Why so little parallelism?
|
From: "Peter "Firefly" Lund" on 17 Jan 2007 03:28 On Wed, 17 Jan 2007, Nicholas King wrote: > Sounds awfully useless to me. There is no magic way of making complex things > simple. The sooner we start properly educating the programmers the better > instead of looking for quick fix solutions and accepting mediocrity. That fancy Leibnitz and his calculus ideas seem awfully useless to me. What's wrong with the method of exhaustion? There is no magic way of making complex things simple. The sooner we start properly educating the geometers the better instead of looking for quick fix solutions and accepting mediocrity. When did notation ever make a difference to anybody? -Peter
From: "Peter "Firefly" Lund" on 17 Jan 2007 03:29 On Tue, 16 Jan 2007, Del Cecchi wrote: > The only way to get past complex problems is to get above them and hide > them. Some problems are hard because the abstractions used are suboptimal. -Peter
From: Ken Hagan on 17 Jan 2007 04:31 On Wed, 17 Jan 2007 02:52:32 -0000, Nicholas King <zeddie(a)internode.on.net> wrote: > There is no magic way of making complex things simple. But there are ways of making simple things complex. Zut!
From: Nick Maclaren on 17 Jan 2007 04:45 In article <Pine.LNX.4.61.0701170919590.31465(a)ask.diku.dk>, "Peter \"Firefly\" Lund" <firefly(a)diku.dk> writes: |> On Wed, 17 Jan 2007, Nicholas King wrote: |> |> > Sounds awfully useless to me. There is no magic way of making complex things |> > simple. The sooner we start properly educating the programmers the better |> > instead of looking for quick fix solutions and accepting mediocrity. |> |> That fancy Leibnitz and his calculus ideas seem awfully useless to me. |> What's wrong with the method of exhaustion? There is no magic way of |> making complex things simple. The sooner we start properly educating the |> geometers the better instead of looking for quick fix solutions and |> accepting mediocrity. |> |> When did notation ever make a difference to anybody? Now, that is an unreasonable analogy! Calculus enables a lot of things that are completely impractical using earlier methods. Once upon a time, I could use geometric methods to prove Pythagoras's theorem, but they would not extend to algebra in general. But I agree with you. Going from locks to transactional memory is like going from the pure labels of Fortran 66 to the control structures of Fortran 90. It doesn't enable anything that couldn't be done before, but it does make it easier to use and much more reliable. However, like that, it makes damn all difference to sorting out the higher level problems. Regards, Nick Maclaren.
From: "Peter "Firefly" Lund" on 17 Jan 2007 04:49
On Wed, 17 Jan 2007, Nick Maclaren wrote: > Now, that is an unreasonable analogy! Calculus enables a lot of things > that are completely impractical using earlier methods. Or impossible. Yeah, I know. Maybe I should have used number systems instead, or the notion of algebraic equations. -Peter |