TUNES website homepage draft
Brian Rice
water@tunes.org
Wed Jul 11 21:50:04 2001
Hi Kyle,
I know I was inflammatory, but let's see if we can make something
useful out of this:
At 8:34 PM -0400 7/11/01, Kyle Lahnakoski wrote:
>Brian Rice wrote:
>
>> Please don't take this the wrong way, but I actually wrote that
>> description with the *intent* of keeping hackers away who don't have
>> a good knowledge of mathematics. ...
>
>Skilled people are required, but skilled people with time on their hands
>are few and far between. They are either enjoying their current
>position, or are being paid well enough that they don't care to change
>things. These people will not gravitate towards TUNES.
I know. But that won't stop me from asking for grant money so that I
and Fare and Tril and whoever else can spend time on it without it
stressing us out too much.
>If the community (as opposed to private funding) is going to build TUNES
>then a lot of time investment is what TUNES needs. Only the people with
>time on their hands (in school or retired) are available to contribute.
>People with time on their hands are often looking for direction. The
>TUNES site should be providing direction to receive their free time in
>return.
That's just it. I don't see how the community could build TUNES at
all without the initial time and theoretical investment. And before
you contradict me, remember that I am the one who does the obscene
amount of research for TUNES: I am speaking from what I've learned
from that, not from some common sense or gut feeling. You would not
believe the countless branches of computer science research that
exhaust lots of niggling differences between semantics or syntax or
libraries in the search for better ways to compute! The smartest
collections of people in the world can't make something even remotely
close to TUNES - how should people be expected to believe in us? I
was laughed at in the University for this very reason! And I fully
empathize with this perspective - it's the most natural one to take!
If you believe in TUNES but don't understand how it technically can
be done, then you have to doubt what you know, all the time, and
challenge it until you know enough to grok the answer!
> > This is a rant, because it's one of the most annoying and
>> disempowering things about newbies that I have seen on the mailing
>> list since it's inception (yes, I have looked at the mailing list's
>> content for the last almost 7 years). This mailing list is not for
>> learning those concepts (we simply don't have the bandwidth, plus
>> many places on the net focus on such things), but TUNES core design
>> simply requires them.
>
>If TUNES is to be done, the web site, mailing lists, and IRC should be
>dedicated to supporting people who have the time to build TUNES. The
>TUNES site must act as an educator; educating the bright-eyed people in
>the ideas and concepts necessary for developing TUNES. The TUNES site
>must be a resource of essential information. Right now the TUNES site
>has a host of links to information that is mostly noise. TUNES needs
>these advanced concepts, but the web is too diluted to be effective.
Okay, I agree. I think finding a person with the talent (who like you
said above was not comfortable where he was) to make this work right
will be quite difficult. What would you suggest to add to my web page
proposal? Or how would you re-organize it?
There's a lot there that I think must remain in order to make sure
that we don't look like fools. But you are right to point out that we
should be viewed as a good place to learn about those things
required. However, my time is quite cramped still, so I don't see how
I could give it enough attention myself. I still haven't truly
updated my Arrow Philosophy paper in 2 years!
>The elders in the group should spend the little time they have on adding
>and refining the web page to handle all the questions from the newbies.
>For example, the language review should pick apart each language,
>showing essential aspects and providing EXAMPLES of the efficiencies
>that can be gained. I can not stress enough the need for examples. A
>good example acts as proof that an aspect is desirable.
The 'elders' are all quite busy themselves. Fare has taken on other
projects; I have employment issues; Tril has a job and an education
to fulfill; Jecel is very much oriented to his company and Self
language issues; Bill Tanksley - I'm not sure about him, he's
interested in a variety of things and I believe that he doubts the
project's aims are even possible (am I wrong, Bill?). Perhaps hcf can
step in; he has the intermediate level of knowledge about TUNES -
doesn't quite get all that Fare and I talk about on first mention,
but grasps a lot of the underlying stuff well enough to help manage
our information. Maybe someone else needs to step in. There are some
people on this list that I wouldn't want to do that because they are
still newbies in their understanding of things, though.
> > TUNES has to be worked out with something beyond common sense,
>> particularly not the sense of the modern programmer, as Tril's
>> manifesto I believe makes clear. This is one of the more common
>> misconceptions (or dare I say arrogances?) of average programmers to
>> think that a knowledge of C and Perl (for example) and their use in
>> any way constitutes the ability to hack TUNES by brute force.
>
>The TUNES site should act as a friendly meeting place and educational
>institution. This type of environment will build a population large
>enough to deal with the many small and boring tasks that lie ahead. A
>community project needs only one person with vision, a few good design
>guys, and many hackers. TUNES has the first two, but the environment is
>not suited for supporting the third.
Of course it isn't! There's nothing to hack, no game plan, not even a
provably-workable idea! All we have is the knowledge that TUNES has
not been definitively shown to be impossible, which is a shaky basis
for coding. *AI* is on better grounds because at least they know what
to imitate. :)
I don't think the environment friendly to hackers will be possible
until the core logic is worked out and explained in excruciating
detail, and that takes time strictly from the first two elements you
mentioned. Which is exactly where we are: I am thinking really hard
and reading books and re-reading TUNES site docs and consulting with
Fare and Tril and imbibing the collected wisdom of Squeakers,
Selfers, functional programmers, Lispers, and Joy people.
> > Bottom line: if you don't grok a concept, Google: "intro category
>> theory" or "intro arrow logic" or "intro type theory", or go buy a
>> book on it.
>
>... is a good example of maintaining an unattractive environment.
Well, then I apologize. But you should *still* go do some research,
being committed as a TUNES member. Keep in mind that I've spent the
last 7 years reading and re-reading the TUNES docs until I could
grasp them in many ways and explain TUNES in many ways. I went out
and bought Dover book after Dover book, and then modern CS and maths
book after modern CS and maths book. None of this knowledge came
easily.
Rhetorically: If you so want to have this place be good for
education, you should be contributing to that yourself! (Again, I
apologize for saying that, but it was rhetorical.) This is a terrible
by-line for open-source projects, even if it is true; I have seen it
many times repeated on the Squeak mailing list when no one has time
to attend to a feature someone asks for. Instead what is supposed to
happen is that we decide on how to handle it, and who will volunteer
to handle it - to collect up the little bits of information that
accumulate on IRC and the web site and to organize them gradually.
This of course demands time, and time to do some really tedious stuff
which has nothing to do with coding. Who's up for it? I believe you
were just lamenting the lack of people who would do so; even if you
weren't, plenty of us have done just that in the past. I know for
certain that several people in a row have volunteered to improve the
Review project over the years, and we have wound up with very little
results. Even Tril is *still* working on the Zope web site management
set-up. It's closer than ever to usable, but it doesn't count until
someone commits to running it properly. We had an incident with Corey
Reece's Wiki where he wound up losing lots of data that myself and
hcf worked on for several days. That was indeed disappointing, and
now Corey doesn't have time for anything related to TUNES it seems.
When will it end? I can't say.
How will it end? Probably with me finishing what I am currently
working on and moving into documentation; unless someone steps up to
work with me on this, or I find a way to get the same done through
educational institutions.
>--
>----------------------------------------------------------------------
>Kyle Lahnakoski Arcavia Software Ltd.
>(416) 892-7784 http://www.arcavia.com
I hope this helps,
~