thinking ahead to 4.5: features or polish?

[Marco Martin]

On Wednesday 04 November 2009, Aaron J. Seigo wrote:
> hi all ..
> 
> i spent some time today thinking about challenges as well as opportunities
> that exist for our little Plasma baby who is growing up at an amazing pace.
> 
> just to give you an idea of where we're at in terms of raw code production,
> here are some sloccount counts:
> 
> kdelibs: 42,121
> runtime: 4,190
> workspace: 83,332
> addons: 55,792
> 
> and we have so many little buds of progress elsewhere too:
> 
> kdereview: 8,365
> media center components: 6,608
> plasmate: 4,518
> networkmanager: 27,029
> lionmail: 2,009
> mobile shell: 1,147
> 
> we have been pounding out feature packed release after feature packed
>  release and 4.4 will be no different in that regard.
> 
> of course, it's not the only way we're growing. were also growing the user
> base as more of our desktop users come online and we start to reach out
>  into new areas of the market such as netbooks and phones. we're also
>  growing in *dum de dum dum* bug count. right now we have 745 open bug
>  reports, not counting wishlist items. that's not as bad as it might sound;
>  it's somewhere around one bug per 240 lines of code. but it's still
>  substantial.
> 
> as i mentioned at T3, plasma has developed a bit of a rhythm from release
>  to release: January brings lots of new features (esp big ones) and some
>  polish, July brings lots of polish and some new features (mostly smaller
>  ones).
> 
> 4.4 will certainly be a big feature release: netbook, remote plasmoids, new
> widget explorer, containment action plugins, lots more javascript in a
>  number of places, nepomuk integration .... but what about 4.5?
> 
> well, i'm seriously considering putting a moratorium on new features in 4.5
> and instead focusing on making it a polishing release.
> 
[...]
> (and just to head off all the "oh yes, please do!" comments from the people
>  on the list here who follow the threads but who aren't active contributors
>  to some part of plasma: let's keep this discussion to those who are
>  working on the code and artwork for now. thanks :)
> 
I am in favor of this, it would look quite good anyways to say "hey, we now 
start in half of the time", if managed well it could be quite well received.

on the other hand there are spots that are too new, thus needing a big amount 
of work, also feature wise, thinking about the netbook, of course, where now a 
big amount of work has been put into having something "shippable", but the 
work to do is still huge.
or also the overview and the nepomuk integration (will they have a first 
shipped version for 4.4? i hope so?)
so we would need a policy like:
* is a feature that can be considered vital for a decent user experience? 
(like completing missing parts of the netbook)
* is something that could be considered the completition of an already present 
half baked feature?
* is the integration of a feature that was put in and still not really used? 
(like using the new animation framework around, making components nepomuk 
activity aware..)
* is something pretty big but that can facilitate with the goals we have for 
4.5? (like rewriting large chunks of old crufty code into something a bit 
smaller and nicer, thinking about the toolboxes for instance)

To summarize, not being something with 0 features as if we were in string 
freeze for 6 months, but something that has a couple of new things with the 
focus on "remember that part you said it had great potentiality? ok, now it 
works"

-- 
Marco Martin