[plt-dev] #lang: the stake in Dracula's heart?

From: Matthew Flatt (mflatt at cs.utah.edu)
Date: Mon Feb 1 18:10:55 EST 2010

I'm not sure that extension registration is different than collection  
installation. For example, the meaning of '#lang foo' depends on the  
'foo' that you have installed.

But ensuring that it's the same would require picking some extesions  
to mean "PLT module". We couldn't keep the current behavior of  
allowing any extension.

On Feb 1, 2010, at 3:26 PM, Eli Barzilay <eli at barzilay.org> wrote:

> On Feb  1, Carl Eastlund wrote:
>> On Mon, Feb 1, 2010 at 5:02 PM, Eli Barzilay <eli at barzilay.org>  
>> wrote:
>>> DrScheme can be made to pretend (relatively easily, I think) that
>>> some "#lang"-less buffer really has the right "#lang indirect
>>> STDIN r6rs" contents, or something like that.  In any case, I
>>> don't consider it important to think about drscheme-specific
>>> problems -- I believe that finding a good solution at the mzscheme
>>> level is bound to lead to a drscheme solution.
>> If mzscheme recognized something like DrScheme tools -- more
>> lightweight and GUI-free, but still a form of library registry --
>> registering file extensions should be easy.
> One of the main things that I dislike with extension registration is
> that it means that running mzscheme on a file makes it do much more
> work, in this case -- digging through all info files (reading
> "collects/info-domain/compiled/cache.ss") for the registration.  This
> is in addition to the (bad) result of making the semantics of files
> depend on my specific environment.  (For example, this is one of the
> bad aspects of having teachpacks specified outside of the file.)
> -- 
>          ((lambda (x) (x x)) (lambda (x) (x x)))          Eli  
> Barzilay:
>                    http://barzilay.org/                   Maze is  
> Life!

Posted on the dev mailing list.