C related, but not a C question..

[clayne]

So I was recently checking out topcoder.com to see what the latest hype
is all about (as I had been listening to an interview on NPR).
Apparently it's a group site where they hold programming competitions,
etc. to come up with the best solutions for a given task (that may have
been farmed out to topcoder via another company). I'm looking around at
the contests, based on algorithms, design, etc. and I'm trying to
figure out "okay, where's the language choices" and eventually come to
figure out that it's 99% C++, Java, or C# only. So a bit miffed I had
to ask why C was not represented.. Check out the general party line to
their perceptions of C:

http://forums.topcoder.com/?module=T...513334&start=0

Let's all be proud to be writing in uncool, old, "legacy" code because
apparently we're "hard to find."

[Richard Heathfield]

clayne said:

Let's all be proud to be writing in uncool, old, "legacy" code

Legacy code, n. (pej.): code written in some language other than the one
preferred by the speaker.

because apparently we're "hard to find."

You didn't seem to have any trouble.

--
Richard Heathfield
"Usenet is a strange place" - dmr 29/7/1999
http://www.cpax.org.uk
email: rjh at above domain (but drop the www, obviously)

[clayne]

Richard Heathfield wrote:

Legacy code, n. (pej.): code written in some language other than the one
preferred by the speaker.

Hah.

because apparently we're "hard to find."

You didn't seem to have any trouble.

In regards to?
One thing, at work, I've actually been told to not write something in C
(that was actually the most appropriate language for it [Solaris memory
statistics, etc.]) and instead write it in Perl, since it would require
others to know more than Perl to support it (even though our developers
al know C/C++). Right, let's use a crappier tool for the task because
everyone else carries the same crappy tools around. Kind of like
installing a door handle with a pipe wrench because that's all one has
ever known.

[William Hughes]

clayne wrote:

Richard Heathfield wrote:

Legacy code, n. (pej.): code written in some language other than the one
preferred by the speaker.

Hah.

because apparently we're "hard to find."

You didn't seem to have any trouble.

In regards to?
One thing, at work, I've actually been told to not write something in C
(that was actually the most appropriate language for it [Solaris memory
statistics, etc.]) and instead write it in Perl, since it would require
others to know more than Perl to support it (even though our developers
al know C/C++). Right, let's use a crappier tool for the task because
everyone else carries the same crappy tools around. Kind of like
installing a door handle with a pipe wrench because that's all one has
ever known.

Hardly. It's more like installing a door handle using slot screws
rather
than phillips screws, so even people without a phillips screwdriver
can work on the door. True, phillips screws are better, but it is
also true that more people have access to something that can be
used to turn slot screws than phillips screws. Whether or not
slot or phillips screws should be used depends on a lot of things,
but using slot screws is not a priori silly.

- William Hughes

[sjdevnull]

clayne wrote:

One thing, at work, I've actually been told to not write something in C
(that was actually the most appropriate language for it [Solaris memory
statistics, etc.]) and instead write it in Perl, since it would require
others to know more than Perl to support it (even though our developers
al know C/C++).

So let me get this straight: you were told to use _Perl_ because of
concerns about others reading and maintaining your code?

Wow.

[clayne]

sj*******@yahoo.com wrote:

So let me get this straight: you were told to use _Perl_ because of
concerns about others reading and maintaining your code?

Wow.

Yep. Due to the department the tool would be written for primarily
being equipped with perl coders. Nope, I'm not going to write a tool to
make lightweight calls into Solaris' process mapping routines in Perl.
In fact, the entire issue came about because the standard interface,
pmap, was too expensive (resource-wise) to use as a source of
monitoring statistics.

The fun continues at "Top" coder land. Apparently C is now a "sub-set"
of C++, and C is a language for heathens:

http://forums.topcoder.com/?module=T...513334&start=0

If it makes you feel any better, star C++ programmer there, dropping
the bombs on C, has made 18,000$ in the last year on just that site
(atleast according to what his profile shows).

[Barry]

"clayne" <cl****@anodized.comwrote in message
news:11**********************@i3g2000cwc.googlegro ups.com...

>
sj*******@yahoo.com wrote:

So let me get this straight: you were told to use _Perl_ because of
concerns about others reading and maintaining your code?

Wow.

Yep. Due to the department the tool would be written for primarily
being equipped with perl coders. Nope, I'm not going to write a tool to
make lightweight calls into Solaris' process mapping routines in Perl.
In fact, the entire issue came about because the standard interface,
pmap, was too expensive (resource-wise) to use as a source of
monitoring statistics.

The fun continues at "Top" coder land. Apparently C is now a "sub-set"
of C++, and C is a language for heathens:

http://forums.topcoder.com/?module=T...513334&start=0

If it makes you feel any better, star C++ programmer there, dropping
the bombs on C, has made 18,000$ in the last year on just that site
(atleast according to what his profile shows).

One of the Fortune companies I used to work for actually wanted to forbade
the use of Perl. :-)

[jacob navia]

clayne wrote:

sj*******@yahoo.com wrote:

>>So let me get this straight: you were told to use _Perl_ because of
concerns about others reading and maintaining your code?

Wow.

Yep. Due to the department the tool would be written for primarily
being equipped with perl coders. Nope, I'm not going to write a tool to
make lightweight calls into Solaris' process mapping routines in Perl.
In fact, the entire issue came about because the standard interface,
pmap, was too expensive (resource-wise) to use as a source of
monitoring statistics.

The fun continues at "Top" coder land. Apparently C is now a "sub-set"
of C++, and C is a language for heathens:

http://forums.topcoder.com/?module=T...513334&start=0

If it makes you feel any better, star C++ programmer there, dropping
the bombs on C, has made 18,000$ in the last year on just that site
(atleast according to what his profile shows).

One of the things a LIKED in there was this:

3) Low level languages are MUCH more costly to maintain and develop then
high level langauges. Why? Harder to find good people in those langauges
- usually translates into higher hiring costs, higher salary, higher
training, etc. So a business will prefer a 'popular' language than a low
level langauge becuase it's easier to find talented people on those.
You see that?

If you program in C you get higher salary than all those "higher
language" guys. It is true that when you look for a C# programmer
you find a dozen to a dime, but when looking for a good C programmer...

[Mabden]

"jacob navia" <ja***@jacob.remcomp.frwrote in message
news:45**********************@news.orange.fr...

clayne wrote:

sj*******@yahoo.com wrote:

>So let me get this straight: you were told to use _Perl_ because of
concerns about others reading and maintaining your code?
Wow.

Indeed. It seems to be a write-only language.

The fun continues at "Top" coder land. Apparently C is now a "sub-set"
of C++, and C is a language for heathens:

Well, I think Stroustrup pretty much stated that in his "Design and
Evolution" book. C++ should never step on C, only enhance it. I think people
who dismiss C++ for C should read that book.

http://forums.topcoder.com/?module=T...513334&start=0

If it makes you feel any better, star C++ programmer there, dropping
the bombs on C, has made 18,000$ in the last year on just that site
(atleast according to what his profile shows).

Is that a Good Thing or a Bad Thing? I would like to get an extra $18K for a
website I created, while working my day job...

One of the things a LIKED in there was this:
3) Low level languages are MUCH more costly to maintain and develop then
high level langauges. Why? Harder to find good people in those langauges
- usually translates into higher hiring costs, higher salary, higher
training, etc. So a business will prefer a 'popular' language than a low
level langauge becuase it's easier to find talented people on those.

You see that?

If you program in C you get higher salary than all those "higher
language" guys. It is true that when you look for a C# programmer
you find a dozen to a dime, but when looking for a good C programmer...

There's someone looking for a good C programmer???

--
Mabden

[Richard Heathfield]

jacob navia said:

<snip>

3) Low level languages are MUCH more costly to maintain and develop then
high level langauges. Why? Harder to find good people in those langauges
- usually translates into higher hiring costs, higher salary, higher
training, etc. So a business will prefer a 'popular' language than a low
level langauge becuase it's easier to find talented people on those.

Actually, the language is neither here nor there. The real issue is that
talent costs money. If you pay peanuts, you get monkeys, and monkeys tend
to use drool-proof languages, because that's all they can understand.

--
Richard Heathfield
"Usenet is a strange place" - dmr 29/7/1999
http://www.cpax.org.uk
email: rjh at above domain (but drop the www, obviously)

[Michael Mair]

Richard Heathfield wrote:

jacob navia said:
<snip>

>>3) Low level languages are MUCH more costly to maintain and develop then
high level langauges. Why? Harder to find good people in those langauges
- usually translates into higher hiring costs, higher salary, higher
training, etc. So a business will prefer a 'popular' language than a low
level langauge becuase it's easier to find talented people on those.

Actually, the language is neither here nor there. The real issue is that
talent costs money. If you pay peanuts, you get monkeys, and monkeys tend
to use drool-proof languages, because that's all they can understand.

This is not entirely fair. Some of the primates may suffer the
restrictions you yourself wrote about in <Io********************@bt.com>
and be restricted by the ape or monkey with the pointy hairs.
Another thing is that you cannot always get the talent you are looking
for. In this case, the system architect or whoever else sits on the
technology decisions may have to go for a solution which is within
the evolutionary horizon of the people who are available and maybe
allow (more or less officially) a "technological spearhead" to move
a little further than the rest to draw them as soon as they are ready.

The same applies to advanced training / continuing education.

If universities and training centres move away from the oh-so-complex,
then this makes me worry much more... -- as it does.
BTW: I hear from time to time things like "We rather hire C programmers
-- they tend to learn and understand much faster", so I think that
a certain correlation is also perceived on the employers' side.
Cheers
Michael
--
E-Mail: Mine is an /at/ gmx /dot/ de address.

[Richard Heathfield]

Michael Mair said:

Richard Heathfield wrote:

<snip>

>Actually, the language is neither here nor there. The real issue is that
talent costs money. If you pay peanuts, you get monkeys, and monkeys tend
to use drool-proof languages, because that's all they can understand.

This is not entirely fair. Some of the primates may suffer the
restrictions you yourself wrote about in <Io********************@bt.com>
and be restricted by the ape or monkey with the pointy hairs.

You are extending my position. Just because someone has been spotted using a
drool-proof language, that doesn't mean they're not bright. As you say,
they may be working under duress. :-) But those who can /only/ use
drool-proof languages are another matter.

Another thing is that you cannot always get the talent you are looking
for.

It's curious, isn't it, in this day of high-tech communication, that it's
still difficult to find what you need? (And it's even /more/ difficult to
find what you /think/ you need.)

In this case, the system architect or whoever else sits on the
technology decisions may have to go for a solution which is within
the evolutionary horizon of the people who are available

A wonderful turn of phrase. :-)

BTW: I hear from time to time things like "We rather hire C programmers
-- they tend to learn and understand much faster", so I think that
a certain correlation is also perceived on the employers' side.

That's a new one on me, but I'm nonetheless glad that you're hearing it.
It's good to know that the corporate world isn't *completely* ignorant of
the importance of intelligence.

--
Richard Heathfield
"Usenet is a strange place" - dmr 29/7/1999
http://www.cpax.org.uk
email: rjh at above domain (but drop the www, obviously)

[clayne]

Richard Heathfield wrote:

That's a new one on me, but I'm nonetheless glad that you're hearing it.
It's good to know that the corporate world isn't *completely* ignorant of
the importance of intelligence.

No, just only 90% so.