[olpc-software] AbiWord, HIG

[Alan Kay]

Hi Stephen --

At 09:50 AM 3/31/2006, Stephen J. Smoogen wrote:
>Ok this is ranging off of olpc..

<snip>

>There are 2 other points that I found in trying to know the field:
>
>Up until the WWW/Altavista search engines it was pretty darn hard to
>know what had been done before. By the time I got into computing
>around 1984, the universities treated their teaching methods like
>religous doctrine or at least like something to make a lot of money
>off of. You heard more about what was thought to be right thinking
>from professors and less about what else was out there. In trying to
>find out what was out there, one was limited to what was on USEnet,
>BBS's or what the local library had on microfiche and you had time to
>try and make your way through. [I fell asleep more often than not
>trying to make it through the local universities microfiche of ACM
>stuff.. and that was because it was pretty much gibberish for a 14
>year old. ] And even now with google you end up with 90% crud and 10%
>useful content and very little guidance of which is which. So knowing
>your field is not easy without some sort of Hacker's Encyclopedia.

This is a very good point, especially compared with what the biologists do. 
(see below)


>The second point is that knowing your field in physics and chemistry
>is not as easy as it seems. It usually takes 40 years before something
>becomes mainstream enough to become textbook fare and be something
>that is acceptable. And this is in fields where data is freely shared
>without patent/profit problems. In most of the fields covered by
>pharmaceutical there is usually a 5-10 year gap in knowledge even
>becoming publishable because of waiting for patents and such to come
>out. And if 3 groups are all researching the same thing.. then 2 of
>the 3 will find their funding die if a clear patent has come out
>unless the other company wants to dive-bomb the product or find
>associated patents.

I agree that both of these are problems (especially for 14 year olds).

But consider one of many counterexamples from Biology. My old friend Bruce 
Alberts (a MB who was head of the National Academy of Sciences for the last 
decade) is also the principal author of the incredible "Molecular Biology 
of the Cell", a (thick) book that, in its 1300 pages, takes you on a voyage 
through what is known about its subject. It is one of my favorite 10 books 
for both style and content. MB is a huge field (see next paragraph) but 
this book gets you to all the important stuff up to a few years prior to 
each of its editions. It would be nice to have something similar for computing.

Another interesting thing about Biology is how it has coped with the 
stifling influences of academia and intellectual bureaucracies. ULCA has 
one Computer Science Department, but (not counting Medical Departments) has 
amazingly 25 full departments of Biology! Every time the field gets bogged, 
the biologists on the fringes start a new department, and NIH has been 
willing to back them.

This should be compared with the opposite tendency in Computing and the 
refusal of NSF, etc., to fund break-away departments.


> >From  a talk by one of the people who invented the nicotine patch..
>his view is that most of the stable science wont be available for
>30-50 years because of stuff having to get past all the patent messes
>so that it really gets investigated  versus profitized.

I worry about this quite a bit in all fields. What a mess in the making!

Cheers,

Alan



>--
>Stephen J Smoogen.
>CSIRT/Linux System Administrator