Stable releases of Bazaar?

[David Ingamells]

Martin Pool wrote:
> On Fri, Jul 25, 2008 at 5:41 PM, David Ingamells
> <david.ingamells at mapscape.eu> wrote:
>   
>> Jonathan Lange wrote:
>>     
>>> On Fri, Jul 25, 2008 at 1:16 PM, Martin Pool <mbp at canonical.com> wrote:
>>>
>>>       
>>>> For people to be happily using stacking with launchpad we need them to
>>>> install 1.6.
>>>>
>>>>         
>>> Or perhaps 1.7. The 1.6 release changes a lot of things and is
>>> therefore likely to have more than its share of bugs.
>>>
>>>
>>>       
>> Has the Bazaar team considered identifying certain releases as stable
>> releases which will enjoy ongoing support with bug fixes being applied?
>>
>> Bazaar is now mature enough with a large enough user base for this to be
>> beneficial. I agree with Jonathan's  remark above (without any criticism of
>> the Bazaar team intended): I am reluctant to move to 1.6 because of the
>> number of major changes that will appear in this release, however I am
>> suffering from some problems with 1.5 (with bzr+ssh://) that are likely
>> solved in 1.6. I therefore have a quandary.
>>
>> I am very aware of the burden added by stable release support as that is one
>> of my roles within my team (I often have to integrate the same fix into 3
>> different releases); the benefit is a stable environment for the users with
>> the ability to control when new features are rolled out.
>>     
>
> Hi David,
>
> It's an interesting point; thanks for raising it.
>
> There would definitely be a benefit for some users in having a stable
> branch which gets only serious bug fixes and no new features.  It is
> always going to be a bit of a compromise as to whether a particular
> change's risk:reward tradeoff is worth putting into there, not to
> mention the tradeoff of putting time into the backport vs new work.
>
> Perhaps you'd be interested in helping maintain a backported stable
> branch - at least in testing it and perhaps nominating/merging things
> to go in?
>
> There is another thread about doing a 1.3.2 to go into hardy-updates
> and perhaps that would be good for the generally available stable
> branch too.
>
>   
You've called me out there ;-)

I've afraid it's a case of "the mind is willing but the day is too short".
I'm already overloaded in my day job, I'm already maintainer of another 
GNU project and my wife complains I sit too long behind the computer. 
Also I am a complete dumbo wrt Python. I've been able to spend a few 
hours benchmarking recently only because half of the office is away on 
annual leave (especially the testers being away makes a big difference 
to my workload ;-) )

So I'll have to say sorry on that one.