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

  • From: Eddy Groen <eddyspeeder@xxxxxxxxx>
  • To: openbeos-cdt@xxxxxxxxxxxxx
  • Date: Thu, 5 Nov 2009 11:58:04 +0100

Dear Creative Design Team enthusiasts,

I just read this list has been asleep for well over two years now. Let's
bump it!

In this email I will explain my views on why a CDT should be regrouped. I
first will try to point out why I think there is a problem that should be
dealt with. Next, I will discuss why the time has come to actually tackle
the problem at hand. Lastly, I will discuss what I see as the proper
solution, in which I feel the actual formation of the CDT should be
coordinated by Ryan.

Please know that this is just a proposal intended to start a discussion. So
fire away at it!


On the main Haiku list, there have been several recent discussions on UI
features of Haiku. The problem of these discussions is, that even though
they evoke the various visions and philosophies of list members, the
discussions are entirely fruitless. Why? Because they yield no implications
to Haiku's design.

At this point the Haiku project does not seem to have anyone with the
responsibility (let alone the power or ability) to take the input, think it
through, research it, and in the end providing a binding solution or
alteration proposal. Ryan mentioned that he and Stippi have been thinking
through Haiku's window design quite early on in the project.

I quickly learnt that Haiku did not have the capacity to properly process UI
design suggestions. Through GSoC and HCD, Informatics students were able to
do a research project through Haiku, and I had hoped for the same thing from
my cognitive ergonomics background. I sent an email inquiring what
opportunities were there, and who is/are specifically involved with design.
( http://www.freelists.org/post/haiku/Masters-thesis-possibility-with-Haiku

In response to my email, Ankur asked me to be a bit more concrete about what
kind of research initiatives I would suggest, and Stippi replied rather
generally and pointed out that there are several (isolated) enhancement
tickets regarding design. That's nice, if a developer feels the need to
reconsider some design he can get down to it. But still there is no driving
force that makes sure these enhancements are discussed and turned into
actual results. Neither is there any guarantee that it will ever be picked
up by anyone, since there is no central coordination for design
improvements. But just like the other UI discussions, the eventual result
was fruitless; the discussion soon disappeared and I had to conclude Haiku
was not yet ready to accept usability students.


So far, the emphasis on the developers' side was not a problem. In fact, I
even believe it was good that it worked this way. We all know that OpenBeOS
was started by very driven BeOS-enthousiasts, many of which I have the
privilege of having met during BeGeistert gatherings. This type of evolution
inevitably meant that during the first years of the project, it was no more
than logical that the developers were the ones having the final say. After
all, it is THEIR project and as long as Haiku is still getting set up, they
should be given every freedom.

But now things are changing. The Haiku project has the benefit that it is
immensely self-aware, the main assertion being that Haiku should leave a
good impression on users. The general trend on the mailing list is that
user-friendliness is a term that remains key in any design decision made.
And that means that there should be a shift towards designers making the
decisions and developers making the technology.

I feel I should briefly discuss Microsoft and Apple as a case study here.
The considerable changes in the appearance of many Microsoft products (e.g.,
the ribbon interface) does suggest that Microsoft has slowly started turning
their system more into the way Apple does it: the designers are the ones
making the call, the developers get the freedom to do anything as long as it
matches with the design.

If Haiku wants to succeed at continuing to leave good impressions to its
users, it should not only be stable, responsive and everything else that
makes the BeOS the BeOS, but it should appeal to people. Discuss whether it
looks "nineties or not" all you want, but minor changes to an interface
within a quickly changing landscape will prove inadequate. Remember that
Haiku's design is still close to interface philosophies prior to OSX,
XP/Vista/7. With BeOS, we have come to accept that the interface is as it
is, but we must heed not to start thinking that it should remain the way it
is, just because it has always been this way.


Recent proposals to just close this list and declare the Creative Design
Team dead and obsolete is not going to solve the problem at hand, that Haiku
is brittle when it comes to usability design. In fact, the inactivity on
this CDT list is just another sign of the problem, and it must not be taken
as a reason to say the solution is to ignore it, hoping the need for it will
just go away. I can tell you right now: it won't. So this is why I am
stepping up, writing a lengthy email explaining just why the CDT should be
revamped and gain importance.

As said in the introduction, I believe this should be coordinated from "up
above", having named Ryan as the one to carry the responsibility (though not
necessarily the one doing all the work in setting it up - feel free to
delegate). Above I have made clear what the problem is, I've pointed out who
should be in control of solving it, but this email would not be complete if
I were not to include my view on the solution. So here is my proposal:

When trying to set up the CDT, the first problem we will run into is that
Haiku is a collective of individuals spread all over the globe. Microsoft
and Apple design teams are all in one place, doing their research and
brainstorming in the same room (and AFAIK, Linux distros do not have such
design teams). They can easily sketch something or demonstrate mockups.
Haiku CDT people will not be so privileged. They must show/share/discuss
ideas over long distances.

So before a CDT is getting shaped up, I propose that the most useful means
of long-distance design communication will get investigated. One way is to
do conference calls over Skype, while sharing designs over means such as
Google Waves (feel free to share better alternatives, I wouldn't know any).
While this investigation is ongoing, team members must be sought.

Fitting this all into the Haiku roadmap, I think a CDT should be set up as
soon as possible. In the first few months, it will already start working on
making design improvements, but while its identity still has to shape up,
the developers remain having the final say. Then Haiku R1 is released,
meaning the goal of recreating BeOS R5 has been attained. From there on, all
and any options to make Haiku future-proof are open. At this time, the CDT
will have to be a steady team, a well-oiled machine, that can be attributed
more power to.

Well, this concludes my proposal for now. Please let me hear your thoughts!


Other related posts: