Queries about new C paper work C11

[lovecreatesbeauty]

1. why no algorithm and data structure library proposed?

i think if C has stack, queue, hash map.. like those in C++/STL,
many C++ application programmers will come to C again.

2. why thread (posix) proposed while socket (posix) not?

3. will this C11 fail like C99?

will it be dead until replaced by the next paper work and no one
implementents it?

 

[Jens Gustedt]

Am 20.05.2012 20:33, schrieb lovecreatesbeauty:

1. why no algorithm and data structure library proposed?

because C usually has a stricter distinctiong between the language and
support libraries

i think if C has stack, queue, hash map.. like those in C++/STL,
many C++ application programmers will come to C again.

there is no need for people coming back to C, it is one of the most used
programming languages of today.

To my personal opinion (and experience) C++ does it best to scare people
off, but which doesn't mean that they will go to use C.

2. why thread (posix) proposed while socket (posix) not?

I think one of the most important additions by C11 (and C++1 BTW) are
atomic operations that definitively should be a language feature. But
atomicity needs a precise description of concurrency and sequencing of
data access. Thus there is the thread interface coming into play. BTW,
this is not the same as POSIX threads: in terms of functionality it is
only a small subset of POSIX and even has some differences.

It seems that for networking code there was no need to place that
insided the language.

3. will this C11 fail like C99?

First C99 didn't fail. It only failed on one platform and for one major
software vendor not playing the game. All others have it.

For C11 only the future will tell, but signs are quite good that it
wouldn't fail. It brings important features and the two major open
source compilers (gcc and clang) are actively integrating it.

will it be dead until replaced by the next paper work and no one
implementents it?

It is basically implemented. Many compilers already have the features as
extensions, only syntax might be a bit different. With a bit of macro
patching around these syntactical differences you may come already quite
close in programming with C11. P99 offers you that glue for gcc and
derivatives (with this respect these are icc, open64 and clang) on POSIX
platforms.


Jens

 

[jacob navia]

Le 20/05/12 20:33, lovecreatesbeauty a écrit :

1. why no algorithm and data structure library proposed?

i think if C has stack, queue, hash map.. like those in C++/STL,
many C++ application programmers will come to C again.

I am developing exactly that: a container library for the C language.

See the state of the project at:

http://code.google.com/p/ccl/

 

[James Kuyper]

1. why no algorithm and data structure library proposed?

i think if C has stack, queue, hash map.. like those in C++/STL,
many C++ application programmers will come to C again.

Is that a good idea? It's still one of the most widely used languages;
if more people came back to it, the field might get overcrowded.

More seriously - C++ already does this; why would becoming more like C++
be a good move? If C has a future as anything other than the predecessor
of C++ (which, with C still going strong several decades after the
invention of C++, seems rather likely), it must be by maintaining a
distinction between C's advantages and those of C++. C's key advantage
in comparison with C++ is simplicity, which makes it easier to learn and
easier to implement. Every complicated feature (like the ones you
suggest) eats away at that advantage, reducing the incentive to move
over to C. That doesn't mean that no complicated features should be
added to C, but only that they need to be well motivated.

....

will it be dead until replaced by the next paper work and no one
implementents it?

Is there a reason you're using "paper work" as if it were a suitable
replacement for "standard version".