Ron, You are right. We discussed this issue before. Wacom doesn't have a serial number to distinguish tablets of the same model in the firmware. I like your by-path rule. But I need to test it a bit more to see how reliable it is. I'll get back to you later.
Ping On 7/10/07, Ron <[EMAIL PROTECTED]> wrote: > On Mon, Jul 09, 2007 at 03:09:34PM -0700, [EMAIL PROTECTED] wrote: > > Both of your suggestions in the OS level help linuxwacom users, that > > is for sure. My concern is how do we deal with a system with 2 > > identical tablets. I know users running 2 Cintiq 21 UX to gain full > > accuracy and flexibility. Right now, the system has trouble to tell > > which Cintiq is the first monitor when the system wakes up from a > > sleep or even a nap :). Do you guys have any suggestion? > > I thought we'd got stuck without a good answer to this previously. > The full thread is at: > http://sourceforge.net/mailarchive/message.php?msg_name=20060418020303.GA16999%40hank.shelbyville.oz > > But the main bits of note to this question were me wondering: > > if there is a serial number or other unique id in the tablet that > we can access, then we can always make a <TabletType>-<SerialNum> > link or similar that people could use. > > To which you confirmed: > > As a short reply to the following discussion, there is no tablet > specific id to distinguish tablets of the same model and size. > > > So at this stage, the only method I'm aware of that we could use > to distinguish two identical devices reliably would be to use the > /dev/input/by-path links to differentiate them by the usb port > they are plugged into (since that's really the only thing the > machine can know about them that is different but potentially > stable). > > They aren't the most user friendly invention in the world, and > will need to be configured for each individual user, but for a > static workstation where the tablets are plugged in once and > left that way, they should always give the same unique device > name for each tablet. Its just a bit brittle if you start > moving things about and has a small initial hand-hack burden. > > > But I guess if we can't convince the hardware crew to add a unique > identifier of some sort to new tablets (and even if we could, that > doesn't help anyone today), then we'll just need to make the by-path > configuration more friendly to users somehow... > > hmm. > > We should be able to concoct a reasonable by-path rule specific > to wacom devices that auto-configuration tools can fairly easily > pattern match, but which will remain the same until the device > is moved to a different port... > > It will probably depend on a relatively recent udev, but starting > from the set of rules I posted earlier, try something like: > > ----- psuedo patch ---- > # Convenience link for the common case of a single tablet. > KERNEL=="event[0-9]*", SYSFS{idVendor}=="056a", SYMLINK="input/wacom" > > + # You probably won't need this IMPORT rule, since an earlier file is > + # likely to have already done it, but the rule that follows it does > + # depend on that having happened to set ID_PATH. > + # IMPORT{program}="path_id %p" > + > + # Port specific link for users of multiple tablets of the same type. > + KERNEL=="event[0-9]*", SYSFS{idVendor}=="056a", ENV{ID_PATH}=="?*", > SYMLINK+="input/by-path/$env{ID_PATH}-wacom" > > # Type-named links for multiple tablets. If you want to use multiple > ----------------------- > > Here this gives me (for one tablet): > /dev/input/by-path/pci-0000:00:02.0-usb-0:4:1.0-wacom -> ../event4 > /dev/input/by-path/pci-0000:00:02.1-usb-0:4:1.0-wacom -> ../event4 > > ... when I plug it into two different usb ports. Should work much > the same if you try it with two tablets. If nobody here can get it > to wreck their system then I can include something like it for the > next upload. > > It's not a great answer, but it adds a constant factor to the device > name that should survive a simple suspend or reboot reliably enough, > and lets people search for *-wacom instead of having to figure out > which *-event devices they want from that lot -- which is a bit more > friendly than having the entire name be like black magic to them... > > If anyone can think of a better way to do this I'd love to hear it. > > Cheers, > Ron ------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ _______________________________________________ Linuxwacom-discuss mailing list Linuxwacom-discuss@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linuxwacom-discuss
