help-gnu-emacs
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: "Anything" package requires w3m command...


From: Elena
Subject: Re: "Anything" package requires w3m command...
Date: Mon, 20 Dec 2010 08:24:44 -0800 (PST)
User-agent: G2/1.0

On Dec 20, 3:17 pm, Sébastien Vauban <wxhgmqzgw...@spammotel.com>
wrote:
> Hi Elena,
>
>
>
> Elena wrote:
> > On Dec 20, 12:40 pm, Sébastien Vauban wrote:
>
> >> --8<---------------cut here---------------start------------->8---
> >> Debugger entered--Lisp error: (error "Install w3m command in `exec-path' 
> >> or set `w3m-command' variable correctly")
> >>   signal(error ("Install w3m command in `exec-path' or set `w3m-command' 
> >> variable correctly"))
> >>   error("Install w3m command in `exec-path' or set `w3m-command' variable 
> >> correctly")
> >>   (progn (error "Install w3m command in `exec-path' or set `w3m-command' 
> >> variable correctly"))
> >>   (if (not (stringp w3m-command)) (progn (error "Install w3m command in 
> >> `exec-path' or set `w3m-command' variable correctly")))
> >>   (when (not (stringp w3m-command)) (error "Install w3m command in 
> >> `exec-path' or set `w3m-command' variable correctly"))
> >>   eval-buffer(#<buffer  *load*<3>> nil "c:/home/sva/src/emacs-w3m/w3m.el" 
> >> nil t)  ; Reading at buffer position 11512
> >>   load-with-code-conversion("c:/home/sva/src/emacs-w3m/w3m.el" 
> >> "c:/home/sva/src/emacs-w3m/w3m.el" nil t)
> >>   orig-require(w3m nil nil)
> >>   (let ((my/require-depth ...)) (orig-require feature filename noerror))
> >>   (cond ((member feature features) (message "%sRequiring `%s' (already 
> >> loaded)" ... feature)) (t (message "%sRequiring `%s'" ... feature) (let 
> >> ... ...) (message "%sRequiring `%s'...done" ... feature)))
> >>   require(w3m)
> >>   (...)
> >>   require(anything-config)
> >> --8<---------------cut here---------------end--------------->8---
>
> > Yes, that's the same problem. I think a fix would be to add "emacs- w3m"
> > directory only when "w3m-command" is defined and available, like this
> > (snippet not tested, I'm busy right now):
>
> > ;; Add emacs-w3m path to load-path only when w3m-command is available.
> > (when w3m-command
> >    (if (executable-find w3m-command)
> >            (add-to-list 'load-path "/path/to/emacs-w3m")
> >            (warn "Executable %s not found, emacs-w3m package not added to 
> > load-
> > path." w3m-command))
>
> In my case, it is a bit different as the load-path is automatically computed
> (recursive scan of all folders under the root one)...

Just like mine is.  However, I've put "troublesome" packages into a
directory at the same level as my auto-scanned root directory (that
is: not a child of the latter).  I've done that with other folders
which are included on a case basis: OS specific packages, "backports"
from newer versions of Emacs, etc.  Moreover, I've tweaked the
procedure which adds directories to leave out those which end with
".BAK" or ".OLD".  That way it's easy to just rename a directory and
have it left out.  I'll post my tweak when I have time.

>
>
>
> >> Now, following Thierry's advice, I've git pull'ed `anything' just 10 min 
> >> ago,
> >> and tried to relaunch my Emacs.
>
> >> I now have a different error (maybe hiding the above one, from a previous
> >> run):
>
> >> --8<---------------cut here---------------start------------->8---
> >> Warning (initialization): An error occurred while loading 
> >> `c:/home/sva/.emacs':
>
> >> error: No buffer named *anything*
>
> >> To ensure normal operation, you should investigate and remove the
> >> cause of the error in your initialization file.  Start Emacs with
> >> the `--debug-init' option to view a complete error backtrace.
> >> --8<---------------cut here---------------end--------------->8---
>
> > I've read something about this issue, but I don't remember what the fix was.
> > If I'm not mistaken, creating "*anything*" buffer in advance fixed the
> > issue. Sorry for not being able to provide definitive answers right now.
>
> Maybe it'd be good that to be checked directly in the `anything' package?  The
> assumption that such a buffer already exists is maybe too strong?  Or I do
> something wrong so that the assumption gets broken?

I don't remember/know whether it is an issue of anything.el.  Maybe
it's a contributed package?  Yes, contacting the maintainer is the
right solution.


reply via email to

[Prev in Thread] Current Thread [Next in Thread]