What happens when your program crashes?

[CBFalconer]

The evidence is that your brain seems damaged falconer!

In one test suite I have
21877 lines in 351807 bytes with 603 files
In another
24390 lines in 727447 bytes with 41 files

and several smaller ones scattered around.

Was the antagonistic reply necessary? I did express an opinion,
but included the possibility of being mistaken.

 

[CBFalconer]

Are you suggesting that each time someone says C is simple, what
they really mean C is simpl_er_ than other mainstream programming
languages?

If anyone is trying to propose 'simplicity' as a reason for
preponderance, it definitely won't wash. Pascal is _much_ simpler
than C. It is also easier to write _correct_ programs in Pascal
than C, but _easier_ to write faulty programs in C than Pascal.

 

[CBFalconer]

Flash Gordon wrote:
.... snip ...

Well, besides your incredible arrogance, personally I wouldn't
pay a PENNY for your time.

As has been pointed out before, debuggers are OT on c.l.c. Besides
being antagonistic, what reason have you got for ignoring
topicality here? Your system operates on lcc, and there is a news
group just for that compiler, where I suspect there are no
topicality barriers.

 

[CBFalconer]

.... snip ...

2) GDB is a monster debugger, like its cousin GCC. Yes,
it is much more bloated anything jacob can ever do,
but that was one of my major selling points:

GCC+GDB is around 30MB of C.
lcc + debugger is around 3MB of C.

Which one would you like to configure and maintain ?

On my system gcc423 occupies a total of roughly 350k bytes. Gdb
6.1 is more expensive, it occupies 1888k bytes. These are versions
for djgpp which DO NOT use any dlls.

As far as I can see the gcc + gdb system uses considerably less
space than lcc-w32 does. In fact, approximately 2/3. In addition,
when I last used lcc-w32 (quite some time ago) it generated object
code about 50% larger than did gcc, of the same vintage. The
distributed install package for gcc is over 4 megabytes, but that
includes comprehensive documentation.

Just thought some of your misinformation should be corrected.

Once more, why are you posting this in c.l.c., where it is all
firmly off-topic?

 

[santosh]

jacob navia wrote, On 20/03/08 20:11:

Flash Gordon wrote:

[...]

Well, besides your incredible arrogance, personally I wouldn't pay a
PENNY for your time.

Indian programmers can also have high salaries didn't you know?

I've worked with them (I spent about 4 months in India with them as
well as time in the UK), have you? I've discussed with managers what
their salaries are, have you? There salary is a *lot* less than the
rate my time is charged out at. Perhaps I should have said, "Indian
developers in India", but I thought it was obvious, just as it was
obvious that I did not mean Native Americans.

I can testify to what Flash is saying. An average s/w developer working
for one of the bigger s/w companies in India has a monthly salary
roughly near 1500USD, give or take a few hundred USD on either side.

<snip>

 

[CBFalconer]

Flash Gordon wrote:
.... snip ...


I can testify to what Flash is saying. An average s/w developer working
for one of the bigger s/w companies in India has a monthly salary
roughly near 1500USD, give or take a few hundred USD on either side.

However I suspect there is no reason to feel sorry for them. In
India that salary probably allows anyone to live like a king, have
servants, etc.

 

[santosh]

On my system gcc423 occupies a total of roughly 350k bytes. Gdb
6.1 is more expensive, it occupies 1888k bytes. These are versions
for djgpp which DO NOT use any dlls.

Are there *any* DJGPP versions that do use DLLs? I thought DJGPP was DOS
based?

As far as I can see the gcc + gdb system uses considerably less
space than lcc-w32 does. In fact, approximately 2/3.

Nevertheless a typical gcc and gdb installation is bigger than
lcc-win32, both the compressed distribution and the uncompressed disk
space occupied.

In addition,
when I last used lcc-w32 (quite some time ago) it generated object
code about 50% larger than did gcc, of the same vintage. The
distributed install package for gcc is over 4 megabytes, but that
includes comprehensive documentation.

You are (I think) talking about one particular port of gcc. The gcc
package supplied at the official website is *quite* a bit bigger than
lcc-win32.

For example the core gcc archive for version 4.3.0 is 23.6 Mb even when
bzip2 compressed.

Note also that the binaries for gcc4.2.3 for DJGPP are about 4.5 Mb. The
documentation is ~10 Mb and source is 45 Mb.

Just thought some of your misinformation should be corrected.

With more misinformation it seems.

Once more, why are you posting this in c.l.c., where it is all
firmly off-topic?

Okay, this is the first line in this post I can concur with.

 

[santosh]

However I suspect there is no reason to feel sorry for them. In
India that salary probably allows anyone to live like a king, have
servants, etc.

Eh? Wherever did I say I felt sorry for anyone? I was merely
corroborating Flash's statements with my own experience.

 

[Richard Heathfield]

CBFalconer said:

If anyone is trying to propose 'simplicity' as a reason for
preponderance, it definitely won't wash. Pascal is _much_ simpler
than C.

I can't agree with that. I've *tried* learning Pascal. Over and over and
over. I don't know what it is about that language, but for some reason it
just won't stick in my head. Perhaps simplicity is in the eye of the
beholder.

It is also easier to write _correct_ programs in Pascal
than C, but _easier_ to write faulty programs in C than Pascal.

For you, maybe. Not for me.

 

[jacob navia]

jacob navia wrote:
... snip ...

On my system gcc423 occupies a total of roughly 350k bytes. Gdb
6.1 is more expensive, it occupies 1888k bytes. These are versions
for djgpp which DO NOT use any dlls.

I was speaking of SOURCE CODE obviously. Not the bare executable

As far as I can see the gcc + gdb system uses considerably less
space than lcc-w32 does. In fact, approximately 2/3. In addition,
when I last used lcc-w32 (quite some time ago) it generated object
code about 50% larger than did gcc, of the same vintage. The
distributed install package for gcc is over 4 megabytes, but that
includes comprehensive documentation.

Just thought some of your misinformation should be corrected.

You are spreading disinformation. The executable of gcc 2.96
(a version very old) in my linux system is 86040. It calls
"cc1" that is 2 680 984 bytes. It uses (probably) /usr/bin/cpp
(87 608 bytes)

This mens the bare executables of gcc add up to 2 854 632 bytes.
The bare executable of lcc-win32 is 632 352 including cpp that
is integrated into that executable.

Under windows gcc 3.2.1 has 87 040, CC1 has 7 772 656 and cpp
has 88 576.

The source distribution of gcc is around 50MB compressed, compared
to lcc-win's 4MB uncompressed.

 

[CBFalconer]

CBFalconer said:



I can't agree with that. I've *tried* learning Pascal. Over and over
and over. I don't know what it is about that language, but for some
reason it just won't stick in my head. Perhaps simplicity is in the
eye of the beholder.


For you, maybe. Not for me.

What Pascal version did you try? Just like C, using something like
the Borland style can throw you way off, because they omit critical
portions and add a multitude of unnecessary 'features'. That's why
I mentioned iso7185.

The other thing that may bug newbies is the i/o. The first item is
always read into the file buffer on file open, at least
figuratively (interactive files use lazy-i/o, which postpones the
read until something occurs that need to know about it. Such as
testing eof, eoln, etc.) This is a portion that is missing from
the Borland products, largely because of omission of f^ and the put
and get functions.

 

[CBFalconer]

Eh? Wherever did I say I felt sorry for anyone? I was merely
corroborating Flash's statements with my own experience.

You didn't. But the general tenure was such that many would feel
that they were grossly underpaid. They may be, but I doubt it.

 

[CBFalconer]

Are there *any* DJGPP versions that do use DLLs? I thought DJGPP
was DOS based?

No. Yes.

Nevertheless a typical gcc and gdb installation is bigger than
lcc-win32, both the compressed distribution and the uncompressed
disk space occupied.

You are (I think) talking about one particular port of gcc. The
gcc package supplied at the official website is *quite* a bit
bigger than lcc-win32.

For example the core gcc archive for version 4.3.0 is 23.6 Mb
even when bzip2 compressed.

Note also that the binaries for gcc4.2.3 for DJGPP are about
4.5 Mb. The documentation is ~10 Mb and source is 45 Mb.

Here is the distribution for DJGPP gcc423b. Note the first 4
files. They are the COMPLETE binary executables. The rest is
basically documentation, and a small portion is updates to the
#include file system. Note that I gave the expanded size of the
binaries, not the compressed distributed size. Also note that the
majority of the package is in the info and man directories, all of
which are online documentation.

I take at least some of it back. There is another 2.5 odd
megabytes hiding in the libexec subdirectory. However, the
distribution is complete (apart from source) with all documentation
etc. The available documentation package basically gives the
ability to generate other formats.

The following listing will have some line wrap problems.

Archive: gcc423b.zip 4527662 bytes 84 files
-rwxa-- 2.3 ntf 119084 bx defX 14-Feb-08 02:08 bin/cpp.exe
-rwxa-- 2.3 ntf 118400 bx defX 14-Feb-08 02:08 bin/gcc.exe
-rw-a-- 2.3 ntf 16446 tx defX 14-Feb-08 02:08 bin/gccbug
-rwxa-- 2.3 ntf 52788 bx defX 14-Feb-08 02:08 bin/gcov.exe
-rw-a-- 2.3 ntf 30058 tx defX 14-Feb-08 02:08
gnu/gcc-4.23/BUGS
-rw-a-- 2.3 ntf 18002 tx defX 14-Feb-08 02:08
gnu/gcc-4.23/COPYING
-rw-a-- 2.3 ntf 26527 tx defX 14-Feb-08 02:08
gnu/gcc-4.23/COPYING.LIB
-rw-a-- 2.3 ntf 23422 tx defX 14-Feb-08 02:08
gnu/gcc-4.23/FAQ
-rw-a-- 2.3 ntf 391168 tx defX 14-Feb-08 02:08
gnu/gcc-4.23/NEWS
-rw-a-- 2.3 ntf 795 tx defX 14-Feb-08 02:08
gnu/gcc-4.23/README
-rw-a-- 2.3 ntf 383 tx defX 14-Feb-08 02:08
gnu/gcc-4.23/README.SCO
-rw-a-- 2.3 ntf 30537 tx defX 14-Feb-08 02:08
gnu/gcc-4.23/bugs.html
-rw-a-- 2.3 ntf 23903 tx defX 14-Feb-08 02:08
gnu/gcc-4.23/faq.html
-rw-a-- 2.3 ntf 12665 tx defX 14-Feb-08 02:08
gnu/gcc-4.23/fixincludes/README
-rw-a-- 2.3 ntf 750 tx defX 14-Feb-08 02:08
gnu/gcc-4.23/fixincludes/README-fixinc
-rw-a-- 2.3 ntf 18002 tx defX 14-Feb-08 02:08
gnu/gcc-4.23/gcc/COPYING
-rw-a-- 2.3 ntf 12566 tx defX 14-Feb-08 02:08
gnu/gcc-4.23/readme.DJGPP
-rw-a-- 2.3 ntf 232585 bx defX 14-Feb-08 02:08 info/cpp.info
-rw-a-- 2.3 ntf 51245 bx defX 14-Feb-08 02:08
info/cppinternals.info
-rw-a-- 2.3 ntf 1519947 bx defX 14-Feb-08 02:08 info/gcc.info
-rw-a-- 2.3 ntf 182042 bx defX 14-Feb-08 02:08
info/gccinstall.info
-rw-a-- 2.3 ntf 1634510 bx defX 14-Feb-08 02:08
info/gccint.info
-rw-a-- 2.3 ntf 26 tx stor 14-Feb-08 02:08
lib/gcc/djgpp/4.23/djgpp.ver
-rw-a-- 2.3 ntf 750 tx defX 14-Feb-08 01:42
lib/gcc/djgpp/4.23/include/README
-rw-a-- 2.3 ntf 3662 tx defX 14-Feb-08 01:42
lib/gcc/djgpp/4.23/include/decfloat.h
-rw-a-- 2.3 ntf 41808 tx defX 14-Feb-08 01:42
lib/gcc/djgpp/4.23/include/emmintrin.h
-rw-a-- 2.3 ntf 5413 tx defX 14-Feb-08 01:42
lib/gcc/djgpp/4.23/include/float.h
-rw-a-- 2.3 ntf 1432 tx defX 14-Feb-08 01:42
lib/gcc/djgpp/4.23/include/iso646.h
-rw-a-- 2.3 ntf 3470 tx defX 14-Feb-08 01:42
lib/gcc/djgpp/4.23/include/limits.h
-rw-a-- 2.3 ntf 4918 tx defX 14-Feb-08 01:42
lib/gcc/djgpp/4.23/include/mm3dnow.h
-rw-a-- 2.3 ntf 2365 tx defX 14-Feb-08 01:42
lib/gcc/djgpp/4.23/include/mm_malloc.h
-rw-a-- 2.3 ntf 26573 tx defX 14-Feb-08 01:42
lib/gcc/djgpp/4.23/include/mmintrin.h
-rw-a-- 2.3 ntf 3897 tx defX 14-Feb-08 01:42
lib/gcc/djgpp/4.23/include/pmmintrin.h
-rw-a-- 2.3 ntf 2606 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/include/ssp/ssp.h
-rw-a-- 2.3 ntf 3612 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/include/ssp/stdio.h
-rw-a-- 2.3 ntf 5859 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/include/ssp/string.h
-rw-a-- 2.3 ntf 2955 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/include/ssp/unistd.h
-rw-a-- 2.3 ntf 4317 tx defX 14-Feb-08 01:42
lib/gcc/djgpp/4.23/include/stdarg.h
-rw-a-- 2.3 ntf 1604 tx defX 14-Feb-08 01:42
lib/gcc/djgpp/4.23/include/stdbool.h
-rw-a-- 2.3 ntf 12695 tx defX 14-Feb-08 01:42
lib/gcc/djgpp/4.23/include/stddef.h
-rw-a-- 2.3 ntf 330 tx defX 14-Feb-08 00:40
lib/gcc/djgpp/4.23/include/syslimits.h
-rw-a-- 2.3 ntf 9096 tx defX 14-Feb-08 01:42
lib/gcc/djgpp/4.23/include/unwind.h
-rw-a-- 2.3 ntf 139 tx defX 14-Feb-08 01:42
lib/gcc/djgpp/4.23/include/varargs.h
-rw-a-- 2.3 ntf 36692 tx defX 14-Feb-08 01:42
lib/gcc/djgpp/4.23/include/xmmintrin.h
-rw-a-- 2.3 ntf 330 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/install-tools/gsyslimits.h
-rw-a-- 2.3 ntf 750 tx defX 14-Feb-08 02:07
lib/gcc/djgpp/4.23/install-tools/include/README
-rw-a-- 2.3 ntf 3662 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/install-tools/include/decfloat.h
-rw-a-- 2.3 ntf 41808 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/install-tools/include/emmintrin.h
-rw-a-- 2.3 ntf 5413 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/install-tools/include/float.h
-rw-a-- 2.3 ntf 1432 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/install-tools/include/iso646.h
-rw-a-- 2.3 ntf 3470 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/install-tools/include/limits.h
-rw-a-- 2.3 ntf 4918 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/install-tools/include/mm3dnow.h
-rw-a-- 2.3 ntf 2365 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/install-tools/include/mm_malloc.h
-rw-a-- 2.3 ntf 26573 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/install-tools/include/mmintrin.h
-rw-a-- 2.3 ntf 3897 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/install-tools/include/pmmintrin.h
-rw-a-- 2.3 ntf 4317 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/install-tools/include/stdarg.h
-rw-a-- 2.3 ntf 1604 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/install-tools/include/stdbool.h
-rw-a-- 2.3 ntf 12695 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/install-tools/include/stddef.h
-rw-a-- 2.3 ntf 9096 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/install-tools/include/unwind.h
-rw-a-- 2.3 ntf 139 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/install-tools/include/varargs.h
-rw-a-- 2.3 ntf 36692 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/install-tools/include/xmmintrin.h
-rw-a-- 2.3 ntf 45 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/install-tools/macro_list
-rw-a-- 2.3 ntf 131 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/install-tools/mkheaders.conf
-rw-a-- 2.3 ntf 66004 bx defX 14-Feb-08 02:09
lib/gcc/djgpp/4.23/libgcc.a
-rw-a-- 2.3 ntf 15280 bx defX 14-Feb-08 02:09
lib/gcc/djgpp/4.23/libgcov.a
-rw-a-- 2.3 ntf 8018 bx defX 14-Feb-08 02:09
lib/gcc/djgpp/4.23/libssp.a
-rw-a-- 2.3 ntf 716 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/libssp.la
-rw-a-- 2.3 ntf 610 bx defX 14-Feb-08 02:09
lib/gcc/djgpp/4.23/libssp_nonshared.a
-rw-a-- 2.3 ntf 746 tx defX 14-Feb-08 02:08
lib/gcc/djgpp/4.23/libssp_nonshared.la
-rwxa-- 2.3 ntf 2623676 bx defX 14-Feb-08 02:09
libexec/gcc/djgpp/4.23/cc1.exe
-rwxa-- 2.3 ntf 91936 bx defX 14-Feb-08 02:08
libexec/gcc/djgpp/4.23/collect2.exe
-rw-a-- 2.3 ntf 26 tx stor 14-Feb-08 02:08
libexec/gcc/djgpp/4.23/djgpp.ver
-rw-a-- 2.3 ntf 13621 tx defX 14-Feb-08 02:07
libexec/gcc/djgpp/4.23/install-tools/fixinc.sh
-rwxa-- 2.3 ntf 88596 bx defX 14-Feb-08 02:07
libexec/gcc/djgpp/4.23/install-tools/fixincl.exe
-rw-a-- 2.3 ntf 3289 tx defX 14-Feb-08 02:07
libexec/gcc/djgpp/4.23/install-tools/mkheaders
-rw-a-- 2.3 ntf 41752 tx defX 14-Feb-08 02:08 man/cat1/cpp.1
-rw-a-- 2.3 ntf 750681 tx defX 14-Feb-08 02:08 man/cat1/gcc.1
-rw-a-- 2.3 ntf 25102 tx defX 14-Feb-08 02:08
man/cat1/gcov.1
-rw-a-- 2.3 ntf 3365 tx defX 14-Feb-08 02:08
man/cat7/fsf-funding.7
-rw-a-- 2.3 ntf 28140 tx defX 14-Feb-08 02:08
man/cat7/gfdl.7
-rw-a-- 2.3 ntf 24210 tx defX 14-Feb-08 02:08 man/cat7/gpl.7
-rw-a-- 2.3 ntf 1056 tx defX 14-Feb-08 02:08
manifest/gcc423b.dsm
-rw-a-- 2.3 ntf 3730 tx defX 14-Feb-08 02:10
manifest/gcc423b.mft
-rw-a-- 2.3 ntf 64 tx defX 14-Feb-08 02:08
manifest/gcc423b.ver
84 files, 8639999 bytes uncompressed, 4513612 bytes compressed:
47.8%

 

[CBFalconer]

I was speaking of SOURCE CODE obviously. Not the bare executable


You are spreading disinformation. The executable of gcc 2.96
(a version very old) in my linux system is 86040. It calls
"cc1" that is 2 680 984 bytes. It uses (probably) /usr/bin/cpp
(87 608 bytes)

Yes, there were some mistakes. I DID miss about 2.5 megabytes of
executable code. However, you have no business speaking about
source code, as it is not available.

 

[Anand Hariharan]

If anyone is trying to propose 'simplicity' as a reason for
preponderance, it definitely won't wash.  Pascal is _much_ simpler
than C.  It is also easier to write _correct_ programs in Pascal
than C, but _easier_ to write faulty programs in C than Pascal.

Prior to your post, I didn't see the thread as taking a direction that
was discussing why C is so ubiquitous. I was extending the discussion
of whether a pair of eyeballs is all that one needs to debug a C
program, or are tools necessary?

That said, the famous essay "Worse is better" does strongly suggest
that C received widespread acceptance not because it was correct,
consistent or complete, but rather because it was simpler than Lisp,
which the essay purported was correct and consistent but difficult to
work with, when compared to C.

- Anand

 

[Richard Heathfield]

CBFalconer said:

What Pascal version did you try?

You mean it comes in VERSIONS?!? Sheesh, haven't they ever heard of
standardisation?

 

[CBFalconer]

CBFalconer said:

You mean it comes in VERSIONS?!? Sheesh, haven't they ever heard
of standardisation?

Taking you seriously, that's why I mentioned ISO7185, which is the
ISO standard. Anything that meets that is OK.

 

[santosh]

jacob navia wrote:

The source distribution of gcc is around 50MB compressed, compared
to lcc-win's 4MB uncompressed.

But that doesn't include full source for lcc-win32, unlike gcc. What
would be the size of download of the full sources for lcc-win32?

 

[santosh]

Taking you seriously, that's why I mentioned ISO7185, which is the
ISO standard. Anything that meets that is OK.

And do we have a fully conforming implementation of ISO 7185 that's
reasonably cross platform as well?

 

[santosh]

You didn't. But the general tenure was such that many would feel
that they were grossly underpaid. They may be, but I doubt it.

Not at all. ~1500USD/month is a salary that many would die for. Compared
to the population of this country, only a small fraction earn that
much. Which is why s/w engineering is currently the most coveted field
of study, to be taken whether or not one really *has* any interest in
CS.