(This is more about publicity than development, but I couldn't find another forum to put it in.)
I've been spreading the word about LXDE (including through @appropedia and @chriswaterguy accounts, on Identi.ca and Twitter). Part of what I'm finding is that many Linux people aren't aware of what it's capable of - and of course there are so many claims about what is fast and what is light, that something more attention-grabbing is needed.
Can we arrange benchmark testing of the different desktops? This could attract a lot of interest among Linux users, and show what LXDE has to offer. Something like the fairly in-depth one described on kde.org (Jan 2007 or earlier, from looking at web.archive.org, but it doesn't include LXDE). Another example of benchmark testing (a very simple test of RAM usage, comparing XFCE and LXDE) is here (reply #9 and #10).
I'm not in a position to do this myself (I have one computer, which I need for work). But if anyone wants to do some benchmark testing, either simple or in-depth, please link or post the results here.
A really proper test would include at least Gnome, KDE, XFCE and LXDE. (Even more thorough would be to add Openbox, a Crunchbang-style implementation of Openbox, and JWM, which is the standard Puppy Linux desktop environment.) But even a simple test like the XFCE/LXDE comparison above is worth doing.
Benchmark testing - proving what LXDE can do.
-
- Posts: 26
- Joined: Wed Dec 10, 2008 6:09 am
-
- Posts: 381
- Joined: Mon Mar 23, 2009 5:14 am
Re: Benchmark testing - proving what LXDE can do.
A big part of lightness on memory usage should be that in an LXDE system the distro doesn't start things that don't need to be running.
A comparison of taking, shall we say, a Gnome desktop, tearing out the Gnome panel, and starting LXPanel instead, without doing any other slimming, will be flawed for that reason.
I am working toward having the minimal set of overhead processes be: init, syslogd, klogd, udevd, dhclient, acpid, cupsd, ntpd, dbus-daemon, pulseaudio, display manager, Xorg, window manager, panel. With some of those, like ntpd, dhclient or pulseaudio, possibly not always wanted.
A comparison of taking, shall we say, a Gnome desktop, tearing out the Gnome panel, and starting LXPanel instead, without doing any other slimming, will be flawed for that reason.
I am working toward having the minimal set of overhead processes be: init, syslogd, klogd, udevd, dhclient, acpid, cupsd, ntpd, dbus-daemon, pulseaudio, display manager, Xorg, window manager, panel. With some of those, like ntpd, dhclient or pulseaudio, possibly not always wanted.
-
- Posts: 26
- Joined: Wed Dec 10, 2008 6:09 am
Re: Benchmark testing - proving what LXDE can do.
Agreed. I'm not sure why you mention it - I am not suggesting such a test.Marty Jack wrote:A comparison of taking, shall we say, a Gnome desktop, tearing out the Gnome panel, and starting LXPanel instead, without doing any other slimming, will be flawed for that reason.
What about my suggestion of comparing Gnome, KDE, XFCE and LXDE? It would need to be done on a distro where each of those desktops has been implemented reasonably well. Debian? Fedora? Mandriva? Not sure which. Then we could even compare Windows on the same machine, for a bit more provocation .
Seems like a very good idea to me, but it obviously hasn't caught the imagination of anyone else here. Ah well.
Re: Benchmark testing - proving what LXDE can do.
A lot depends on the distro used, for example Lubuntu starts more services and uses more ram than Lxde on a Debian system. Things like prelink/preload can also have an effect on how much ram is used and how fast apps start.
89mb of ram in use not bad after being up for a day.
$ inxi -F
Less than 30 second boot using Slim, all apps start instantly including apps like OpenOffice, GIMP, Firefox.
Code: Select all
python ps_mem.py
Private + Shared = RAM used Program
80.0 KiB + 14.0 KiB = 94.0 KiB klogd
88.0 KiB + 22.5 KiB = 110.5 KiB gksu-run-helper
108.0 KiB + 18.5 KiB = 126.5 KiB irqbalance
120.0 KiB + 31.5 KiB = 151.5 KiB gnome-pty-helper
140.0 KiB + 15.0 KiB = 155.0 KiB sh
132.0 KiB + 23.5 KiB = 155.5 KiB atd
140.0 KiB + 27.5 KiB = 167.5 KiB init
148.0 KiB + 28.0 KiB = 176.0 KiB syslogd
144.0 KiB + 47.5 KiB = 191.5 KiB ck-launch-session
160.0 KiB + 64.5 KiB = 224.5 KiB hald-addon-input
164.0 KiB + 65.0 KiB = 229.0 KiB hald-addon-acpi
224.0 KiB + 24.5 KiB = 248.5 KiB dirmngr
192.0 KiB + 58.5 KiB = 250.5 KiB hald-runner
200.0 KiB + 54.5 KiB = 254.5 KiB gam_server
200.0 KiB + 69.0 KiB = 269.0 KiB lxsession
252.0 KiB + 36.5 KiB = 288.5 KiB dbus-launch
236.0 KiB + 85.5 KiB = 321.5 KiB cron
328.0 KiB + 54.0 KiB = 382.0 KiB bitlbee
300.0 KiB + 106.5 KiB = 406.5 KiB su
288.0 KiB + 145.0 KiB = 433.0 KiB hald-addon-storage (2)
384.0 KiB + 124.0 KiB = 508.0 KiB menu-cached
492.0 KiB + 23.5 KiB = 515.5 KiB dhclient
412.0 KiB + 120.0 KiB = 532.0 KiB getty (6)
656.0 KiB + 82.0 KiB = 738.0 KiB ntpd
536.0 KiB + 338.0 KiB = 874.0 KiB dbus-daemon (2)
328.0 KiB + 684.5 KiB = 1.0 MiB udevd (3)
996.0 KiB + 208.0 KiB = 1.2 MiB gnome-keyring-daemon
1.2 MiB + 186.0 KiB = 1.3 MiB console-kit-daemon
1.3 MiB + 124.0 KiB = 1.4 MiB hald
1.4 MiB + 141.5 KiB = 1.5 MiB cupsd
1.2 MiB + 423.5 KiB = 1.6 MiB gksu
1.4 MiB + 371.0 KiB = 1.7 MiB slim
2.7 MiB + 63.0 KiB = 2.7 MiB bash
1.7 MiB + 1.2 MiB = 2.9 MiB parcellite
2.6 MiB + 555.5 KiB = 3.2 MiB openbox
3.3 MiB + 40.0 KiB = 3.3 MiB preload
2.6 MiB + 1.3 MiB = 3.9 MiB x-terminal-emul
4.1 MiB + 1.7 MiB = 5.8 MiB lxpanel (deleted)
14.6 MiB + 4.5 MiB = 19.1 MiB pcmanfm (2)
29.1 MiB + 1.6 MiB = 30.7 MiB Xorg
---------------------------------
89.0 MiB
=================================
Private + Shared = RAM used Program
$ inxi -F
Code: Select all
System: Host craigevil Kernel 2.6.35-rc6.dmz.2-liquorix-686 i686 (32 bit) Distro Debian GNU/Linux squeeze/sid
CPU: Single core Intel Pentium 4 (-HT-) cache 1024 KB flags (lm nx sse sse2 sse3) bmips 5585.26
Clock Speeds: (1) 2792.632 MHz (2) 2792.632 MHz
Graphics: Card nVidia G96 [GeForce 9400 GT] X.Org 1.7.7 Res: 1280x1024@50.0hz
GLX Renderer GeForce 9400 GT/PCI/SSE2 GLX Version 3.3.0 NVIDIA 256.35 Direct Rendering Yes
Audio: Card Intel 82801FB/FBM/FR/FW/FRW (ICH6 Family) High Definition Audio Controller driver HDA Intel BusID: 00:1b.0
Sound: Advanced Linux Sound Architecture Version 1.0.23
Network: Card Intel 82562ET/EZ/GT/GZ - PRO/100 VE (LOM) Ethernet Controller driver e100 v: 3.5.24-k2-NAPI at port dcc0 BusID: 03:08.0
Disks: HDD Total Size: 160.0GB (21.1% used) 1: /dev/sda WDC_WD1600JS 160.0GB
Partition: ID:/ size: 103G used: 32G (33%) fs: ext3 ID:swap-1 size: 5.25GB used: 0.00GB (0%) fs: swap
Info: Processes 99 Uptime 20:42 Memory 252.5/2015.9MB Runlevel 5 Client Shell inxi 1.4.12
Debian Sid LXDE Kernel liquorix CPU Pentium IV 2.80GHz GeForce 9400 GT
Debian - "If you can't apt-get something, it isn't useful or doesn't exist"
Giant Debian sources.list | Debian upgrade script smxi | sysinfo script inxi
Debian - "If you can't apt-get something, it isn't useful or doesn't exist"
Giant Debian sources.list | Debian upgrade script smxi | sysinfo script inxi