Latest pretest builds fine on windows but...

Latest pretest builds fine on windows but...

[Eric Lilja]

...I did notice a couple of these:
Compiling url/url-handlers.el

In toplevel form:
url/url-handlers.el:277:1:Error: Cannot open doc string file 
"c:/cvsemacs/emacs/etc/DOC"

Compiling url/vc-dav.el
Loading subst-ksc...
Loading subst-gb2312...
Loading subst-big5...
Loading subst-jis...

In toplevel form:
url/vc-dav.el:30:1:Error: Cannot open doc string file 
"c:/cvsemacs/emacs/etc/DOC"

Anything to worry about?

Also, all those warnings about functions that are not known to be 
defined also worry me sometimes, but I know little of lisp. And those 
have always been around and I've been building cvs versions for three 
years now I think.

Building native windows version on XP using latest MinGW make and the 
cygwin tools. Haven't tried the resulting executable yet.

- Eric

Re: Latest pretest builds fine on windows but...

[Richard Stallman]

It looks like the error you got comes from the following kludge:

    (defmacro url-handlers-create-wrapper (method args)
      `(defun ,(intern (format "url-%s" method)) ,args
	 ,(format "URL file-name-handler wrapper for `%s' call.\n---\n%s" method
		  (or (documentation method t) "No original documentation."))
	 (setq url (url-generic-parse-url url))
	 (when (url-type url)
	   (funcall (url-scheme-get-property (url-type url) (quote ,method))
		    ,@(remove '&rest (remove '&optional args))))))

For the long term, we should get rid of that, or else do something
that ensures it will work.

How did it happen that your build recompiled url-handlers.el?
If you were building from CVS, that would be understandable,
but a pretest, like a release, should contain up-to-date .elc files
and building it should not recompile them.

It also puzzles me that you were compiling this with an Emacs
that could not get its DOC file.