On Sun, Jan 9, 2011 at 11:49 AM, Cedric Sodhi <[email protected]> wrote: > On Sun, Jan 09, 2011 at 11:26:05AM -0600, Chris Bagwell wrote: >> On Sun, Jan 9, 2011 at 11:13 AM, Cedric Sodhi <[email protected]> wrote: >> > Hi, this is a "wish" i had for the driver that I'd consider very useful. >> > I've recently written a simple bash script which scripts the >> > functionality of the button, an example the button on the wheel changes >> > which keycodes the wheel sumit and hence which function in cotrnols. >> > >> > If anyone is interested I'd like to share it. >> > >> > The problem is, where even that is kind of cumbersome, more complicated >> > interaction because a real pain in the arse. Since the driver will only >> > emit keyboard events, everything will go through X. >> > >> > Right now, I'm using my WM (openbox), to respond to the global keys to >> > control the behaviour. For example the touchring button emits Ctrl + >> > Shift + Alt + T which openbox catches and runs my script on, to change >> > the mappings and give feedback through osd_cat. >> > >> > And whereas I could do the same for all buttons and find the most exotic >> > keybindings to never collide with running programs, it becomes quite >> > cumbersom to continuously remap them, map the bindings in Openbox and so >> > on. The amount of required keybindings which are used NOWHERE else on >> > the system grows exponentially with functionality. >> > >> > All these problems would immediately be solved if, instead of >> > keybindings, actual programs could be bound to keys. It would greatly >> > simply certain methods and give us a great flexibilty to write our own >> > complex behaviour without any effort. >> >> Launching scripts from inside a driver is pretty much a non-starter I >> suspect. >> >> One idea that quickly comes to me is perhaps xsetwacom could be >> extended to allow at least a hand full of the special keys defined in >> X11/XF86keysym.h. Those "internet keyboard" ones are basically meant >> to launch scripts and not be used directly by end applications. So >> although conflict can occur, they will be smaller I think and user can >> use standard GUI tools (Gnome Keyboard Shortcuts for example) to bind >> to your scripts as needed. >> > > Might you elaborate on why this would be a "non-starter". I don't see > any per se reason why this would not be possible, useful and as-secure > as the current driver. >
Well, I'm not the best to speak in this area but its red flags to me for an app (X and its input drivers) that runs as root to run arbitrary scripts. I'm sure its dropping permissions and all that good stuff but still a red flag. > Internet or keyboard or not, in general you might also find ultra-exotic > keycombos on normal keyboards that will only collide with 2% chance. > However, that option of starting programs makes a lot of things > significantly easier. > > If you argue against the very point that starting a program is > "insecure" or whatsoever, please explain why in particular and also > consider that I might aswell bind a keysequence that opens a terminal > (Usally Alt + F1) and executes an arbitrary command. Once your at this level, all the apps are again running as logged in user and have only those permissions. Don't get me wrong here. I fully understand and can related to your request. I just think there are enough security and philosophical issues (should input drivers be the termination point of input events?) that it will not happen. ------------------------------------------------------------------------------ Gaining the trust of online customers is vital for the success of any company that requires sensitive data to be transmitted over the Web. Learn how to best implement a security strategy that keeps consumers' information secure and instills the confidence they need to proceed with transactions. http://p.sf.net/sfu/oracle-sfdevnl _______________________________________________ Linuxwacom-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/linuxwacom-devel
