Charles Lepple
2008-Jan-13 14:45 UTC
[Nut-upsuser] new features on Testing branch (was: Belkin F6H375 not seen by nut 2.2.1)
[moving this thread to nut-upsdev] On Jan 12, 2008 5:54 PM, Alexander I. Gordeev <lasaine at lvk.cs.msu.su> wrote:> On Sun, 13 Jan 2008 01:30:37 +0300, Charles Lepple <clepple at gmail.com> > wrote: > > when you figure this out, can we make sure that we know which > > changesets would need to be back-ported to branches/Testing (since > > that is where we would create version 2.2.2+, if needed)? Apologies in > > advance, I do not know much about this particular driver. > > > > Hmm, can we add new features to this branch or only bugfixes are allowed? > There is only one new feature in the trunk, reconnect support. Everything > else is code cleanup and bugfixing. I'd rather backport everything.I'm cc:ing Arnaud, (Arjen, feel free to weigh in as well) but when a driver is developed based on experimentation rather than a well-defined protocol specification, I personally think that the advantages of "releasing early and often" outweigh the disadvantages of keeping the changes only in the trunk. It looks like the megatec_usb driver in 2.2.0 and 2.2.1 does not support the Phoenixtec protocol, but that shouldn't affect current users of megatec_usb. Also, given that the reconnection feature has been somewhat tested in other drivers, it is not as risky as introducing something completely new. -- - Charles Lepple
Charles Lepple
2008-Jan-13 16:28 UTC
[Nut-upsdev] new features on Testing branch (was: Belkin F6H375 not seen by nut 2.2.1)
apparently I can't manage simple skills like changing a list address. ---------- Forwarded message ---------- From: Charles Lepple <clepple at gmail.com> Date: Jan 13, 2008 9:45 AM Subject: [Nut-upsuser] new features on Testing branch (was: Belkin F6H375 not seen by nut 2.2.1) To: "Alexander I. Gordeev" <lasaine at lvk.cs.msu.su> Cc: nut-upsuser at lists.alioth.debian.org [moving this thread to nut-upsdev] On Jan 12, 2008 5:54 PM, Alexander I. Gordeev <lasaine at lvk.cs.msu.su> wrote:> On Sun, 13 Jan 2008 01:30:37 +0300, Charles Lepple <clepple at gmail.com> > wrote: > > when you figure this out, can we make sure that we know which > > changesets would need to be back-ported to branches/Testing (since > > that is where we would create version 2.2.2+, if needed)? Apologies in > > advance, I do not know much about this particular driver. > > > > Hmm, can we add new features to this branch or only bugfixes are allowed? > There is only one new feature in the trunk, reconnect support. Everything > else is code cleanup and bugfixing. I'd rather backport everything.I'm cc:ing Arnaud, (Arjen, feel free to weigh in as well) but when a driver is developed based on experimentation rather than a well-defined protocol specification, I personally think that the advantages of "releasing early and often" outweigh the disadvantages of keeping the changes only in the trunk. It looks like the megatec_usb driver in 2.2.0 and 2.2.1 does not support the Phoenixtec protocol, but that shouldn't affect current users of megatec_usb. Also, given that the reconnection feature has been somewhat tested in other drivers, it is not as risky as introducing something completely new. -- - Charles Lepple
Arnaud Quette
2008-Jan-14 08:41 UTC
[Nut-upsuser] new features on Testing branch (was: Belkin F6H375 not seen by nut 2.2.1)
fellows, 2008/1/13, Charles Lepple <clepple at gmail.com>:> [moving this thread to nut-upsdev] > > On Jan 12, 2008 5:54 PM, Alexander I. Gordeev <lasaine at lvk.cs.msu.su> wrote: > > On Sun, 13 Jan 2008 01:30:37 +0300, Charles Lepple <clepple at gmail.com> > > wrote: > > > when you figure this out, can we make sure that we know which > > > changesets would need to be back-ported to branches/Testing (since > > > that is where we would create version 2.2.2+, if needed)? Apologies in > > > advance, I do not know much about this particular driver. > > > > > > > Hmm, can we add new features to this branch or only bugfixes are allowed? > > There is only one new feature in the trunk, reconnect support. Everything > > else is code cleanup and bugfixing. I'd rather backport everything. > > I'm cc:ing Arnaud, (Arjen, feel free to weigh in as well) but when a > driver is developed based on experimentation rather than a > well-defined protocol specification, I personally think that the > advantages of "releasing early and often" outweigh the disadvantages > of keeping the changes only in the trunk. It looks like the > megatec_usb driver in 2.2.0 and 2.2.1 does not support the Phoenixtec > protocol, but that shouldn't affect current users of megatec_usb. > > Also, given that the reconnection feature has been somewhat tested in > other drivers, it is not as risky as introducing something completely > new.As I've told several times, I'm ok with Charles' above statement. Testing is primarily focused on bugfixes and improvements. But new drivers are welcome, along with drivers enhancements that doesn't affect the current behavior or support perimeter. This is the "release early, release often" principle that allows us to improve our hardware support and to please our users. Arnaud -- Linux / Unix Expert R&D - MGE Office Protection Systems - http://www.mgeops.com Network UPS Tools (NUT) Project Leader - http://www.networkupstools.org/ Debian Developer - http://people.debian.org/~aquette/ Free Software Developer - http://arnaud.quette.free.fr/