I’ll cut right to the chase. I have a 2.5 year old Toshiba M400 laptop which is also a tablet PC, and the rotate options don’t work ‘properly’, and they never have. Many others have similar laptops, and theirs “don’t work” either. I’d like this to be fixed, but don’t know how to achieve that.
I didn’t buy the laptop because it was a tablet PC. I bought it because at the time it had the highest resolution (1400×1050) at a low physical size (12.1″) with built in 3G. If anyone ever asks “do you use it as a tablet PC” the answer is “no”. If the tablet features worked I’d probably use it as one more often.
The issues with this laptop:-
The screen doesn’t autorotate when you swivel it round.
This means you have to use the graphical screen rotation tool either before you swivel the screen or after. Either way you either end up with the desktop screen upside down, or having to man-handle the laptop so it’s up the right way, or use the UI upside down for a moment. This is naff. It should automagically do this as it does on Windows.
There is a graphical tool for rotating the screen, but no graphical tool for rotating the tablet.
The tablet is an input device which has no graphical tool for rotating. Ideally the screen rotation tool should ‘know’ that the PC is a tablet and should rotate the input device. Alternatively a separate graphical tool should exist which has options for rotating the tablet input device.
The screen and tablet don’t rotate together
If I do rotate the screen using the graphical tool I end up with the tablet and screen out of sync with eachother. Using the tablet pen the directions are inverted (left is right, up is down etc) whilst the screen is rotated. There is a command line tool in the wacom-tools package which can rotate the tablet, but no GUI, and nothing that keeps the screen and input device in sync.
I attended a sesion called Better touchscreen support at the recent Ubuntu Developer Summit but it was sadly only concerning touch screens, not tablets. I did raise these issues in the session but nobody really has any time or inclination to fix them, despite there being at least two devices in the room of this type.
Over the last 2 years I have spoken to various developers who work in the areas covered by this and nobody really seems interested in fixing this. I get the usual “feel free to submit a patch”. I’m not a coder 
There are various scripts and fudges online that can achieve some of this, but none of it is integrated into the existing tools and none of it is shipped with Ubuntu. This makes me sad. Anyone fancy helping?
14 Comments
Well, I have one too. I got it because it’s a tablet pc. Linux pretty much sucks with tablet pc’s, windows 7 on the other hand works flawlessly and have many enhancements, but still, I’ll get rid of windows soon anyway.
There’s also the issues of a graphical tool *remembering the damn settings* for configuring stylus properties and buttons etc
I’d help with testing at least, i’ve got an old M200 which I rotate with a shell script linked from the rotate button and another which sets-up the stylus buttons on login. If anything is developed it would be good if it was simple to add in an acceleometer backend in case ubuntu is used on any MID type devices which have them.
@popey — if you have a command line tool for rotating the tablet, presumable xrandr can rotate the screen part for you. Those should be connectable to a key at least so you get one key rotate of them both. There is also likley a key event on rotate, xev would be your friend for finding that out. If thats true you should be able to use the gnome hotkeys thing to trigger the script on button press and bingo. So:
1) file a LP bug with all the info you have in it for a) commands to roate pad, b) commands to rotate screen,
2) xev output from turning the screen round
You never know where you might get
.
Yeah, it’s a combination of two commands, xrandr and xsetwacom. It’s all pretty well documented on this wiki page:-
https://help.ubuntu.com/community/X61T#Setup the Tablet Rotate Button
But I’ll certainly file a bug.
Isn’t this just another case where the hardware is closed, and the manufacturers are in hock to Microsoft? With the relatively high cost and low market penetration of Tablets, is this an item that’s unlikely to get fixed?
Actually no. It’s all well known hardware inside, most of them are. Mine has Intel video card (so xrandr to rotate works fine) and a wacom tablet which is well supported out of the box. It’s just missing the glue to tie all the config together.
I have a Lenovo X200 Tablet and similar problems. Rotation actually worked with Intrepid Ibex and the Tabuntu package, which is a collage of scripts handling rotation and a daemon monitoring the “rotation state”.
However, starting from Jaunty, tablet rotation no longer works and the reason is “xsetwacom”. Why? Because the wacom tools only work with static xorg.conf configurations but Jaunty uses Hal instead — xsetwacom is simply unable to find any devices. In principle, xinput should be able to fully replace xsetwacom in this case, but currently lacks support for inversion and swapping of pointer axes.
Long story short, it actually became _more_ troublesome to fully support the tablet mode of notebooks. This is sad, even more so as all necessary drivers and scripts are readily available.
I had a Toshiba Portege M200 and everything had to be setup manually, wacom, xorg and a rotate script. I’ve now got a Toshiba portege m700 which has touchscreen and stylus. I got it working at one point but it breaks every kernel update.
I’m hoping that updating to Karmic will automagically sort everything out, but I am doubtful.
I think that multitouch with fingers will take over but in the end a combination with stylus will win (see Microsoft courier project) because we can not write with fingers and touchkeyboars suck. So supporting tablet PCs is not something futile.Count me in for testing (hp 2510p tablet pc. Love xournal and stuff.
I’ve got a Toshiba Tecra m7. It’s fairly similar to yours, even more similar to the thinkpads. My idea was to attach to the ACPI LID event but for some reason it emits an event when you convert from tablet to laptop mode but not when you convert from laptop mode to tablet.
Honestly at this point the problem isn’t programming so much as browsing tech sheets for how the manufacturer does it. They do have an information website but I’ve seen nothing of use.
A few months ago I wrote a daemon which allows configuration of input devices with Xinput through DBus (i.e. from any language that have DBus bindings).
The Wacom driver in Lucid will support Xinput properties (as, for example, touchpads do). This means that I could implement a simple feature in my daemon which makes the daemon look for (RandR) screen rotation events and rotate the tablet accordingly (as tablet’s rotation will be an Xinput property). This of course would be an option in a simple UI written in, say, Python (or you could simply modify the xml configuration file of the daemon).
I hope to have the time to do this :-/
This sounds like a job for the 100 paper cuts team
Don’t forget us lefties. We can’t scroll down on tablets without covering the page contents with our hands. It’s terrible. Widnows decided not to fix this in Vista (don’t know about Win7) because it would break too many apps that hard code layouts and dimensions to have a ‘move scrollbars to the left side’ option. I say, break ‘em. This is one area of usability where gnome could shine.
If you consider filing a bug, please have a look at Bug #159026.