On Wed, May 20, 2015 at 5:29 PM, Matt Rice <[email protected]> wrote: > On Tue, May 19, 2015 at 10:26 PM, Jonathan S. Shapiro <[email protected]> > wrote: >> On Mon, May 18, 2015 at 9:52 AM, Keean Schupke <[email protected]> wrote: >>> >>> >>> On 18 May 2015 17:07, "Jonathan S. Shapiro" <[email protected]> wrote: >>> > >>> > On Sun, May 3, 2015 at 7:36 AM, Keean Schupke <[email protected]> wrote: >>> >> >>> >> To clarify I am suggesting a single, multi equals let as the way it >>> >> should be done, not necessarily the way other people do it. >>> > >>> > Clear, but not what we are going to do. >>> >>> Then I would vote for no lets at the top level at all. >> >> >> Right. I think I said that elsewhere already. Top-level forms don't enclose >> a scope, so they shouldn't overload the LET keyword. > > Egads, I think I finally understand what you mean by top-level let now > e.g. In the following code there is a top-level let which is never > closed by a subsequent 'in', static let moduleInstance = new > 'PriorityQueueModule () > > and thus the let remain open throughout the top-level scope > > https://github.com/jack-pappas/experimental-functors/blob/master/PriorityQueues/Functors.fs > > I'm not very fond of this, > > to try and restate my opinion above (in the idea that i'm not > fundamentally opposed to multi-let statements: > > let x = 0 > let y = 1 > in (x, y) // fine by me... > > let x = 0 > let y = x + 1 // evil! > in (x,y) > > let x = 0 > let y = x + 1 // I don't understand why this would be chosen over > multi-equals let, but I guess? > in (x,y) > > The F# stuff, seems to me to be breaking a very fundamental idea that > things which depend on the sequentially constructed computations above > them should be indented > to the right as well > > so when I see something to the effect of: > let x = 0 > > static member foo() > x + 1 > > I think the indentation scheme is all messed up, and it'd be better to > consider > let DECL of DECL end > let DECL in EXPR end
Apologies for the self response, It occurred to me that in SML this in general is already expressed through the notion of 'local' local DECL in DECL end I don't see it used or taught very much though _______________________________________________ bitc-dev mailing list [email protected] http://www.coyotos.org/mailman/listinfo/bitc-dev
