On Sun, Aug 02, 2009 at 10:21:50PM +0200, Henning Glawe wrote:
> > BTW: Hopefully the current kernel provides enough support for devices
> > which implement 802.15.4 funcionality in the hardware (i.e. see the
> > drivers/ieee802154/fakehard.c in the mainline kernel). Maybe you can
> > look into implementing basic support for your remote? If you have any
> > questions regarding driver, you can ask them here.
> 
> Yes, that's exacty what I am up to; I have a small test program in userspace
> ready, which is setting up the dongle correctly and is able to reveive
> keypresses from the SRM7500.
> But before going into the complications of accessing the remote's LCD, I'm
> going to write a 'proper' dongle driver for your 802.15.4 stack.
> 
> 
> > The major part that is still missing is stable API MIB access
> > (via SET/GET). We don't have a common view, how this should be
> > implemented in a clean way.
> 
> I was also thinking about this, evenmore because the current API doesn't
> implement all the MIB set commands I need for my project:
> a generic 'mlme-set.request' would be nice to have, instead of one command
> for each attribute. but on the other hand, introducing this with a char
> pointer as 'value' makes it annoying to use, as most of the MIB attributes
> are either one or two bytes long...

Yep. There are only 2 or 3 MIB elements (IIRC) that are greater than 2
bytes.

I thought about adding special union (of bool, byte, word, etc.).
However I never implemented this, as I was burried beneath other topics
and direct MIB management wasn't very important to us (ATM of course).

I'll think about this, but I don't promise that I'll come with a
solution right now.

Another problem that I can think about is a COMM-STATUS.indication
primitive. Although it's there in standard, I'd strongly oppose
imlementing it as is. Maybe it's me, but I find it a bit confusing to
have one single callback for all problems. Maybe I'll change my mind,
but currently you'd better leave it out.

> 
> Do you have a reference about the state of discussion?

No, unfortunately. Most of our discussions are either spoken-only as
most (well, practically all ATM) do seet in one room, so no real need
or much paperwork/typeing, sorry.

-- 
With best wishes
Dmitry


------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with 
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
_______________________________________________
Linux-zigbee-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-zigbee-devel

Reply via email to