[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [rdiff-backup-users] What is the current rdiff-backup version?
From: |
Frank Crawford |
Subject: |
Re: [rdiff-backup-users] What is the current rdiff-backup version? |
Date: |
Sun, 08 Dec 2013 14:08:25 +1100 |
On Sun, 2013-12-08 at 02:34 +0000, Edward Ned Harvey (rdiff-backup)
wrote:
> > From: rdiff-backup-users-bounces+rdiff-
> > address@hidden [mailto:rdiff-backup-users-
> > address@hidden On Behalf Of Frank
> > Crawford
> >
> > Firstly, what version of rdiff-backup do most people use? There is the
> > stable 1.2.8 and unstable 1.3.3, but both date back to 2009. From what
> > I can see most distributions use the stable version, so has anyone
> > extensively tested the 1.3.3 release, and is it really stable enough to
> > promote to say 1.4?
>
> Well, as you said, the version currently considered stable is 1.2.8. The
> 1.3.x releases, back when Andrew was working on it... As far as I can tell,
> the milestone was more symbolic than anything, but nobody's run the
> regressions in a very long time, and I have to presume they didn't call it
> stable because there were some features in development, or regressions that
> weren't passing...
>
> In order to call any particular rev "stable" I think we'll agree some
> substantive unit testing must pass. Which implies figuring out how to run
> the tests. And writing tests to test previously untested (or un-passed)
> features.
The only addition to that I'd add would be that if a large group of
people have been running it for a number of years, you may also consider
making it stable. However, if hardly anyone has run it, then it would
still be a dev version.
> I certainly have no deep conviction for a new release to be called 1.3.4 vs
> 1.4.0. Given the lapse in development, it might make sense to go to 1.4.0,
> but ultimately I think it depends on the work that's taking place. I think
> 1.3.4 implies testing & bugfix for existing features. I think 1.4.0 implies
> new features. And as soon as either one passes what we generally consider a
> solid regression sequence, we can call it "stable."
>
I thought of going to 1.4.0 really to indicate additional work now going
on. Of course we would then need to continue on with that work.
> > Secondly, a couple of people have mentioned their own private patches.
> > Are these collected together anywhere, and if not, should we do that?
> > We may even be able to agree to merge them into the mainline.
>
> I don't really know anything about that...
>
Looking at the Fedora RPM there seem to be two patches included. I
don't know how big or small they are, but they exist. I know Joe Steele
mentioned a patch attached to a bug report he sent in a few years ago.
There may be a few more around.
Regards
Frank