C++ market

[cpisztest]

Are you guys finding that more and more you are given those "Here is this mess from 10-20 years ago that needs fixing" projects as opposed to the "Let's make this new thing" projects? Or is it just me?

It seems more and more that there isn't new code being written and only old code to be fixed in C++ and the new stuff, to my detriment, is being done in Java and .NET, save some server stuff where performance is critical, some scientific, and games.

Of course there is always some amount of fixing others code, and debugging stuff, etc. in any job, and it is also dependent upon experience level, but it seems that is all there is. I wanna make my own bugs and stop fixing other people's!

 

[Stefan Ram]

Are you guys finding that more and more you are given those
"Here is this mess from 10-20 years ago that needs fixing"
projects as opposed to the "Let's make this new thing"
projects? Or is it just me?

10 - 20 years ago it was written:

»Today's C++ programs will be tomorrow's unmaintainable
legacy code. As [GWS 94] says of C++:

"The seeds of software disasters for decades to
come have already been planted and well fertilised."«

retrieved in 2007 from:
http://burks.brighton.ac.uk/burks/pcinfo/progdocs/cppcrit/index005.htm

It seems more and more that there isn't new code being
written and only old code to be fixed in C++ and the new
stuff, to my detriment, is being done in Java and .NET, save
some server stuff where performance is critical, some
scientific, and games.

C++ has lost more than 3 % in the last year from 9.196 %
down to 5.986 on TIOBE, C++'s rank: 1994:2, 2004:3, and 2014:4.

 

[woodbrian77]

Are you guys finding that more and more you are given those "Here is this mess from 10-20 years ago that needs fixing" projects as opposed to the "Let's make this new thing" projects? Or is it just me?



It seems more and more that there isn't new code being written and only old code to be fixed in C++ and the new stuff, to my detriment, is being done in Java and .NET, save some server stuff where performance is critical, some scientific, and games.



Of course there is always some amount of fixing others code, and debugging stuff, etc. in any job, and it is also dependent upon experience level, but it seems that is all there is. I wanna make my own bugs and stop fixing other people's!

In that case you should either start a company or join a small
company. There aren't many jobs at big companies that will be
satisfying for you.


Brian
Ebenezer Enterprises - Where the Spirit of the L-RD is
there is liberty.
http://webEbenezer.net

 

[Jorgen Grahn]

Are you guys finding that more and more you are given those "Here is
this mess from 10-20 years ago that needs fixing" projects as opposed
to the "Let's make this new thing" projects? Or is it just me?

Depends very much on what business you're in. For me it has always
been like that, since I work with long-lived products (telecom servers).

And it's natural for most development to be about extending mature
products. Unfortunately that often means the code is a mess ...

/Jorgen

 

[Victor Bazarov]

[..]I wanna make my own bugs and stop fixing other people's!

How true that is! And where our code shall be in 20 years? In the
hands of other programmers who will likely have the same attitude, if
we're lucky. Or in a virtual trash bin. I'd rather do everything to
achieve the former...

V

 

[Jorgen Grahn]

[..]I wanna make my own bugs and stop fixing other people's!

How true that is! And where our code shall be in 20 years? In the
hands of other programmers who will likely have the same attitude, if
we're lucky. Or in a virtual trash bin. I'd rather do everything to
achieve the former...

Incidentally, the best way to keep out of the trash bin is to do
maintenance programming. Spend 6 months trying to write something
new, and lots of disasters can happen along the way. If you spend 2
weeks finding and fixing a bug you can assume someone will benefit.

/Jorgen

 

[Stefan Ram]

Incidentally, the best way to keep out of the trash bin is to do
maintenance programming. Spend 6 months trying to write something
new, and lots of disasters can happen along the way. If you spend 2
weeks finding and fixing a bug you can assume someone will benefit.

»Single worst strategic mistake you could ever make?
Rewriting code from scratch«

Things You Should Never Do, Part I - Joel on Software

 

[Melzzzzz]

On Tue, 13 May 2014 15:12:03 -0700 (PDT)

Are you guys finding that more and more you are given those "Here is
this mess from 10-20 years ago that needs fixing" projects as opposed
to the "Let's make this new thing" projects? Or is it just me?

Examples? You can always start project by yourself
There is huge code base written, so writing from scratch
is really not practical

It seems more and more that there isn't new code being written and
only old code to be fixed in C++ and the new stuff, to my detriment,
is being done in Java and .NET, save some server stuff where
performance is critical, some scientific, and games.

What new stuff? Same business applications all over again ;0?

Of course there is always some amount of fixing others code, and
debugging stuff, etc. in any job, and it is also dependent upon
experience level, but it seems that is all there is. I wanna make my
own bugs and stop fixing other people's!

Then start new project! It is really difficult to get job
to write something new when all companies have already code base...

 

[Melzzzzz]

On 13 May 2014 22:22:44 GMT

C++ has lost more than 3 % in the last year from 9.196 %
down to 5.986 on TIOBE, C++'s rank: 1994:2, 2004:3, and 2014:4.

And you really believe that C is first, Java second and Objective-C
third?

 

[David Brown]

On 13 May 2014 22:22:44 GMT

And you really believe that C is first, Java second and Objective-C
third?

TIOBE must be the most pointless indicator of programming language usage
or popularity ever invented. It is based mainly on the number of google
searches people make on a language.

So every time someone searches for the letter "C", they increase C's
popularity as a programming language - even if it is "what element has
the chemical symbol C?".

Every time someone thinks "I'm going to write an IPlod application and
get rich quick", searches for the languages used by Apple, and runs away
screaming when they see the syntax for Objective C, the popularity of
Objective C increases.

Every time someone uses C++ to write a program that works, nothing
happens to the C++ ratings on TIOBE.

At best, TIOBE gives a rough rating to "number of people considering a
language" times "proportion of users who need help".

 

[woodbrian77]

Incidentally, the best way to keep out of the trash bin is to do
maintenance programming.

There's nothing wrong with maintenance work ...
but playing it safe gets old eventually.

Spend 6 months trying to write something
new, and lots of disasters can happen along the way. If you spend 2
weeks finding and fixing a bug you can assume someone will benefit.

New things are more difficult, but the rewards are
greater when things do work out. Fear-based decision
making isn't the way I want to live. Dreams are
important and I encourage people to prepare themselves
for good stuff. In some cases that may mean starting
a company so you have a vehicle ready when you need it.

Brian
Ebenezer Enterprises - Nothing ventured, nothing gained.
http://webEbenezer.net

 

[James Kanze]

Are you guys finding that more and more you are given those
"Here is this mess from 10-20 years ago that needs fixing"
projects as opposed to the "Let's make this new thing"
projects?

There's some of that, but...

Most of the work I do at present is on new code. The code has
to integrate into existing code, but it's either completely new
functionality, or its a complete rewrite, throwing away large
parts of the existing code.

Or is it just me?

It seems more and more that there isn't new code being written
and only old code to be fixed in C++ and the new stuff, to my
detriment, is being done in Java and .NET, save some server
stuff where performance is critical, some scientific, and
games.

C++ probably isn't the ideal language for a quick and showy
front end. But it's still widely used for the code behind the
quick and showy front end, which does the actual work. And
while it's rarely totally new, there are almost always new
features which require new code, and old code which has reached
the end of its lifetime, and gets totally rewritten (without
much regard to what was there before). Remember, code has
a maximum lifetime of about 10 years; if you have to change
something in code older than that, you're usually better off
rewriting it from scratch.

 

[Jorgen Grahn]

There's nothing wrong with maintenance work ...
but playing it safe gets old eventually.

I wouldn't call it "playing it safe", but perhaps we have different
definitions. I mentally included roughly any changes to a system
which already has users.

I could just as well call it "taking responsibility".

/Jorgen