[openbeos] Re: OBOS Poll: Multiple Monitor Support
- From: Craig Turner <craig@xxxxxxxxxx>
- To: openbeos@xxxxxxxxxxxxx
- Date: Wed, 09 Jan 2002 10:35:45 +1030 (CST)
I wrote..
> Also there's the issue of the dock - do you display it once for each
> workspace? I like the idea of a plain one-workspace-to-monitor approach,
> with minimal changes. That is - each workspace has a dock, each has its
> standard icons, etc. I'd just be concerned that configuring different icon
> setups on different workspaces because messy. Only my opinion of course,
> but if this went to voting I'd vote for a simple to understand and
> implement solution :).
Actually I will comment further on this, because I think there's a broader
principle. I remember a while back that the issue of themes came up for
BeOS - some users were saying how fantastic themes would be, that sort of
thing. That made me quite annoyed and here's why..
There's a user interface principle that I follow, but can't site from
anybody else :) That is, write the system so that it needs minimal
configuration, and so the user *can't* spend endless hours tweaking, etc.
I rememeber the OS/2 desktop and the amount of playing you needed to do to
get it usable. It's a source of annoyance for administrators - because
every system in the office gets wierd configuration, somebody installs
litestep, that sort of thing. It's a source of slackness in programmers
making good design decisions: "I'll do it this way and not worry about it,
and if the user doesn't like it they can wade through configurations to
make it right". And it's annoying when you're a user who just likes to get
things done, particularly when you have to do a reinstall.
For this reason, I'm skeptical of suggestions to make things
over-configurable when that energy could be put into writing drivers or
improving the efficiency of core systems. Over configuration is the linux
way of doing things, where using the operating system becomes a laborious
game. I believe BeOS embodies a principle of simple elegance. The
interface excludes things which arn't obvious to the user. It may be
possible to implement these new features in a way that doesn't force
itself on the user.
I don't want to rain on parades, moreso because I'm new here :) I may
misunderstand the purpose of the list. If this is just brainstorming then
I apologise. But if we're trying to form a list here of definitive stuff
we're aiming for in R2, I'm inclined to argue against features which arn't
core to advancing the operating system.
- C
Other related posts:
