The future of selections and masks in Krita

[Boudewijn Rempt]

On Tue, 1 Aug 2006, Thomas Zander wrote:

> Maybe you can find that book and read about it to see its some smart guy 
> who thought this up, and its not just some weird idea I came up with ;)
> "The inmates are running the asylum"

I've got that book (and read it, of course) and also the companion volume
"About Face" -- and, what's more, I've also read the discussions between
Alan Cooper and Kent Beck. There are a couple of basic problems here:

* Our persona's are not really complete yet. That's something that'll
come with time, but it's too early to say "no, that's not going to be
coded because A nor B nor C will us it".

* Our application sphere is completely warped by Photoshop. Photoshop
is not designed, but accreted, and still people expect that every
raster image app is a Photoshop clone. In Cooper terms, every
current user of a graphics app is an apologist -- and therefore
the worst possible person to ask about interaction design.

* Cooper does not deal with what Steve McConnel calls "software
research". Cooper's work is not about innovation: it's about
carving out a chunk of market share by producing a application
that works better, but has less features than the established
applications. McConnell (spelling?) deals with that problem
quite nicely in his book "Rapid Development".

* Developing for persona's is ultimately market-driven: producing
an application that will do well in the market so you can pay
developers to do something else than work for themselves. Krita
developers are not paid; removing the scratch-your-itch motivation
and not replacing it with the money motivation will leave a void.
Trying to do the best we can for our users is fine, _but_ at least
some of us did not start working on Krita because of altruistic
motives (giving the people who are too cheap for pro apps a free
alternative), but because we needed an application that didn't
exist yet. Or because we wanted to give ourselves an application
that we were too cheap to buy ourselves. That's me, actually.
I wanted to have Fractal Painter on my Linux Box, or something
close to that, so I could paint a good campaign map. Sure, I care
about the users who send me nice mails, and sure, I get depressed
about the oiks who make the web forums an unhealthy place to hang out.
But I don't care about users to the exclusion of myself: to do so
would remove a lot motivation.

And in the end, without motivation, no code gets written, and
without code, any interaction design is worthless.

* Finally, we don't have an interaction design yet for selection
handling. An interaction design is quite a bit more involved
than what we have now: and I am not sure that anyone present in
this conversation has the experience to create a full-featured
interaction design document.

To conclude: yes, persona's are useful to focus our thoughts. My
own thinking about masks, selections and so on have been developed
a lot through this discussion, and I think that Bart's summary is
better than what I could write. But our persona's are not done,
and they are not yet suitable for deciding what to do and what
not to do.

Boudewijn