After about a 1 month pause from work on URandR I decided to improve the code a bit (reusing some code from Envy) and started implementing Xinerama-style dual screen. There is still some work to do and I think I might have to change a great deal of the provisional GUI I’m using in order to improve its usability.
Currently Xinerama-style dual screen requires editing your xorg.conf so as to set the Virtual resolution (i.e. the highest resolution you can use) (something you’ll only have to do once). The reason for performing such operation (in Keith Packard’s words) is the following:
The current [Intel] driver cannot reallocate the frame buffer, so whatever size you start with is the maximum the screen can ever become, and that this amount of physical memory is tied down for the whole X server run.
This is only a temporary “problem” (which the devs are going to fix). Of course URandR can calculate the Virtual resolution you need and, after checking whether you need it or not, add it to your xorg.conf (this is something I have already implemented).
Expect a few new screenshots in the next future.
Alberto,
I’ve been watching the Ubuntu randr stuff improve, and I’m anxiously awaiting the day when it will work on my hardware (Intel 965Q), so that I can rotate my nice wide display, which I wish was a nice tall display.
Rotating in Tribe 4 pretty much crashed the display, but Tribe 5 is almost useful. There’s bitblt splatter all over the place, but the basic rotation and resize of the dimensions seems to work (if compiz is disabled). It sort of functions, at least for a few minutes.
Anyway, keep up the good work!
@Mike
if you’re using the “intel” driver (instead of the old “i810” driver) with randr then you might have found a bug.
I don’t work on drivers therefore I can’t help you, however I’m sure that, if you report the problem on launchpad.net, Ubuntu’s devs will work to fix the bug.
I cannot find the classes module on launchpad. Did you forget to commit it?
./randrgtk.py
Traceback (most recent call last):
File “./randrgtk.py”, line 25, in
from classes import randrback
ImportError: No module named classes
@Manuel
you’re right. It lacks a lot of files. I’ll fix it soon.
Ok, all the code is available now.
Thanks for reporting
Thanks, that was fast. There are still two bugs I found:
Output devices can have a number appended like VGA-0 or similar. This should be reflected in the regular expression, because you have to place the correct name in the “xrandr –output” command.
Next is the detection of the refresh frequency. On my radeon 9200 is shows like “1280×1024 75.0 + 76.0 75.0 71.9 59.9*” that means a refresh rate of 59.9 is used.
@Manuel
1) the devices can be:
VGA, LVDS, TV, TMDS-1 (2,3, etc.)
2) can you post the output of this command, please?
xrandr -v
xrandr -v
Server reports RandR version 1.2
xrandr
Screen 0: minimum 320 x 200, current 1280 x 800, maximum 2048 x 2048
VGA-0 connected (normal left inverted right)
1280×1024 75.0 + 76.0 75.0 71.9 59.9
1280×960 59.9
1152×864 75.0 74.8
1024×768 75.1 70.1 60.0
832×624 74.6
800×600 72.2 75.0 60.3
640×480 75.0 72.8 66.7 60.0
720×400 70.1
LVDS connected 1280×800+0+0 (normal left inverted right) 0mm x 0mm
1280×800 59.9*+ 60.0
1280×768 60.0
1024×768 60.0
800×600 60.3
640×480 59.9
S-video disconnected (normal left inverted right)
so my external device is called VGA-0 but in your code the name is hardcoded to VGA.
VGA-0 and S-video???
Is it an idiosyncrasy of the ATI driver? I will ask in the Xorg mailing list
I would like to know:
1) which version of Ubuntu/Debian you’re using
2) the output of this command:
sudo aptitude show xserver-xorg-video-ati
3) how did you install the ati driver?
xserver-xorg-video-ati 1:6.7.192-1
xserver-xorg-core 2:1.3.0.0.dfsg-12ubuntu2
Ubuntu gutsy and the xserver-xorg-video-ati is from https://www.die-welt.net/index.php/blog/187/ATI_radeon_opensource_driver_with_XRandr_1.2_support_for_Debian_unstable
Currently there is no randr 1.2 support for radeon in ubuntu.
Its a Acer Aspire 2000 with Radeon 9200 and external connectors for vga and s-video/composite.
I have plugged in my Radeon 9200, installed the driver from GIT on Debian sid and had the same output.
Some devs in the mailing list explained me that the name schemes depend on the driver.
This means that I will have to do additional work to support ATI cards :-/
Will-you add the support of the ATI Proprietary Linux x86 Display Driver 8.41.7?
http://ati.amd.com/support/drivers/linux/linux-radeonhd.html
It’s write that this driver ad the complete support of OpenGL Radeon HD 2000 series and they will add the support of all card who go out beteween Radeon 9500 and this serie on the next ATI Proprietary Linux x86 Display Driver 8.42.X.
@pierluc
I think I will add the support for driver 8.42.x when it’s released. I would rather not add driver 8.41.7 since AMD itself clearly warned its users against the following problems:
Caution: AMD recommends that this release of the AMD Proprietary Linux software driver not be used for distribution packages. Distributors should continue to use the AMD Proprietary Linux driver version 8.40.4
Caution: The ATI Catalyst™ Linux software suite 8.41.7 does not support ATI Workstation products. AMD recommends using the AMD Proprietary Linux software driver version 8.40.4.
Caution: The AMD Proprietary Linux driver version 8.41.7 is not recommended for the ATI Mobility and Integrated Products. Using this driver on any of the following products may result in stability issues. AMD recommends using the AMD Proprietary Linux software driver version 8.40.4.
Caution: The AMD Proprietary Linux driver version 8.41.7 is not recommended for use on ATI Radeon X1900 series of products and lower. Using this driver on any of the following products may result in stability issues. AMD recommends using the AMD Proprietary Linux driver version 8.40.4
Can you create a envy-beta, a software to installe beta driver. When a new driver go out and it is very better we can use envy-beta to install them. But, if we would not have problem we use envy and we install the last stable driver.
I’ve tried doing Xinerama for my dad’s laptop with the Intel driver with somewhat degree of success — some screen corruption here and there, now and then. Maybe I didn’t set it up correctly… I also tried Cloned output but I didn’t manage to get that. Might you happen to know how to set up Cloned output successfully on the Intel driver, Alberto?