Nut upsdev - Jun 2016 - updating the NUT website

Quick question, mostly for Arnaud and Daniele: How do we want to handle updates
to the website between NUT releases?

For instance, we have a new NodeJS client, and updates to the VMware package,
and those aren't tightly coupled to the released versions of NUT. However,
if we add them to the NEWS file in the NUT repository, then that pulls in any
other changes from that branch.

Does it make sense to create a NUT branch based off of the 2.7.4 release, and
put all of the NEWS/download updates there? We can then merge that back into
master as needed.

-- 
- Charles Lepple
https://ghz.cc/charles/

hmm..
In my opinion, since we try to keep the master branch of our main git
repo (nut) always up-to-date (i.e. we don't do development in a
separate branch and only merge it back into master when readying a new
release) and in a working condition, we should not keep the website
frozen between releases, we should, instead, fetch and publish updates
when needed: the website will then reflect the status of our
officially official (albeit not yet part of a set-in-stone release)
code -- if anyone wants a snapshot of the website at the time of a
given release, they it can be built from tags.
(..and, being the personification of laziness, I'd rather prefer to
avoid any possible conflict that a separate branch could generate when
merged back into master)

Hi folks,

Dan is right, and this approach is the best for us to minimize the overhead.
There may however be some points of improvement to give the right
visibility to users, and avoid the confusion of "things being visible on
the website though not yet published as a release".
not much details from my side on this for now, though...

cheers,
Arno


2016-06-07 0:50 GMT+02:00 hyouko at gmail.com <hyouko at gmail.com>:
> hmm..
> In my opinion, since we try to keep the master branch of our main git
> repo (nut) always up-to-date (i.e. we don't do development in a
> separate branch and only merge it back into master when readying a new
> release) and in a working condition, we should not keep the website
> frozen between releases, we should, instead, fetch and publish updates
> when needed: the website will then reflect the status of our
> officially official (albeit not yet part of a set-in-stone release)
> code -- if anyone wants a snapshot of the website at the time of a
> given release, they it can be built from tags.
> (..and, being the personification of laziness, I'd rather prefer to
> avoid any possible conflict that a separate branch could generate when
> merged back into master)
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.alioth.debian.org/pipermail/nut-upsdev/attachments/20160607/f9f73f19/attachment.html>