Re: [j-devel] Re: require loads source, not compiled file

[Sam S. <sd...@gn...>]

> * Peter Graves <cr...@ne...t> [2004-08-03 10:38:20 -0700]:
>
> On Tue, 03 Aug 2004 at 12:22:55 -0400, Sam Steingold wrote:
>> > The compiled file should be loaded if socket.abcl exists in the
>> > directory pointed to by *LISP-HOME*. Is that the case in your
>> > situation?
>>
>> no it's not.
>> note that the loading is done from the JAR file, not disk.
>
> Well, if the .abcl file were in the jar, it would get loaded (I think).
> The relevant code is in Load.java, starting around line 148.

good.

> Why is socket.abcl not in the directory pointed to by *LISP-HOME* in
> your case?

because the top-level make did not compile it, I guess.
now, _why_ didn't top-level make did not compile it?

>> shouldn't make compile all those lisp files and put only the compiled
>> versions into the JAR?
>
> As things stand, make doesn't even do COMPILE-SYSTEM.

it should...

> It's easier for me to work on the compiler if the .abcl and .cls files
> are out in the open where I can get at them for debugging purposes, so
> I haven't been in any hurry to teach COMPILE-FILE how to jar up its
> output.

well, the j.jar rule saves *.class _and_ *.lisp
which is obviously wrong (these should never be together: either
*.java+*.lisp or *.class+*.abcl).

-- 
Sam Steingold (http://www.podval.org/~sds) running w2k
<http://www.camera.org> <http://www.iris.org.il> <http://www.memri.org/>
<http://www.mideasttruth.com/> <http://www.honestreporting.com>
Between grand theft and a legal fee, there only stands a law degree.