On Oct 25, 2009, at 10:51 AM, Anders Eriksson wrote:
> jarod@... said:
>> On Aug 31, 2009, at 8:16 AM, Anders Eriksson wrote:
>>> jarod@... said:
>>>> Was reasonably obvious why after reading over bit of code...
>>>> Setting the ir protocol requires a device with control urb
>>>> endpoints, which the ffdc devices don't have. I've fixed
>>>> the code in cvs so that it'll print a warning when someone tries to
>>>> set the proto on a device that doesn't support it, and
>>>> just fall back to imon proto.
>>> Hmmm. That's alarming.
>>> I've got an ffdc + Harmony525 and the only combo I got working
>>> was in MCE mode. That is, ir_protocol=1 and install MCE emulation on
>>> harmony. This is on a cvs lirc as of ca 3 weeks ago. >
>>> Any testing needed?
>> Yes, actually... Another user on the mythtv mailing lists
>> mentioned this
>> weekend that they had an ffdc device that they used exclusively
>> in mce
>> mode. It seems to be yet another case of SoundGraph putting the
>> same device
>> ID on wildly differing hardware. :\
>> *HOWEVER*, I believe the check added looks for a tx control
>> endpoint, and
>> if not found, then things fall back to imon proto. Since it
>> appears your
>> devices *does* have a tx control endpoint (lest the ir proto
>> setting never
>> would have took), it *should* still get past that check and
>> behave as
>> expected, but confirmation of that would be good.
> Well. I just upgraded to 0.8.6, and the code there concluded that my
> does not support mce mode. Under 0.8.5 it works fine in mce mode, so
> in some version is not telling the truth. Any HW dumps I can provide
> help sort this out? In the interim, I'll have to revert to 0.8.5 to
> keep the
> family happy...
Crap, just took another look at the code... No, your device doesn't
have a tx control endpoint. But setting the ir proto is NOT dependent
on that, I somehow got my wires crossed -- non-tx-control devices use
the usb_tx_buf via usb_fill_int_urb(). I'll fix that in cvs and my git
tree in the morning.
>> On a semi-related note, I'll add that I'm using my own imon
>> receiver with a
>> harmony 880 programmed as an Antec Veris or Antec Fusion remote
>> or some
>> such thing (i.e., native imon protocol), and its actually working
>> well. Most keys repeat more than I'd like, but someone from the
>> team is actually talking to Logitech about tightening up their
>> Antec remote
>> profile so this doesn't happen anymore. So currently, with the
>> aid of a ~/
>> .lircrc with some relatively high repeat= lines for each button,
>> I can use
>> both my harmony 880 and the stock RM200 remote simultaneously
>> (just can't
>> hold any buttons on the harmony for very long or I get extra
> That aspect is the thing that works consistently in mce mode on my
> 525. I
> allways get 3 repeats on all buttons. In imon mode, I got all sorts
> of repeats.
I've got some ideas on how to tighten things up a bit, but haven't got
around to actually implementing any of them yet, been swamped lately
(hrm, maybe that's actually the norm...)