This message is the promised explanation message for my very emotional
response to Jims first Freiburg Draft proposal.
I think I owe Jim an explanation as I attacked him heavily, sorry about
that.
I think the key message of this mail can be abbreviated as "yes, when
you want to do smthng then please do it professionally. "
If I may say so: at the last meeting with Kispro my impression was that
I am not alone with this opinion. If I was I would have given up the
Oudeis project. But I haven't. It's just the modus vivendi or modus
of working that I criticised and would like to see changed.
Ok, here it comes:
>
> Freiburg proposal
>
> Have an online presentation that parallels our face-to-face
> presentations, so that people who would like to participate but who
>
> cannot make it to Freiburg can have access to papers, some
> discussion,
> and even a live show.
>
[...]
Hmmm....
Hmmmm.......
Hmmmmm........
Personally, I feel I have to make a short statement about this concept and
to the idea list in general. All following opinions are copyrighted by
me and ought to be thrown back at me and at nobody else.
First of all I didn't quite understand Jims concept which either doesn't speak
for my intelligence, concentration or I am missing background info -- or:
it wasn't formulated precisely enough. The only piece of information I
got from it was that only the MOO is to be used in Freiburg.
Second of all Jim let me state again something I have been saying very often:
Computer technology (if you plan to use it) isn't as cheap as everybody
thinks. "Houshold computer technology" (PC, modem, Windows95) might be cheap.
But whole systems are a bit more complex. This is because it takes a lot of
time to get things running. An it takes ten times longer when the formulation
of the problem which is to be solved by means of comp. tech. is not precise.
Time is money.
Don't misunderstand me: presently I am not interested in earning a lot.
I am interested in good projects. Good projects require good planning.
The first thing I noticed in the concept is that the money aspect which
is a major major part in good planning was missing.
Third of all I can say from my experience that a big mistake in many
projects involving comp. tech. is that new ideas pop up from time to time.
These "new good ideas" will kill the project. Let me explain that in more
detail. Whenever computer technology is involved and especially when
some programming has to be done the designer of such a system will
model the problem or the real world. A new idea might not fit into
the model. People in computer sciences are frantically trying to develop
new programming languages and systems ("Java", "Object oriented languages")
to avoid such a situation but it is inherent to modeling.
So for all projects involving comp. tech. I suggest "sticking to
one and only one concept". Jim, your ideas seem new to me. I haven't seen
them in the oudeis concept so far (which of course could be my fault
as I might have been missing something).
Freiburg: MOO only performance. Ok, if you wish so. No big technical problem.
Nothing new to invent or program. But still (as can seen by reading
the last two paragraphs): someone will nevertheless have to make sure
that
a) a MOO client is up and running on each PC
b) the projector is working:
I dont know if you know how much trouble we had with
a simple projector in Linz. To be precise: I had
2^8 (that is two to the eigth) possibilities to set up one
single adapter. Now I wasn't sure if even this adapter
was the cause why it wasn't working. We were under time
pressure. Please change a stupid dip-switch and afterwards
reboot the PC afterwards 256 times (=2^8) and check which
combination works. That takes time!
c) The TCP/IP connection is working for each PC.
In Linz we had a nice problem: _we_ that is the oudeis
people found out that the cable connecting the Sky Cafe with
the Hub at the cellar of the building wasn't working.
We all depended on that connection heavily as it was our only
one.
Next problem: The jack where you plugged in the cable for
the netork connection had a "Wackelkontakt" (i.e. if you
wiggle the cable then sometimes the network connection will
work sometimes it won't)
Next problem we had in Linz: If you pulled out the cable
on the side of one specific PC then the network card of this
PC got confused. So the whole machine had to be rebooted
everytime.
Next problem we had in Linz: There was a firewall in the
AEC. Which means that nobody from the outside internet world
would have been able to reach our nice little RealAudio Server
in the Sky Cafe. (so that nobody gets confused: this
problem was solved in the beginning so nobody noticed it
afterwrds). If we hadn't found this problem before the
performance there would have been no transmission of any kind
to Vienna or Station Manor...
d) etc. This list is by far not complete. So the first thing to
make sure is that you have diagrams of the setup and then
go thru _everything_ and make sure that you don't miss anything
which _could_ go wrong (do a Fault Tree, error analysis).
This takes time again. And time is money.
Enough of that. This is not to pat ourselves but to demonstrate the nice
little detail problems that in my humble opinion _have_ to be addressed
beforehand in every project.
Now out of this experience:
I had the feeling when I read your Draft proposal, Jim, that nobody
learned from the Linz / Vienna Ars Electronica experience. And that's when
I got mad. I am sorry for exploding in the middle of a mailing list.
I admit this was not a very good idea, but I simply had the feeling
"Jesus, what can you do? This endless discussions will go on and on and on
and nobody ever learns. They will always find somebody who will be willing
to work 14-16 hours a day dealing happily with all the 'nice little
problems' as described above. I don't want to be their nice and friendly tech
guy who will solve problems that only arise when nobody thinks ahead"
Soooo.... I am _very_ interested in continuing doing something for
Oudeis (whatever that is at the moment), BUT!!: the whole project has
to be run professionally. If not, then I am off to my theoretical
computer sciences, to my real time systems and embedded systems, to
my mathematics, to my commercial programming where I earn double, three
or four times as much, to etc.
And finally then I would be off to do some more computation of prime
numbers ;-)))))
Hope this clarified something. If not please let me know. I would like
to make me position as clear as possible.
cu,
Aaron.
"mr. TCP/IP and mr. C Compiler writer" ;-)