[...]

Whisper syntax may not have problems according to order of evaluation, but it's
a really inconvenient notation since it's clearly not what opCall was "created"
for.
Seems nothing but UNNATURAL notation and really looks like a hack to me. That's
the original reason of this thread. Someone was confused by the language abuse
"whisper syntax" is. So, I can't agree for it to be encouraged. (Plus I don't
like it at all :-) )

OTOH, Manfred, you can't be serious. I really don't care what the specs states for because it may be still erroneous or incomplete (someday it'll stop to be this way when it grows up to a mature documentation).

Here and in the China 'x.a().b().c()' can't be evaluated in some odd way like if
it was another algebraic expression (being of course that these are non-static
methods). It's clearly equal to 'c(b(a(x)))' and if it's not, think I'll have to
start over again my CS course of studies because I misconceived everything since
the beginning (it's just a speech, no way i'd do that :-) ).

Tom

In article <dng1u9$78g$1@digitaldaemon.com>, Tom says...
>
>[...]
>
>Whisper syntax may not have problems according to order of evaluation, but it's
>a really inconvenient notation since it's clearly not what opCall was "created"
>for.
>Seems nothing but UNNATURAL notation and really looks like a hack to me. That's
>the original reason of this thread. Someone was confused by the language abuse
>"whisper syntax" is. So, I can't agree for it to be encouraged. (Plus I don't
>like it at all :-) )
>

I can't keep this back any longer...

So you don't like it.  That summarizes your opinion.

Abuses opCall? No it doesn't.  What do you think opCall is for?  Innovation? Originality?  D is about being creative and "whisper" syntax is trying out something new.  If Walter wanted to pigeon-whole people into a singular programming paradigm, he would have eliminated features like operator overloading and templates.

C++ does even worse, in my opinion, when it uses "<<" and ">>" for streams. "Whisper" on the other hand uses syntax that is common to all aspects of programming expressions: "()"

Unnatural notation?  What is natural/unnatural about it? From my perpsective it looks expressive and clean... and quite natural.  If it's unnatural from what your used to, well that's not surprising. New ideas have a nasty habit of being that way.

All your criticisms have no bearing on the argument discussed in this thread . And contrary to what your topic title intends, your opinions -- as is the case of most subjective opinions -- does nothing but fuel the already raging fire here: You chastise Kris for creating an orignal idea;  you chastise Manfred for not making any sense.  Do you think you are going to clear up this issue this way?  You certainly won't gain the respect of either individual, I assure you.

In short, give people some space.  You may choose not to like these new ideas, but be open to innovation and let others make up there minds.  Forcing your opinion on people won't help.  You're certainly free to give your opinions on the matter, but providing statements without clearly thought out reasons is of no benefit to your readers.

So describe why it's unnatural.  By what standard do you make this assessment? Why is it opCall abuse?  Is there a rule somewhere that states this is so?  Why does it look like a hack?  Have you examined the mango source?  Have you seen if the code looks like a hack in this area? (I assure you it does not).

The root of this whole issue _might_ merely be D Language documentation/specification, which is Walter's realm.  Otherwise I see no other problems with it... and I think we should continue to _encourage_ all sorts of innovation to take D to the limit of it's abilities.  This process alone will be the means of guaranteeing that the language specification is purged of all ambiguities.

-JJR

In article <dng5un$aem$1@digitaldaemon.com>, John Reimer says...
>
>In article <dng1u9$78g$1@digitaldaemon.com>, Tom says...
>>
>>[...]
>>
>>Whisper syntax may not have problems according to order of evaluation, but it's
>>a really inconvenient notation since it's clearly not what opCall was "created"
>>for.
>>Seems nothing but UNNATURAL notation and really looks like a hack to me. That's
>>the original reason of this thread. Someone was confused by the language abuse
>>"whisper syntax" is. So, I can't agree for it to be encouraged. (Plus I don't
>>like it at all :-) )
>>
>
>I can't keep this back any longer...
>
>So you don't like it.  That summarizes your opinion.
>
>Abuses opCall? No it doesn't.  What do you think opCall is for?  Innovation? Originality?  D is about being creative and "whisper" syntax is trying out something new.  If Walter wanted to pigeon-whole people into a singular programming paradigm, he would have eliminated features like operator overloading and templates.

Indeed. I'd go even farther and assert that Walter even endorsed the "whisper" syntax (back in 2003). See http://www.digitalmars.com/d/archives/18945.html

jcc7

J C Calvarese wrote:

[...]
> Indeed. I'd go even farther and assert that Walter even endorsed the "whisper" syntax (back in 2003). See http://www.digitalmars.com/d/archives/18945.html

... and my comments on whisper syntax not beeing safe and how to enhance the language to make them useful reach as far back as http://www.digitalmars.com/d/archives/26477.html

In article <dng5un$aem$1@digitaldaemon.com>, John Reimer says...
>
>In article <dng1u9$78g$1@digitaldaemon.com>, Tom says...
>>
>>[...]
>>
>>Whisper syntax may not have problems according to order of evaluation, but it's a really inconvenient notation since it's clearly not what opCall was "created" for.

I was clearly wrong in this one.

>>Seems nothing but UNNATURAL notation and really looks like a hack to me.

Just unnatural would have been enough in this sentence :)

>> That's
>>the original reason of this thread. Someone was confused by the language abuse "whisper syntax" is. So, I can't agree for it to be encouraged. (Plus I don't like it at all :-) )

People may encounter whisper syntax confusing, as the original post states I think. Not an abuse, don't know what was happening to me when I wrote about the abuse. I'm not in my best moment.

>
>I can't keep this back any longer...

Don't restrain your anger for so long :). It was not my intention to offend you nor anybody in this matter. Though reading my post again makes me realize it was a little too rough and I beg your pardon.

>So you don't like it.  That summarizes your opinion.
>
>Abuses opCall? No it doesn't.  What do you think opCall is for?  Innovation? Originality?  D is about being creative and "whisper" syntax is trying out something new.  If Walter wanted to pigeon-whole people into a singular programming paradigm, he would have eliminated features like operator overloading and templates.

Sorry on this, not an abuse, my mistake.

>C++ does even worse, in my opinion, when it uses "<<" and ">>" for streams.

Why?!

>"Whisper" on the other hand uses syntax that is common to all aspects of programming expressions: "()"

And so what?

Maybe the fact that it's a new syntax upsets me a little, this could be what I feel about whisper. It seems unnatural maybe just because of this, I admit it, maybe I should open my mind in this issue, i'll do my best.

>Unnatural notation?  What is natural/unnatural about it? From my perpsective it looks expressive and clean... and quite natural.  If it's unnatural from what your used to, well that's not surprising. New ideas have a nasty habit of being that way.
>
>All your criticisms have no bearing on the argument discussed in this thread.

So how is that?

>And contrary to what your topic title intends, your opinions -- as is the case of most subjective opinions -- does nothing but fuel the already raging fire here: You chastise Kris for creating an orignal idea;  you chastise Manfred for not making any sense.  Do you think you are going to clear up this issue this way?  You certainly won't gain the respect of either individual, I assure you.

You're right that this would not dim the fire of the discussion. That wasn't my intention at all. I was wrong with the title of the post, it wasn't to be taken so seriously and "chastise" IS NOT the word to describe my intentions over Kris's & Manfred's opinions/posts. Wasn't my intention to be so rude in my criticism, just imagine a smile in my face while I was writing it :D. I hope they doesn't take it that seriously.

>In short, give people some space.  You may choose not to like these new ideas, but be open to innovation and let others make up there minds.  Forcing your opinion on people won't help.  You're certainly free to give your opinions on the matter, but providing statements without clearly thought out reasons is of no benefit to your readers.
>
>So describe why it's unnatural.  By what standard do you make this assessment? Why is it opCall abuse?  Is there a rule somewhere that states this is so?  Why does it look like a hack?  Have you examined the mango source?  Have you seen if the code looks like a hack in this area? (I assure you it does not).

I didn't say Mango looks like a hack nor that the code is a hack. I don't really know and I don't care at all. I just criticized the whisper notation in general.


>The root of this whole issue _might_ merely be D Language documentation/specification, which is Walter's realm.  Otherwise I see no other problems with it... and I think we should continue to _encourage_ all sorts of innovation to take D to the limit of it's abilities.  This process alone will be the means of guaranteeing that the language specification is purged of all ambiguities.

I have to agree, at last :).

Tom
PS: I'll be a lot more careful with my opinions and with the way I express them in the future.

Tom,

Your response to my response took me by surprise.

Thank you for bearing with me also.  You showed much more forbearance than I expected from reading your original post.  I obviously misjudged the person to whom I was responding.  Obviously my annoyance with the whole thing wasn't helping me either. :-D

Take care,

John Reimer

"J C Calvarese" <technocrat7@gmail.com> wrote
> Indeed. I'd go even farther and assert that Walter even endorsed the
> "whisper"
> syntax (back in 2003). See
> http://www.digitalmars.com/d/archives/18945.html


Well, I'll be dipped in dogshit ... had no idea this had happened at all. It's funny reading through it, since all the same issues and options occured within mango.io early on (no surprise there!); right down to the notion of using '~' and the '[' ']' operators.

Turned out that reading needs a bit more hand-holding than writing, and using opCall allows for more than one parameter per call. That's why Mango settled on it as a 'standard'.

Got to hand it to you ... if anyone ever needs to find anything at all D related, just ask JCC ;-)

(a private jest)

In article <dnghut$ljl$1@digitaldaemon.com>, Kris says...

>Got to hand it to you ... if anyone ever needs to find anything at all D related, just ask JCC ;-)
>
>(a private jest)
>
>

I'll say!  Justin is pratically a bloodhound!  I don't know how he does it.

-JJR

Manfred Nowak wrote:
> J C Calvarese wrote:
> 
> [...]
> 
>>Indeed. I'd go even farther and assert that Walter even endorsed
>>the "whisper" syntax (back in 2003). See
>>http://www.digitalmars.com/d/archives/18945.html 
> 
> 
> .... and my comments on whisper syntax not beeing safe and how to enhance the language to make them useful reach as far back as
> http://www.digitalmars.com/d/archives/26477.html  

OK, we will not discuss whisper syntax because it is safe, but i would like to make a comment on your syntax proposal.

I like it and i think it could be usefull, but you made a fatal mistake in those posts that you didn't explain why would anyone want it. This NG likes examples.

If i was proposing something like that (like I am now) i would have said:

Often i find myself wanting to pass a group of varargs to a function.
How about this new syntax:

int compareGroupsOfNumbers(int[] nums1...; int[] nums2...)
{
  if(nums1.length > nums2.length) ....
}

Then you could call this method this way:
compareGroupsOfNumbers(1,2,3,4; 5,6,7,8);

This syntax change doesn't interfer with other function delcarations and calls in the language,  and there could be a few simple rules about using multiple varargs lists (for example:
* if there is only one vararg list, everything like before)
* if there are more than one or there is just one but not the last argument then it must be followed by ;

PS I really like this idea, why not revive it at least as a possible 2.0 feature?

Ivan Senji wrote:

[...]
> I like it and i think it could be usefull, but you made a fatal mistake in those posts that you didn't explain why would anyone want it. This NG likes examples.

I described the usefulness in four postings, starting with http://www.digitalmars.com/d/archives/26867.html


[...]
> PS I really like this idea, why not revive it at least as a possible 2.0 feature?

This is a must for 1.0 unless D is able to detect the ambiguities reported in http://www.digitalmars.com/drn-bin/wwwnews?digitalmars.D.bugs/4713

-manfred