What about release plans for 2.2.0?

[Christian Stimming]

What do we do with our release plans for a 2.2.0 release? We planned  
that release for the upcoming weekend [1], but currently the  
development has slowed significantly. Even more importantly, the PR  
Planning [2] has made almost zero progress, but we would need at least  
a week of work there before we have enough PR material for the Windows  
release.

To recall the rationale for a 2.2.0 release: We planned this release  
to turn our current "trunk" into a stable branch right before the GSoC  
students can start to integrate their work into trunk. The plan would  
be to have yet another stable series after all GSoC projects have  
finished successfully (end of summer) and we've integrated the new  
features into the overall product sufficiently (end of the year). This  
is IMHO the reason for the 2.2.0 at this time because if we didn't  
have it before GSoC, we won't have it before probably the end of the  
year, which in turn would be quite a very long time since 2.0.0 (July  
2006).

The only major new feature for 2.2.0 is the completed Windows port,  
and my hope is that this would attract significantly more users (and  
eventually developers) for the future. However, on the downside the  
release on Windows will also bring us many many new bugreports, and  
even worse, many of those bugreports that most of us won't be able to  
reproduce because most of us don't have MS Windows readily available.  
Hence, the question is not only whether we have Trunk ready for a  
stable release by now, but also: Are we ready for another 6-8 weeks of  
MS Windows bug squashing once 2.2.0 bugs are being reported? I, for  
one, have to admit I probably won't have much time; I can work on 1-2  
actual bugs per months right now, but not more than that. I don't know  
how the upcoming months look for you...

Basically two options:

* Continue as planned: Have 2.2.0 ASAP, and just try our best to fix  
the newly reported Windows bugs in the weeks after that (implying  
several 2.2.x releases in the upcoming weeks). We probably can't make  
it this weekend as originally planned, which is why I proposed another  
2.1.4 release this weekend in the wiki page.

Pro: This will give a whole new momentum to our userbase and PR  
coverage, and we also get rid of some 2.0.x bugs that already have  
been fixed in trunk (although I can't recall any from the top of my  
head).

Con: We risk being drowned in Windows bugs that we can't easily fix,  
which might lead to considerable frustration for bugreporters and for  
us as well. (Note that the small number of new features IMHO isn't a  
marketing problem at all because "Port to Windows" is a big deal  
already. From a MS Windows user point of view, we're basically a brand  
new application that didn't exist before!)

* Delay 2.2.0 until after GSoC has finished.

Pro: No overwhelming by too many reported Windows bugs. Marketing the  
new release will be even easier because apart from the Windows port,  
we will have some shiny new GSoC features available.

Con: We can't tell by now the point in time when the GSoC  
contributions will have stabilized enough so that we can call this a  
stable release. GSoC ends by end of August. We really don't know  
whether a 2.2.0 would be possibly by September or rather by December,  
which in turn can easily turn into beginning of 2008. Also, some users  
might be enticed to switch to 2.1.x releases just because the last  
stable release will already be so old by the end of this year.

More thoughts? Decisions?

I'm unsure, really. I'd prefer to have 2.2.0 by now, but I have to  
admit I won't have much time dealing with Windows bugs after that. If  
there is a majority to have 2.2.0 by now, I'd happily deal with the PR  
work; however, if we prefer to delay it, then I'd also happily delay  
the PR work as well.

Regards,

Christian

[1] http://wiki.gnucash.org/wiki/Release_Schedule
[2] http://wiki.gnucash.org/wiki/PR_planning