On Thu, Jan 16, 2003 at 12:02:06PM -0600, Bryan W. Headley wrote: > HID guys, > > I took over support of the Aiptek tablet HID driver from Chris Atenasio. > Things have been difficult insofar as I don't completely understand the > HID spec, and I've been working from an incomplete hardware spec. (Other > than that, everything is good!) > > 1. I have some new Aiptek product IDs for usbutils. My problem is, > there is not a clearcut definition of which tablets use which product > Ids. Which is why, if I want to know which model is attached, I have to > ask the tablet for the size of it's drawing area. Their Windows driver > has the same problem.
In that case the IDs may be related to firmware revision or similar. The usb.ids file will list what it can - worst case just 'Aiptek Tablet'. Send the IDs to me. > 2. I have questions about how the HID interrupt callback routine > works (aka, wacom_irq). I'm currently getting stylus information passed > to me. However, that information (encapsulated in report 1) doesn't have > information about the macro keys. Those come from a different report. Then you'll get report 2 or some other. By the way, the 'best' way to implement the Aiptek support would be to fit it into the HID driver (hid-core.c et al) in a similar way the HID ff drivers fit in (lgff, hid-ff) in 2.5. Because the Aiptek table is HID after all, with some extensions. > I don't see anything in the driver the specifically registers > report 1 to come in through the interrupt callback. But assuming that I > could have reports 1 and 4 use interrupt callbacks, how do I know which > report is being served? E.g., is the report id embedded in the urb > structure? You don't register for any specific report, you should be getting all of what the device sends.. > 3. Is there any way to communicate to the driver? For example, I want > to programmatically ask for relative coordinates, or absolute > coordinates. What's the best way? Right now, I'm thinking about making a > file available somewhere in the procfs. Is that good? Why? Absolute coordinates should be enough - they can be trivially converted into relative ones by the application. -- Vojtech Pavlik SuSE Labs ------------------------------------------------------- This SF.NET email is sponsored by: SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See! http://www.vasoftware.com _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
