High load on linux64-1.0.0.

[bakters]

Computer specs : Thinkpad X61, Intel Core 2 Duo 2 GHz, Intel GM965/GL960 graphics card.

Operating system : 64 bit Gentoo flavor (Funtoo, actually. Conservative and stable, nothing crazy).

Game options : Disabled everything possible, ASCI tileset, Windowed mode, 15 fps framerate.

What happens : When the game window has focus, the game uses 60% of my processor power while displaying a perfectly static screen. Nothing happens, nothing moves, yet my fan is running at full speed and I drain battery. When I switch focus to another screen, the load drops to 10%.

Expected behavior : Displaying a static screen should not require any processing power.

[darkgod]

Check that your video card has a correct driver that supports OpenGL.
No opengl means the CPU has to do the drawing which leads exactly this effect

[bakters]

Code: Select all

$ glxinfo | ack rendering
direct rendering: Yes

I can attach a full output, not sure how to do that here, though. It's not a software rendering mode

Code: Select all

 $ ack glx /var/log/Xorg.0.log
[  1672.427] (II) LoadModule: "glx"
[  1672.510] (II) Loading /usr/lib64/xorg/modules/extensions/libglx.so
[  1672.537] (II) Module glx: vendor="X.Org Foundation"

Anyways, what rendering is supposed to happen when nothing happens? Between turns it's a static display, until a bird flies or something like that.

I'm willing to work. I'll patch the game, run tests, do "crashdums" - whatever you ask.

Edit: I just installed this game on a desktop machine. Dual core Athlon 64 4200+, Nvidia GT 430. Everything off, ASCI tileset, 30 fps and processor load is about 25%, when I switch focus out of the game only 4% (GUI feels laggy too, but this may be a feature, I don't know).

Nvidia drivers are actually as good as you can get. Codeweavers tested them to consistently perform about 30% better than on Windows in GPU bound tests and about 10% faster in CPU bound tests, so no question about having crappy drivers in this case. Tome4 is still very busy doing nothing.