Subject: Re: [linux-audio-dev] mouse wheel behavior and RFC: human interface guidelines
From: Lee Revell (rlrevell_AT_joe-job.com)
Date: Sun Aug 22 2004 - 00:47:51 EEST
On Sat, 2004-08-21 at 17:31, John Check wrote:
> On Saturday 21 August 2004 04:51 pm, Lee Revell wrote:
> > On Sat, 2004-08-21 at 16:36, John Check wrote:
> > > On Saturday 21 August 2004 04:24 pm, Lee Revell wrote:
> > > > On Sat, 2004-08-21 at 16:14, John Check wrote:
> > > > > On Saturday 21 August 2004 02:10 pm, Pete Bessman wrote:
> > > > > > I guarantee you that the last thing on 99.8% of users' minds when
> > > > > > they're adjusting a horizontal volume slider is "This is kind of
> > > > > > like reading a book, which goes left to right and top to bottom;
> > > > > > and if we assume an association between beginnings and ends, then
> > > > > > it follows that left and up are vaguely synonymous. Therefore, to
> > > > > > decrease the volume with my mousewheel, I MUST spin up."
> > > > >
> > > > > Nope, they try one way, then the other, then make a mental note if it
> > > > > doesn't fit the expectation.
> > > >
> > > > Which is why is HAS to work the same way in every app. You can't
> > > > expect users to make mental notes of differences like this from app to
> > > > app.
> > >
> > > I agree, but I think "should" is a more realistic position than "has to".
> >
> > Sorry, I disagree. This kind of thing is MADDENING for users and makes
> > Linux look fly-by-night.
> >
>
> That presumes we have a monopoly on inconsistency. ;)
> Having a consistent interface will definitely work to our advantage, but
> I'm talking about near term when I say things like that. Being as I'm not
> going to be the one coding, I can't say "must" or "has to".
>
It's easy enough (well it should be, I would imagine you can set this
behavior when you create a GTK slider, if not then it's a completely
reasonable feature request) for a coder to fix that I think we can
safely say it must work a certain way (in order to meet the Linux Audio
HIG that I guess I am proposing).
Right now, I would have the HIG state that this behavior must be
globally configurable. Globally for the sake of the above mentioned
consistency, and configurable because obviously one way is intuitive to
some people, and the other way to others.
> > Of course, you can't force people to follow human interface guidelines
> > in the free software world. All we can do is carefully develop them
> > based on what users want, then if you choose not to comply you are
> > explicitly stating that you are hubristically choosing to ignore the
> > users need for a consistent experience.
> >
>
> And in that case one gets what one deserves. From what I can see so far
> WRT proprietary stuff, there are a lot of little annoyances that add up to a
> lot of opportunity for LA.
>
Yes, see above. The list of annoyances people have with the 'state of
the art' in proprietary software is a major motivating factor for many
Linux audio developers. It should be the same way for users, but we are
not there yet.
Lee
Lee
This archive was generated by hypermail 2b28 : Sun Aug 22 2004 - 00:51:41 EEST