how to communicate capabilities

hello dietrich.

spawning this into a different thread....

On 2015-04-19 01:59, Dietrich Schulten wrote:
> PS: the fact that container types proliferate is not a good sign IMO. I
> wonder why this is so and if it can or should be prevented. But that is
> a different issue.

i think this is a very astute and important observation.  you could ask 
this a bit more provocatively: is there "the one true model" for a given 
problem, and we should work on discovering/defining it and then it's 
just a matter of time until everybody shifts to the one true model?

to me, that's a fallacy that happens in many places. models are driven 
by needs, and since needs differ, models also will differ. one important 
need is to be interoperable which helps to keep the proliferation at 
bay, but that still doesn't mean that there is one model that will work 
well for everybody.

this means that we have to design things so that we don't assume that 
the one true model will emerge, and then when everybody uses it, things 
will work well. we should be able to support both scenarios, so that a 
world where two or three or four ways to do the same thing still is 
something that does not clash with our designs.

it just seems to me that this is another good angle to look at when 
considering whether the model should be represented in the uniform 
interface. i think it absolutely should, so that people can talk about 
it when using the uniform interface.

cheers,

dret.

-- 
erik wilde | mailto:dret@berkeley.edu  -  tel:+1-510-2061079 |
            | UC Berkeley  -  School of Information (ISchool) |
            | http://dret.net/netdret http://twitter.com/dret |

Received on Sunday, 19 April 2015 15:48:21 UTC