From: "mitchell_laks" <mlaks2000@yahoo.com>
Subject: Re: faster emacs startup, can I postpone package loading?
Date: 2 Jan 2006 18:38:25 -0800 [thread overview]
Message-ID: <1136255905.045258.105640@g14g2000cwa.googlegroups.com> (raw)
In-Reply-To: mailman.21176.1136250882.20277.help-gnu-emacs@gnu.org
> >
> > First, you'd have track down, where all the loading is happening.
> Perhaps this happens in a file like /usr/share/emacs/site-start/site-init.el
Thank you! Bingo!
The directories /usr/share/emacs and /usr/share/emacs21 are interesting
ls /usr/share/emacs/site-lisp/
a2ps cogre ftnchek.el preview-latex
aplus-fsf-el csh-mode.el gri-mode.el semantic
aribas debian-startup.el kdesdk-scripts speedbar
auctex dictionaries-common maxima tramp
autoconf ecb mgp yacas
cedet-common ede mpqc
cedet-contrib eieio noweb-mode.el
chktex.el ess octave2.1-emac
There is a file debian-startup.el and the directories correspond to the
things that emacs loaded during startup.
While in
/usr/share/emacs21/site-lisp/
ls /usr/share/emacs21/site-lisp/
a2ps cedet-common debian-startup.elc ede gri-mode.elc
octave2.1-emacsen speedbar tramp
aribas cedet-contrib dictionaries-common eieio maxima
preview-latex subdirs.el yacas
auctex cogre ecb ess mpqc
semantic tex-site.el
There, in the ../emacs21/.. directory
the file is a compiled file .elc.
I had noticed that when I installed on debian a program that has a
emacs component, alot of compiling goes on.
Also emacs seems to be known as emacs21 on debian, so maybe that is the
right location.
But I can't read debian-startup.elc but I can read the
debian-startup.el in ../emacs/ .. :).
Seems to refer to the load sequence ... :( very lispy :(
;; debian-startup.el --- Debian specific emacsen startup code.
;; Copyright (C) 1998 Rob Browning
;; Maintainer: Rob Browning <rlb@defaultvalue.org>
;;; Commentary:
;; This file contains startup code needed by all the various flavors
;; of Emacs for a Debian system.
;; next follow incomprehensible code then
redone code because of comments that
"elisp doesn't appear to do tail recursion"
which seems unlikely from what I know of modern lisps...
; Rewritten to less elegant -- non recursive version because elisp
;; doesn't seem to handle tail recursion :<
(defun debian-unique-strings (strings)
"Takes a list of strings and returns the list with *adjacent*
duplicates removed."
(let ((result '()))
(while (consp strings)
(if (not (string= (car strings) (car (cdr strings))))
(setq result (cons (car strings) result)))
(setq strings (cdr strings)))
(nreverse result)))
(defun debian-run-directories (&rest dirs)
"Load each file of the form XXfilename.el or XXfilename.elc in any
of the dirs, where XX must be a number. The files will be run in
alphabetical order. If a file appears in more than one of the dirs,
then the earlier dir takes precedence, and a .elc file always
supercedes a .el file of the same name."
etc etc.
> Next, you can try Jari Aalto's tiny-load: http://tiny-tools.sourceforge.net/
> It loads packages during idle time. I use it in
> http://emacro.sourceforge.net/
Q: Am I going to have to rewrite this thing
?? :((( ?? using tiny-tools? Over my head...
Has anyone else smarter than me already done this?
I clearly am way over my lisp head!
I thought maybe I was doing something really stupid in configuration,
such as
when I found that emacs was doing reverse dns lookup till it timed out
at 10 seconds which I fixed by adjusting /etc/hosts in my machine with
no dns.
Maybe there just a package or two that is screwed up that is delaying
my loading?
After all it is only loading 47 or so things. Emacs works fine on
another machine with only 16 things to load and 4seconds is a lot
longer >> 3* almost nothing.. :).
Here in this excerpt of the full *Messages* buffer given above,
we see some funny things
[2 times]
Error loading 55ecb etc.
vis:
Loading semantic-idle (source)...done [2 times]
Loading ede-speedbar (source)...done
Loading 52semantic (source)...done
Loading 53cedet-contrib (source)...done
Loading 53cogre (source)...done
Loading 55ecb (source)...
"/usr/share/emacs21/site-lisp/cedet-common/" added to `load-path'
Error while loading 55ecb
Loading semantic-el (source)...done
For information about the GNU Project and its goals, type C-h C-p.
call-interactively: Beginning of buffer [14 times]
call-interactively: End of buffer
Dear Bruce your sourceforge site which I am looking at now is amazing.
I had no idea what a high level of emacs wizardry exists out in the
wild!
Thanks
Mitchell
next prev parent reply other threads:[~2006-01-03 2:38 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <mailman.21176.1136250882.20277.help-gnu-emacs@gnu.org>
2006-01-03 1:12 ` faster emacs startup, can I postpone package loading? BRUCE INGALLS
2006-01-03 18:03 ` Kevin Rodgers
2006-01-03 2:38 ` mitchell_laks [this message]
2006-01-03 3:34 ` Ian Zimmerman
2006-01-04 0:10 BRUCE INGALLS
-- strict thread matches above, loose matches on Subject: below --
2006-01-03 11:05 David Reitter
2006-01-02 20:27 mitchell_laks
2006-01-02 21:40 ` Thorsten Bonow
2006-01-06 13:56 ` kgold
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
List information: https://www.gnu.org/software/emacs/
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1136255905.045258.105640@g14g2000cwa.googlegroups.com \
--to=mlaks2000@yahoo.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).