Here's a lively debate over in comp.lang.c++.moderated people might be interested in:

http://groups.google.com/group/comp.lang.c++.moderated/browse_thread/thread/60117e9c1cd1c510/c92f7fd0dc9fedd1?lnk=st&q=safer+better+c%2B%2B&rnum=1&hl=en
#c92f7fd0dc9fedd1

Walter Bright wrote:
> Here's a lively debate over in comp.lang.c++.moderated people might be
> interested in:
> 
> http://groups.google.com/group/comp.lang.c++.moderated/browse_thread/thread/60117e9c1cd1c510/c92f7fd0dc9fedd1?lnk=st&q=safer+better+c%2B%2B&rnum=1&hl=en
> #c92f7fd0dc9fedd1

Wow.  That thread has positively exploded in the past few days.  Looks like I've got some catching up to do :-)


Sean

In article <dnhvoi$dja$1@digitaldaemon.com>, Walter Bright says...
>
>Here's a lively debate over in comp.lang.c++.moderated people might be interested in:
>
>http://groups.google.com/group/comp.lang.c++.moderated/browse_thread/thread/60117e9c1cd1c510/c92f7fd0dc9fedd1?lnk=st&q=safer+better+c%2B%2B&rnum=1&hl=en
>#c92f7fd0dc9fedd1
>

That didn't format correctly in my browser, so here's a tiny version for folks should they have the same problem:

http://tinyurl.com/9hl3r


- EricAnderton at yahoo

In article <dnhvoi$dja$1@digitaldaemon.com>, Walter Bright says...
>
>Here's a lively debate over in comp.lang.c++.moderated people might be interested in:
>
>http://groups.google.com/group/comp.lang.c++.moderated/browse_thread/thread/60117e9c1cd1c510/c92f7fd0dc9fedd1?lnk=st&q=safer+better+c%2B%2B&rnum=1&hl=en
>#c92f7fd0dc9fedd1
>

Interesting!  One choice quote, if I may:

----------------------
From: Kai-Uwe Bux <jkherci...@gmx.net> - Find messages by this author Date: 7 Dec 2005 11:00:04 -0500
----------------------
Peter Most wrote:
> Hello everybody,

> I would like to get some opinions about whether it would make sense to change the C++ design principles, so it would become safer for beginner?

[snipped: suggestion to swap at() and operator[] in std::vector]

I do not think this is a good idea. I would rather have operator[] do something like:

reference operator[] ( size_type pos ) {
assert( pos < this->size() );
return ...;
}

so that if DEBUG is defined I will have bug detection, and I do not have a performance penalty in production code.
----------------------

The rest of the thread reminds me of the debate of '04, where folks* were going back and forth over Error vs Exception and wether or not the current approach is valid.

(*well, okay, it was mostly Matthew)

- EricAnderton at yahoo

In article <dnhvoi$dja$1@digitaldaemon.com>, Walter Bright says...
>
>Here's a lively debate over in comp.lang.c++.moderated people might be interested in:
>
>http://groups.google.com/group/comp.lang.c++.moderated/browse_thread/thread/60117e9c1cd1c510/c92f7fd0dc9fedd1?lnk=st&q=safer+better+c%2B%2B&rnum=1&hl=en
>#c92f7fd0dc9fedd1
>

After reading some of the threads, and seeing mention of D in there basically just get ignored, it becomes apparent that one could come up with the better mousetrap and still the very people who should take notice won't even acknowledge the existance of something that quite probably is not only better but readily available.

Take me for example, I first noticed D two years ago while heavily involved in C++ development. Primarily four thoughts came to mind back then, and it took me a year of ignorance before I finally started to look seriously at D:

1) Garbage collected, ergo slow (wrong!).
2) The OOP support seems a lot like Java, ergo slow (wrong!).
3) If I already know (or think I know) a lot about C++, and D doesn't offer any
features that I can't hack together in some way in C++, why switch?
4) If I need 'safer', why D instead of Java or C#?

Ultimately the first two assumptions ended up being just plain wrong (on several levels) and are probably more important to me than many people anyway.

But the last two I think still have some merit w.r.t. why people will not take the leap and try D.

The arguments in favor of D for 3) and 4) such as:
- productivity
- power
- speed
- improved safety
- unittest
- DbC
- C library compatibility
- built-in AA's, first class arrays, foreach, etc., etc...
- C-like support for naked pointers
- Built-in inline assembler spec.
all rolled into one language only become apparent after trying D.

So it seems D is stuck in the proverbial 'catch 22' situation here, where the real merits of the language only become apparent after learning how to use it, yet those who need to try it won't because of their (understandable) assumptions...

How can D break out of this situation?

"Dave" <Dave_member@pathlink.com> wrote

> How can D break out of this situation?

Years ago I would have waxed on about the merits of grass-roots, technical superiority, clarity of focus and so on. That's all so wrong ;-) When D is deemed "ready", what it needs is widespread, persistent, and memorable ...

marketing


I don't know how that can be done on a small budget, even with the web. Do you know? Or is there some obvious way to get that sort of thing funded? Through licensing perhaps?

Kris wrote:
> "Dave" <Dave_member@pathlink.com> wrote
> 
> 
>>How can D break out of this situation?
> 
> 
> Years ago I would have waxed on about the merits of grass-roots, technical superiority, clarity of focus and so on. That's all so wrong ;-) When D is deemed "ready", what it needs is widespread, persistent, and memorable ...
> 
> marketing
> 
> 
> I don't know how that can be done on a small budget, even with the web. Do you know? Or is there some obvious way to get that sort of thing funded? Through licensing perhaps?
> 
> 

It would be nice if everyone just 'got it,' but most people want to stay with the crowd. Grass roots works but it is slow because it is not mass spoon fed into the population like marketing is.

I don't think it is time to worry about publicity now, if people started taking D seriously today they might be put off by lack of compiler options and an 'incomplete' language. Better leave it to the adventerous and those who do understand benefits.

Here are some things I forsee as having a big enough impact for the masses to start noticing D.

1) Stable 1.0 compiler, for those who know about D but don't want the risk of a beta compiler

2) When the time is ready, another Slashdot article.

3) Acceptance amonst linux distro's.

4) A book or two

If D doesn't start catching fire after that, then I'd start to worry why people are not accepting it.

Of course, lets not hope this slightly modified matrix quote is true :o

"Did you know that the first version of C++ was designed to be perfect? Where none suffered, where everyone would be happy. It was a disaster. No one would accept it. Entire programmers were lost. Some believed we lacked the means to describe your perfect language. But I believe that, as a species, human beings define their reality through suffering and misery. The perfect language was a dream that your primitive cerebrum kept trying to wake up from. Which is why C++ was redesigned to this: the peak of its popularity." - Agent Smith

In article <dni880$11rv$1@digitaldaemon.com>, Kris says...
>
>"Dave" <Dave_member@pathlink.com> wrote
>
>> How can D break out of this situation?
>
>Years ago I would have waxed on about the merits of grass-roots, technical superiority, clarity of focus and so on. That's all so wrong ;-) When D is deemed "ready", what it needs is widespread, persistent, and memorable ...
>
>marketing
>
>
>I don't know how that can be done on a small budget, even with the web. Do you know? Or is there some obvious way to get that sort of thing funded? Through licensing perhaps?
>
>


How about a programming challenge? Develop a programming problem that requires the contestant to program against a D language API.

clayasaurus wrote:

> Here are some things I forsee as having a big enough impact for the masses to start noticing D.

See also: http://prowiki.org/wiki4d/wiki.cgi?HelpDProgress

--anders

On Sun, 11 Dec 2005 23:28:12 +0000, BCS wrote:

> In article <dni880$11rv$1@digitaldaemon.com>, Kris says...
>>
>>"Dave" <Dave_member@pathlink.com> wrote
>>
>>> How can D break out of this situation?
>>
>>Years ago I would have waxed on about the merits of grass-roots, technical superiority, clarity of focus and so on. That's all so wrong ;-) When D is deemed "ready", what it needs is widespread, persistent, and memorable ...
>>
>>marketing
>>
>>
>>I don't know how that can be done on a small budget, even with the web. Do you know? Or is there some obvious way to get that sort of thing funded? Through licensing perhaps?
>>
>>
> 
> 
> How about a programming challenge? Develop a programming problem that requires the contestant to program against a D language API.

A programming challenge is a good idea.
What about setting up some bounties for writing some code modules for the
D std library!

Or maybe make a website where people can make bounty jars for there
favorite code module and then let everyone interested in the module
contribute with donations.