[Nut-upsuser] Megatec and Belkin F6H550-UPS

[Arjen de Korte]

>> We don't need this, since there would be no way to autodetect this UPS
>> anyway.
>
> IMHO, it would be very useful to still gather this information from
> people and create
> a database of such UPS on website/wiki or even in configs, so people could
> just
> peek them and use, without doing all this scary experiments with full
> discharging.

I understand that you're volunteering to setup such a database/wiki?
Please do, this has been discussed many times before, but so far nobody
has taken up this task... :-)

> It is just matter of user-friendliness and simplicity. Just imagine -
> you're forcing
> every user of such UPS to do this experiment, spending time setting up
> hardware
> for this, capturing logs, etc, while we could minimize this useless
> time spending
> by taking values from one user and publishing them.

For your information, although the idea behind the battery charge
calculation was made with the best intentions, it was flawed from the
start. The way the battery charge is calculated by the megatec driver is
better than nothing (but not much) and really can't reliably be used to
guesstimate the runtime remaining. We've wasted far too much time on this
already and I really think it is time to lay this finally to rest.
Therefor, the values have little to no use to us anymore.

> Jayson, please, post results of your test here. Even if developers won't
> go with
> creating database I described, your values still will be available for
> people through
> mailing list archives. :)

I would say, don't waste your time on determining these values. For most
(if not all) applications, the values need to be determined on a
system-by-system basis to have some meaning.

If one really needs a somewhat reliable guesstimate of the runtime
remaining, you should steer clear of UPSes that use the Megatec protocol,
since these systems are lacking some important things that are required to
make this of any use. I have written about this several times in the past,
both on the nut-user and nut-devel mailinglists.

Best regards, Arjen