On 13/04/2013 3:48 AM, alex wrote:
> On 9 April 2013 11:27, Ross Bencina <rossb-lists_at_audiomulch.com> wrote:
>> "time-separable programming" vs "non-time-separable programming" perhaps?
>>
>> What are the options for injecting time back in? Seems like that got thrown
>> out when Alonzo showed up. And monads just treat it like a bag on the side.
>
> I wasn't really thinking in terms of modelling time in a program, but
> recognising that computer programming can be a form of use. That is,
> the timeline in which a program is written can be the same time in
> which it is used. This does mean that timestamps in revision control
> systems are relevant to execution too, so I guess then the RCS ends up
> being part of the model of time.
I was also thinking not so much as "modelling time in a program" but
modelling the whole process as a program. I.e. the program is a system
of communicating agents/systems, at least one of whom is the programmer.
The point I was trying to make was that if your formal system treats i/o
and time as something to isolate and exclude from "the program" then
perhaps you're going to have serious trouble concieving of a bigger
picture that incorporates human agency, time, and things like RCS into
the programming model.
Ross.
Received on Fri Apr 12 2013 - 21:56:00 BST