5 ms·

Dr. Kay, if you're still following... then with singular respect and gratitude for your life-changing work and ideas, I would like to ask you one question. Is

by gavinpc 10y ago

Dr. Kay, if you're still following... then with singular respect and gratitude for your life-changing work and ideas, I would like to ask you one question. Is there a good way to use bad systems? Such as the web, which you describe as a “broken wheel,” lacking even a fraction of (e.g.) Englebart's vision. Or Linux, which you call “a budget of bad ideas.” (And no small budget, at that.) Or the iPad (and all common tablets, I assume), whose interface you call “brain dead.”[0] What should we do with these things? Are they dead ends? Are they good for anything? Can they not be salvaged incrementally? Here in the Hacker News community, where I am happy to see that my enthusiasm about your work and message is strongly shared, there is yet a huge amount of energy being poured into the wrong end of the low-pass filter, or, as you call it, “the world.” I know that we are not averse to learning curves, but maybe there is too much sunk cost to question what's already “working”? What should we do? One answer is to use bad systems to simulate better ones. But—when this is even feasible—it's always done at the cost of performance, and VPRI's publications make no secret of that. A proof-of-concept does not equal a product. And at any rate, most of us are not researchers. Because of this apparent dilemma, the exhiliration that I always feel when I hear you speak or read your writing is always tainted with a sense of despair. Is there any enlightened way to use today's systems (for example, as application developers), or should all of our efforts be directed at fixing (or indeed replacing) the systems themselves? Thank you again, for all that you've done and continue to do. [0] https://www.youtube.com/watch?v=gTAghAJcO1o&t=28m https://www.youtube.com/watch?v=gTAghAJcO1o&t=28m https://archive.org/details/130124AlanKayFull https://archive.org/details/130124AlanKayFull And others. I believe that these quotes are representative and not misused out of context.

alankay 10y ago
I just found your comment. One answer wrt e.g. Javascript is to use it as a "machine code" and just put a whole better designed thing on top of it. Or try to get the web world to get behind Native Client or something better that allows a protected sandbox to be really developed into new facilities. Another answer is to not to go back to Engelbart, but to at least start with large ideas like his, and try to think of the Internet and Personal Computing as something much more than a convenience -- but more as a "lifter" of humanity. What would that mean? Another ploy would be to simply think about what is needed close to the end-users and to follow that back towards the plug in the wall (one hint is that there is no need to encounter a 60s style "operating system" -- so what would be much better in a real Internet universe?) The main heuristic is to posit "I don't really know what I'm doing, and I'm not a strong enough thinker, and 'You can't learn to see until you admit you are blind, and etc." This is my starting place for trying to do real thinking i.e. we aren't as smart as we need to be -- so we have to put real work and method into thinking and design. Tony Hoare had a good observation. He said that debugging is harder than programming, so don't use all your abilities in programming, or you'll never get the program working. We can extend that into design. Design is difficult, but being able to assess one's designs is even harder -- leave something in reserve to avoid simply making things because we can.
- gavinpc 10y ago
  Thanks for your insights, Dr. Kay. I appreciate your taking the time to reply. For context, I was recently struggling with these questions in trying to rationalize a Shakespeare project (on current-day systems) as being good for humanity. In the section "why new media," I rely on your and Bret Victor's ideas as a standard for making that argument.[0] Thanks again. [0] http://gavinpc.com/project_willshake.pdf http://gavinpc.com/project_willshake.pdf

by gavinpc 10y ago

alankay 10y ago

I just found your comment. One answer wrt e.g. Javascript is to use it as a "machine code" and just put a whole better designed thing on top of it. Or try to get the web world to get behind Native Client or something better that allows a protected sandbox to be really developed into new facilities. Another answer is to not to go back to Engelbart, but to at least start with large ideas like his, and try to think of the Internet and Personal Computing as something much more than a convenience -- but more as a "lifter" of humanity. What would that mean? Another ploy would be to simply think about what is needed close to the end-users and to follow that back towards the plug in the wall (one hint is that there is no need to encounter a 60s style "operating system" -- so what would be much better in a real Internet universe?) The main heuristic is to posit "I don't really know what I'm doing, and I'm not a strong enough thinker, and 'You can't learn to see until you admit you are blind, and etc." This is my starting place for trying to do real thinking i.e. we aren't as smart as we need to be -- so we have to put real work and method into thinking and design. Tony Hoare had a good observation. He said that debugging is harder than programming, so don't use all your abilities in programming, or you'll never get the program working. We can extend that into design. Design is difficult, but being able to assess one's designs is even harder -- leave something in reserve to avoid simply making things because we can.

gavinpc 10y ago
Thanks for your insights, Dr. Kay. I appreciate your taking the time to reply. For context, I was recently struggling with these questions in trying to rationalize a Shakespeare project (on current-day systems) as being good for humanity. In the section "why new media," I rely on your and Bret Victor's ideas as a standard for making that argument.[0] Thanks again. [0] http://gavinpc.com/project_willshake.pdf http://gavinpc.com/project_willshake.pdf