bloated vs. lightweight WM's

[purslow at sympatico.ca]

[ there's been an interesting debate on Gentoo-user today
re how much memory etc various WM/DE's actually use.  below is a sample ]

On Mon, 2004-05-24 at 13:16, davidgn at servidor.unam.mx wrote:
> Quoting "Lloyd H. Meinholz" <lloyd.meinholz at bmpcoe.org>:
>> console (no X) - 179 MB, 127 procs
>> gnome 2.6      - 522 MB, 232 procs
>> kde 3.2.2      - 513 MB, 215 procs
>> xfce4 4.0.5    - 515 MB, 209 procs
>> openbox 3.2    - 438 MB, 197 procs
>> iceWM 1.2.13   - 441 MB, 202 procs
> I would have measured X alone too, maybe with a single "xterm" running

Similar to your suggestion, I tried it with just X running, and with X +
evolution, firefox and idea. All tests are with server apps running
(apache, tomcat, openldap, mysql, postfix). Here's my new results:

Without evolution, firefox and idea running:

console (no X) - 179 MB, 127 procs
X only         - 198 MB, 129 procs

With evolution, firefox and idea running:

DE/WM                TOTAL MEM/PROC USAGE     DE/WM MEM/PROC USAGE
==================   ====================     ====================
X                  - 335 MB, 158 procs
gnome 2.6          - 522 MB, 232 procs        187 MB, 74 procs
kde 3.2.2          - 513 MB, 215 procs        178 MB, 57 procs
xfce4 4.0.5        - 515 MB, 209 procs        180 MB, 51 procs
openbox 3.2        - 438 MB, 197 procs        103 MB, 39 procs
fluxbox 0.9.9      - 454 MB, 202 procs        119 MB, 44 procs
blackbox 0.65.0-r3 - 451 MB, 202 procs        116 MB, 44 procs
iceWM 1.2.13       - 441 MB, 202 procs        106 MB, 44 procs

Again, the base for my tests:

* My workstation is a P4 2.3 GHz, with 1GB memory.
* All tests were performed with the following servers running: apache,
tomcat, openldap, mysql, postfix.
* 3 big X applications running: evolution, firefox, idea.

The reason I did the test the way I did was to use a real world scenario
I was familiar with (my own). I have been hearing for a long time about
using a WM performing better than big DE's and that just wasn't my
experience (at least in the environment I am working in) and I wanted to
find out what the difference actually was.

-- 
========================,,============================================
SUPPORT     ___________//___,  Philip Webb : purslow at chass.utoronto.ca
ELECTRIC   /] [] [] [] [] []|  Centre for Urban & Community Studies
TRANSIT    `-O----------O---'  University of Toronto