Worst ideas/features in programming languages?

Timon Gehr timon.gehr at gmx.ch
Sun Jan 9 10:05:18 UTC 2022

On 1/8/22 02:39, Walter Bright wrote:
> I'm concerned about what would happen to the following.
> All these are considered equivalent currently:
> (a,b,c)
> (a,(b,c))
> ((a,b,c))
> ((a,b),(c))
> ((a),(b),(c))
> ...

x is equivalent to (x). (Always, and a design that changes that is broken.)

Therefore, `(a,b,c)`, `((a,b,c))` and `((a),(b),(c))` are the same, a 
triple with components a, b, and c.

> [Of course, using the __tuple syntax.]
> Does your proposal change that? I can't quite be sure.

`(a,(b,c))` is not the same as `((a,b),c)` in my proposals, those are 
always pairs, where the second and first component, respectively, are 
also pairs. Both are distinct from `(a,b,c)`, which is always a triple.

However, I don't propose to _change_ any existing auto-expanding 
behavior, I want to add tuple syntax and actual tuple behavior for 
something similar to a Phobos tuple. (I.e., based on a `struct` template.)

More information about the Digitalmars-d mailing list