[Openmcl-devel] require and alias files
gb at clozure.com
Sat Aug 31 00:40:47 UTC 2002
On Fri, 30 Aug 2002, John DeSoi wrote:
> In my init file I have:
> (pushnew #4p"lisplib:_modules;" *module-search-path*)
> The _modules folder has aliases to all of the module files. In
> calling require, OpenMCL just returns the module string without
> following the alias and loading the file. Bug?
I think that the behavior of "just returning the module string"
is what happens when you successfully load a 0-length file; in the
Darwin layer, a MacOS alias isn't distinguishable from a 0-length
file. (The information that opaquely identifies the target of
an alias is stored in the alias file's resource fork.)
I made an alias of "ccl:bindarwin;eval.dfsl" in the Finder, then
in the shell, I did:
[~/ccl/bindarwin] gb at dervish> ls -l eval.dfsl*
-rw-r--r-- 1 gb wheel 75622 Aug 18 20:03 eval.dfsl
-rw-r--r-- 1 gb wheel 0 Aug 30 17:44 eval.dfsl alias
Obviously, the bit of information that distinguishes a MacOS alias
>from a regular file is stored -somewhere-, but I don't believe that
it's stored anywhere where a Darwin program (like ls or openmcl or
cat or ...) can easily get at it.
To answer your question: I don't think that it's a good thing that
someone trying to do what you're doing should have it fail silently
and mysteriously, but I don't think that it's a bug that a Darwin
program doesn't recognize file attribute information that isn't
available to it.
> John DeSoi, Ph.D.
> Openmcl-devel mailing list
> Openmcl-devel at clozure.com
Openmcl-devel mailing list
Openmcl-devel at clozure.com
More information about the Openmcl-devel