[openbeos] Re: Visual Design for R1?
- From: "Michael Phipps" <mphipps1@xxxxxxxxxxxxxxxx>
- To: openbeos@xxxxxxxxxxxxx
- Date: Wed, 03 Sep 2003 20:08:54 -0400
I consider look and feel to be very much in the same catagory as API - R1 ==
R5.
I know that a whole field of people (Hi, Helmar) will groan and think that this
is very unreasonable. I believe that there is no other reasonable alternative.
Reason #1 - You will never get consensus on *ANY* other look and feel. This has
been an issue with BeOS from the very beginning. Some BeOS people are
ex-Amigans. They want assign (the CLI command), screen functionality and
locale. Some were former Mac people (although most have gone back to the 7
color koolaid) - esp within Be, they liked the API and the icons done just so.
Others were Unix people. Their contribution was the POSIX layer. Where a group
has an opinion, nothing will change it. Personally, I like Stuart's work.
Others do not. Some people like Aqua. Others hate it. Some like Win XP's
Crayola layout. Others don't. Look at the mass hysteria over YellowTab's
screenshots. The *last* thing we need is to divide people any more than we have
to ***RIGHT NOW***.
Reason #2 - features take time. I don't care if it is 1 line of code or 100. 1
bitmap or 100. It takes time. And that is time that we *should* be using to get
R1 done.
Reason #3 - whatever we do for R1 would be half baked. This is the same reason
that there will not be any internationalization, resizable GUI, etc support -
when it is done, it should be The Right Thing. Not some half done thing. In
this thread, there has been some discussion of how Stuart's work can't be
implemented with the current Decorator API. That is a prime example of a half
baked API. Be had them too - they hid theirs. As OSS, we don't. But that
doesn't mean that they are ready for prime time, either. I would be strongly
against documenting Decorator and putting it in the neuvo-BeBook until we are
*sure* that it is what we want. I suspect that DW would agree. If we were CSS,
no one would even know about it.
Reason #4 - THIS IS NOT THE RELEASE TO CONQUER THE WORLD. Let's be honest,
guys, R5 was barely ready for the world in 2000. Our R1, coming in oh, let's
just be wild and assume 2004, isn't going to measure up in every way to other
OS's out there. Does that mean that we shouldn't release it? Of course not. Nor
does it mean that we should wait until we can out-do every OS in existance. R1
is to be many things. One is proof of concept - that we (as a group) can write
an OS. One is replacement for R1 - fixing some of the obvious bugs/issues (1
gig memory limit, mmap, networking). One is a shipping OSS product. R2 or R3 or
sometime down the road should be more "earth shaking". But, right now, we have
out collective hands overflowing getting the basics down.
Reason #5 - we don't know how many apps we will break. This sort of thing *HAS*
to be done in a controlled way. How many apps out there hard code the grey
color? A new look and feel could break that. How many sets of documentation out
there specify the yellow tabs? Everyone needs advanced warning of what will
change and what it will change to. That is the hallmark of a *product*, not a
*project*.
As for what form the final product (post-R1) will take, I don't know. No one
does. I would be hesitant to leave the current look and feel until we have
something that is drastically better. Looking like
KDE/Gnome/MacOSX/XP/Win2K/QNX/whatever is not necessarily an improvment. There
should be rhyme and reason to what we do. That is an interesting expression -
rhyme - poetry and reason - logical thought. That is exactly what is needed -
rhyme and reason. It has to look nicer and work better or it doesn't make sense
to bother. Rounded corners is not a compelling story, to me. I know that
probably makes Helmar feel all justified (Hi, Helmar!). I didn't say, though,
that design is a bad thing. In fact, I would say the opposite - that *just*
making it pretty is like a monkey in a dress - it doesn't fool anyone. What
most of the people who have seen Stuart's work *don't* know is the discussion
that he and I have had about how what he drew should *work*. That is what is
really intere!
sting. A lot more replicants. A lot more systemic thought. Not just rounded
curves, but some real changes in Tracker/Deskbar.
Anyway - enough pontification. The look is R5 unless someone (offlist, please)
can change my mind.
Other related posts:
