Jonathan, thank you for your feedback.
Jonathan Leffler wrote [to just the ttm list]:
Sounds reasonable to me...there are languages that do that already -
Pascal and family.
var
x : int;
y, z : float;
For declared-name/data-type pairs I had already been doing this too; eg:
var x : Int
var y : Float
var z : Float
... and I have no intention of changing the ordering there.
I don't see an explicit side-by-side comparison of the 'current' and
'suggested' notations. I think that would have been helpful.
I'll repeat the relevant portions from my first message, with minor changes,
demonstrating the flip of a material-kind/declared-name flip.
The current way:
function cube (Int <-- topic : Int) {
topic exp 3
}
A possible new way:
cube ::= function (Int <-- topic : Int) {
topic exp 3
}
or:
let cube ::= function (Int <-- topic : Int) {
topic exp 3
}
While the exact syntax can possibly change, I would be happy with the new syntax
given above as-is. Note that the last 2 are essentially the same format, such
that "let" is an optional keyword meant to aid readability but there is no
ambiguity without it.
And of course, Muldis D lets you format your whitespace however you want.
Note also that the newer format might resemble an EBNF grammar token somewhat,
like my existing uses of ::= ... such as this is intentional, the idea that this
being a format for declaring things that has precedents.
Thank you in advance for any further feedback.
-- Darren Duncan
_______________________________________________
muldis-db-users mailing list
muldis-db-users@mm.darrenduncan.net
http://mm.darrenduncan.net/mailman/listinfo/muldis-db-users
