duplicity-team team mailing list archive

[Kenneth Loafman <kenneth@xxxxxxxxxxx>]

On Wed, Nov 20, 2013 at 7:45 AM, Michael Terry <mike@xxxxxxxxxxx> wrote:

> On 20 November 2013 05:49, <edgar.soldin@xxxxxx> wrote:
>
>> any reason why you sent this to the "inner circle" list but not to
>> duplicity talk for a wider audience?
>>
>
> Oh, I just figured this was maintainer talk.  Wasn't trying to hide
> anything.  We can move to duplicity-talk if that's better.
>

Works either way for me.  We might get more interest on the main list.
 Maybe.

Ken's decision, but i doubt he'd like the additional work load. aside from
>> the fact that contributors would have to provide two branches henceforth.
>>
>
> I imagine contributors would only ever develop against trunk (0.7).  In
> the rare occasion where a critical bug should be backported, we can
> directly apply the branch to the 0.6 branch and push out a new release.
> I'd hope this would be an at-most once a year thing.
>

At one time 0.7 was slated for PAR inclusion.  There was a partial
development on it, but it's seen no activity in a couple of years.


> btw. Ken, we should remove the 'beta' status from the website! duplicity
>> far from perfect is definitely not beta anymore.
>>
>
> Likewise, what's our criteria for dropping the leading 0 in the version?
> :)  Maybe we can jump from 0.6 to 1.0.  Or if that sounds too much like
> we're declaring "Mission Accomplished" maybe go from 0.6 to 7.0...  :)
>

It's time to go to either 0.7 or 1.0.  If we've really fixed the data
corruption bugs, then maybe 1.0.


> do future imports really work with any old python version? or are there
>> versions that e.g. didn't have the 'unicode_literals' .. so that using
>> future imports will enforce minimal minor versions for the python runtime
>> e.g. you have to install python 2.6.10 'cause 2.6.1 didn't have it at that
>> time?
>>
>
> Python wouldn't introduce new features like that in a point release.
> 'unicode_literals' and 'print_function' were both introduced in 2.6.
> That's the main reason to want the bump.
>

I thought one of the problems was librsync not having a presence on 3.x.
 There may be others missing.


> Ken: any idea about all the branches on launchpad? what to do with them?
>> keep them for reference?
>>
>
> Which branches?  They should all be able to stay around...
>
> There are a bunch of branches not owned by us.  The owners would have to
take the time to remove them, I think.

Some of them are already merged in, but Launchpad did not remove the
reference.  Has not been worth the time to chase down why.

...Ken