[openbeos-cdt] Re: CDT regrouping: why and how?

  • From: Ryan Leavengood <leavengood@xxxxxxxxx>
  • To: openbeos-cdt@xxxxxxxxxxxxx
  • Date: Fri, 6 Nov 2009 01:07:56 -0500

OK there is a lot of input from you guys here so I'm not going to
quote anyone for the sake of brevity. But I'll try to address what has
been brought up:


I agree that a lot of discussions about UI end up just being everyone
giving their opinion and then nothing happening. I think part of the
problem is that it is just difficult to go back and process 60+
message threads to get the good stuff (even if you originally read
through all the messages.) But it is also the case that any
suggestions generally have to be implemented by a developer, and if no
developers are interested it just won't happen. I'm certainly willing
to do some development for UI improvements, but I'm just one person
and I'm already busy on other important projects (mainly the browser.)

But if you and other's think I should be the coordinator of a new CDT,
I will certainly consider it. But frankly, I'm unsure of how viable
the idea of a CDT is. Go back and read the archives of this list.
There was a lot of bickering, endless discussions of names (at least
that was eventually resolved), and apparently most of the discussions
were held in a forum which I assume is now long gone. There were also
several "leaders" of the CDT and in general I really don't think the
CDT produced much. Stuart McCoy (who was the second and I guess last
leader of the CDT) did produce the original Haiku logo, which Stephan
has updated. I honestly think that was about all that came out of the
CDT which is still around (and Stephan has heavily modified it.)

In fact on that note I think it is safe to say that Stephan has pretty
much been the REAL CDT, as far as actually getting something done. Now
zuMi and myself and a few others have made icons, and other developers
have done UI programming, but in general most of Haiku's looks are due
to Stephan. I think he has done a great job, and the point of this is
saying that in the end ONE PERSON really made the Haiku UI what it is
today, when before a huge CDT pretty much produced nothing. So I'm
just not so big on this team idea, based on the past.

Of course things are a bit different now. Haiku is much further along,
we already have a pretty decent UI to build off of, as well as good
icon guidelines (from Stephan) and the beginnings of a human interface
guide (http://factory.haiku-os.org/documentation/HIG/) written by
DarkWyrm a while ago.

So maybe a new CDT could actually work now. I'm willing to give it a
try, but with some more focus. The previous CDT was involved with a
lot (too much IMO), and some of those areas are now pretty well
covered. For example Jorge Mare has pretty much got the web-site
design covered. I do not think we should worry about that at the
moment. Most of the Haiku icons have been created, and we have a good
set of icon guidelines, so in general we should not worry about that
(beyond improving the guidelines and creating any needed icons.)
DarkWyrm's HIG is a start, but has not really been vetted with all the
developers, so we should actually think about that. But mainly the
focus should be on the UI, with usability being one aspect, and
graphics design the other (with the former always being more important
than the latter.)

With that said, I agree having a better way to interact would be good.
I would love to use Google Wave, but I have not yet gotten an invite
(and like a dummy I missed the invites Google was giving out to GSoC
mentors!) Plus not many people are on Wave yet, so it would become a
barrier to entry. Unless of course we set up our own Wave server
(something I am seriously considering.) I will look into that option
at least.

But for now we may have to settle for email, IRC and maybe a Wiki
somewhere. Though I might be open to an occasional Skype meeting.


I definitely agree that BeOS got a lot of things right. Obviously none
of us would be here if we thought otherwise. That is also why
newcomers see a lot of resistance when they suggest changes to
"classic" BeOS behavior. If I end up being the new "CDT lead" I will
certainly ensure that any UI improvements are true improvements and
not just changes for the sake of change. Of course I won't be alone
since I'm sure the other developers would speak up for any unwelcome

On that note I hope newcomers will use Haiku for a while and get
comfortable with the system before suggesting changes. Because no one
will listen if you are just suggesting changes because Haiku does not
work like what you are used to on other systems.


User friendly definitely should not mean "dumbed down". Users are not
novices for long, so it is naive and eventually NOT user friendly to
design UIs for novices. This will just end up annoying and slowing
down intermediate and advanced users (which in the long run are much
more common than novices.) But I'm sure I'm just repeating stuff most
of you already know. Or I hope so :)

As for whether the Haiku programmers would accept designers dictating
the interface: probably not. Mainly because we aren't getting paid to
do this and we just plain won't work on something if we don't like the
idea. That just means that the designers need to do a good job and get
consensus before something will be implemented. Which is probably why
most of the improvements to the Haiku UI so far have been done by a
combination designer/programmer (mainly Stephan, and some others and
myself to a lesser extent.)

I don't necessarily agree that everyone who has used BeOS or Haiku for
a while will think the usability is perfect. And if many do, that
means that the usability must be pretty good. The fact that people can
adapt to an interface obviously does not automatically mean that
interface is good, but that also doesn't mean the inverse is true. In
other words people adapting to an interface doesn't mean it is bad. It
might be just fine :)

As for fears of cloning another OS, most of the Haiku people I know
are open to taking good ideas from other places. Our GUI layout system
is highly inspired by the one from Qt, with a little Java thrown in.
Our network stack has a lot of BSD influence. The Ctrl-Alt-arrow
combos to move around workspaces in GNOME was copied in Haiku, as well
as the pretty general Linux combos to allow moving a window by left
clicking any part of it (I believe by holding down Ctrl?) I'm sure
there are many more examples.

The whole idea of people not liking new interfaces because they are
used to old ones has been pretty much destroyed by the iPhone and
similar interfaces. Also by the Nintendo Wii. People are open to new
interfaces if they are good. Unfortunately many designers think their
interfaces are better than they really are.

You imply that normal people cannot compete with academics in
interaction design. While I think education is great, it does not
automatically mean someone should be listened to above others. I'll
take experience over education any day. Now despite that I'm certainly
open to more academic takes on improving Haiku's interface.

As for any beliefs about Haiku not being used by normal people, all I
can say is: we aren't Linux. Of course right now Haiku is an alpha and
our main target is developers. But that does not mean we don't want
normal people using Haiku. Of course we do. But right now Haiku can be
buggy so it isn't the best for the average person who just wants to
use their PC.

As for your last list, I agree those are valid concerns and they
should definitely be something that Haiku (and/or the new CDT) should
focus on.

Finally let me say that while this discussion is good, we are all
making things difficult with such long emails. Let's try to start
focusing and address more specific topics to make it easier for
everyone to contribute and read.


Other related posts: