[Nut-upsdev] about nut revision 1289

[Alexander Gordeev]

On Monday 08 June 2009 22:38:26 Arjen de Korte wrote:
> Citeren Alexander Gordeev <lasaine at lvk.cs.msu.su>:
> >> But in the remote case those lines really do something, removing them
> >> will at most cause some glitches in some single model that only one
> >> person has. So I guess if they are causing problems with megatec_usb
> >> (and USB models are much more popular nowadays), it is better to just
> >> remove them.
> >
> > Thanks!
> > So I'll remove them and we'll see if anyone complains. :)
>
> I hereby do (and also exercise my authority of Prime NUT Senior
> Lieutenant to enforce this... :-)
>
> 1) What functionality is there in megatec_usb that isn't also present
> in blazer_usb? Remember that the first is scheduled to be replaced by
> the latter in the future.
>
> 2) If the above isn't feasible for now, consider reworking the megatec
> driver to be split in an interface independent core driver and two
> subdrivers that handle the serial and USB commnunication respectively.
> Something along the lines of the blazer core and blazer-ser and
> blazer-usb subdrivers. By doing so, you cleanly separate these
> interfaces, while keeping the common stuff neatly packages in one place.

Sure, you are right and blazer_* should replace megatec* and I don't want to 
reinvent the wheel redoing what you have already done. But anyway I wanted to 
fix this bug just because it exists. We can do this as well by throwing 
megatec* away. :) I'm ok with both ways. But IMO we shouldn't throw it away 
until the next release so that we can prepare the users to the transition.

-- 
  Alexander