On Tue, 15 Aug 2006 10:32:19 +0200
"Stefan Tibus" <sjti_AT_gmx.net> wrote:
> On Mon, 14 Aug 2006 17:54:57 +0200 Denis Grelich <denis_AT_grelich.de>
> wrote:
> > On Mon, 14 Aug 2006 15:41:31 +0200 "Stefan Tibus" <sjti_AT_gmx.net>
> > wrote:
> > > First of all, there is some personal taste involved, which no
> > > wm will ever be able to guess.
> >
> > Personal taste comes only into play when you select your tool. You
> > can't configure that much on a screwdriver or a hammer. Neither on a
> > car. Yet your taste is very important when you choose if you either
> > want a screwdriver with a big, round handle or small one of
> > plastic, or if you buy a VW Polo or a Mercedes.
> >
> > Now think about it for a second. How many super-customizable tools
> > with exchangeable handles and stuff did you throw away already or
> > did you put away on the shelf because they're all crappy and flaky?
> > Good tools are simple tools.
> Ok. How many screw drivers do you know that try to guess
> where you want to put your next screw themselves?
> That's the main point, why something like wmii will need
> customization. Because it tries to arrange the windows
> automatically for the user. But what is good for the user?
> That's not what I meant. Using your car metaphor, wmii tries to be
> a car with new steering controls and automated driving. But driving
> styles can be very different and it will need customization options
> in order to fit these. It's just not taking the same way through
> the city centre I'd choose, because it doesn't now my hometown as
> well as I do and I can't feed in the roadmap... ;-)
It really depends on what and how far you automate. The goal is to
automate tasks that are the same most of the time and that you would do
anyway. Like, on a car, that the blinker jumps back when you center the
steering wheel again, or, even further, automatic gear shift. It does
in 90% of all cases what you would have done anyway. In the ten or so
percent where it doesn't, you just re-set the blinker or switch to
Tip-Tronic. Or you adapt to the machine and try to do what you wanted
to do in a different way, suited for the automation.
> There's also taste and habits involved. Other wm's leave this
> completely to the user or are somewhat dynamic but stick
> to fixed layouts more or less. wmii tries to be better in not
> imposing fixed layouts. It is good. But often enough the result
> is just not want. (I surely prefer fixed layouts on views for
> repeating or fixed tasks. Yet I like dynamic layouts for other
> tasks. - I need a good mixture of both.)
One huge problem of fixed layouts is when something unforeseen happens.
Like, I want to to copy&paste some stuff from another window into my
MUA, and thus put the window into my mail view. Then, the fixed layout
is rendered completely useless until I re-order it by hand, which is
highly annoying. That's not something that should happen.
Anyway, while you are working with dynamic layouts, they're fixed too.
When you start the application, you, or the automatic, arrange the
windows in some way and you keep it that way throughout the day, until
a new window comes up or you close something. But I have to admit that
the automatism must always create the same result for a given
situation, and should act absolutely predictable in new situation. Or
else the application gets in control of you.
Now weigh the time you have to spend doing repetitive tasks on both
approaches. With dynamic layouts, if you're not happy with the layout
you get after starting your MUA (or any other multi-window application,
see http://www.%c3%bcberl33t.net/tmp/sylpheed_multiple_windows.png), you
re-arrange it once, and it is, hopefully ;), not tampered with too much
by the mechanisms of dynamic window management. When a window opens in a
fixed layout, you have to re-arrange all the windows quite often
until you can work with them. I prefer the former.
> > Honestly, are there any applications that you use every day that
> > you /really/ like? I, honestly, didn't find a good MUA, or a usable
> > Browser or even text editor yet. I suppose they have to be written
> > from scratch anyway. Hopefully,the 10kloc project will provide us
> > with those ;)
> I'm waiting for those since 15 years or more. Do you really expect
> them to emerge within the next two years or so? And what should I
> do in the meantime? - That's just no answer you provided here.
We're all working on a better future. I stick with crappy apps too, atm.
> > And besides, 90% of your programs are OSS anyway. So what's the
> > problem?
> I have other things to do than write my every-day use software
> myself. I'm no programmer, I'm a user.
But someone might still do those changes to popular OSS applications.
If you do not want to do this, okay, then you have to wait until
someone else does it. Like with every feature.
> > > (and the hints provided by the programmer may still not result in
> > > what the user wants)
> > Again, a good craftsman learns to use his tools well, not the other
> > way round. A tool is by definition /not/ smart, it mustn't try to
> > adapt to the user. It is very simple to adapt to a tool though. And
> > if the program does not fit your needs, it's simply the wrong one.
> Hm. Seems most programmers aren't good craftsmen then...
> And then, wmii definitely tries to be smart regarding the dynamic
> window arrangement - so it's no tool according to your definition?
Smart means to guess what a user wants. That's never what wmii tried to
do, because it is futile, in principle. Wmii is made to automate things,
but always in predictable ways that users can adapt to. The
results /sometimes/ look similar, but they're not the same things.
Greetings,
Denis
This archive was generated by hypermail 2.2.0 : Sun Jul 13 2008 - 16:12:55 UTC