Re: [CEDET-devel] problem building cedet in cygwin on windows
Brought to you by:
zappo
From: Nate S. <nat...@gm...> - 2012-07-13 15:20:33
|
Folks, Ignore the "shell feature not provided" problem. When I move the cedet CVS package to someplace OUTSIDE of the emacs installation, things build OK to completion. With the cedet package in my site-lisp directory (before I built it, anyway,) what seems to happen is, when "shell" is required, emacs picks up "ede/shell" instead. In fact, if I start emacs (without a .emacs file) and do describe-function on "shell", the *Help* buffer starts with > shell is an interactive compiled Lisp function in `shell.el'. but the hyperlink for shell.el opens up ede/shell.el. The INSTALL file does not say CEDET should be installed where it would get picked up in load-path, but it doesn't say it should NOT go to such a place either. Lesson is: CVS CEDET needs to stay away from load-path. Eric - 1 correction to the INSTALL file: under 3) configuration, it says to add (load-file "~/cedet-VERSION/common/cedet.el") but -- as you know -- cedet.el is under lisp/cedet nowadays. Thanks! 2012/7/12 Nate Schley <nat...@gm...> > Austin: > > That script needs to be archived! It solved the problem! > > however, > > I completed the autoloads process, but upon make -C cedet, I have a new > problem that baffles me. > > In toplevel form: > cedet-android.el:35:13:Error: Required feature `shell' was not provided > make[2Makefile:38: recipe for target `cedet-android.elc' failed > ]: *** [cedet-android.elc] Error 1 > make[2]: Leaving directory > `/cygdrive/c/progra~1/gnu/emacs-24.1/site-lisp/cedet/lisp/cedet' > make[1Makefile:29: recipe for target `cedet' failed > ]: *** [cedet] Error 2 > make[1]: Leaving directory > `/cygdrive/c/progra~1/gnu/emacs-24.1/site-lisp/cedet/lisp' > Makefile:43: recipe for target `compile' failed > make: *** [compile] Error 2 > > shell.el is a standard package right in the standard lisp directory that > emacs is supposed to inherently have in its load-path. > Again, I'd appreciate any thoughts from anyone. > > -- > Thanks, > > Nate > -- Thanks, Nate |