Ok, so I thought that the original problem might be that the original
make from the cedet root directory failed. I have attached the output
of that to this email in a file called make_output.txt.
That is the output of simply calling make in the cedet directory.
Along another line I tried to byte compile the directory. That also
produced many errors and is attached in a file called
I produced that by running C-u 0 byte-recompile-directory on the cedet
root dir. This was done after the previous make had already failed.
I did this with a completely minimal .emacs to remove the possibility
that it was some of my own config that was screwing it up.
My .emacs file now contains the following:
(add-to-list 'load-path "~/.emacs.d/packages/jde-2.3.3/lisp")
(add-to-list 'load-path "~/.emacs.d/packages/elib-1.0")
(setq semantic-load-turn-useful-things-on t)
;; Load CEDET
;; end .emacs
I also tried commenting out (require 'jde) until after startup, calling
M-x toggle-debug-on-error and then using C-x C-e after the jde line to
run it. That output is attached in debug_on_error.txt.
Just for kicks I ran
grep -rI semantic-format-prototype-tag at the root of all my emacs
packages. That output is in grep_prototype_tag.txt.
If there is anything else I can do to help. Please let me know. I am
completely dumbfounded here.
Thanks a lot for all the responses.
On Thu, 2004-03-18 at 15:06, Eric M. Ludlam wrote:
> >>> John Russell <jorussel@...> seems to think that:
> >I just upgraded all of my old cedet packages (eieio, semantic, speedbar,
> >etc) to the single cedet package, cedet-1.0betac. However, now whenever
> >I try to open a java file with jde-mode 2.3.3 I get the following error.
> >File mode specification error: (void-variable
> It appears that the java parser defines an overload for
> semantic-format-prototype-tag, which does not exist. wisent-java
> needs to define semantic-format-tag-prototype.
> David may have had some other devious plan, but I'm not sure what it
> is. I do not know who is calling it. The existing implementation
> should be innocuous.
> To find out why, try:
> M-x toggle-debug-on-error RET
> then cause the error again.
> >and the buffer stays in text-mode.
> >Also, the function comment-dwim returns this error. Sorry for the
> >semantic-edits-change-function-handle-changes: Invalid function:(macro .
> >#[nil"\300\207" [(setq semantic-parse-tree-state (quote needs-update))]
> >("/home/jorussel/.emacs.d/packages/cedet-1.0beta1c/semantic/semantic.elc" . 7842)])
> This almost seems like a compilation issue. The macro above is byte
> compiled, but being executed normally. Not sure why that is a
> problem. Same stack trace thing would help.
> >My guess is that these are all related, but I'm not sure what else I am
> >supposed to do. Is there anything else I need to do to set this up
> >properly besides the
> >(setq semantic-load-turn-useful-things-on t)
> >;; Load CEDET
> >(load-file "~/cedet-VERSION/common/cedet.el")
> That is the right thing to do.
> >from the INSTALL file. BTW the load file is pointing to a place in my
> >load-path and cedet.el loads properly so that's not the problem.
> >Thanks for any help.
> >PS I really like the bundled packaging. It makes upgrading much
> >easier. Thanks.
> [ ... ]
> Yay! Now if we could just get it to work... ;)