Achieve Pidsley Level (Rape my RAM)

Forum rules
We don't support installations in VirtualBox, VMWare, qemu or others. We ignore posts about WINE, PlayOnLinux, Steam and Skype. We don't support btrfs, lvm, UEFI, side-by-side installations with GPT or dualboot with anything newer than Windows XP.
Google your problem first. Check the Wiki. Read the existing threads. It's okay to "hijack" an existing thread, yes! If your problem is not yet covered, open a new thread. To get the quickest possible help, mention the exact release codename in your post (uname -a is a good idea, too). Due to the lack of crystal balls, attach the output of lspci -nnk if you encounter hardware problems.
User avatar
DebianJoe
Frame Buffer
Posts: 1915
Joined: Mon Jul 01, 2013 5:41 am
Location: emacs.d

Re: Achieve Pidsley Level (Rape my RAM)

Unread post by DebianJoe » Sat Sep 28, 2013 5:37 am

In the quest to find out the best way to compare gtk engines, I ran across GtkPerf and some tests on speed that had already been done, albeit on totally different hardware. Although this is geared more towards speed than how large to footprint of each one is.

The other option I'm considering is to use `ps` to test how much memory is allocated to a process, to see if there are trends in running the same programs, but with alternating gtk engines. It's not going to be as simple to create a "Pidsley-score" for this particular query.
|>>BBQ Roaster, Alpha Branch<< | >> clinky << | >> X11 must die << |
Thanks BASIC

User avatar
DebianJoe
Frame Buffer
Posts: 1915
Joined: Mon Jul 01, 2013 5:41 am
Location: emacs.d

Re: Achieve Pidsley Level (Rape my RAM)

Unread post by DebianJoe » Sat Sep 28, 2013 6:30 am

Okay, so here's some rendering numbers for the sake of seeing how this does play out:

Mist
GtkPerf 0.40 - Starting testing: Sat Sep 28 13:21:48 2013

GtkEntry - time: 0.04
GtkComboBox - time: 0.59
GtkComboBoxEntry - time: 0.47
GtkSpinButton - time: 0.09
GtkProgressBar - time: 0.06
GtkToggleButton - time: 0.10
GtkCheckButton - time: 0.09
GtkRadioButton - time: 0.13
GtkTextView - Add text - time: 0.34
GtkTextView - Scroll - time: 0.11
GtkDrawingArea - Lines - time: 0.55
GtkDrawingArea - Circles - time: 1.12
GtkDrawingArea - Text - time: 1.02
GtkDrawingArea - Pixbufs - time: 0.14
---
Total time: 4.87

Murrine
GtkPerf 0.40 - Starting testing: Sat Sep 28 13:23:27 2013

GtkEntry - time: 0.04
GtkComboBox - time: 0.68
GtkComboBoxEntry - time: 0.57
GtkSpinButton - time: 0.09
GtkProgressBar - time: 0.07
GtkToggleButton - time: 0.11
GtkCheckButton - time: 0.09
GtkRadioButton - time: 0.13
GtkTextView - Add text - time: 0.34
GtkTextView - Scroll - time: 0.11
GtkDrawingArea - Lines - time: 0.55
GtkDrawingArea - Circles - time: 1.19
GtkDrawingArea - Text - time: 1.05
GtkDrawingArea - Pixbufs - time: 0.15
---
Total time: 5.18

CleanIce
GtkPerf 0.40 - Starting testing: Sat Sep 28 13:27:00 2013

GtkEntry - time: 0.06
GtkComboBox - time: 0.62
GtkComboBoxEntry - time: 0.70
GtkSpinButton - time: 0.07
GtkProgressBar - time: 0.03
GtkToggleButton - time: 0.07
GtkCheckButton - time: 0.07
GtkRadioButton - time: 0.07
GtkTextView - Add text - time: 0.34
GtkTextView - Scroll - time: 0.11
GtkDrawingArea - Lines - time: 0.70
GtkDrawingArea - Circles - time: 1.15
GtkDrawingArea - Text - time: 1.04
GtkDrawingArea - Pixbufs - time: 0.13
---
Total time: 5.16

Crux
GtkPerf 0.40 - Starting testing: Sat Sep 28 13:27:45 2013

GtkEntry - time: 0.04
GtkComboBox - time: 0.77
GtkComboBoxEntry - time: 0.63
GtkSpinButton - time: 0.11
GtkProgressBar - time: 0.10
GtkToggleButton - time: 0.17
GtkCheckButton - time: 0.11
GtkRadioButton - time: 0.17
GtkTextView - Add text - time: 0.34
GtkTextView - Scroll - time: 0.15
GtkDrawingArea - Lines - time: 0.56
GtkDrawingArea - Circles - time: 1.11
GtkDrawingArea - Text - time: 1.03
GtkDrawingArea - Pixbufs - time: 0.14

Total time: 5.44


......so, MB is right, the Mist engine does actually render faster than other engines that I have tested. I noticed that I didn't get the same results as the posted tester when it came to using ClearIce. Mist was significantly faster in benchmarks than the other options I tested.

*edit: now removing gtk-engines/gtk-themes/gtk-etc from this install. Next question?
|>>BBQ Roaster, Alpha Branch<< | >> clinky << | >> X11 must die << |
Thanks BASIC

machinebacon
Baconator
Posts: 10253
Joined: Thu Sep 16, 2010 11:03 am
Location: Pfälzerwald
Contact:

Re: Achieve Pidsley Level (Rape my RAM)

Unread post by machinebacon » Sat Sep 28, 2013 7:46 am

Thanks Joe, nice to have it in numbers, and surprising to see how well (relatively) the Murrine engine performs.

'ps' does anyway stand for Pidsley Score ;) Well, what could be next? The shells maybe?
sh/csh/fish/dash/ksh/zsh/bash - and I am pretty sure there'll be some surprises.
..gnutella..

User avatar
DebianJoe
Frame Buffer
Posts: 1915
Joined: Mon Jul 01, 2013 5:41 am
Location: emacs.d

Re: Achieve Pidsley Level (Rape my RAM)

Unread post by DebianJoe » Sat Sep 28, 2013 7:49 am

Getting on the shells as we speak, will post ASAP. ;)
|>>BBQ Roaster, Alpha Branch<< | >> clinky << | >> X11 must die << |
Thanks BASIC

User avatar
DebianJoe
Frame Buffer
Posts: 1915
Joined: Mon Jul 01, 2013 5:41 am
Location: emacs.d

Re: Achieve Pidsley Level (Rape my RAM)

Unread post by DebianJoe » Sat Sep 28, 2013 8:25 am

echo $SHELL | {unbloat()}

It's worth noting that these are NOT configured shell numbers. I didn't import my personal .zshrc or .bashrc, which I'm sure would make at least a bit of difference. This is a simple "chsh <whatever_shell>" cold reboot test, with an instant "free" on boot. I try to remove as much user input as possible.

Shell Pidsley-Scores

Code: Select all

zsh  = 0
bash = 424
fish = 1960
ksh  = 2020
csh  = 2936
dash = 3068
sh   = 3300
edit: whoa, whoa....did fish beat bash? Ummmmm....that doesn't make sense at all.
|>>BBQ Roaster, Alpha Branch<< | >> clinky << | >> X11 must die << |
Thanks BASIC

User avatar
GekkoP
Emacs Sancho Panza
Posts: 5878
Joined: Tue Sep 03, 2013 7:05 am

Re: Achieve Pidsley Level (Rape my RAM)

Unread post by GekkoP » Sat Sep 28, 2013 8:29 am

sh is impressive, using it as login shell I can drop up to 6-7MB of RAM.

Post Reply