[Nut-upsuser] Problems with Unitek Alpha 1000 Ps

[Arjen de Korte]

Citeren Jordi Moreno <jmoreno at cim.es>:

> I'm attaching usbsnoop.log.tgz. I've tried to keep the log as little  
> as possible, only logging Winpower agent's start.
>
> I hope it will help...

It does. This looks like a Q1/megatec protocol UPS, unlike what I  
expected based on earlier megatec logs:

out: 46 0d 4d 6f f1 cf 11 88
      F  \r
in : 23 32 33 30 2e 30 20 30 30 30 20 30 31 32 2e 30 20 35 30 2e 30 0d 00 00
      #  2  3  0  .  0  sp 0  0  0  sp 0  1  2  .  0  sp 5  0  .  0  \r

out: 51 31 0d 6f f1 cf 11 88
      Q  1  \r
in : 28 32 31 36 2e 30 20 32 31 38 2e 30 20 32 31 31 2e 30 20 30 30 30 20 34
      (  2  1  6  .  0  sp 2  1  8  .  0  sp 2  1  1  .  0  sp 0  0  0  sp 4
      39 2e 39 20 31 33 2e 36 20 32 34 2e 30 20 30 30 30 30 31 30 30 30 0d 00
      9  .  9  sp 1  3  .  6  sp 2  4  .  0  sp 0  0  0  0  1  0  0  0  \r

So based on this log, I would have expected that the UPS would be  
supported by the megatec_usb driver out of the box. Going back to your  
first post, it looks like the megatec_usb driver only read the last 8  
bytes of the answer to the Q1\r command. The reason could be the  
somewhat dodgy way of reading the reply from the interrupt report in  
this driver. It really doesn't handle unexpected lengths correctly.  
Could you try the latest bleeding edge version from the SVN trunk  
instead? There is a new driver (blazer_usb) that deals with this better.

Best regards, Arjen
-- 
Please keep list traffic on the list