Opening Malfunction

[GoogleCodeExporter]

I do not know why, but sometimes, when you open the program, it doesn't detect 
the fingers. In some cases, I must close and reopen the application until 8 or 
10 times to work properly. Other times, however, works well from the beginning.

My operating system is Lion (Mac OS X 10.7.2), and the computer is an iMac 
mid-2011. I have 0.3.0 MagicKeyboard version installed. Do you know why this 
happens?

Greetings!

Original issue reported on code.google.com by JoCol...@gmail.com on 24 Oct 2011 at 1:45

[GoogleCodeExporter]

I got some other reports about this, too.  Has it always happened?  Or only 
with specific versions?

I haven't been able to track down the cause of this yet.  It happened once or 
twice to me as well, but it never re-happened when I went to debug it.

One thing to blame might be that I'm using some undocumented (private) 
multitouch API, and there may be something I should be doing but I'm not.  The 
problem here is to find out what and why, as there's no documentation 
whatsoever, and things may change without notice across OSX versions.

I'll be looking for more aggressive device detection options for the next 
releases.

Original comment by emanuele...@gmail.com on 24 Oct 2011 at 2:38

• Changed state: Accepted
• Added labels: Milestone-Release0.3

[GoogleCodeExporter]

It has happened from the beginning, but I thought it was because I had a 
problem with any other application as MagicPrefs, but after uninstalling both 
and restarting, the problem still exists, even without having installed 
MagicPrefs.

And the number of times I close and reopen the application is not always the 
same. As an user, I can't find the pattern.

Original comment by JoCol...@gmail.com on 24 Oct 2011 at 3:03

[GoogleCodeExporter]

Hello,
I've got the same problem (10.7.3, mid 2011 imac, current release of 
magickeyboard).
But even doing some intensive tracing inside the code I wasn't able to track 
the reason down. However, every time the problem arises the registered callback 
is never called, so it seems, the failure happend while registering the 
callback by 'MTRegisterContactFrameCallback' or starting to report 
'MTDeviceStart'.
However, I found that appending an 'NSLog' after the calls to 'MT..' solves the 
problem... (at least on my machine, and.. never ask why...)
"
MTRegisterContactFrameCallback([deviceList objectAtIndex:i], callback); 
//assign callback for device
MTDeviceStart([deviceList objectAtIndex:i], 0); //start sending events
NSLog(@" ");
"

PS: Cool piece of useful code!
I'm right at the beginning to write software on and for my imac (after around 
20 years of Windows/C++ development) and a bit annoyed about the lack of the 
most often needed '[]{}'-keys on my wireless keyboard. So I started to design a 
new keyboard definition/layout for magickeyboard with just these few missing 
keys. When I'm successful I will of course upload this new layout!

Greetings from Germany!

Original comment by AundH.Al...@googlemail.com on 18 Feb 2012 at 12:15

[GoogleCodeExporter]

Hi, AundH.Albers.  Now that you mention it, I once encountered the same problem 
when I was at the very beginning writing this app.  For apparently no reason at 
all, on one of my (back then) two machines, it stopped working, and it didn't 
restart working until I added a dummy NSLog around there (not sure on the exact 
position, but it was for sure around that area, and I can bet it was the same 
as yours).

I later removed the NSLog line because it restarted working even without it, 
out of nowhere.  Thinking it was either a silly error elsewhere that I had 
fixed without even noticing, or that I was just a little crazy and had made up 
the entire problem, I took it off.  I guess that's what I get for using 
undocumented private APIs after all, heh.

I'll just add that dummy line in the next release I publish (it's probably 
gonna be around two weeks from now, when the college semester is finally over 
and I can restart working on hobby projects).  The line won't do any hurt, and 
if it helps getting the app to work on some computers, it's only good.

About the lack of [, ], {, }  keys, I totally share your pain.  Italian 
keyboards lack the curly brackets as well, and the square ones are a little 
annoying to type (AltGr / right alt), not to mention the lack of backtick (`) 
and tilde (~).
My workaround was quite more radical than yours: one day, really annoyed, I 
decided to learn blind-typing on an italian (hardware) keyboard, with the 
US-International (software) layout.  As a result, now all of my computers use 
some variant of the US Intl layout (I currently only own laptops and keyboards 
with US Intl physical layout), and I don't think I'm ever gonna go back to the 
national one.

If you'll be so kind to upload your layout once it's done, I'll be glad to 
include it in a future version of the app.  A nice gift to all the programmer 
fellows out there who don't want to go the radical way as I did :P

Original comment by emanuele...@gmail.com on 19 Feb 2012 at 12:35

[GoogleCodeExporter]

This issue was closed by revision 2f153878bb35.

Original comment by emanuele...@gmail.com on 16 Apr 2012 at 11:44

• Changed state: Fixed