btnx: Send keyboard and mouse combination events with mouse buttons

[tim71]

I have no clue what could be causing this. btnx shouldn't be in any way linked to boot time framebuffer sizes. Maybe the framebuffer option causes something in the input system to change?

To be exact - this is not a "boot-time only" thing - if you change to tty1...6 with Ctrl+Alt+F1...F6 then the console screen has different resolution, then default one.

P.S. Sometimes btnx even worked without a problem with vga=791 (1024 x 768 resolution), but mostly not, if I recall corectly. Without this option in GRUB menu all works without problems.

[daou]

btnx-config-0.1.3 tries to fix the crash bug during button detection.
Instead of freeing the communication pipes during child exit, the parent simply flags them as obsolete and continues. A timer then periodically checks for any obsolete pipes and frees as necessary. And a whole bunch of checks have been added.

[daou]

I'd like translate it to italian. How can I do?

If you are referring to btnx-config, sure. But first I need to get the program fully functional. Once that is done, I'll try to add support for translations.

[rzj]

btnx-config-0.1.3 tries to fix the crash bug during button detection.
Instead of freeing the communication pipes during child exit, the parent simply flags them as obsolete and continues. A timer then periodically checks for any obsolete pipes and frees as necessary. And a whole bunch of checks have been added.

Tried all the buttons three times through, no crashing. Seems stable to me.

Ray

[imperialline]

Thumb wheels:

Hi daou,

Having played around with the btnx and the MX revolution in the past few days on AMD64, I have found that everything works beautifully except the thumbwheels' behaviors.

I use xev and also do some debuggings under gdb and here are what I have found:

1) Clicking on the Thumb Wheel Forward produces button 9 and 3 event presses and releases. So it will genrate 2 separate actions based on 2 distinct button 9 and button 3 clicks
2) Click on the Thumb Wheel Backward generates button 8 and 2 event presses and releases. So it will genrate 2 separate actions based on 2 distinct button 8 and button 2 clicks.
4) Sometimes one click generates a spurious sequence of (1) or (2) mentioned above although btnx recognizes as one click
4) Btnx appears to do eveverything correctly but somehow the we still have the problem. I guess that this could be a mouse driver problem.

I.

[tim71]

Update - I think, that I managed to find at least some kind of a solution. I got the lomoco package and issued

sudo lomoco --no-sms

which disabled the smart-scroll and other additional events for some of the additional keys - needs more testing though...

Had to add this line to /etc/rc.local because it needs to be re-applied after restart.

Now I found a bit more elegant way to do the same thing - just add the following line to /etc/default/lomoco

LOGITECH_MOUSE_DISABLE_CC=yes

and the rc.local entry is no longer needed

[daou]

btnx-0.2.14 adds support for MS Intellimouse Explorer 2.
Thanks to glave for doing the hexdumping.

[casell]

Sorry daou i didn't have much time in this period!

I tried btnx-config a little. and i obviously have something to say!

1) I got an error (handled) with btnx daemon on, it would be good that the config apps shuts down the daemon or, even better in my opinion, have the config app ignore btnx "fake" devs

2) Then please transform the keycode selectbox in a combo so one can search in it and save his/her eyes!

3) Add an accelerator to the top button in all the mouse detection "wizard" or even better prevent other things to steal the focus! The workflow for me should be:
a) Focus on start button detection
b) Focus on stop button detection (here is to decide cause it may be cause of human's error, maybe it's better to add an accelerator only for the Stop action, so the user need to think about stop)
c) Focus on text box
d) At click to Add button, give back focus to start button detection

4) Minor one: in main window the Configuration File Clear Button has ALT+C as shortcut and guess what is the shortcut for the Close Button (yes it's ALT+C)

5) In the last version the logo path is hardcoded in widgets.c line 52

I was thinking about making a QT/KDE version of btnx-config in Python for those poor Kubuntu users (like me ), but i'm going to wait until feature and appearence freeze to make it similar to GTK one.

While waiting and testing I'll try to figure out a way to get correct mouse in case of same VID & PID.

Bye and good work!

[daou]

Sorry daou i didn't have much time in this period!

I tried btnx-config a little. and i obviously have something to say!

1) I got an error (handled) with btnx daemon on, it would be good that the config apps shuts down the daemon or, even better in my opinion, have the config app ignore btnx "fake" devs

2) Then please transform the keycode selectbox in a combo so one can search in it and save his/her eyes!

3) Add an accelerator to the top button in all the mouse detection "wizard" or even better prevent other things to steal the focus! The workflow for me should be:
a) Focus on start button detection
b) Focus on stop button detection (here is to decide cause it may be cause of human's error, maybe it's better to add an accelerator only for the Stop action, so the user need to think about stop)
c) Focus on text box
d) At click to Add button, give back focus to start button detection

4) Minor one: in main window the Configuration File Clear Button has ALT+C as shortcut and guess what is the shortcut for the Close Button (yes it's ALT+C)

5) In the last version the logo path is hardcoded in widgets.c line 52

I was thinking about making a QT/KDE version of btnx-config in Python for those poor Kubuntu users (like me ), but i'm going to wait until feature and appearence freeze to make it similar to GTK one.

While waiting and testing I'll try to figure out a way to get correct mouse in case of same VID & PID.

Bye and good work!

Excellent feedback! Thanks .

1) Good observation. This has already been fixed last night. btnx now reserves a vendor ID of b216, and the mouse detection ignores it. However, I haven't released the latest btnx-config (0.1.4) that does this (must have btnx-0.2.14 to register the correct btnx vendor and product IDs).

2) I've been thinking about this since I first made the combobox . I'll change it, but it is a lower priority at the moment.

3) I'll do as you suggest.

4) Actually, this is also dependent on what language you are running btnx with. It's impossible to predict if some buttons will have accelerator overlaps or not.

5) It will be fixed soon. And, as I'm writing this, I'm looking at making .desktop files and correct icons+paths for btnx icons.

Some other things that are in 0.1.4: internationalization support, tooltips, a new delay value (not supported by btnx yet), and button deletion. And some other minor bug fixes.
I need to polish up the makefiles for 0.1.4 before I'll release it (the SVN, of course, is open to anyone to get the latest. If someone does, please use "make debug" to build it. "make all" doesn't work yet).

And I'm glad you think btnx-config is worth porting to QT. It probably won't be simple, so I commend you for it! Actually, I started making some wrapper macros for things like setting label and button texts (SET_LABEL_TEXT()). These might be really useful if you are porting to QT. Unfortunately, only the newest code bits use them. If you think these will help in QT, I could try updating the code to use more wrappers (and make more wrappers) in GTK specific calls. The look will be frozen when btnx-config-0.2.x is released, so be on the lookout for it.

[Spudgun]

Worked great, and now I can tilt my mouse wheel left and right to go back and forward in Firefox, hurrah!