GoogleCodeExporter
commented
9 years ago My apologies. Did not mean to label as "Defect".
Pierre.Original comment by J.Pierre...@gmail.com on 25 Sep 2013 at 10:57
Open GoogleCodeExporter opened 9 years ago
GoogleCodeExporter
commented
9 years ago My apologies. Did not mean to label as "Defect".
Pierre.Original comment by J.Pierre...@gmail.com on 25 Sep 2013 at 10:57
GoogleCodeExporter
commented
9 years ago Hello Pierre,
it is no problem, I can adjust the labels...
First of all, I'd need to know what platform are you on - Windows, Linux or Mac?
Next question is, do you have access to a Windows machine, or do you have
access to Linux box with Windows instaled in a virtual machine? Unfortunately
I'm not aware of any reasonably working USB dumper for Mac, however should you
find one, it should work as well. The thing is, that what I need is a dump of
the USB communication between the device and the driver...
For Linux and Windows you can use the Wireshark (on Linux directly, on windows
you have to install also the usbpcap - see
http://wiki.wireshark.org/CaptureSetup/USB ).
On windows, you can also use the trial version of usblyzer (
http://www.usblyzer.com ).
Let me know, should you need any help with setting that up...
Kind regards,
Michal
PS. Unfortunately I can't CC on the mail you specified - the CC in the form
allows to CC only project members and project member mail must be linked to
google account...Original comment by f.jo...@email.cz on 26 Sep 2013 at 6:13
GoogleCodeExporter
commented
9 years ago Alright. Sorry for the delay. Forgot to check this page. Here is a USBLyzer
file for a short run. Let me know if you need anything more than this, and if
so, just specify details.
Regards,
Pierre.Original comment by yolu...@gmail.com on 19 Oct 2013 at 11:08
Attachments:
GoogleCodeExporter
commented
9 years ago Hello Pierre,
it is OK, I had some work on my other project, so no problem...
Thank you for the dump, it gave me some idea (the protocol seems to be
identical to TrackIR4)...
I'd like to ask you for the following:
Close the SmartNav software and unplug the device; plug the device in and start
the capture, then start the SmartNav software (it will record the firmware
upload and the starting sequence). When the SmartNav runs, please move the
marker all the way to the bottom, then all the way up, left and right (the
exact sequence doesn't matter, I just need you to put the marker to all the
extremes of device's field of view). When this is done, please stop the
SmartNav software again and stop the capture (so I can see the shutdown
sequence).
I'm not sure, what controls does the SmartNav3 features (threshold?, led
brightness?, grayscale mode?, anything else?) - but if you could do a short
dump for each one of them, just start dumping, move the control all the way to
one side, then to the other and back, where it was and stop the dump...
Thank you very much,
MichalOriginal comment by f.jo...@email.cz on 20 Oct 2013 at 8:15
GoogleCodeExporter
commented
9 years ago Here is the first capture, as well as screen shots of the software tabs. Do you
need off/on and levels up/down for each button and slider of each tab?
Regards,
Pierre.Original comment by yolu...@gmail.com on 23 Oct 2013 at 1:17
Attachments:
GoogleCodeExporter
commented
9 years ago Hello Pierre,
thank you for the dump, I'll play with it later in the evening...
Looking at the screenshots, the only thing that is of interest is the Tracking
tab - just try unchecking and checking the "Camera LEDs On" checkbox, then try
that light filtering checkbox (chances are some other control shows there; if
not just check/uncheck it).
Thank you,
MichalOriginal comment by f.jo...@email.cz on 23 Oct 2013 at 10:53
GoogleCodeExporter
commented
9 years ago Here is the tracking tab capture. First turned the LED off (which opens a
warning dialog, clicked ok) then back on. Then turned filtering on and moved
the slider to each extreme a few times. Shut down the control window in
SmartNav and stopped the capture.
Regards,
Pierre.Original comment by yolu...@gmail.com on 23 Oct 2013 at 10:28
Attachments:
GoogleCodeExporter
commented
9 years ago Hello Pierre,
thank you for the dump... I was able to extract informations needed to support
the device and I started the coding. One more thing I'd need to know - what
system/distro are you using (MacOSX, Linux) and in case of Linux, is it 32 or
64 bit one?
I need that information so I can send you an experimental build so you can try
if it does work.
Kind regards,
MichalOriginal comment by f.jo...@email.cz on 24 Oct 2013 at 4:46
GoogleCodeExporter
commented
9 years ago Hello Pierre,
the work is going well and the code seems to be working well.
However, I'd ask you for one more dump, so I can be sure everything works fine
- could you dump marker moving to all extremes (left, right, up, down), if
possible to get it out of the tracker's field of view each time?
Thank you,
Michal
PS. One more question - does smartnav3 have those sip/puff or foot pedal switch
inputs, like smatrnav4? If so, do you have any of them?Original comment by f.jo...@email.cz on 25 Oct 2013 at 7:39
GoogleCodeExporter
commented
9 years ago Hi there Michal,
Here is the data requested. Note that I'm running SmartNav in "Absolute" mode,
which maintains the cursor's relative position when you move past the edges of
the screen boundary. "Relative" is the default for some reason, but it's
terrible to use; you constantly have to center the pointer, as it loses track
of your seated position.
I'm using Windows 7, 64 bit and Linux Mint, 64 bit (still running from a liveCD
only, but it is the version I am going to install soon - still sorting out my
new system). SmartNav3 has the foot pedal input, but it's not enabled on my
version of the device. It's probably just a firmware addition, but the enabled
version was more expensive and I didn't need it.
Regards,
Pierre.
Original comment by yolu...@gmail.com on 25 Oct 2013 at 9:00
Attachments:
GoogleCodeExporter
commented
9 years ago Hello Pierre,
I just wanted to let you know how things are going...
I think I have the SmartNav3 support mostly ready (unfortunately it took me
more time than I expected due to fixing of a bug that appeared on ubuntu
13.10); right now I'm working on a mouse emuation, adding the absolute mode and
fixing few problems I found along the way.
So, if all goes well, I should be able to provide you with the first testing
package still this weekend...
Kind regards,
MichalOriginal comment by f.jo...@email.cz on 1 Nov 2013 at 6:30
GoogleCodeExporter
commented
9 years ago Hello Pierre,
I just uploaded the testing package - please download it from
http://www.linuxtrack.eu/repositories/universal
and choose linuxtrack-0.99.7.1-64.zip.
Install it according to this page:
http://code.google.com/p/linux-track/wiki/UniversalPkgInstall
Please note, that the command to install the udev file will be
sudo cp /opt/linuxtrack-0.99.7.1/share/linuxtrack/51-TIR.rules
/lib/udev/rules.d/
After this command, the device needs to be unplugged/replugged (or system
rebooted) to make the rule effective.
When done, run the ltr_gui and set TrackIR as the tracking device on the first
pane,
on the second pane set the model to "single point reflective" and try, if the
tracking works. If it does, you can try running mickey - the mouse emulator.
To make it work, you have to have the /dev/uinput writeable - easiest way
(holds until restart!) is this:
sudo setfacl -m u:${USER}:rw /dev/uinput
Hopefully it will work for you as well.
Kind regards,
MichalOriginal comment by f.jo...@email.cz on 3 Nov 2013 at 10:18
GoogleCodeExporter
commented
9 years ago Hi,
did you have a chance to try it out? Did it do something?
Kind regards,
MichalOriginal comment by f.jo...@email.cz on 6 Dec 2013 at 5:40
GoogleCodeExporter
commented
9 years ago Enlisting help for instructions. Will be testing tonight or this weekend.
Sorry for delay.
Pierre.Original comment by J.Pierre...@gmail.com on 14 Dec 2013 at 1:15
GoogleCodeExporter
commented
9 years ago Ok Michal, I've got it mostly set up. It's only in a VM (Oracle VirtualBox
4.3.4) at the moment, but I've got the SmartNav recognized in Linux. If
necessary, I can run it with a fresh install, but I need to get that set up
first.
I've run through all your instructions and the GUI is up and running. The
SmartNav is tracking fine, but I can't get Mickey to move the mouse pointer.
Could this be a problem with the VM?
Let me know if I can continue in VM or if I need full install. Either way, any
ideas what to try next?
Sorry about the delay again, but I accidentally lost my HDD structure and had
to reinstall everything and due to my inexperience in Linux, I had to find time
to enlist my brother's help.
Thanks,
Pierre.Original comment by J.Pierre...@gmail.com on 14 Dec 2013 at 3:07
GoogleCodeExporter
commented
9 years ago Hello Pierre,
it is OK, no problem...
Now, when you say, that it is tracking fine, if you go to the 3D view (cockpit
model), the view does move according to your head movements?
To be honest, I've never tried it in the virtual machine, but on the other
hand, it should work...
What you can try is this: close Mickey, if it runs and go to the /dev/input
directory. Note the files called eventX, where X is a number. Now start the
Mickey - did another event file pop up?
Now, if you take a look at the end of the /var/log/Xorg.0.log, is there
anything about "Linuxtrack's Mickey"?
If you can't make sense of it, then please package the /tmp/linuxtrack*.log
files and the /var/log/Xorg.0.log attach them here or send them to my email and
I'll try to find out what might be wrong...
Thank you,
MichalOriginal comment by f.jo...@email.cz on 14 Dec 2013 at 12:12
GoogleCodeExporter
commented
9 years ago Hello Michal,
Sorry about the sporadic work, but I've got some of the info you asked for.
The 3d cockpit view works fine with the tracking. Follows movements perfectly.
Mickey does make the event file appear. It disappears when Mickey closes.
Getting a message in the Xorg file about input driver missing. Might be problem
with VM. I'm attaching the log files, but if you think it's a Virtual Machine
issue, then I'll do a full install ASAP and try again.Original comment by J.Pierre...@gmail.com on 19 Jan 2014 at 1:46
Attachments:
GoogleCodeExporter
commented
9 years ago Hello Pierre,
it is OK, I still have other things to keep me occupied, so take your time...
The log seems to indicate that the X server doesn't know what to do with the
device, which is quite strange. I'll try to reproduce the problem and report
back with my findings (it might take me few days though).
Kind regards,
MichalOriginal comment by f.jo...@email.cz on 19 Jan 2014 at 8:34
GoogleCodeExporter
commented
9 years ago Hello Pierre,
I'm sorry it took me so long, but finally I have found the answer.
To let the mickey work in the virtualbox, you have to turn off the mouse
integration - either in the menu, or by pressing the host+I (host key is right
Ctrl by default). The thing is, that the virtual machine "overrules" mickey -
when you turn the integration off, it will let mickey do its job. I also had to
turn off the 3D acceleration, as the cursor was not visible with it - seems to
be a bug in Virtualbox.
If I may, I'd advise you to try the VMware player instead - it is free, and the
VMware doesn't seem to have many of those shortcommings...
Kind regards,
MichalOriginal comment by f.jo...@email.cz on 1 Feb 2014 at 11:14
GoogleCodeExporter
commented
9 years ago Thanks for the solution. I've tried it and it works. The cursor tracks the
movement well. However, the movement is a bit restricted. I've tried running
the calibration in Mickey and increasing the axes sensitivity in the LinuxTrack
GUI, but it still requires pretty extreme head movements to get the cursor
across the screen. Is this as it's supposed to be or am I missing some setup
procedures?
Thanks,
Pierre.Original comment by J.Pierre...@gmail.com on 12 Mar 2014 at 6:57
GoogleCodeExporter
commented
9 years ago Hello Pierre,
you can try this: when calibrating, use only the comfortable head movement
range. Then hopefully the you will have the full range of movement without a
need for uncomfortable movements.
Kind regards,
MichalOriginal comment by f.jo...@email.cz on 13 Mar 2014 at 8:07
GoogleCodeExporter
commented
9 years ago Hi Michal,
You're right, that worked well. Just calibrating with slight head movements
really increases the sensitivity. The movements are quite good now.
I would say that it works well. It's not really the easiest to setup, but it
does work well, and my inexperience with Linux probably didn't help in getting
started. Is there anything else for me to test or does this make it ready for
release?
Regards,
Pierre.Original comment by J.Pierre...@gmail.com on 5 Apr 2014 at 1:19
GoogleCodeExporter
commented
9 years ago Hello Pierre,
I'm glad it finally works to your liking...
If I may, I'd like to ask you, if you could reiterate what caused you the
problems, what you found counter intuitive or not obvious, so I can add it to
the documentation.
Given the fact that I developed the thing, causes me sometimes to forget to
mention certain things, that seem obvious to me, but they really are not. That
makes feedback from others so valuable and helps me to improve everything.
As for release, I'm revising the documentation and external Linuxtrack
interface. When that is done, I think I will finally release the 1.0 version.
Thank you for your help,
Michal Original comment by f.jo...@email.cz on 5 Apr 2014 at 6:07
GoogleCodeExporter
commented
9 years ago Hello there Michal,
Let me first apologize that it has taken me so long to get back to you. We had
a new baby at beginning of March and things have been busy. Secondly, it may
very well be my limited Linux knowledge that caused me more difficulty than
usual. I'm not very experienced with terminal commands and so the "enter your
name/file path here" style of example commands give me some trouble. It would
have helped if you had written "For example, your system name is SYSTEM and the
file path is user/dev/whatever" And then written the examples accordingly.
Anyhow, I did find that having to modify several individual files in order to
be able to run the software was a bit complicated (perhaps I'm spoiled by
Windows). It would be nice to have this process scripted so that the system
finds and opens the files automatically and asks you for the access permission
whenever required. Also, having two separate piece of software in order to
accomplish one task could be improved by having the first piece of software
access the other (even if this is just by including a button in the linux-track
GUI that lets you open Mickey). As it is, having to open two terminal windows
to turn on what is essentially one software tool is a bit tedious. Furthermore,
it would be nice if the software could be accessed through the program menus
rather than having to remember the terminal commands to get it started.
Also, as to the functionality, it might be a good idea to mention in the
documentation the proper method to calibrate the cursor movements (as it was in
my case, where you have to start with very small movements in order to get a
decent speed on the cursor).
Finally, I just want to say that I greatly appreciate your work in the software
and that these comments above are only feedback intended to help you improve
it. I want to be frank in providing you with constructive criticism, but I'm
very thankful of all the work you have done.
Regards,
PierreOriginal comment by yolu...@gmail.com on 14 Jun 2014 at 8:07
GoogleCodeExporter
commented
9 years ago Hello Pierre,
it is OK; I know what you mean - we have three and half year old twin boys and
a new baby in about two weeks...
I appreciate your feedback very much and I'll try to improve the documentation
in that regard.
Hopefully things will get easier when I reintroduce native packages for
Ubuntu/Fedora (and Debian) - all of them should introduce menu items for the
ltr_gui, Mickey and wiimote server. Just as a sidenote - you can create a
launcher for the binaries yourself - but you'll have to consult your desktop's
environment documentation to find exactly how (most environments let you do
that with right click on the desktop and creating shortcut/launcher)...
Kind regards,
MichalOriginal comment by f.jo...@email.cz on 15 Jun 2014 at 1:04
Original issue reported on code.google.com by
J.Pierre...@gmail.comon 25 Sep 2013 at 10:55