From: Ulrich Hobelmann
Subject: Re: A Semicolon-Syntax
Date: 
Message-ID: <3vb99nF15arf9U1@individual.net>
Stefan Ram wrote:
> < < setq a b >
>   < setq c d >>

That's fine for sequences, but what about stuff like

((find-function key) a b) in Lisp?

My point is that your above list has another list as its first element.

> < setq a b;
>   setq c d >

Ok, now you have a separator character.  Is this just for sequences, or 
could you write:
<+ + a b; - c d; 5>
?

I think it'd get unreadable quickly.

>   Where the meaning of ";" is to mark the part preceding it
>   (up to "<" or a previous ";") as a sublist, i.e.,
> 
> < a b c d; e f g; h i >
> 
>   means
> 
> << a b c d >< e f g >< h i >>

But all of this just for sequences?

How much sequential code do you really write in Lisp?  Is it that bad to 
have it read
(progn
   (foo bar baz)
   (do something else))
?

>   Possibly this could be extended to "," so that:

Uuuh, more complicated.  You'll quickly end up with something as ugly as 
  C.  If you want a non-lispy syntax, look at SML for inspiration, or at 
Haskell.  They're not perfect, but they're better than the rest, IMHO.

Of course that means no more macros, easy quoted data structures ...

-- 
The road to hell is paved with good intentions.

From: Ulrich Hobelmann
Subject: Re: A Semicolon-Syntax
Date: 
Message-ID: <3vb9c4F15arf9U2@individual.net>
Ulrich Hobelmann wrote:
> Stefan Ram wrote:
>> < < setq a b >
>>   < setq c d >>
> 
> That's fine for sequences, but what about stuff like
> 
> ((find-function key) a b) in Lisp?

Sorry, of course in CL this'd be
(funcall (find-function key) a b)

-- 
The road to hell is paved with good intentions.
From: Ulrich Hobelmann
Subject: Re: A Semicolon-Syntax
Date: 
Message-ID: <3vbifdF148ernU1@individual.net>
Stefan Ram wrote:
>> Ok, now you have a separator character.  Is this just for sequences, or 
>> could you write:
>> <+ + a b; - c d; 5>
>> ?
> 
>   This could be written and would mean:
> 
> < < + + a b > < - c d > < 5 >>

No, what I meant is:
(+ (+ a b) (- c d) 5)

So I'm using the ";" to avoid having to write (), which I assume was 
your purpose of introducing the ";".  In this case it's rather 
unreadable (as we can see by the misunderstanding already).

>   Actually I do not use this for Lisp, but for example, I have
>   a kind of Java-preprocessor converting
> 
> quit = < < &verbose ["quitting"] >< &System.exit 99 >> 
> 
>   to
> 
> public static void quit ()
> { verbose( "quitting" ); 
>   System.exit( 99 );  }
> 
>   With the new semicolon-syntax I could convert
> 
> quit = < < &verbose ["quitting"] >< &System.exit 99 >> 
> 
>   to
> 
> quit = < &verbose ["quitting"]; &System.exit 99 > 
> 
>   and the last line looks more readable to me.

Ok, for a heavily imperative language like Java that might make sense...

-- 
The road to hell is paved with good intentions.
From: matteo d'addio 81
Subject: Re: A Semicolon-Syntax
Date: 
Message-ID: <1133601284.130096.255460@f14g2000cwb.googlegroups.com>
Stefan Ram ha scritto:

>   Using the "." this could be written as:
>
> < progn .
>   foo bar baz;
>   do something else >
>
>   Where material preceding the first "." is added to the top
>   list.

What about:

< progn:
  foo bar baz;
  do something alse >

Looks good.

matteo
From: Ulrich Hobelmann
Subject: Re: A Semicolon-Syntax
Date: 
Message-ID: <3vdepcF153mvaU1@individual.net>
matteo d'addio 81 wrote:
> What about:
> 
> < progn:
>   foo bar baz;
>   do something alse >

This kind of reminds me of Smalltalk/ObjC-style function names (like 
insert:bla atPlace:bla), with the ";" having similar function to the "," 
in ObjC function with variable argument lists.

It'd be interesting to have a Lisp with that kind of syntax (i.e. named 
arguments); sure, we have keywords, but they're different (even though 
more powerful sometimes).  I know all you working Common Lispers are 
already groaning by now, so I stop :D

-- 
Majority, n.: That quality that distinguishes a crime from a law.
From: Ulrich Hobelmann
Subject: Re: A Semicolon-Syntax
Date: 
Message-ID: <3vdo5sF159ea4U1@individual.net>
Stefan Ram wrote:
>   The syntax I am using ("Unotal") already has named entries.

Ah, only heard of it.  Looks interesting...

-- 
Majority, n.: That quality that distinguishes a crime from a law.
From: Giorgos Keramidas
Subject: Re: A Semicolon-Syntax
Date: 
Message-ID: <864q5apdxy.fsf@flame.pc>
On 2 Dec 2005 16:18:44 GMT, ···@zedat.fu-berlin.de (Stefan Ram) wrote:
> Ulrich Hobelmann <···········@web.de> writes:
>>Stefan Ram wrote:
>>> < < setq a b >
>>>   < setq c d >>
>>That's fine for sequences, but what about stuff like
>>((find-function key) a b) in Lisp?
>
>   If the special semicolon syntax can not be used,
>   one can always use the canonical syntax without ";",
>   i.e.,
>
> < funcall < find-function key > a b >

What would this mean then?

  < funcall ; < find-function-key > a b >

is it equivalent to this

  (funcall nil ( find-function-key ) a b )

or this?

  (funcall ( find-function-key ) a b )

or even this?

  (funcall nil ( ( find-function-key ) a b ) )

:-(