Re: Not enough parallelism in programming
- From: "Robert Myers" <rbmyersusa@xxxxxxxxx>
- Date: 12 Sep 2005 08:50:31 -0700
Greg Lindahl wrote:
> In article <Fb6dncT1wJR5_bjeRVn-iw@xxxxxxxxxxxxxxx>,
> David C. DiNucci <dave@xxxxxxxxxx> wrote:
> >That you can express some aspect of a program as a graph is very
> >different from saying that the program is the graph or vice versa.
>
> But I never said programs were graphs! That's a distinction you
> brought up. In this case the programs are C++ with a couple of
> keywords added. You could also run a real graph language if you care
> to invent one. I don't care about the disctinction you're trying
> to make.
>
You're making a distinction that's just too cute (a language that
"does" executable graphs, as opposed to a language whose programs are
written as graphs, presumably with a formal graph grammar, although I
did not make that clear).
SF>>> OK, so if programmers can think in 2-D, what do they need in a
language
SF>>> to reasonably express those thoughts?
RM>>"Executable graphs" is the obvious answer.
GL> So what don't you like about existing languages that do executable
graphs? One such is Mentat.
If the original point was *thinking* in a particular way, then programs
written in C++ that happen to be analyzable as graphs is a bizarre
example to put forward, if the point of my original proposal was to
think in graphs (and that was my point).
RM
.
- References:
- Re: Not enough parallelism in programming
- From: Greg Lindahl
- Re: Not enough parallelism in programming
- From: Greg Lindahl
- Re: Not enough parallelism in programming
- Prev by Date: Re: interrupting for overflow and loop termination
- Next by Date: Re: interrupting for overflow and loop termination
- Previous by thread: Re: Not enough parallelism in programming
- Next by thread: Re: Not enough parallelism in programming
- Index(es):
Relevant Pages
|