Re: asdf documentation/*central-registry* bugreport

Hannu Koivisto <[email protected]> writes:

> The documentation does not seem to match the implementation in the
> case of *CENTRAL-REGISTRY*.  README says: "A ssytem directory
> designator is either a function designator for a function of one
> argument" but SYSTEM-DEFINITION-PATHNAME does (if (and (symbolp
> dir) (fboundp dir)) (funcall dir name) (eval dir)) which, as far as
> I can see, does not work for function objects (which are function
> designators).

Oops, yes.  Thanks for spotting this.

> FWIW, I think that the usage of functions in *CENTRAL-REGISTRY*
> is somewhat suboptimal.  The current method of returning a
> directory needlessly restricts the system files to the ".asd"
> extension.

Yeah, I'd tend to agree, and am considering deprecating it altogether
in favour of something that saw some element of thought in its design. 
Maybe along the lines of a list of functions that are called in order
and are responsible for loading or otherwise generating the system
definition, with the default contents of the list being the single
function that does "look through *central-registry* and load the first
systemname.asd we find"

> Also, I think it would be good to allow functions to
> say that "the latest version is now in the core".  I would probably

You mean "I have taken care of loading or otherwise generating the
system definition for you, please don't load anything at all"?  I can
see the need for the feature; I don't think *central-registry* is the
place for it (see my previous paragraph)

Suggestions welcome.


-dan

-- 

  http://ww.telent.net/cliki/ - Link farm for free CL-on-Unix resources 


-------------------------------------------------------
This sf.net email is sponsored by: See the NEW Palm 
Tungsten T handheld. Power & Color in a compact size!
http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0001en