[kde-guidelines] Goal of the HIG
Philipp Stefan
sogatori at gmail.com
Sun Jul 27 12:10:20 UTC 2014
Hello everyone,
I'm not sure if this is the right place to voice my concern as I want to
discuss our general approach to to HIGs.
Important disclaimer: I don't want to criticize anyone here. You are all
some doing important work and I'm confident that we can bring KDE
Software forward. Also, this applies not to all HIGs, like the layout
patterns etc.
I like this proposal, I think it includes everything that needs to be
said and a bit more. My concern is, however, that we create more and
more guidelines for the developers to read. I fear that we increase the
load of the developers instead of easing it. Yes, we write guidelines,
but we have more and more guidelines for them to read. We can, of
course, keep the guidelines as short as possible, but that won't help us
much as long as we create ever more of them. Another thing we have to
work against is that KDE doesn't have long standing history of high
quality usability design. This kind of quality is also a learned skill,
and I think we are not there yet.
Another point why other platform have good designed applications is IMHO
that they try to take as much weigh off the developers.
Of course all the HIGs are work in progress and always in a state of
change, but I think we really should focus more on making guides and not
just guidelines. I think we should make a guide that shows step by step
how make a good designed, integrated Application. The guide should cover
everything from point 1 of basic considerations over wording and
animations to integrating into Plasma in a linear way. Of course we can
always reorganize the HIGs in a way to fit this ideal or write such a
guide, but I think we have to keep this concept in mind to come up with
something that really helps developers.
The way we currently organize things is a bit overwhelming for me
personally. There is no way to progress through the Usability HIGs
without having the page opened and open new tabs. It really resembles
the Wikipedia style much more than it does resemble a guide you'd find
in a book (which is what I think we should strive for).
Another big thing is that we leave much to the developers which then
balloons up our guidelines. I think ideally we'd have a e.g. pre-made
search widget with a predefined behaviour so that we only have to tell
developers how to feed it with information and where in the application
to integrate it.
I think the "putt the button here, the icon there and it should behave
this way" approach has failed us before. Personally I think that design
guidelines and design patterns should not be as separated as they
currently are. Take animations for example (I know it isn't finished
yet, great work btw! :), I don't think we should tell developers how to
animate certain widgets. Their behaviour should be as pre-defined as
possible, so that we only have to give them general guidelines, with
examples of DOs and DON'Ts.
I'm sorry if this doesn't sound too coherent, but I have problems
thinking clearly right now. Anyway, I hope you get what I mean. I don't
think what we're doing is wrong, but I think it could be improved when
we would approach our work a bit differently. Naturally this won't
happen in this or next cycle. To say it cynically, throwing a bunch of
text at people won't change much.
Phil
More information about the kde-guidelines
mailing list