-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Mon, 14 Oct 2002 03:28, Lars Doelle wrote: > I'm currently specifying the comminication protocol for a usb controlled TV > set. It appears as if something along the lines of the monitor class could > be used. It is for channel setting, loudness, power control, and all that > kind of stuff. > > After a glance over the monitor class, it seems to be the most standard > conformant way. Is that the right way to go? Or, do you consider the > monitor class unsuited for my intention and advice to better do some free > style vendor specific interface? Vendor specific isn't usually need with HID - there are more than enough options to choose from! Please understand that the "monitor class" isn't really an exclusive domain. It is a set of additional pages for HID. There isn't really any reason why you couldn't use some additional pages from HUT if absolutely needed.
> For the userland, I thought the hiddev driver might be fine, but a look > into the source showed that "write" is implemented as "return -EINVAL". > That's strange for me. Is there any deep intention behind it or is it just > an omission i should fill in? Or, as another alternative should i better go > and do a user space driver via libusb instead? You'd likely be better off wth hiddev, although I'm possibly not unbiased. You can't really just "write" an event to the HID interpreter. You have to set it up first, hence an ioctl() interface. I recently posted a reference to some code that does hiddev output (to control LEDs on a keyboard). Ah, here we go: http://www.frogmouth.net/hid-doco/examples/hiddev-misc/write-events.c There is some comments in the code (which is really ugly - it was a 30 minute hack of some other code to read hiddev descriptors) explaining the ioctl()s you need. > What basically confuses me most is that the HID class appears throughout > being considered as a input-_only_ thingy in the linux drivers. Now how i > understand hid, it is well intended to control a hid device, also. Again my > question whether not having implemented writable devices is a casual > omission or something really intended due your deeper usb understanding. It does output, just not write(). Brad - -- http://linux.conf.au. 22-25Jan2003. Perth, Aust. I'm registered. Are you? -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.0.6 (GNU/Linux) Comment: For info see http://www.gnupg.org iD8DBQE9qd++W6pHgIdAuOMRAuU4AKCRtr2gFPIS6gxi7uhiMb0kjzyGRQCgvdjm 0SZ40I2+jMu9U/uF5/XtBUg= =3KF/ -----END PGP SIGNATURE----- ------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
