[Openmcl-devel] Feature request: DEFINE-TOPLEVEL-COMMAND should be external

[Tim Bradshaw]

On 24 Feb 2010, at 18:10, Ron Garret wrote:

> Obviously I know you know this.  What I'm really dancing around is  
> the question of why you think it's important for this to be part of  
> the standard distribution.

Because I would be interested in it becoming an interface which is  
more committed / documented than what I'd associate with an internal  
definition in a package.  I can obviously arrange life so that I don't  
have to type the "CCL::" easily, but I can't commit to the interface  
because I don't control it.  (And I care about this because I often  
end up with quite large sets of toplevel commands I've defined in  
implementations I use, because that's how I tend to work.)

So I guess what I was really asking is: is this an interface which  
could become committed (in whatever sense is suitable)?

(Having now written a couple of commands more than I had when I wrote  
the original message, one thing that would be nice in such an  
interface is to have a sort-of EVALQUOTE option.  For instance I have  
a command called :ed which calls ED (but remembers arguments) and it  
would be nice to be able to say ":ed foo" rather than ":ed 'foo".   
I'll try and work out how to do that.)

The above all sounds like I'm doing the standard awful "I didn't pay  
for this tool but I *insist* you add feature x *right now*" whining  
you hear so often - it's not meant to be at all.  I obviously don't  
particularly expect anyone to do anything: I just think it would be an  
interesting thing to make en exported interface.

--tim