On Jul 7, 2013, at 6:09 AM, Alf Høgemark wrote:

> Do other NUT users miss knowing exactly how the upsc output will look for a 
> UPS model ?
> Does anyone have opinions on where we could record such information ? Should 
> it be one page per model, or one page per driver with a section for each 
> model ?

It would be nice to collect this information somehow. At the moment, we only 
really have upsc information for the UPS models which have been problematic, 
plus a few more emails where users have sent HCL updates.

The exact organization of the content probably doesn't matter, so long as the 
output is cross-referenced with a driver. (In the Thinkwiki example, their use 
of categories seemed backwards to me at first. However, it is useful to know 
whether a given laptop supports a particular feature, and the category listing 
at the bottom provides that nicely.)

I would be curious to hear what others think about a wiki or similar 
user-editable resource. The NUT HCL was never meant to fill that role, but a 
completely free-form wiki might not be as useful as a custom web site. In 
either case, we would need to have some decent spam protection. The NUT Trac 
instance was hit fairly hard a while ago.

Mike brings up an important point, in that NUT can only relay the information 
that the UPS provides. If that information is incorrect, this sort of 
repository could flag those variables or instant commands as being improperly 
implemented by the UPS.

Perhaps we could do something like those online UPS sizing wizards: ask the 
user which features matter most, and show a filtered list of suggested UPS 
models.

The biggest issue in my mind is that UPS vendors frequently change things out 
under the hood without changing the model name. Sometimes it is for the better 
- Powercom shifted to a standard USB HID PDC interface (which was easier to 
support in NUT). However, APC apparently dropped support for a lot of variables 
from their HID PDC interface in favor of their proprietary Microlink protocol 
(see apcupsd.com for more information).

Maybe we can steer people away from bad UPS models or vendors, but in the end, 
if I were purchasing a new UPS, I would want a no-questions-asked, open box 
refund policy from the distributor.

Alf: I like this idea, but I don't have the free time to help with this for a 
while. (We still need to get the first Git-based source release out the door.) 
Feel free to take this idea and run with it.

-- 
Charles Lepple
clepple@gmail




_______________________________________________
Nut-upsuser mailing list
Nut-upsuser@lists.alioth.debian.org
http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/nut-upsuser

Reply via email to