 'conditional/code.e' could modify
> the 'User' variable and therby control further includes.
I have a simpler suggestion (rather like goto replaces all jumps by other
names): dynamic loading/unloading. Like in mirc (the toy non-language that's
extremely versatile). It seems to me one of the strengths of an interpreted
language is you *can* point the core of the interpreter at anything and say
"run this script", be it a variable or a new include file. And if you can point
it
to a new include file, the an unload command becomes useful too, and since
all this loading/unloading is going on, the $isalias() to peek into the
interpreter's table of what is currently loaded will be useful too. And if you
have functions by the same name in more than one include file, they are
executed in order of loading. And you can do loading at any point in the
program flow. Same for events. Simple, no? You are the programmer, you
tell the language how to do the job, the interpreter shouldn't be fighting you
over it.
Kat
|
Not Categorized, Please Help
|
|
