Re: question about linker

On 07/12/2024 15:57, David Brown wrote:

On 06/12/2024 19:41, Bart wrote:

My language:
>
   println =a,=b                  # 13 characters, 0 shifted
>

 Since some of us live in a world with more than one person - indeed, more than one country - you might like to know that "=" is shifted on many keyboards.

I want to make my comment both fit on one line, and line up with the other, so I had to leave out '(on my keyword)'. I'd hoped that was implied.
Yes, I design my stuff around a UK English layout. That's what I've used for 30 years, since I got my first portable PC. Before that, I either had to bring my own keyboard when visiting clients, or try to work with unfamiliar layouts on clients' computers, usually in NL or CH.

 

(And that space is optional!)

 The biggest irritation I have with your style, whether in your language or in C, is the lack of space.  It's a big key - it's easy to press.

Spaces were left out to do a fairer comparison. Spaces aren't a big deal, they are fairly easy to type, I even left one in in my example because it looked too odd otherwise, and I was still ahead anyway.

Good spacing habits (in the broadest sense) is the single biggest factor to readability of code, text, hand-writing, and pretty much anything else.
 The language Forth is not known for being easy to read amongst non- experts, but one thing it gets right is insisting that there are spaces between tokens.

That is misleading. Forth REQUIRES those spaces otherwise it wouldn't work; each line would be read as one giant token. It's not because it wants to instill good habits.
If you wrote C in the same style, it would look pretty weird:
    f ( ) ;
    f ( x , y , z ) ;
    ++ p . x ;
    q = ( void ) p ;
    label :
    f ( x , y , z ) ;
    a [ i + 1 ] ;

In C (assume the existence of stdio.h):
>
   printf("A=%lld B=%f\n",a,b);   # ~28 characters, 8 shifted
>
Enough said.

 I don't know what you are getting at, so I don't know if enough has been said or not.  If you are trying to suggest that the only thing you think is important about language syntax is how you print out two variables in as few keystrokes as possible, then I suppose you have succeeded in that goal, and can therefore rest your case.

Pretty much everything in C requires more typing and tends to need more punctuation. It is the latter that slows me down.
Some things may require more characters in my syntax but they are unshifted alphabetics which are quick to type. If working with someone else's codebase, I'm constantly have to shift case.
I know this won't cut any ice, as nothing I can possibly write will ever make the slightest bit of difference, but in C you write stuff like this:
    void F(uint32_t a1, uint32_t a2, uint32_t a3, uint32_t a4, uint32_t a5, uint32_t a6, uint32_t a7, uint32_t a8) {
while I would just write:
    proc F(u32 a1, a2, a3, a4, a5, a6, a7, a8) =
It's night and day.
(BTW how long does it take to verify that all of those parameters are of the same type? This comes from a real example.)
I'll read the rest of your post at some time in the future, after I've had a stiff drink.