[Qgis-user] Stress about release plans

[AntonioLocandro]

Why QGIS is in such a need to release every 4/6 months? It's obvious the
project doesn't have enough resources and I find a bit troublesome that
developers in the period add tons of new features yet bug squashes are
relatively small and left basically for the last week to be covered by
funding. 

Just a thought here but if billion dollar company having thousands of
workers releases just one major version (with actual new features and
refactoring) and one or two minor releases (mainly bug fixes) why should
QGIS try to do something like the current situation.

I think there should be a reality check by QGIS PSC, developers and users
about what is wanted and what can be actually met with resources, adding
more funding won't solve the isssue, getting more people involved will. But
getting more people involved when they haven't yet master one version when
you already have the next one (let's be honest it takes people time to get
used to software) 

QGIS should settle on one major release per year containing new features and
code refactoring and maybe one minor release containing mostly bug squashing
(have you seen the now getting longer and longer as we speak bug list?) ,
this will mean documentation can be updated more easily and should be
documented as they are added to the major release

Major release 2.6 
Minor  release 2.6.1 (6 months from 2.6 this version contaings only bug
fixes)
Major release 2.8 (12 months from 2.6 this version contains new features)

developers should document what they do at least a small summary. I disagree
that having features added without documentation is better than not having
them because undocumented unsupported features are not as widely used.

One final thought, do the QGIS PSC have statistics of which version are the
biggest funders (which I assume will be the biggest users that can put $$$
in the project) using? I will be very worried if statistics showed them
using <=2.0 in great proportion which means they find newer versions either
not well tested/new bugs or breaking their workflow

My very long two cents



Régis Haubourg wrote
> 
> Larry_S wrote
>> I propose just extending the current 4 month release scenario to ** 6
>> months **, i.e. just two releases a year. With all dev and freeze cycles
>> expanding proportionally.
> Again, I fully agree with Larry, and all this was already discussed in 2.4
> prerelease period.
> 
>  By myself, I contribute as much as I can (fund features, bugsquashing,
> fund doc translation by trainers, testing, translating at night when
> babies sleep.. ). I still try to find long lasting ways of sponsoring and
> having a support that help us in release candidate debug sprint . 
> Anyway, I will never be able to hire here another GIS administrator to be
> able to test next release when I still work on deploying the current one.
> I think the "stress" in this post title is what I felt when not being able
> to participate to 2.4 tests and debug, and that trainers feel when they
> upgrade their courses. 
> 
> 6 months would be far enough, and is still a very fast release cycle. When
> I see how stable is 2.4, I think we can be confident that in a 6 month
> cycle, we will be ok, and we will get back some workers on testing,
> bugfixing, and documenting.. 
> 
> Régis





--
View this message in context: http://osgeo-org.1560.x6.nabble.com/Stress-about-release-plans-tp5152189p5152805.html
Sent from the Quantum GIS - User mailing list archive at Nabble.com.