Is c# so easy? Easier than c? Think again...

[jbailo]

Subject: dotnet is a farce.

You know, I was always impressed with
the way one could write a http program
with c# or java. To me, it showed how
'superior' these languages are because
they made it so 'easy' to
set up connection points, open a socket,
send the request.

But, I never really know how to do it in
c! So I wasn't comparing it with anything.
Then, I just received one of my four new
books yesterday, 'Advanced Linux Programming'
from New Riders, and there was some sample
code for opening a socket and reading
http data.

IT'S THE SAME THING AS IN C# !!!!!!

If anything, it's less and more elegant.

Tell me that:

connect(socket_fd, &name, sizeof( struct sockaddr_in))

a SINGLE line that sets up the connection followed
by a while loop to read the data is not
THE SAME EXACT SYNTAX as in c# for doing the same thing.

See -- UNIX has always had 'ASSEMBLIES' -- or SYSCALLS

-- or OPERATING SYSTEM LEVEL LIBRARIES AND METHODS

DOT NET IS A FARCE (!)(!)(!)(!)(!)

[Miha Markic]

Please, do continue posting stuff like this.
It is very refreshing.

Miha

"jbailo" <ja*****@earthlink.net> wrote in message
news:31******************************@news.teranew s.com...

Subject: dotnet is a farce.

[Fergus Henderson]

"jbailo" <ja*****@earthlink.net> writes:

Subject: dotnet is a farce.

You know, I was always impressed with
the way one could write a http program
with c# or java. To me, it showed how
'superior' these languages are because
they made it so 'easy' to
set up connection points, open a socket,
send the request.
That is mainly a matter of the relative strengths of the libraries,
rather than the language.

The _language_ advantages of C# or Java compared with C are mainly the
following: greater determinism and safety (i.e. the absence of most of
the unspecified and undefined behaviour which is rife throughout C),
automatic memory management (garbage collection), better support for
encapsulation and OOP, and exception handling.
But, I never really know how to do it in c! ....IT'S THE SAME THING AS IN C# !!!!!!
Apparently you still don't know.
Tell me that:

connect(socket_fd, &name, sizeof( struct sockaddr_in))

a SINGLE line that sets up the connection followed
by a while loop to read the data is not
THE SAME EXACT SYNTAX as in c# for doing the same thing.
It's not. That single line of C code will *silently ignore any failures*.
I'm pretty sure the corresponding C# or Java code would throw an
exception, ensuring that the failure will either be dealt with by the
caller or eventually reported to the user.
DOT NET IS A FARCE (!)(!)(!)(!)(!)

It didn't take much for you to leap to that conclusion, did it?

--
Fergus Henderson <fj*@cs.mu.oz.au> | "I have always known that the pursuit
The University of Melbourne | of excellence is a lethal habit"
WWW: <http://www.cs.mu.oz.au/~fjh> | -- the last words of T. S. Garp.

[ToddT]

pls, don't feed the troll.
On Wed, 05 Nov 2003 15:02:22 GMT, Fergus Henderson <fj*@cs.mu.oz.au>
wrote:

"jbailo" <ja*****@earthlink.net> writes:

Subject: dotnet is a farce.

You know, I was always impressed with
the way one could write a http program
with c# or java. To me, it showed how
'superior' these languages are because
they made it so 'easy' to
set up connection points, open a socket,
send the request.

That is mainly a matter of the relative strengths of the libraries,
rather than the language.

The _language_ advantages of C# or Java compared with C are mainly the
following: greater determinism and safety (i.e. the absence of most of
the unspecified and undefined behaviour which is rife throughout C),
automatic memory management (garbage collection), better support for
encapsulation and OOP, and exception handling.

But, I never really know how to do it in c!

...

IT'S THE SAME THING AS IN C# !!!!!!

Apparently you still don't know.

Tell me that:

connect(socket_fd, &name, sizeof( struct sockaddr_in))

a SINGLE line that sets up the connection followed
by a while loop to read the data is not
THE SAME EXACT SYNTAX as in c# for doing the same thing.

It's not. That single line of C code will *silently ignore any failures*.
I'm pretty sure the corresponding C# or Java code would throw an
exception, ensuring that the failure will either be dealt with by the
caller or eventually reported to the user.

DOT NET IS A FARCE (!)(!)(!)(!)(!)

It didn't take much for you to leap to that conclusion, did it?

[mlw]

Fergus Henderson wrote:

That is mainly a matter of the relative strengths of the libraries,
rather than the language.

The _language_ advantages of C# or Java compared with C are mainly the
following: greater determinism and safety (i.e. the absence of most of
the unspecified and undefined behaviour which is rife throughout C),
automatic memory management (garbage collection), better support for
encapsulation and OOP, and exception handling.

I would use C++ over .NET or Java anytime.

..NET and Java do not have pointers, therefor it will always be more
difficult to perform various operations. They have a common runtime
interpreter, yea, sure, there is a JIT compiler, but this is an
optimization of the classic tokenized interpreter.

I like "real code." When I see something I don't understand, I want to look
at the assembly output. I want to know and be able to audit what is being
run. Runtime environments are not reliable. When newer version come out,
older APIs are depricated. Bugs in the JIT can be a cause for major
difficulties, just as in a real compiler, but the difference is that you
can inspect the code and alter the code generation as nessisary.

The biggest part of puzzle is portability. You are only as portable as the
makers of your virtual machine. Think about trying to get the same version
of Java for all your supported platforms, if you can, that is.

Virtually every platform has a C/C++ compiler, the same can not be said
about Java, and certainly not .NET.

What, by the way, is the "unspecified and undefined behavior" in C or C++?

[Fergus Henderson]

mlw <ml*@nospam.no> writes:

I would use C++ over .NET or Java anytime.

.NET and Java do not have pointers,
FYI .NET does have pointers. Please go and read some more technical
details about .NET, and criticize what it actually is, rather than some
figment of your imagination.
What, by the way, is the "unspecified and undefined behavior" in C or C++?

See Annex J of the 1999 C standard. It lists more than a hundred
different aspects of C which have either unspecified, undefined, or
implementation-defined behaviour. There are also other aspects of C
that result in undefined behaviour which are not specified in that list.

--
Fergus Henderson <fj*@cs.mu.oz.au> | "I have always known that the pursuit
The University of Melbourne | of excellence is a lethal habit"
WWW: <http://www.cs.mu.oz.au/~fjh> | -- the last words of T. S. Garp.

[john bailo]

Wed, 05 Nov 2003 10:02:46 -0600, ToddT,
<lo********************************@4ax.com>

pls, don't feed the troll.

that's right -- call argument you can't defeat a 'troll'

go home to your mama and eat some warm cookies and
milk, because your technical skills are so weak
you can't argue the benefits of your own language.

cry baby
cry baby
cry baby


On Wed, 05 Nov 2003 15:02:22 GMT, Fergus Henderson <fj*@cs.mu.oz.au>
wrote:

"jbailo" <ja*****@earthlink.net> writes:

Subject: dotnet is a farce.

You know, I was always impressed with
the way one could write a http program
with c# or java. To me, it showed how
'superior' these languages are because
they made it so 'easy' to
set up connection points, open a socket,
send the request.

That is mainly a matter of the relative strengths of the libraries,
rather than the language.

The _language_ advantages of C# or Java compared with C are mainly the
following: greater determinism and safety (i.e. the absence of most of
the unspecified and undefined behaviour which is rife throughout C),
automatic memory management (garbage collection), better support for
encapsulation and OOP, and exception handling.

But, I never really know how to do it in c!

...

IT'S THE SAME THING AS IN C# !!!!!!

Apparently you still don't know.

Tell me that:

connect(socket_fd, &name, sizeof( struct sockaddr_in))

a SINGLE line that sets up the connection followed
by a while loop to read the data is not
THE SAME EXACT SYNTAX as in c# for doing the same thing.

It's not. That single line of C code will *silently ignore any failures*.
I'm pretty sure the corresponding C# or Java code would throw an
exception, ensuring that the failure will either be dealt with by the
caller or eventually reported to the user.

DOT NET IS A FARCE (!)(!)(!)(!)(!)

It didn't take much for you to leap to that conclusion, did it?

[John Young]

I agree with your comment, Miha.
I used to find these groups boring sometimes, but now 'J "The Muppet" Bailo'
is on the scene, I can have a good laugh when I get back from work after
doing some real programming! You know bailo, real programming is creating
software that works reliably, doesn't take years to create, and works for
more than 5% of computer users (eg. Windows!). Get real bailo, users seem
to want Windows software so that is what we give them.
If users decide that they want Linux software, then we will write that!

In the meantime, I'm writing software for the OS that users want to use (oh,
and pay me for !).

Hehh

John

"Miha Markic" <mi***@spin.NO.SPAM.si> wrote in message
news:OM*************@TK2MSFTNGP11.phx.gbl...

Please, do continue posting stuff like this.
It is very refreshing.

Miha

"jbailo" <ja*****@earthlink.net> wrote in message
news:31******************************@news.teranew s.com...

Subject: dotnet is a farce.

Another thing, MC++ is getting an overhaul to dump thise mess __keywords

And another feature ... CUSTOM ATTRIBUTES..

with 2.0 its a rich environment, while you are mopping up yer maintenance
shti ill be releasing rich solutions before you even start emliminating
youre bugs and doing boring things like housekeeping.

"john bailo" <ja*****@earthlink.net> wrote in message
news:ba******************************@news.teranew s.com...

Wed, 05 Nov 2003 10:02:46 -0600, ToddT,
<lo********************************@4ax.com>

pls, don't feed the troll.

that's right -- call argument you can't defeat a 'troll'

go home to your mama and eat some warm cookies and
milk, because your technical skills are so weak
you can't argue the benefits of your own language.

cry baby
cry baby
cry baby

On Wed, 05 Nov 2003 15:02:22 GMT, Fergus Henderson <fj*@cs.mu.oz.au>
wrote:

"jbailo" <ja*****@earthlink.net> writes:

Subject: dotnet is a farce.

You know, I was always impressed with
the way one could write a http program
with c# or java. To me, it showed how
'superior' these languages are because
they made it so 'easy' to
set up connection points, open a socket,
send the request.

That is mainly a matter of the relative strengths of the libraries,
rather than the language.

The _language_ advantages of C# or Java compared with C are mainly the
following: greater determinism and safety (i.e. the absence of most of
the unspecified and undefined behaviour which is rife throughout C),
automatic memory management (garbage collection), better support for
encapsulation and OOP, and exception handling.

But, I never really know how to do it in c!
...
IT'S THE SAME THING AS IN C# !!!!!!

Apparently you still don't know.

Tell me that:

connect(socket_fd, &name, sizeof( struct sockaddr_in))

a SINGLE line that sets up the connection followed
by a while loop to read the data is not
THE SAME EXACT SYNTAX as in c# for doing the same thing.

It's not. That single line of C code will *silently ignore any failures*.I'm pretty sure the corresponding C# or Java code would throw an
exception, ensuring that the failure will either be dealt with by the
caller or eventually reported to the user.

DOT NET IS A FARCE (!)(!)(!)(!)(!)

It didn't take much for you to leap to that conclusion, did it?

[The Ghost In The Machine]

In comp.os.linux.advocacy, jbailo
<ja*****@earthlink.net>
wrote
on Wed, 05 Nov 2003 13:58:49 GMT
<31******************************@news.teranews.co m>:

Subject: dotnet is a farce.

You know, I was always impressed with
the way one could write a http program
with c# or java. To me, it showed how
'superior' these languages are because
they made it so 'easy' to
set up connection points, open a socket,
send the request.

But, I never really know how to do it in
c! So I wasn't comparing it with anything.
Then, I just received one of my four new
books yesterday, 'Advanced Linux Programming'
from New Riders, and there was some sample
code for opening a socket and reading
http data.

IT'S THE SAME THING AS IN C# !!!!!!

If anything, it's less and more elegant.

Tell me that:

connect(socket_fd, &name, sizeof( struct sockaddr_in))

a SINGLE line that sets up the connection followed
by a while loop to read the data is not
THE SAME EXACT SYNTAX as in c# for doing the same thing.

See -- UNIX has always had 'ASSEMBLIES' -- or SYSCALLS

-- or OPERATING SYSTEM LEVEL LIBRARIES AND METHODS

DOT NET IS A FARCE (!)(!)(!)(!)(!)

You're confusing a *lot* of issues here. For starters,
are you referring to:

[1] the programming of a client which can use HTTP to
transfer data from client to server and back to client?
[2] the programming of a server which is HTTP-aware?
[3] something else?

I know Java has java.net.URL and java.net.URLConnect;
those help a *lot*. I suspect C# has something
similar. Java, however, also has java.net.Socket and
java.net.ServerSocket, if one really wants to bypass Java's
URL handling -- and ServerSocket is required if one wants
to process incoming server requests anyway, at some level.

I can't say I know whether C# has something similar but they
might be able to either have their own variant of URL or
they can simply call WinInet or HTTP subpackages (both available
in Windows). WinInet is roughly equivalent to URL/URLConnection,
albeit more complicated (and with a somewhat procedural focus, as
opposed to Java's method-based one -- one passes in handles
of the appropriate type into these routines).

--
#191, ew****@earthlink.net
It's still legal to go .sigless.

[Eric Matteson]

"jbailo" <ja*****@earthlink.net> wrote in message news:<31******************************@news.terane ws.com>...

Subject: dotnet is a farce.

You know, I was always impressed with
the way one could write a http program
with c# or java. To me, it showed how
'superior' these languages are because
they made it so 'easy' to
set up connection points, open a socket,
send the request.

But, I never really know how to do it in
c! So I wasn't comparing it with anything.
Then, I just received one of my four new
books yesterday, 'Advanced Linux Programming'
from New Riders, and there was some sample
code for opening a socket and reading
http data.

IT'S THE SAME THING AS IN C# !!!!!!

If anything, it's less and more elegant.

Tell me that:

connect(socket_fd, &name, sizeof( struct sockaddr_in))

a SINGLE line that sets up the connection followed
by a while loop to read the data is not
THE SAME EXACT SYNTAX as in c# for doing the same thing.

See -- UNIX has always had 'ASSEMBLIES' -- or SYSCALLS

-- or OPERATING SYSTEM LEVEL LIBRARIES AND METHODS

DOT NET IS A FARCE (!)(!)(!)(!)(!)

I like the C language the best because it makes the most sense.
Both C and C++ support goto label, which is similar to jmp label
in assembley language.
goto statements are needed in most non-trivial programs.
I hate JAVA because JAVA does not have required goto statements.
The simple traditional C language is the best high-level language
that I know of.
er*************@netzero.net

[Jeff Relf]

Hi jbailo , You say :
" IT'S THE SAME THING AS IN C# !!!!!! "

Exactly ... And it's running the same code too .

Today's libraries are designed to run under any language .

But unlike scripts , C++ code can run stand alone .

And C++ doesn't treat you like an idiot ,
You can hack into structures as much as you want ...
And it has a nicer debugger too .

Does anyone know where I can find a Win98 debugger
that steps from statement to statement ?
( Instead from line to line )

Does anyone know where I can find a Win98 compiler
who's macros aren't line oriented ?

[mlw]

Jeff Relf wrote:

Hi jbailo , You say :
" IT'S THE SAME THING AS IN C# !!!!!! "

Exactly ... And it's running the same code too .

Today's libraries are designed to run under any language .

But unlike scripts , C++ code can run stand alone .

And C++ doesn't treat you like an idiot ,
You can hack into structures as much as you want ...
And it has a nicer debugger too .

Does anyone know where I can find a Win98 debugger
that steps from statement to statement ?
( Instead from line to line )
Just drop in to assembly mode. Trace the instructions.

Does anyone know where I can find a Win98 compiler
who's macros aren't line oriented ?

What do you mean "line oriented?" As a little foot note, the compiler,
strictly speaking, doesn't handle the macros, the preprocessor does. Many
people create their own pre-preprocessors for various languages, sort of
like embedded SQL compilers.

[William Ryan]

Fergus:

Why is it that so many people persist in this myth of No Pointers? I use
them all the time in the Compact Framework and a few legacy apps. Then as I
just finish debugging something which uses pointers, Bailo and or on of his
few Homies pipes up with this myth again.

How can they program in C, C++ and Java and be this clueless? (I truly
wonder how many memory leaks and dangling pointers exist in their 'elite'
C++ code)

Cheers,

Bill
"Fergus Henderson" <fj*@cs.mu.oz.au> wrote in message
news:3f********@news.unimelb.edu.au...

mlw <ml*@nospam.no> writes:

I would use C++ over .NET or Java anytime.

.NET and Java do not have pointers,
FYI .NET does have pointers. Please go and read some more technical
details about .NET, and criticize what it actually is, rather than some
figment of your imagination.

What, by the way, is the "unspecified and undefined behavior" in C or

C++?
See Annex J of the 1999 C standard. It lists more than a hundred
different aspects of C which have either unspecified, undefined, or
implementation-defined behaviour. There are also other aspects of C
that result in undefined behaviour which are not specified in that list.

--
Fergus Henderson <fj*@cs.mu.oz.au> | "I have always known that the pursuit The University of Melbourne | of excellence is a lethal habit"
WWW: <http://www.cs.mu.oz.au/~fjh> | -- the last words of T. S. Garp.

[mlw]

William Ryan wrote:

(I truly

wonder how many memory leaks and dangling pointers exist in their 'elite'
C++ code)

Well, that is always a matter of design and understanding what you are
building. Rules on how and where memory is used. Knowing when to use
automatics, alloca, or malloc.

A well designed program does not call malloc unless absolutely nessisary.
When it does need to allocate memory from the heap, it is not done in the
critical path and it is done in a sufficiently centralized place.

In fact "malloc" should never be called directly in a non-trivial program.
Like socket calls and file operations, memory managment should be designed
in from the start.

Why do this when Java and .NET have garbage collection? Well, because memory
management is probably one of the most important jobs for the software
designer for stability, efficiency, and performance.

[Jeff Relf]

Hi mlw , You say :
" What do you mean ' line oriented ? ' "

You must be a young whipper snapper .

In the early days ,
some editors would only edit one line at a time .

They were " Line Oriented " .

My preprocessor , in MS Dev C++ 6.0 ,
won't let me do something like this :

int X; #define Mac1(X) X++; #define Mac2(X) X--; Mac1(X);

If there a better preprocessor ?

[Jeff Relf]

[ Reposted , This time to the right groups ]

Hi mlw , You say :
" What do you mean ' line oriented ? ' "

You must be a young whipper snapper .

In the early days ,
some editors would only edit one line at a time .

They were " Line Oriented " .

My preprocessor , in MS Dev C++ 6.0 ,
won't let me do something like this :

int X; #define Mac1(X) X++; #define Mac2(X) X--; Mac1(X);

If there a better preprocessor ?

[Jeff Relf]

Hi Eric Matteson , You say :
" The simple traditional C language is
the best high-level language that I know of . "

As an old timer myself , I tend to agree with you .

But you can can convert all of your old C programs
into C++ programs in just a few minutes .

Why do this ?

Because I've found that
there is more documentation and support for C++ .

If you stick with C you'll be left behind .

COM , DirectX , etc. have little support for C .

For example :
#include <dsound.h> // DSound.LIB
#include <ddraw.h>
// Dxerr8.lib DDraw.LIB DInput8.lib
// DXguid.LIB DSound.LIB WinMM.LIB
#include <dinput.h>
#include <Dxerr8.h> // MAKE_DDHRESULT( 150 )
#include <olectl.h>

For example , below is some C++ code that
will open .JPGs , .GIFs , .BMPs etc. :

Notice the IPicture *IPic; , OleLoadPicturePath();
and the IPic->get_Handle((OLE_HANDLE*)&BM);

IPicture *IPic; typedef HBITMAP BM_T;

BM_T _BM( int Ty, char *P) { ushort PP[400]; char B[400];
Str( B, "%s\\%s", Working[ Ty], P);
mbstowcs( PP, B, 400 ); short S= 0; BM= 0;
if ( OleLoadPicturePath(
PP, 0, 0, 0, IID_IPicture, (void**)& IPic)) return 0;
if ( IPic->get_Type( & S )<0 || S != PICTYPE_BITMAP||
IPic->get_Handle(( OLE_HANDLE*)& BM)< 0|| !BM) Rel( IPic)
return BM; }

[jbailo]

posterzz (!) Jeff Relf <__**********@NCPlus.NET>, with the izzo say:

Hi Eric Matteson , You say :
" The simple traditional C language is
the best high-level language that I know of . "

As an old timer myself , I tend to agree with you .

But you can can convert all of your old C programs
into C++ programs in just a few minutes .

Why do this ?

c b aced by Czzzzz.

[Jeff Relf]

Hi jbailo , You say : " c b aced by Czzzzz "

What is Czzzzzzzzzz ?

[jbailo]

posterzz (!) Jeff Relf <__**********@NCPlus.NET>, with the izzo say:

Hi jbailo , You say : " c b aced by Czzzzz "

What is Czzzzzzzzzz ?

Day Is C#

Day Is C++++++

Day Is C

I uzzzz Czzzzz

Czzzz assbackass, RPNzzz,
In-Fixzzzing-Notashunzzzz.

you don't know what hood is good
until you is programmzzz in Czzzzzzz

Sa Czzzz
Izn, Sa c#
Izn, Sa c
Izn, Sa c++++++

Sa Czzzzzzz

Exclamation POINT (!)

[Jeff Relf]

Hi Fergus Henderson , You say :
" That single line of C code will
_ silently ignore any failures _ . "

I hope the hell it does !

Unless you're debugging the code ,
you never want your code to complain or halt .

For example , unless your debugging ,
dividing by zero should be a non issue .

[Cable Speed Test]

"Eric Matteson" <er*************@netzero.net> wrote in message
news:ff**************************@posting.google.c om...

I like the C language the best because it makes the most sense.
Both C and C++ support goto label,

'sides typing GOTO reminds you of all the fun you had with BASIC...

[keyur shah]

The line of code u are talking about is the precise way of doing it....
nothing less or more...

and we anyway need a loop to access out whatever have been received thru
the port...

Keyur Shah
732-423-0745

*** Sent via Developersdex http://www.developersdex.com ***
Don't just participate in USENET...get rewarded for it!

[Gary van der Merwe]

You are crazy

"jbailo" <ja*****@earthlink.net> wrote in message
news:56******************************@news.teranew s.com...

posterzz (!) Jeff Relf <__**********@NCPlus.NET>, with the izzo say:

Hi jbailo , You say : " c b aced by Czzzzz "

What is Czzzzzzzzzz ?

Day Is C#

Day Is C++++++

Day Is C

I uzzzz Czzzzz

Czzzz assbackass, RPNzzz,
In-Fixzzzing-Notashunzzzz.

you don't know what hood is good
until you is programmzzz in Czzzzzzz

Sa Czzzz
Izn, Sa c#
Izn, Sa c
Izn, Sa c++++++

Sa Czzzzzzz

Exclamation POINT (!)

[mlw]

Jeff Relf wrote:

[ Reposted , This time to the right groups ]

Hi mlw , You say :
" What do you mean ' line oriented ? ' "

You must be a young whipper snapper .
Not likely, I worked on a Sun One and a vax 11/780. You know, the old
machines. I have been in the field for 20 years.

In the early days ,
some editors would only edit one line at a time .

They were " Line Oriented " .

My preprocessor , in MS Dev C++ 6.0 ,
won't let me do something like this :

int X; #define Mac1(X) X++; #define Mac2(X) X--; Mac1(X);

If there a better preprocessor ?

In your example, I don't see what the would be a benefit. Macro definitions
have no effect in compiled code. The are filtered out from the compile
stage so it is unnessisary to have them terminated with a semicolon.

In general, your line is confusing at best, because you have two elements in
a line which have no real effect on the code.

They are not "line oriented" per say, like old cobol, but the parser does
read one line at a time. You can make it span lines with a back slash.

[Linønut]

["Followup-To:" header set to comp.os.linux.advocacy.]
On 2003-11-06, Cable Speed Test <do********@here.com> wrote:

"Eric Matteson" <er*************@netzero.net> wrote in message
news:ff**************************@posting.google.c om...

I like the C language the best because it makes the most sense.
Both C and C++ support goto label,

'sides typing GOTO reminds you of all the fun you had with BASIC...

goto's are fun!

if (istrouble)
goto hell;

hell:

--
Windows desktops and servers can find a safe haven on a
GNU/Linux/FreeBSD network!

[mlw]

Jeff Relf wrote:

Hi Eric Matteson , You say :
" The simple traditional C language is
the best high-level language that I know of . "

As an old timer myself , I tend to agree with you .

But you can can convert all of your old C programs
into C++ programs in just a few minutes .

Why do this ?

Because I've found that
there is more documentation and support for C++ .

If you stick with C you'll be left behind .

[snip]

I'm an old C guy, and I use C++ almost exclusively. I use it mostly like a
better C with classes. There is nothing in C that can't be done in C++, but
the converse is not true.

Yea, a lot of people will say, "C++ lets you make bad programs," this may be
true, but strict C allows really ugly stuff, especially when you try to do
the stuff C++ allows but C does not. Look at the GTK compared to QT.

[mlw]

Jeff Relf wrote:

Hi Fergus Henderson , You say :
" That single line of C code will
_ silently ignore any failures _ . "

I hope the hell it does !

Unless you're debugging the code ,
you never want your code to complain or halt .

For example , unless your debugging ,
dividing by zero should be a non issue .

divide by zero will almost always cause a hardware fault, and it will be
pushed back up to the process, where you will have to deal with it
asyncronously.

[Ian Hilliard]

On Wed, 05 Nov 2003 19:21:38 -0800, Jeff Relf wrote:

[ Reposted , This time to the right groups ]

Hi mlw , You say :
" What do you mean ' line oriented ? ' "

You must be a young whipper snapper .

In the early days ,
some editors would only edit one line at a time .

They were " Line Oriented " .

My preprocessor , in MS Dev C++ 6.0 ,
won't let me do something like this :

int X; #define Mac1(X) X++; #define Mac2(X) X--; Mac1(X);

If there a better preprocessor ?

If reliable code is your aim, you should avoid macros as they are not type
safe. It is much better to use inlines. The second advantage of inlines is
that you can also step into the inline code with the debugger.

Ian

[The Ghost In The Machine]

In comp.os.linux.advocacy, Jeff Relf
<__**********@NCPlus.NET>
wrote
on Wed, 5 Nov 2003 21:23:01 -0800
<fok67qfa9gnl$.dlg@__.Jeff.Relf>:

Hi Fergus Henderson , You say :
" That single line of C code will
_ silently ignore any failures _ . "

I hope the hell it does !

Unless you're debugging the code ,
you never want your code to complain or halt .

For example , unless your debugging ,
dividing by zero should be a non issue .

Is that what the Yorktown was doing? :-)

--
#191, ew****@earthlink.net
It's still legal to go .sigless.

[Jeff Relf]

Hi Spooky , You say :
" Is that what the Yorktown was doing ? :-) "

The Yorktown ? What is that ?

It sounds more like a city than a piece of code .

[Peter Köhlmann]

Jeff Relf wrote:

Hi Spooky , You say :
" Is that what the Yorktown was doing ? :-) "

The Yorktown ? What is that ?

It sounds more like a city than a piece of code .

For me it sounds much more like a warship. And it is
It is the one which had to be *towed* back into harbour after Windows shut
down the complete network, leaving the ship without propulsion. Another
example of windows superiority: Divide by zero and you disable complete
ships. Good, sound OS design
--
Those who do not understand Unix are condemned to reinvent it, poorly.
-- Henry Spencer

[Jeff Relf]

Hi mlw , You say :
" divide by zero will almost always cause a hardware fault ,
and it will be pushed back up to the process ,
where you will have to deal with it asynchronously . "

In non-debug mode , That's not true at all .

By default , Microsoft's old sixteen bit compilers
did throw an exception for a floating point divide by zero ,
Even when you weren't in debug mode .

But that drastic error was fixed with the 32 bit stuff .

The following code will only toss an exception
if you compile for _ Debugging _ :

int XXX= 1, YYY= 0; XXX /= YYY;

Non-debug code _ Should _ never halt or complain .

[mlw]

Jeff Relf wrote:

Hi mlw , You say :
" divide by zero will almost always cause a hardware fault ,
and it will be pushed back up to the process ,
where you will have to deal with it asynchronously . "

In non-debug mode , That's not true at all .

By default , Microsoft's old sixteen bit compilers
did throw an exception for a floating point divide by zero ,
Even when you weren't in debug mode .

But that drastic error was fixed with the 32 bit stuff .

The following code will only toss an exception
if you compile for _ Debugging _ :

int XXX= 1, YYY= 0; XXX /= YYY;

Non-debug code _ Should _ never halt or complain .

It "should" because the result is undefined.

[Jeff Relf]

Hi mlw , You say :
" It "should" because the result is undefined . "

Well you better try to get Microsoft
to change it's default behavior then !

I'm using MS Dev C++ 6.0 on Win98 .

Build some divide by zero errors in release mode ,
( not debug mode )
try both floating point and integers ...
No interrupts will happen .

[Jeff Relf]

Hi mlw , You say :
" I don't see what the would be a benefit . "

I assume you don't know of a preprocessor
that is not line oriented then ... Oh well .

How archaic .

[Paul Hovnanian P.E.]

The language isn't the entire issue.

First of all, C# is supposedly an object oriented language. (I say
'supposedly' because I've never looked at it carefully. I'm only
repeating what the press has reported on it.) Therefore, it should be
compared to other object oriented languages, including those that
enforce OOP or simply allow it as an optional programming paradigm. Some
of these are: Java, C++, Perl, etc.

Then, you have to answer some important questions about each language,
such as: number of platforms supported, library support, o/s system APIs
for each platform, portability, multiple vendor support, etc.

The biggest problem I can see with C# is that it was designed by
Microsoft as a language coupled to an environment (.NET) that would
forever bind developers to Windows. With the development of the Mono
project, this is no longer true. So, as a developer, how much time and
money are you willing to spend when Microsoft drops C# and comes out
with their next strategic move? Ask the J++ developers how they are
feeling right about now.

--
Paul Hovnanian mailto:Pa**@Hovnanian.com
note to spammers: a Washington State resident
------------------------------------------------------------------
Dedicated to the unrestricted propagation of worthless
information across the Internet.

ECMA and ISO isnt a windows machine.

GTK# isnt windows specific, Mono isnt windows specific.

Have you ever read any of the specifications?

While you are dilly dallying about fancy footing and whining I will be to
market before you. Thats what counts, thats what made microsoft what it is,
they where there when needed.

Be mr.fancy pants all you want. I will be there first :D

"Paul Hovnanian P.E." <Pa**@Hovnanian.com> wrote in message
news:3F***************@Hovnanian.com...

The language isn't the entire issue.

First of all, C# is supposedly an object oriented language. (I say
'supposedly' because I've never looked at it carefully. I'm only
repeating what the press has reported on it.) Therefore, it should be
compared to other object oriented languages, including those that
enforce OOP or simply allow it as an optional programming paradigm. Some
of these are: Java, C++, Perl, etc.

Then, you have to answer some important questions about each language,
such as: number of platforms supported, library support, o/s system APIs
for each platform, portability, multiple vendor support, etc.

The biggest problem I can see with C# is that it was designed by
Microsoft as a language coupled to an environment (.NET) that would
forever bind developers to Windows. With the development of the Mono
project, this is no longer true. So, as a developer, how much time and
money are you willing to spend when Microsoft drops C# and comes out
with their next strategic move? Ask the J++ developers how they are
feeling right about now.

--
Paul Hovnanian mailto:Pa**@Hovnanian.com
note to spammers: a Washington State resident
------------------------------------------------------------------
Dedicated to the unrestricted propagation of worthless
information across the Internet.

[Paul Hovnanian P.E.]

di********@discussion.microsoft.com wrote:

ECMA and ISO isnt a windows machine.

GTK# isnt windows specific, Mono isnt windows specific.

Have you ever read any of the specifications?

While you are dilly dallying about fancy footing and whining I will be to
market before you. Thats what counts, thats what made microsoft what it is,
they where there when needed.

Be mr.fancy pants all you want. I will be there first :D

I'm not principally a developer. I'm a customer. Over the past 10 years
I have yet to see anyone successfully port applications we use to
Windows. They have been available on SunOS, HPUX and AIX for years. And
the port to Linux was trivially simple, according to the developers.
They believe that it won't be much more difficult for OS X. The code I
do write is to tie these applications together, so I need something that
can interoperate between multiple systems.

One other thing you have to consider when selling apps. to _informed_
customers: We can calculate our cost application ownership quite easily.
We don't so much care how easy it was for the vendor to crank out. In
fact, in my experience, platforms that support quick and dirty s/w
development tend to get exactly that kind of product built for them. And
then we, the end users, have to pay.

--
Paul Hovnanian mailto:Pa**@Hovnanian.com
note to spammers: a Washington State resident
------------------------------------------------------------------
On a clear disk, you can seek forever.

[Dead Dave, Founder of Wendy's]

Paul Hovnanian P.E., you should try our classic single cheddar burger,
just 99 cents (!)

We don't so much care how easy it was for the vendor to crank out. In
fact, in my experience, platforms that support quick and dirty s/w
development tend to get exactly that kind of product built for them. And
then we, the end users, have to pay.

Finally (!)

You know when the books on 'RAD' and
'Extreme Programming' appeared, I thought
they were good ideas.

Now I think they are code words for 'crappy
software' and 'employer exploitation of
contract programmers'. Note - that some
of the RAD ideas (books) were written
by miblosoft people.

Some of the best, most useful and most
stable software systems in the world
were not 'thrown together' but built up
painstakingly over decades, building in
the best ideas.

You know, one of the problems of a group
like this is that when posters say 'software'
they almost always mean desktop applications.

But desktop applications do not run our
society. It's the big information systems
that do. The air traffic controller
software, the NORAD software, the traffic
control software, railroad routing, airplane
reservation systems, police databases,
IRS systems, Social Security information
systems, ADP payroll -- and so on...

[Sohail]

On Wed, 05 Nov 2003 18:57:43 +0100, discussio wrote:

with 2.0 its a rich environment, while you are mopping up yer
maintenance shti ill be releasing rich solutions before you even start
emliminating youre bugs and doing boring things like housekeeping.

I think the main thing with the MS guys is that they reimplement the same
program 300 times in two years. Look at MS, they will keep releasing
office 2000 until 2010.
Database file #1: termpaper.bib
Illegal end of database file---line 22 of file termpaper.bib
: }
Database file #1: termpaper.bib
Illegal end of database file---line 22 of file termpaper.bib
: }

So yes you may come out with a program faster than another will come out
with one that is sort of similar. However programs developed for the
"other" platforms will likely need very few feature changes because there
is an inherent skill and intelligence level required which directly
affects design and extensibility.

Tell me one windows program that you cannot port to *n.x systems. Windows
is an ok desktop (GNOME/KDE are better) but as a system, it SUCKS. There
is no way in hell I would be willing to have my company infrastructure
and/or business run on windows.

Windows RAD development has come to the point where Joe Blow stupid
programmer can do whatever the hell he wants and as long as it doesnt
break when he steps through it, it works and s/he will think it is
production level.

Part of this has to do with stupid non technically-inclined management too
but dont get me started.

Oh yeah, btw every *good* developer knows that maintenance is more than
the majority of time required for any product. So go on and release your
"rich", market-hype filled, problem-making solutions.

[Sohail]

On Sun, 09 Nov 2003 21:11:57 +0000, Sohail wrote:

Database file #1: termpaper.bib
Illegal end of database file---line 22 of file termpaper.bib
: }
Database file #1: termpaper.bib
Illegal end of database file---line 22 of file termpaper.bib
: }

heh. I must have clicked the middle button :)

Divide by ZERO!!!

Well, you are now tarnishing me with the same brush that you wipe yer anti
ms arsse with yet you know nothing of my skill, laughable troll.
"Sohail" <so****@NOSPAMsohailsomani.com> wrote in message
news:pa****************************@NOSPAMsohailso mani.com...

On Wed, 05 Nov 2003 18:57:43 +0100, discussio wrote:

with 2.0 its a rich environment, while you are mopping up yer
maintenance shti ill be releasing rich solutions before you even start
emliminating youre bugs and doing boring things like housekeeping.

I think the main thing with the MS guys is that they reimplement the same
program 300 times in two years. Look at MS, they will keep releasing
office 2000 until 2010.
Database file #1: termpaper.bib
Illegal end of database file---line 22 of file termpaper.bib
: }
Database file #1: termpaper.bib
Illegal end of database file---line 22 of file termpaper.bib
: }

So yes you may come out with a program faster than another will come out
with one that is sort of similar. However programs developed for the
"other" platforms will likely need very few feature changes because there
is an inherent skill and intelligence level required which directly
affects design and extensibility.

Tell me one windows program that you cannot port to *n.x systems. Windows
is an ok desktop (GNOME/KDE are better) but as a system, it SUCKS. There
is no way in hell I would be willing to have my company infrastructure
and/or business run on windows.

Windows RAD development has come to the point where Joe Blow stupid
programmer can do whatever the hell he wants and as long as it doesnt
break when he steps through it, it works and s/he will think it is
production level.

Part of this has to do with stupid non technically-inclined management too
but dont get me started.

Oh yeah, btw every *good* developer knows that maintenance is more than
the majority of time required for any product. So go on and release your
"rich", market-hype filled, problem-making solutions.

[Peter Köhlmann]

di********@discussion.microsoft.com top-posted:

Well, you are now tarnishing me with the same brush that you wipe yer anti
ms arsse with yet you know nothing of my skill, laughable troll.

Considering your language "skills", you would be lucky to know what a
computer language is. So you are probably a VB-"programmer". Or a
"c#-dimwit". One of those people not allowed to clean the toilets in real
IT-shacks

< snip >
--
Microsoft's Guide To System Design:
Let it get in YOUR way. The problem for your problem.

[Sohail]

On Sun, 09 Nov 2003 23:48:58 +0100, discussio wrote:

Well, you are now tarnishing me with the same brush that you wipe yer
anti ms arsse with yet you know nothing of my skill, laughable troll.

Youre just trying to FUD your way out of the validity of your argument. MS
just teaches you how to do it well.

[Sohail]

On Mon, 10 Nov 2003 01:04:17 +0000, Sohail wrote:

On Sun, 09 Nov 2003 23:48:58 +0100, discussio wrote:

Well, you are now tarnishing me with the same brush that you wipe yer
anti ms arsse with yet you know nothing of my skill, laughable troll.

Youre just trying to FUD your way out of the validity of your argument. MS
just teaches you how to do it well.

heh... of course I meant validity of my argument :)

Your argument just consists of BS

[Sohail]

On Sun, 09 Nov 2003 23:48:58 +0100, discussio wrote:

Well, you are now tarnishing me with the same brush that you wipe yer anti
ms arsse with yet you know nothing of my skill, laughable troll.
"Sohail" <so****@NOSPAMsohailsomani.com> wrote in message
news:pa****************************@NOSPAMsohailso mani.com...

On Wed, 05 Nov 2003 18:57:43 +0100, discussio wrote:

> with 2.0 its a rich environment, while you are mopping up yer

Is it just me or do all the MS dumbasses *always* top post?

[Peter Köhlmann]

Sohail wrote:

On Sun, 09 Nov 2003 23:48:58 +0100, discussio wrote:

Well, you are now tarnishing me with the same brush that you wipe yer
anti ms arsse with yet you know nothing of my skill, laughable troll.
"Sohail" <so****@NOSPAMsohailsomani.com> wrote in message
news:pa****************************@NOSPAMsohailso mani.com...

On Wed, 05 Nov 2003 18:57:43 +0100, discussio wrote:

> with 2.0 its a rich environment, while you are mopping up yer

Is it just me or do all the MS dumbasses *always* top post?

No. Just a certain number of them. A lot of windows dimbulbs now try to look
less stupid. Then they blow it with using OE...
--
Another name for a Windows tutorial is crash course

[Kadaitcha Man]

It's just you.

Sohail wrote:

On Sun, 09 Nov 2003 23:48:58 +0100, discussio wrote:

Well, you are now tarnishing me with the same brush that you wipe
yer anti ms arsse with yet you know nothing of my skill, laughable
troll.
"Sohail" <so****@NOSPAMsohailsomani.com> wrote in message
news:pa****************************@NOSPAMsohailso mani.com...

On Wed, 05 Nov 2003 18:57:43 +0100, discussio wrote:

with 2.0 its a rich environment, while you are mopping up yer

Is it just me or do all the MS dumbasses *always* top post?

--
If you want to be me, lop your dick off first.
http://kadaitcha.kicks-ass.org:83/smoking.aspx