Ubuntu on Eee PC T101MT

[tlimsisnw]

I tried Hangman228's recipe to invert the camera, but got this error:

make -C lib all
make[1]: Entering directory `/home/johnson/Downloads/v4l-utils-0.7.92-test/lib'
make -C libv4lconvert all
make[2]: Entering directory `/home/johnson/Downloads/v4l-utils-0.7.92-test/lib/libv4lconvert'
cc -Wp,-MMD,"libv4lconvert.d",-MQ,"libv4lconvert.o",-MP -c -I../include -fvisibility=hidden -fPIC -DLIBDIR=\"/usr/lib\" -DLIBSUBDIR=\"libv4l\" -I../../include -D_GNU_SOURCE -g -O1 -Wall -Wpointer-arith -Wstrict-prototypes -Wmissing-prototypes -o libv4lconvert.o libv4lconvert.c
/bin/sh: cc: not found
make[2]: *** [libv4lconvert.o] Error 127
make[2]: Leaving directory `/home/johnson/Downloads/v4l-utils-0.7.92-test/lib/libv4lconvert'
make[1]: *** [all] Error 2
make[1]: Leaving directory `/home/johnson/Downloads/v4l-utils-0.7.92-test/lib'
make: *** [all] Error 2

Anyone else got this? Sorry, I'm new to ubuntu and linux. Grateful for some pointers.

[Hangman228]

@tlimsisnw : I think you have ton install build-essential before run commands

@all : Have you tried the new unify ? It works quite good on my computer and it's good for tactile! But there is many bugs at the moment...

[Plippo]

Unfortunately, I wasn't able to find the SVN revision mentioned in the MyPaint bug (MyPaint has switched to GIT, so I haven't found any SVN at all) so I tried to create my own patch for this bug, with more or less success: It works for some simple brushes, but not for all. Anyway I've created a package with the modified version so my wife and anyone else who's interested can try it out, I guess a few working brushes are better than no working brushes. The package is available here: http://www.philmerk.de/dwl/deb/mypaint-resistive.deb - have fun!

@Hangman228: I've tried unity, interesting concept, but at the moment too many bugs for me, e.g. dragging a new link to the bar crashes it. Also I don't like that the maximized windows have title bars, I like it better how it's in UNE now with the title integrated into the panel.

@dusjagr: Strange that the device ids change for you, probably because you're using a mouse. I'll look if I can update the rotation script to automatically determine the ID. The speed problem is not a problem with the touchscreen but with the graphics driver, Intel have screwed up their Linux driver a lot in the last two years. Maybe there's some workaround to make it quicker, I don't know.

@Confuseling: I also guess the EFI is only for the boot booster, but I don't really know. I left all default partitions intact, only reduced the size of the Windows partition to 40 GB and created my Ubuntu partitions in the available space.

[Plippo]

Oh, and @dusjagr: Strangely for me automatic switching between speakers and headphones works. Have you installed any special ALSA package? Or maybe ASUS has built in different audio chipsets in different models. For me, lspci -v shows the following audio device:

00:1b.0 Audio device: Intel Corporation N10/ICH 7 Family High Definition Audio Controller (rev 02)
    Subsystem: ASUSTeK Computer Inc. Device 83ce
    Flags: bus master, fast devsel, latency 0, IRQ 22
    Memory at f7cf8000 (64-bit, non-prefetchable) [size=16K]
    Capabilities: <access denied>
    Kernel driver in use: HDA Intel
    Kernel modules: snd-hda-intel

[mrspacklecrisp]

Plippo, I can't thank you enough.

Mypaint is working very well, but this error interrupts me after a minute of drawing. After, I draw invisible lines that don't appear until I minimize the window, then maximize it again.

Traceback (most recent call last):
  File "/usr/share/mypaint/gui/tileddrawwidget.py", line 203, button_press_cb(self=<TiledDrawWidget object at 0xb82ee3c (GtkDrawingArea at 0x9c350a0)>, win=<TiledDrawWidget object at 0xb82ee3c (GtkDrawingArea at 0x9c350a0)>, event=<gtk.gdk.Event at 0x9ebe1a0: GDK_2BUTTON_PRESS x=403.60, y=236.01, button=1>)
                    self.last_painting_pos = x, y
                self.doc.stroke_to(dtime, x, y, pressure)
  variables: {'y': ('local', 236.01098901098902), 'x': ('local', 403.59926739926738), 'dtime': ('local', None), 'self.doc.stroke_to': ('local', <bound method Document.stroke_to of <lib.document.Document instance at 0xb82d08c>>), 'pressure': ('local', None)}
  File "/usr/share/mypaint/lib/document.py", line 109, stroke_to(self=<lib.document.Document instance>, dtime=0.0, x=403.59926739926738, y=236.01098901098902, pressure=None)
            l.surface.begin_atomic()
            split = self.brush.stroke_to (l.surface, x, y, pressure, dtime)
            l.surface.end_atomic()
  variables: {'self.brush.stroke_to': ('local', <bound method Brush.stroke_to of <lib.brush.Brush; proxy of <Swig Object of type 'Brush *' at 0x9ebe230> >>), 'pressure': ('local', None), 'split': (None, None), 'l.surface': ('local', <lib.tiledsurface.Surface; proxy of <Swig Object of type 'TiledSurface *' at 0x9ebe218> >), 'y': ('local', 236.01098901098902), 'x': ('local', 403.59926739926738), 'dtime': ('local', None)}
  File "/usr/share/mypaint/lib/mypaintlib.py", line 134, stroke_to(self=<lib.brush.Brush; proxy of <Swig Object of type 'Brush *' at 0x9ebe230> >, *args=(<lib.tiledsurface.Surface; proxy of <Swig Object of type 'TiledSurface *' at 0x9ebe218> >, 403.59926739926738, 236.01098901098902, None, 0.0))
        def set_mapping_point(self, *args): return _mypaintlib.Brush_set_mapping_point(self, *args)
        def stroke_to(self, *args): return _mypaintlib.Brush_stroke_to(self, *args)
        def get_state(self): return _mypaintlib.Brush_get_state(self)
  variables: {'_mypaintlib.Brush_stroke_to': ('global', <built-in function Brush_stroke_to>), 'stroke_to': (None, None), 'self': ('local', <lib.brush.Brush; proxy of <Swig Object of type 'Brush *' at 0x9ebe230> >), 'args': ('local', (<lib.tiledsurface.Surface; proxy of <Swig Object of type 'TiledSurface *' at 0x9ebe218> >, 403.59926739926738, 236.01098901098902, None, 0.0))}
TypeError: in method 'Brush_stroke_to', argument 5 of type 'float'

I'm having the same issue dusjagr is, and that output matches mine, but in 9.10 it would automatically switch for me (although unplugging the headphones wouldn't)

[Confuseling]

@Plippo - well, I went the slightly more reckless route (heck, when you have a restore disc little is truly reckless), and deleted partition 2, turning it into two 30 gig root partitions, a 4 gig swap, and the rest home.

Grub went on one of the roots initially, and I set the bootable flag, but it wouldn't work, so I used a rescue CD to put grub on the MBR and have had no notable problems since - can still use Express Gate (whether that counts as an advantage or not...), but did admittedly lose the F9 restore function.

The drivers work excellently as all have said, so thanks to you and StCh.

@anyone: What I'm more concerned about now though is that I think my unit may be defective. Once a system is up and running nothing seems to be able to find ethernet - ifconfig on Ubuntu and Debian reports only loopback and wireless, lspci can't see anything either, Express Gate has the 'LAN' option greyed out... Even Win7 device drivers just says 'wifi adapter' under networking.

The really bizarre thing is that the Debian Squeeze netinst ISO downloads and installs everything perfectly through ethernet. It's just that once you boot the installed system - or any installed system - it can't see it...

Is anyone else having problems? Thanks.

[Plippo]

@Confuseling: I have no problems with Ethernet, eth0 shows up in ifconfig. But once after I had installed Ubuntu, I had the problem that my Bluetooth device disappeared. It turned out that it has been automatically disabled in BIOS for some reason. U reactivated and everything worked again. Maybe that happened to your Ethernet, too? Maybe for network booting the BIOS ignores this setting...

[hanoc]

Hi,

first of all I want to thank everyone in this thread and specially Plippo for their help.

I bought the T101MT just some hours ago and in just 3 hours I got an ubuntu netbook edition running and the touchscreen working.

1. Screensaver

EDIT: SOLVED

I followed the instructions to get the rotate button recognized and I got it as xf86screensaver. As Plippo said the first time i clicked the button it launched the screensaver. The problem is it's still doing it and the screen does not rotate (it does rotate correctly if I use the command directly).

EDIT: Sorry, the screen saver did not let me see the "reassing notice". I tried again, reasigned, and it works.

2. Longpresses
I guess it's not possible to perform different actions with long presses of a button, isn't it?
It would be wonderful to be able to launch matchbox keyboard and to rotate the screen while in tablet mode and we only have one button in the screen.

3. Virtual keyboard
Additionaly, are there any alternatives to matchbox-keyboard? does anyone if I can configure it to get displayed when I try to write in any field while on tablet mode? I know this should be explained elsewhere but my google skills are eluding me right now. I would appreciate even if anyone has any idea how to ask google for it.

Again, thanks to all of you, it's great to see the amount of work that there's been on lucid and the t101mt in so little time.

Thanks!

Edit: Out out, you demons of stupidity!

[Plippo]

I've created a new version of the MyPaint package that should fix the bug mentioned by mrspacklecrisp. It's available at the old place, http://www.philmerk.de/dwl/deb/mypaint-resistive.deb

[Plippo]

@hanoc:
To 1.: Have you set the screensaver button for the rotation action you created by clicking at the »Shortcut« column in the respective row in the Keyboard Shortcuts application and then pressing the rotation button? This set the shortcut for me and disabled launching of the screensaver.
EDIT: Great that it has been solved

To 2.: As the button sends its press events repeatedly, that should be possible in theory.

To 3.: I'm using CellWriter as a virtual keyboard. It's mainly a handwriting recognition software, but it also has a keyboard. It comes with a notification area icon so you can start it by just pressing the icon.