On Tuesday 13 March 2007 12:53, Eric Baum wrote: > Are you suggesting that there is in no sense a decision made that > there is a new font to be learned (and possibly reserving physical space).
Definitely not reserving space. I'm not even sure that the new capability would be in a physically different place than the old, rather a case of retraining the old modules. > In other words, you think that I learn individual characters locally, > not as a whole font? I think that what gets changed here is a small set of intermediate transforms that are used by all the word concepts. Fast reading goes word for word, not letter for letter. We're not assuming you're having any difficulty recognizing an A by itself. I just really have a hard time accepting that there is anything that can swap around major chunks of structure in the way I took your metaphor to imply. Changes in some modules that are used by lots of others could have that apparent effect (i.e. if out of all the companies in New York, just one decided to change how phone numbers worked -- but it was the phone company). Otherwise -- and this is really just a conjecture -- I'm betting changes are local and largely independent. (To answer Ben's question yet another way, that's a good way to control the size of the search space.) > Do you think I do that for a new language? I'm pretty sure that I at least do it one word at a time. Last year I drove all the way across Austria and halfway back before I finally realized that those signs I kept seeing: "Einbahn", meant "One Way." Eric Drexler tells the story of being in Japan and becoming frustrated because he couldn't even recognize a string of characters on a sign as one he had on paper and was specifically looking for. So he forced himself to "read" a japanese book for some hours and slowly formed enough pattern memory for the characters to be able to identify a word as the one he was looking for. Whereas I had a head start, since German uses a familiar alphabet, and thus I could drive around wondering why every little town had an Einbahn Street. > Maybe what you are saying is that there is a decision to split and > learn a new font, and apportion the same input output etc to it, > but this decision is in some sense made locally or maybe by some > emergent distributed computation. If so, I'd like more details on what > you have in mind if possible. But surely whatever mechanism you > attribute, it wants to be able to do similar kinds of splits > and adjustments at numerous different points in the program, > so it seems to be much like an instance creation. Maybe the point is, > it is like an instance, but what is a market system is the programmer? > I have no problem with that, but it doesn't seem to explain the > aversion to the OOP metaphor. As above, I just can't see it happening that way. It's like a market or evolution: no saltations. Optima are Pareto, not global. (Of course there are actually saltations in genetics but they're due to the same kind of wide influence of a single local change I mentioned above.) > I would like to understand your point better. What exactly about > objects is missing here, and why is it important? I guess I'm just thinking in a totally different paradigm. My model of what's going on is a lot more like Actors than OOP. Josh ----- This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?list_id=303
