www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - [OT] interview with Stroustrup

reply =?iso-8859-1?q?Knud_S=F8rensen?= <12tkvvb02 sneakemail.com> writes:
Hi 

I stumbled over this interview with Stroustrup


http://artlung.com/smorgasborg/Invention_of_Cplusplus.shtml
Nov 15 2005
next sibling parent reply "Zz" <junkie noware.com> writes:
Not real.

"Knud Sřrensen" <12tkvvb02 sneakemail.com> wrote in message 
news:pan.2005.11.15.18.59.07.75576 sneakemail.com...
 Hi

 I stumbled over this interview with Stroustrup


 http://artlung.com/smorgasborg/Invention_of_Cplusplus.shtml 

Nov 15 2005
next sibling parent Sean Kelly <sean f4.ca> writes:
No, but pretty funny.  And a lot of the comments are fairly accurate 
regarding some of the code I've been asked to maintain over the years. 
I like C++, but the range of programming techniques it supports doesn't 
make it the ideal language for poorly managed teams, or teams of 
unskilled programmers--both of which are quite common.

Zz wrote:
 Not real.
 
 "Knud Sřrensen" <12tkvvb02 sneakemail.com> wrote in message 
 news:pan.2005.11.15.18.59.07.75576 sneakemail.com...
 Hi

 I stumbled over this interview with Stroustrup


 http://artlung.com/smorgasborg/Invention_of_Cplusplus.shtml 


Nov 15 2005
prev sibling parent James Dunne <james.jdunne gmail.com> writes:
Zz wrote:
 Not real.
 
 "Knud Sřrensen" <12tkvvb02 sneakemail.com> wrote in message 
 news:pan.2005.11.15.18.59.07.75576 sneakemail.com...
 
Hi

I stumbled over this interview with Stroustrup


http://artlung.com/smorgasborg/Invention_of_Cplusplus.shtml 


Surprisingly accurate though, no? =P
Nov 15 2005
prev sibling next sibling parent reply clayasaurus <clayasaurus gmail.com> writes:
Knud Sřrensen wrote:
 Hi 
 
 I stumbled over this interview with Stroustrup
 
 
 http://artlung.com/smorgasborg/Invention_of_Cplusplus.shtml

Oh great. Now everytime I go to program C++, it actually /feels/ like the language was designed to deceive me.
Nov 16 2005
next sibling parent Georg Wrede <georg.wrede nospam.org> writes:
clayasaurus wrote:
 Knud Sřrensen wrote:
 
 Hi
 I stumbled over this interview with Stroustrup


 http://artlung.com/smorgasborg/Invention_of_Cplusplus.shtml

Oh great. Now everytime I go to program C++, it actually /feels/ like the language was designed to deceive me.

Welcome to the club!
Nov 16 2005
prev sibling parent reply Tomás Rossi <Tomás_member pathlink.com> writes:
In article <dlgvac$tql$1 digitaldaemon.com>, clayasaurus says...
Knud Sřrensen wrote:
 Hi 
 
 I stumbled over this interview with Stroustrup
 
 
 http://artlung.com/smorgasborg/Invention_of_Cplusplus.shtml

Oh great. Now everytime I go to program C++, it actually /feels/ like the language was designed to deceive me.

Are you kidding? Didn't you believe it right? It's a fake... not real! It's just a joke, don't be afraid. :) PS: Somebody will have to put an after comment to that interview, telling it's just a fake or we'll see many programmers killing themselves in the next days. I would commit suicide if that article were true :P Tom
Nov 17 2005
parent reply huh <huh_member pathlink.com> writes:
No way! Its real.

In article <dlhnif$j3e$1 digitaldaemon.com>, Tomás Rossi says...
In article <dlgvac$tql$1 digitaldaemon.com>, clayasaurus says...
Knud Sřrensen wrote:
 Hi 
 
 I stumbled over this interview with Stroustrup
 
 
 http://artlung.com/smorgasborg/Invention_of_Cplusplus.shtml

Oh great. Now everytime I go to program C++, it actually /feels/ like the language was designed to deceive me.

Are you kidding? Didn't you believe it right? It's a fake... not real! It's just a joke, don't be afraid. :) PS: Somebody will have to put an after comment to that interview, telling it's just a fake or we'll see many programmers killing themselves in the next days. I would commit suicide if that article were true :P Tom

Nov 17 2005
parent reply Tomás Rossi <Tomás_member pathlink.com> writes:
In article <dlhq0n$pns$1 digitaldaemon.com>, huh says...
No way! Its real.

Real my but! http://www.alledegodenavnevaroptaget.dk/interview.html
In article <dlhnif$j3e$1 digitaldaemon.com>, Tomás Rossi says...
In article <dlgvac$tql$1 digitaldaemon.com>, clayasaurus says...
Knud Sřrensen wrote:
 Hi 
 
 I stumbled over this interview with Stroustrup
 
 
 http://artlung.com/smorgasborg/Invention_of_Cplusplus.shtml

Oh great. Now everytime I go to program C++, it actually /feels/ like the language was designed to deceive me.

Are you kidding? Didn't you believe it right? It's a fake... not real! It's just a joke, don't be afraid. :) PS: Somebody will have to put an after comment to that interview, telling it's just a fake or we'll see many programmers killing themselves in the next days. I would commit suicide if that article were true :P


Tom
Nov 17 2005
parent reply clayasaurus <clayasaurus gmail.com> writes:
Tomás Rossi wrote:
 In article <dlhq0n$pns$1 digitaldaemon.com>, huh says...
 
No way! Its real.

Real my but! http://www.alledegodenavnevaroptaget.dk/interview.html
In article <dlhnif$j3e$1 digitaldaemon.com>, Tomás Rossi says...

In article <dlgvac$tql$1 digitaldaemon.com>, clayasaurus says...

Knud Sřrensen wrote:

Hi 

I stumbled over this interview with Stroustrup


http://artlung.com/smorgasborg/Invention_of_Cplusplus.shtml

Oh great. Now everytime I go to program C++, it actually /feels/ like the language was designed to deceive me.

Are you kidding? Didn't you believe it right? It's a fake... not real! It's just a joke, don't be afraid. :)



I know :-P I'm not afraid, but I never really thought 'what if' the language was designed to be deceptive. Now every time I go to C++, I'll have to wonder about it.
PS: Somebody will have to put an after comment to that interview, telling it's
just a fake or we'll see many programmers killing themselves in the next days.
I would commit suicide if that article were true :P



If it were true, then we could get everyone to switch to D.
 
 Tom

Nov 17 2005
parent Tomás Rossi <Tomás_member pathlink.com> writes:
In article <dli3m8$11vs$1 digitaldaemon.com>, clayasaurus says...
Tomás Rossi wrote:
 In article <dlhq0n$pns$1 digitaldaemon.com>, huh says...
 
No way! Its real.

Real my but! http://www.alledegodenavnevaroptaget.dk/interview.html
In article <dlhnif$j3e$1 digitaldaemon.com>, Tomás Rossi says...

In article <dlgvac$tql$1 digitaldaemon.com>, clayasaurus says...

Knud Sřrensen wrote:

Hi 

I stumbled over this interview with Stroustrup


http://artlung.com/smorgasborg/Invention_of_Cplusplus.shtml

Oh great. Now everytime I go to program C++, it actually /feels/ like the language was designed to deceive me.

Are you kidding? Didn't you believe it right? It's a fake... not real! It's just a joke, don't be afraid. :)



I know :-P I'm not afraid, but I never really thought 'what if' the language was designed to be deceptive. Now every time I go to C++, I'll have to wonder about it.
PS: Somebody will have to put an after comment to that interview, telling it's
just a fake or we'll see many programmers killing themselves in the next days.
I would commit suicide if that article were true :P



If it were true, then we could get everyone to switch to D.

The thing is that if this were true, not only C++ would be disreputable but every (C++)-like language and maybe every OO-language (D included). I think that if this was the case, i'd retire from programming (and studying CSs) and travel to a distant cave in a distant mountain and live like an hermit :P. Lucky for us it's just a scam :) Tom
Nov 17 2005
prev sibling next sibling parent Dejan Lekic <leka entropy.tmok.com> writes:
All I can say - my gratitude to the writer of that, so called, "interview".
IMHO that person (who wrote the text) should be a SciFi writer. :)

-- 
...........
Dejan Lekic
  http://dejan.lekic.org
  
Nov 17 2005
prev sibling parent reply taras <taras_member pathlink.com> writes:
In article <pan.2005.11.15.18.59.07.75576 sneakemail.com>,
=?iso-8859-1?q?Knud_S=F8rensen?= says...
Hi 

I stumbled over this interview with Stroustrup


http://artlung.com/smorgasborg/Invention_of_Cplusplus.shtml

Well, it is rather an old joke :-) But it is very interesting now precisely the author describes all the shortcomings of C++(and trully, I see absolutely no advantage of that language). I don't agree on the OO criticism thought. When I first stumbled across D I was really impressed. I think it is the first time in the world that a c clone was really useable :-) If we don't count java(with it's crappy standart library). Well, I don't get why everyone is coding in C and C++ anyways... Only the 10min compiling times tend to kill me...
Nov 17 2005
next sibling parent reply BCS <BCS_member pathlink.com> writes:
I saw that one a few years ago and thought it funny then (is hummer has become
more cynical the more I work with C++). But It also got me thinking "How does D
address these shortcomings of C++?" The parody might make a good check list of
this to do/avoid doing with D. Has D address them? Are we at risk of falling
into any of those traps? I haven't re read the interview yet (I will shortly)
but maybe we should make a list of its criticisms and look into them.

In article <dlic7d$1gcd$1 digitaldaemon.com>, taras says...
In article <pan.2005.11.15.18.59.07.75576 sneakemail.com>,
=?iso-8859-1?q?Knud_S=F8rensen?= says...
Hi 

I stumbled over this interview with Stroustrup


http://artlung.com/smorgasborg/Invention_of_Cplusplus.shtml

Well, it is rather an old joke :-) But it is very interesting now precisely the author describes all the shortcomings of C++(and trully, I see absolutely no advantage of that language). I don't agree on the OO criticism thought. When I first stumbled across D I was really impressed. I think it is the first time in the world that a c clone was really useable :-) If we don't count java(with it's crappy standart library). Well, I don't get why everyone is coding in C and C++ anyways... Only the 10min compiling times tend to kill me...

Nov 17 2005
parent reply clayasaurus <clayasaurus gmail.com> writes:
BCS wrote:
 I saw that one a few years ago and thought it funny then (is hummer has become
 more cynical the more I work with C++). But It also got me thinking "How does D
 address these shortcomings of C++?" The parody might make a good check list of
 this to do/avoid doing with D. Has D address them? Are we at risk of falling
 into any of those traps? I haven't re read the interview yet (I will shortly)
 but maybe we should make a list of its criticisms and look into them.

alias might be more dangerous than typedef, since you can alias anything, type's and functions and vars.
 
 In article <dlic7d$1gcd$1 digitaldaemon.com>, taras says...
 
In article <pan.2005.11.15.18.59.07.75576 sneakemail.com>,
=?iso-8859-1?q?Knud_S=F8rensen?= says...

Hi 

I stumbled over this interview with Stroustrup


http://artlung.com/smorgasborg/Invention_of_Cplusplus.shtml

Well, it is rather an old joke :-) But it is very interesting now precisely the author describes all the shortcomings of C++(and trully, I see absolutely no advantage of that language). I don't agree on the OO criticism thought. When I first stumbled across D I was really impressed. I think it is the first time in the world that a c clone was really useable :-) If we don't count java(with it's crappy standart library). Well, I don't get why everyone is coding in C and C++ anyways... Only the 10min compiling times tend to kill me...


Nov 17 2005
parent reply Tomás Rossi <Tomás_member pathlink.com> writes:
In article <dlirse$27fq$1 digitaldaemon.com>, clayasaurus says...
BCS wrote:
 I saw that one a few years ago and thought it funny then (is hummer has become
 more cynical the more I work with C++). But It also got me thinking "How does D
 address these shortcomings of C++?" The parody might make a good check list of
 this to do/avoid doing with D. Has D address them? Are we at risk of falling
 into any of those traps? I haven't re read the interview yet (I will shortly)
 but maybe we should make a list of its criticisms and look into them.

alias might be more dangerous than typedef, since you can alias anything, type's and functions and vars.

I'd rather compare alias with #define (a more restrictive #define). Typedef of D is much more nice than C++'s one. Tom
Nov 17 2005
parent reply clayasaurus <clayasaurus gmail.com> writes:
Tomás Rossi wrote:
 In article <dlirse$27fq$1 digitaldaemon.com>, clayasaurus says...
 
BCS wrote:

I saw that one a few years ago and thought it funny then (is hummer has become
more cynical the more I work with C++). But It also got me thinking "How does D
address these shortcomings of C++?" The parody might make a good check list of
this to do/avoid doing with D. Has D address them? Are we at risk of falling
into any of those traps? I haven't re read the interview yet (I will shortly)
but maybe we should make a list of its criticisms and look into them.

alias might be more dangerous than typedef, since you can alias anything, type's and functions and vars.

I'd rather compare alias with #define (a more restrictive #define). Typedef of D is much more nice than C++'s one. Tom

I was saying that alias may be worse than C's typedef, since C's typedef is equivilent to an alias, but you can use alias on function names as well and I'm guessing you could alias alias's, never tried though. I'm betting you could really obfuscate code with D's alias if you wanted to. D's typedef, on the other hand, gets it right.
Nov 19 2005
parent reply Tomás Rossi <Tomás_member pathlink.com> writes:
In article <dlnn4u$n2f$1 digitaldaemon.com>, clayasaurus says...
Tomás Rossi wrote:
 In article <dlirse$27fq$1 digitaldaemon.com>, clayasaurus says...
 
BCS wrote:

I saw that one a few years ago and thought it funny then (is hummer has become
more cynical the more I work with C++). But It also got me thinking "How does D
address these shortcomings of C++?" The parody might make a good check list of
this to do/avoid doing with D. Has D address them? Are we at risk of falling
into any of those traps? I haven't re read the interview yet (I will shortly)
but maybe we should make a list of its criticisms and look into them.

alias might be more dangerous than typedef, since you can alias anything, type's and functions and vars.

I'd rather compare alias with #define (a more restrictive #define). Typedef of D is much more nice than C++'s one. Tom

I was saying that alias may be worse than C's typedef, since C's typedef is equivilent to an alias, but you can use alias on function names as well and I'm guessing you could alias alias's, never tried though. I'm betting you could really obfuscate code with D's alias if you wanted to. D's typedef, on the other hand, gets it right.

Guess you could. But you cannot claim on D for programmers faults. I mean, you could obfuscate code with any language if you want (maybe not only with alias, typedefs or #defines). Yes, I agree that alias could be misused so to be irksome as hell. It's unavoidable from the language creator point of view because alias, #define and typedef are indispensable for things to look nice (sometimes). Tom
Nov 19 2005
parent reply Munchgreeble <Munchgreeble_member pathlink.com> writes:
In article <dlnnuu$nua$1 digitaldaemon.com>, Tomás Rossi says...
In article <dlnn4u$n2f$1 digitaldaemon.com>, clayasaurus says...
I was saying that alias may be worse than C's typedef, since C's typedef 
is equivilent to an alias, but you can use alias on function names as 
well and I'm guessing you could alias alias's, never tried though. I'm 
betting you could really obfuscate code with D's alias if you wanted to. 
D's typedef, on the other hand, gets it right.

Guess you could. But you cannot claim on D for programmers faults. I mean, you could obfuscate code with any language if you want (maybe not only with alias, typedefs or #defines). Yes, I agree that alias could be misused so to be irksome as hell. It's unavoidable from the language creator point of view because alias, #define and typedef are indispensable for things to look nice (sometimes).

This is another area where I'm just so impressed with D. Not only are typedefs fixed, but all forms of aliasing in the language now _have_ to be marked with the alias keyword - brilliant. This means people can use aliasing if they want to, but only in a way where aliases are easy to identify (both to the author and to the rest of the team). Previously aliases have been hard to spot - now they're not. Now you can't get away with kidding yourself, you have to call a spade a spade. Brilliant =) And of course if you want to outlaw aliasing as part of your coding standard (e.g. for safety related applications) it's dead easy to enforce: you just ban the use of the alias keyword - a one line script can check for violations. I really am quite stunned with the quality of the ideas that have gone into this language. Phenomenal! Munch
Nov 19 2005
next sibling parent reply James Dunne <james.jdunne gmail.com> writes:
Munchgreeble wrote:
 In article <dlnnuu$nua$1 digitaldaemon.com>, Tomás Rossi says...
 
In article <dlnn4u$n2f$1 digitaldaemon.com>, clayasaurus says...

I was saying that alias may be worse than C's typedef, since C's typedef 
is equivilent to an alias, but you can use alias on function names as 
well and I'm guessing you could alias alias's, never tried though. I'm 
betting you could really obfuscate code with D's alias if you wanted to. 
D's typedef, on the other hand, gets it right.

Guess you could. But you cannot claim on D for programmers faults. I mean, you could obfuscate code with any language if you want (maybe not only with alias, typedefs or #defines). Yes, I agree that alias could be misused so to be irksome as hell. It's unavoidable from the language creator point of view because alias, #define and typedef are indispensable for things to look nice (sometimes).

This is another area where I'm just so impressed with D. Not only are typedefs fixed, but all forms of aliasing in the language now _have_ to be marked with the alias keyword - brilliant. This means people can use aliasing if they want to, but only in a way where aliases are easy to identify (both to the author and to the rest of the team). Previously aliases have been hard to spot - now they're not. Now you can't get away with kidding yourself, you have to call a spade a spade. Brilliant =) And of course if you want to outlaw aliasing as part of your coding standard (e.g. for safety related applications) it's dead easy to enforce: you just ban the use of the alias keyword - a one line script can check for violations. I really am quite stunned with the quality of the ideas that have gone into this language. Phenomenal! Munch

Not saying the 'alias' keyword isn't great, but what _other_ way is there to define an alias that you can think of?
Nov 19 2005
parent Munchgreeble <"a" b.com \"munchgreeble xATx bigfoot xDOTx com\"> writes:
 Not saying the 'alias' keyword isn't great, but what _other_ way is 
 there to define an alias that you can think of?

Pointers. But then you can't easily avoid that and at least you have to put a little star in front of your identifier, which flags up to the reader that it could be an alias. Or is that not what you were thinking of? Munch
Nov 21 2005
prev sibling parent JT <JT_member pathlink.com> writes:
In article <dlo7h0$195f$1 digitaldaemon.com>, Munchgreeble says...
In article <dlnnuu$nua$1 digitaldaemon.com>, Tomás Rossi says...
In article <dlnn4u$n2f$1 digitaldaemon.com>, clayasaurus says...
I was saying that alias may be worse than C's typedef, since C's typedef 
is equivilent to an alias, but you can use alias on function names as 
well and I'm guessing you could alias alias's, never tried though. I'm 
betting you could really obfuscate code with D's alias if you wanted to. 
D's typedef, on the other hand, gets it right.

Guess you could. But you cannot claim on D for programmers faults. I mean, you could obfuscate code with any language if you want (maybe not only with alias, typedefs or #defines). Yes, I agree that alias could be misused so to be irksome as hell. It's unavoidable from the language creator point of view because alias, #define and typedef are indispensable for things to look nice (sometimes).

This is another area where I'm just so impressed with D. Not only are typedefs fixed, but all forms of aliasing in the language now _have_ to be marked with the alias keyword - brilliant. This means people can use aliasing if they want to, but only in a way where aliases are easy to identify (both to the author and to the rest of the team). Previously aliases have been hard to spot - now they're not. Now you can't get away with kidding yourself, you have to call a spade a spade. Brilliant =) And of course if you want to outlaw aliasing as part of your coding standard (e.g. for safety related applications) it's dead easy to enforce: you just ban the use of the alias keyword - a one line script can check for violations. I really am quite stunned with the quality of the ideas that have gone into this language. Phenomenal! Munch

Yeah its been designed from top to bottom with *easy parsing* in mind. This opens up the posibility of tools the quality of which we have never seen with c++. We have yet to fully unleash the power of this language.
Nov 19 2005
prev sibling parent reply Tomás Rossi <Tomás_member pathlink.com> writes:
In article <dlic7d$1gcd$1 digitaldaemon.com>, taras says...
In article <pan.2005.11.15.18.59.07.75576 sneakemail.com>,
=?iso-8859-1?q?Knud_S=F8rensen?= says...
Hi 

I stumbled over this interview with Stroustrup


http://artlung.com/smorgasborg/Invention_of_Cplusplus.shtml

Well, it is rather an old joke :-) But it is very interesting now precisely the author describes all the shortcomings of C++(and trully, I see absolutely no advantage of that language).

Despite the shortcomings, there's advantages in many MANY things.
I don't agree on the OO criticism thought. When I first stumbled across D I was
really impressed. I think it is the first time in the world that a c clone was
really useable :-)

You're overstating a little too much.
 If we don't count java(with it's crappy standart library).
Well, I don't get why everyone is coding in C and C++ anyways... Only the 10min
compiling times tend to kill me... 

They code in C and C++ because these languages are powerful and mature languages for many tasks. Plus, you could not write C++ or D code (mean using most part of it's features, not only inline asm) that do things in the exact way you want they to happen. With C++/D you're leaving a lot of work to the compiler; C is much more ASM straightforward so it will never die. That's why C is the language OS developers choose. Hope D can achieve the same results in this field in the future. Also D IMHO is still immature in the sense that it isn't recommendable to employ it on serious big projects (yet). Some features will be added, some bugs will be fixed, a long road we'll have to walk yet :) (hope this all happens soon). Regards (and dreaming some day D becomes the standard) Tom
Nov 17 2005
parent taras <taras_member pathlink.com> writes:
They code in C and C++ because these languages are powerful and mature languages
for many tasks. Plus, you could not write C++ or D code (mean using most part of
it's features, not only inline asm) that do things in the exact way you want
they to happen. With C++/D you're leaving a lot of work to the compiler; C is
much more ASM straightforward so it will never die. That's why C is the language
OS developers choose. Hope D can achieve the same results in this field in the
future.

Tom

I agree fully that C is a very good tradeoff for assembler. And I never doubted it that C is bad for OS programming. It is just that C isn't good for anything else besides the low-level programming(I use it too for this). And C++ isn't good at anything. Well, let's not start flame wars around here :-) I can only tell my optnion. I grew with delphi(now freepascal) and when I tried to switch to C++, I just was not able to do it. The syntax was so messy and the whole code structure so unclear... I myself tend to believe in virtual machines that will wrap around the hardware and act like OS, much like microsofts .NET. This will remove the porting problems. What I would like to have in D are precompiled metapackeges(modules) taht will contain the parsed header information about types and functions + links to the libraries, with some version checking mechanisms. This is much like what pascal is doing. This way the compilation will be much faster and error-free. I think the header files are one of greatest C/C++ problems, even if they make them so incredibly flexible. -taras
Nov 17 2005