The definition of systems programming

In Uncategorized on November 24, 2009 at 2:40 pm

All of this is gathered from this LtU post.

So far my impression is that it looks like a decent set of tradeoffs overall, though it’s not clear to me why this is a "systems language" in a way that cannot be satisfied by higher-level languages.

comment link

"Systems programs" means "programs where the constant factors are important".

comment link

I would also add, "precise control over representation".

comment link

I like the BitC literature’s observation that systems programming takes place in a context in which representations are prescriptive rather than descriptive. That nailed it for me in a way that no previous attempts at definition had. It also gave me a pivot on which to swing the observation that static typing had previously broken down into two camps: the prescriptivists ("types define how bits are laid out in memory") and the descriptivists ("types define legal operations on terms of that type"), with a lot of people who don’t see the value in static typing essentially being prescriptivists, whereas "typeful programming," etc. comes from the descriptivist camp. What’s genius—and, so far, as far as I can tell, overlooked—about BitC is that it ends the divorce between prescriptive and descriptive static typing.

comment link

Page last updated on November 24, 2009

%d bloggers like this: