[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: [RFR] templates://nut/{nut.templates}



Hi Justin,

2009/3/18 Justin B Rye <jbr@edlug.org.uk>
Arnaud Quette wrote:
> a 2nd one to have in mind for the Desc field, since I haven't had time to work
> on this specific point:
> prior to 2.4, NUT was only support UPSs. As of 2.4, it also supports PDUs
> (either through the nut-snmp package or nut-powerman-pdu). You might want to
> integrate that point.

Perhaps by amending the boilerplate like this:

   NUT is a client/server monitoring system that allows computers to share
-   uninterruptible power supply (UPS) hardware. The server monitors the UPS
-   and notifies its clients when the UPS is on or has a low battery.
+   uninterruptible power supply (UPS) and power distribution unit (PDU)
+   hardware. The server monitors the UPS and notifies its clients when the
+   UPS is on or has a low battery.

As a knock-on effect, nut-powerman-pdu wouldn't need to expand "PDU":
   This package provides powerman-pdu, which allows NUT clients to communicate
-   with the PowerMan daemon to support Power Distribution Units.
+   with the PowerMan daemon to support PDUs.

But meanwhile it occurs to me that "when the UPS is on" in the
boilerplate may not be the best way to put it.  The UPS is never
"off"; the question is whether it's... in use?  Active?  What's the
word?

you're misreading, but the sentence is not so well formed too:
you should read "is on battery or has a low battery".

and an UPS can be off. you can even monitor some UPSs when they're off.
in this case, it's simply that the UPS delivers no power on its outputs, but it still can be running and reachable.
that can be quite useful since it's a more reliable way to wake up a bunch of computers than WOL (wake on lan)

I propose the following, which is more generic:

   NUT is a client/server monitoring system that allows computers to share
-   uninterruptible power supply (UPS) hardware. The server monitors the UPS
-   and notifies its clients when the UPS is on or has a low battery.
+   uninterruptible power supply (UPS) and power distribution unit (PDU)
+   hardware. Clients access the hardware through the server, and are notified
+   whenever the power status changes.

I also agree with your PDU expansion remark.

> A last one for the spelling case of NUT: please use Network UPS Tools with
> capitalised first letter.

(In synopsis lines, that is.)  The idea is that as well as being the
software's full name, "network UPS tools" also happens to work as a
self-description, and descriptive phrases in the package synopsis
don't need capital letters.  On the other hand of course
capitalising it would be an extra hint that the words are an
expansion of the acronym.

I suppose it also makes a difference to translators; are they
expected to preserve the English initialism, or even perhaps the
entire English name, or do they translate the short description into
"outils d'ASI de réseau" or whatever it is?

yep, that's the point:
this is really (both NUT and the expansion Network UPS Tools) the name of the project, and should not be translated.

but I'm fine with your remark and Christian's one, that is having the synopsis with low cap. and the boilerplate with the upper

>> Please read /usr/share/doc/nut/UPGRADING.gz for the upgrading procedure.
>
> (This is outside d-l-e jurisdiction, but shouldn't it be somewhere
> that apt-listchanges will see?)
>
> note that it's also referenced in README.Debian.

Yes, and that file _isn't_ monitored by apt-listchanges.  Still, I
imagine it's mentioned in the changelog.

right, my fault there. I should already be using NEWS.Debian for long.
I'll start using it with the 2.4 branch on the next upload.

>>  Package: nut-cgi
>>  Recommends: apache | httpd
>
> (Surely httpd-cgi?)
>
> ? not sure to understand your remark here.

Compare the web server "boa" (which "Provides: httpd, httpd-cgi")
with "micro-httpd" (which only "Provides: httpd").  Will nut-cgi
work on the web servers that don't support CGI?  My guess is no.

oh, great. I've completely missed that one. thanks for pointing it.
yep, httpd-cgi better fits the purpose here.

cheers,
-- Arnaud

Reply to: