Nut upsdev - Dec 2006 - megatec over USB - using Andrey Lelikov's approach

Hi All,

finally, I've got time and resources for development. Unfortunately I  
couldn't
afford playing with my UPS because it was used permanently, but now it's  
ok.

I've checked out cureent trunc and I see that Andrey Lelikov's changes  
haven't
been applied yet. What's wrong with them?
I think, he made a great job to simplify future support for this kind of  
UPSes.
His patch is outdated now but if you don't mind I can try to merge it to  
the
current code. Or maybe Andrey wants to do it himself?

If you consider this change of the megatec driver harmfull I'll write  
separate
driver, basing on his code.

--
   Alexander

On 12/30/06, Alexander I. Gordeev <lasaine@lvk.cs.msu.su>
wrote:
> Hi All,
>
> finally, I've got time and resources for development. Unfortunately I
> couldn't
> afford playing with my UPS because it was used permanently, but now
it's
> ok.
>
> I've checked out cureent trunc and I see that Andrey Lelikov's
changes
> haven't
> been applied yet. What's wrong with them?
> I think, he made a great job to simplify future support for this kind of
> UPSes.
> His patch is outdated now but if you don't mind I can try to merge it
to
> the
> current code. Or maybe Andrey wants to do it himself?
>
> If you consider this change of the megatec driver harmfull I'll write
> separate
> driver, basing on his code.
The changes aren't harmful, but like Arjen suggested, the best
approach is to make the USB communications layer have the same source
level API as the serial layer. Then, the driver only has to be linked
with either "serial.o" or, say, "usb.o" to create the two
different
versions. This has the added benefit of maybe being useful for other
drivers as well.

-- 
Carlos Rodrigues

After Andrey sent his patch, Arjen requested some changes that Andrey
agreed to but never made. So the patch is in limbo. -- Peter

Alexander I. Gordeev wrote:
> 
> Hi All,
> 
> finally, I've got time and resources for development. Unfortunately I  
> couldn't
> afford playing with my UPS because it was used permanently, but now
it's
> ok.
> 
> I've checked out cureent trunc and I see that Andrey Lelikov's
changes
> haven't
> been applied yet. What's wrong with them?
> I think, he made a great job to simplify future support for this kind of  
> UPSes.
> His patch is outdated now but if you don't mind I can try to merge it
to
> the
> current code. Or maybe Andrey wants to do it himself?
> 
> If you consider this change of the megatec driver harmfull I'll write  
> separate
> driver, basing on his code.
> 
> --
>    Alexander
> 
> _______________________________________________
> Nut-upsdev mailing list
> Nut-upsdev@lists.alioth.debian.org
> http://lists.alioth.debian.org/mailman/listinfo/nut-upsdev
>