[Nut-upsdev] Megatec_USB on OpenBSD 4.3 - no input interrupt endpoint

[ng-sup01]

Hello again,

   and thanks for your quick replies.

Arjen de Korte wrote:
>
>>     And, finally, here's a clip of my ups.conf file:
>
> [...]
>
>> [atlantisland]
>>     driver = usbhid_ups
>>     port = /dev/usb0
>>     desc = "Atlantis Land 1501"
>
> Use the correct the driver, 'usbhid_ups' != 'megatec_usb'. Looking at 
> the output on your Linux system, you'd need the latter.
>
Yes, and don't post if you didn't get a good night's sleep. This morning 
(around 3AM) in despair, I was fooling around with the .conf, trying the 
usbhid_ups driver as well. The ups.conf file reflects the result of said 
ramblings...
>>    Please note I tried all possibile values for "port": /dev/uhid0,
>> /dev/uhidev0 and /dev/usb0
>
> The value of 'port' is irrelevant, as long as it's not empty. See the 
> FAQ.
Missed that one.
>
> Best regards, Arjen
>
> PS  This question would have been more appropriate on the nut-upsusers 
> list.
Posted here because I remember seeing prior messages concerning same UPS 
that were moved from upsusers to upsdev: but I'm new, and might be very 
wrong. Please tell me what to do ;)


Henning Brauer wrote:

  > * ng-sup01 <ng-sup01 at mclink.it> [2008-09-12 12:24]:

> > uhidev0 at uhub0 port 1 configuration 1 interface 0 "UIS_ABLER USB to 
> > Serial" rev 1.00/1.00 addr 2
> that's the problem, uhiddev must not claim the device and leave it to
> ugen. 

   But the same version of NUT (2.2.2), compiled on an Ubuntu machine, works fine, while on OpenBSD I get this problem. Is this possible?

> you must add a quirk like

> { USB_VENDOR_APC, USB_PRODUCT_APC_UPSPRO500,       ANY,   {UQ_BAD_HID }},

Will try and post results.


Thanks to everybody for now!


--Vic