C++ UFCS update

Timon Gehr via Digitalmars-d digitalmars-d at puremagic.com
Wed Feb 24 14:55:11 PST 2016


On 24.02.2016 23:10, Ola Fosheim Grøstad wrote:
> On Wednesday, 24 February 2016 at 21:56:12 UTC, Timon Gehr wrote:
>> The ultimate validity of the premise does not matter for what I
>> objected to. You were trying to paint Walter's position internally
>> inconsistent in a very specific way that just does not hold water.
>
> Sigh. I am not trying to paint anything, not even a bikeshed. I am
> merely observing the following:
>
> 1. I've never had any issues related to "<<" for iostream.
> ...

Note that I am not the one complaining about overloaded tokens.

> 2. I am rather annoyed by the inconsistent use of symbols in D.
> ...

It's not inconsistent. It's overloading at the lexical level that gets 
resolved at the level of the grammar.

> 3. Objecting to "<<" for iostream while using symbols incosistently
> yourself IS ironic.
>
>
> What is and isn't ironic, isn't a matter of «validity of the premise».

You misunderstand what that part of the sentence refers to. (It's what I 
quoted.)

> It is a matter of interpretation. Not your interpretation. Mine.
> ...

You might want to keep those aspects to yourself or express their 
subjectivity more explicitly if you don't want them to be mistaken for a 
point being made that resembles points being made using the same wording.

> Also, the usability issues concerning symbols isn't related to
> mathematical definitions, but is related to the mnemonics of the
> symbols, or the interpretation of them, by a human being. Not by a machine.
>
> As such a useful mnemonic for "<<" is that it is for moving stuff to the
> left. Which could work equally well for a stream as it does for bits.
>
> There is no apparent overlap between the mnemonics for "!" in the
> context of templates or bools. Same with "~", which in the context of C
> means "flip the bits",

In the context of C, "*" means either "dereference" or "multiply". "&" 
means either "take address" or "bitwise and".

> with the wave being a mnemonic for flipping.
> ...

Uh...

> I find it ironic that iostream is more consistent with the mnemonics of
> the symbols than D is. Deal with it.  Don't defend it. Fix it.
>

I'm afraid I personally don't care too much about any of that.

Also, I'm not defending anything. I was attacking annoying rhetoric.
Not anymore though. I'm done here.


More information about the Digitalmars-d mailing list