> finally I completed the first version of USBHOTKEY, a small utility that allows you to implement a script similar to autohotkey, just under Linux/X11.
Looks good, Lars, I'm excited about working with it. Unfortunately I haven't had a lot of success compiling software. It rarely works for me. I got the SVN sources for libhid and it configured but failed miserably to compile. I can't find an RPM for it. Any idea why I might get stuff like this? Thanks.
hid_wrap.c:2642: error: conflicting types for 'arg5' hid_wrap.c:2593: error: previous declaration of 'arg5' was here hid_wrap.c:2642: error: initializer element is not constant hid_wrap.c:2642: warning: data definition has no type or storage class hid_wrap.c:2643: error: syntax error before '}' token hid_wrap.c:2644: warning: type defaults to 'int' in declaration of 'result' hid_wrap.c:2644: error: conflicting types for 'result' hid_wrap.c:2594: error: previous declaration of 'result' was here hid_wrap.c:2644: error: initializer element is not constant hid_wrap.c:2644: warning: data definition has no type or storage class hid_wrap.c:2646: warning: type defaults to 'int' in declaration of 'resultobj' hid_wrap.c:2646: error: redefinition of 'resultobj' hid_wrap.c:1828: error: previous definition of 'resultobj' was here hid_wrap.c:2646: error: initializer element is not constant hid_wrap.c:2646: warning: data definition has no type or storage class hid_wrap.c:2647: error: syntax error before '{' token hid_wrap.c:2659: error: syntax error before '*' token hid_wrap.c:2659: error: syntax error before '*' token hid_wrap.c:2661: error: redefinition of 'arg1'
> Looks good, Lars, I'm excited about working with it. Unfortunately I > haven't had a lot of success compiling software. It rarely works for > me. I got the SVN sources for libhid and it configured but failed > miserably to compile. I can't find an RPM for it. Any idea why I might > get stuff like this? Thanks.
> hid_wrap.c:2642: error: conflicting types for 'arg5'
This is caused by the python wrapper. I guess you don't have python installed (like me, I get the same errors).
Try:
./configure --disable-swig make
It worked for me, for both the svn version and the tgz version. I put that hint in the README for the next release -- Lars Krueger http://LarsKrueger.homestead.com/files mailto:lars_e_krue...@gmx.de
> This is caused by the python wrapper. I guess you don't have python installed (like me, I get the same errors).
That's odd, I definitely have python... maybe I don't have dev headers?
> Try:
> ./configure --disable-swig > make
Yup, that works, thanks. Now to configure usbhotkey... it says it can't find libhid (which I just compiled and installed). How do I specify its location (/opt/libhid)? Thanks.
> Yup, that works, thanks. Now to configure usbhotkey... it says it > can't find libhid (which I just compiled and installed). How do I > specify its location (/opt/libhid)? Thanks.
Try LDFLAGS="-L/opt/libhid" ./configure
Maybe you need LDFLAGS="-L/opt/libhid" CFLAGS="-I/opt/libhid" ./configure
On Mar 30, 4:27 pm, "Lars Krueger" <lars_e_krue...@gmx.de> wrote:
> Try > LDFLAGS="-L/opt/libhid" ./configure
> Maybe you need > LDFLAGS="-L/opt/libhid" CFLAGS="-I/opt/libhid" ./configure
No, that didn't work. I recompiled libhid without specifying prefix and that worked (usbhotkey configure found it). Now it can't find imlib. I installed imlib-devel and that worked. For fun I installed ruby-devel, too. Now it configures OK but doesn't compile. Here's the entire make response:
make all-recursive make[1]: Entering directory `/opt/downloads/usbhotkey-0.1' Making all in images make[2]: Entering directory `/opt/downloads/usbhotkey-0.1/images' Making all in ag5 make[3]: Entering directory `/opt/downloads/usbhotkey-0.1/images/ag5' make[3]: Nothing to be done for `all'. make[3]: Leaving directory `/opt/downloads/usbhotkey-0.1/images/ag5' make[3]: Entering directory `/opt/downloads/usbhotkey-0.1/images' make[3]: Nothing to be done for `all-am'. make[3]: Leaving directory `/opt/downloads/usbhotkey-0.1/images' make[2]: Leaving directory `/opt/downloads/usbhotkey-0.1/images' make[2]: Entering directory `/opt/downloads/usbhotkey-0.1' gcc -DHAVE_CONFIG_H -I. "-I/usr/lib/ruby/1.8/i386-linux" -g -O2 -MT main.o -MD -MP -MF .deps/main.Tpo -c -o main.o main.c mv -f .deps/main.Tpo .deps/main.Po gcc -g -O2 -o usbhotkey main.o -lXtst -lImlib -lhid -lruby /usr/bin/ld: cannot find -lXtst collect2: ld returned 1 exit status make[2]: *** [usbhotkey] Error 1 make[2]: Leaving directory `/opt/downloads/usbhotkey-0.1' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/opt/downloads/usbhotkey-0.1' make: *** [all] Error 2
I'm going out of town this weekend, so I'll be back on this Sunday or Monday. Thanks for your help, Lars.
So, I got usbhotkey installed. However I encountered this error when running it (had to use sudo):
usb_set_debug: Setting debugging level to 1000 (on) usbhotkey: Loading script /usr/local/share/usbhotkey-0.1/ag5.rb usbhotkey AG5 script starting usbhotkey: Connecting to 0510:100b Check that you have permissions to write to 002/001 and, if you don't, that you set up hotplug (http://linux-hotplug.sourceforge.net/) correctly. USB error: could not claim interface 0: Device or resource busy usbhotkey AG5 script started usbhotkey: Starting event loop
I've checked my AG5's info via lsusb -vvv, and the information match those in the script, so it should be connecting to the correct device. What else can cause this error? I've setup Xorg to use evdev for the AG5's mouse, could this be the culprit (although the mouse is on a different end-point)?
Do it as root. It scans your installed libraries and makes a cache file to find them later on.
To your other question: The error can't find -lXtst
means that you don't have the X test extension installed. Your distribution should provide a package containing it. If not, check the forum/howto pages of your distribution. -- Lars Krueger http://LarsKrueger.homestead.com/files mailto:lars_e_krue...@gmx.de
> usb_set_debug: Setting debugging level to 1000 (on) > usbhotkey: Loading script /usr/local/share/usbhotkey-0.1/ag5.rb > usbhotkey AG5 script starting > usbhotkey: Connecting to 0510:100b > Check that you have permissions to write to 002/001 and, if you don't, > that you set up hotplug (http://linux-hotplug.sourceforge.net/) > correctly.
This is a bogus error message of libhid. It basically means you can't access a part of your USB tree from your current user. If it gets past this message it also means that you don't actually need to access it.
> USB error: could not claim interface 0: Device or resource busy
Again, bogus message. Some device (I think it's the hub in the AG5) can't be opened.
> usbhotkey AG5 script started > usbhotkey: Starting event loop
This is the important message here. If you get this far, everything works and you can pass events to X11. You should see the keymap display in the lower right of your screen.
> This is the important message here. If you get this far, everything works and you can pass events to X11. You should see the keymap display in the lower right of your screen.
Ah, I did notice something on the lower right corner of my screen, but it appears that the images cannot load correctly, I get blank, white buttons instead of the png images. The Ruby code point to the correct paths so I'm not quite sure what's wrong. After trying to type this message using my AG5, I also noticed that the bindings worked, so everything seems fine except for the button layout image display.
> The Ruby code point to the correct > paths so I'm not quite sure what's wrong.
Ca'nt help you there. I copied the code from xteddy. Maybe your imlib does not have png support. You may try to change the filenames and convert to, say, JPG. -- Lars Krueger http://LarsKrueger.homestead.com/files mailto:lars_e_krue...@gmx.de
I just found out via Ubuntu 6.10's package repository, that there's imlib2 and there are already ruby extensions written for it as well. Is there any reason not to use imlib2? Considering that it's possible to install the imlib2 dependency via a ruby gem, this can simplify the installation procedure a bit.
Can't say I "forgot" since I never knew... anyway, I tried it and it didn't work. Maybe I'll go back to try to compile libhid with swig so I can code in Python. Perhaps I'll have to wait 'til libhid comes out in RPM (followed shortly by usbhotkey.rpm :-) )
> >Considering that it's possible to install the imlib2 dependency via a > ruby gem
Since the image display codes from xteddy (displays a cuddly teddy bear on your screen, nothing else), you might try to install it and maybe imlib is installed automatically.
> Since the image display codes from xteddy (displays a cuddly teddy bear on your screen, nothing else), you might try to install it and maybe imlib is installed automatically.
I do have imlib installed from the package repository, I'm just wondering why isn't imlib2 used instead, since that is suppose to supersede imlib. Plus there's already ruby extensions written for imlib2 so you don't need to write and compile your own (even though you're only using a few functions).
I still haven't figure out why my imlib won't load the images correctly (the transparent part works, but the opaque part only comes in white. I've tried using jpeg and I just get a white box), I'll dig in some more when I get the time.
I'm also thinking about moving the key bindings map to a YAML file, as Ruby have excellent support for YAML (http://yaml4r.sourceforge.net/), and YAML is also much more readable than XML.
|
