Re:2019's Fastest Growing Programming Language Was C, Says T...
by ncmncm on Saturday January 11, 2020 @03:24PM
(#22022052)
Tiobe again.
Things just do not flip-flop that much in the programming world. What does happen is the indicators Tiobe uses are unreliable, seemingly selected as giving the most churn.
Generally almost any measure is more reliable than Tiobe.
We can say reliably that attendance at costly C++ conferences is soaring. New such conferences are appearing frequently. Each ISO C++ Standard meeting has had higher attendance than any previous one, for the past few years.
Re:2019's Fastest Growing Programming Language Was C, Says T...
by coldtea on Saturday January 11, 2020 @09:06PM
(#22023950)
If you take account quality and maturity of compilers, number of alternative implementations, support, quality of tooling, availability of architectures, number of practicing programmers, libs, existing interfaces and OSes supporting it, and so on, there are tons of places...
Re:2019's Fastest Growing Programming Language Was C, Says T...
by ncmncm on Sunday January 12, 2020 @02:16AM
(#22024985)
There are many languages that can be transcribed to C. So, if you have C, you have every other such language, only a makefile entry away.
Furthermore most environments that seem to have only C are a Gcc target, and then you have access to all the other languages Gcc can compile, without a detour to C.
Re:2019's Fastest Growing Programming Language Was C, Says T...
by clktmr on Sunday January 12, 2020 @08:54AM
(#22026263)
Even if it's possible, the change is not worth it.
Transcribing leads to a horrible debug experience. Immature languages maintained by a single/few are a no-go. Your team already knows C and is good at it, finding new programmers is easy.
You don't want another language, even if it's supported by GCC. If you want anything from C, then it's more safety guarantees. And C is flexible enough to run a better type system on top of it, e.g. look at what the kernel community is doing with `sparse`.
Re:2019's Fastest Growing Programming Language Was C, Says T...
by nickysielicki on Saturday January 11, 2020 @03:36PM
(#22022121)
I mean I buy the argument, a lot of people are trying to make things in that space and a lot of people don't know what they're doing, so they find themselves using Keil or the Arduino IDE and so they're writing C99 because it gets them from zero to one. But who cares?
Trends among all programmers don't matter. It's much more important to ask yourself what trends can be observed among prominent top-tier development teams, and on a timeframe much more than a single year, because those people do know what they're doing and take a year or two to evaluate languages. I suspect the list would look a lot different. I can't imagine many teams are starting new projects today and saying, "You know what went wrong on our last project? We used C++ instead of C."
My best guess at the fastest growing programming language? Some variant of RTL. Hardware acceleration is huge right now. It's either that or Rust.
Re:2019's Fastest Growing Programming Language Was C, Says T...
by ncmncm on Saturday January 11, 2020 @07:39PM
(#22023534)
Meaning Verilog and VHDL?
Rust gets a lot of attention here on HN, and the percentage growth in number of (non-production) coders, relative to any previous period, is high, but the absolute numbers are still very, very tiny, whoever you count.
If we count new production coders, any growing, established language is adding enormously more. C++, Javascript, maybe C#. The most surprising of these is C++, because it should already have penetrated everywhere it was viable, but paid conference attendance doesn't lie.
And, yes, Verilog etc. are probably adding quite a lot more than a year ago, again relatively.
Re:2019's Fastest Growing Programming Language Was C, Says T...
by elfexec on Saturday January 11, 2020 @08:27PM
(#22023777)
I find it hard to believe that the language with the fastest growth rate is a major established language like C with a large user base.
Fastest growth rates usually come from new languages with hype. A language which grows from 1 user to 10,000 has a growth rate that is impossible for C to match.
Things just do not flip-flop that much in the programming world. What does happen is the indicators Tiobe uses are unreliable, seemingly selected as giving the most churn.
Generally almost any measure is more reliable than Tiobe.
We can say reliably that attendance at costly C++ conferences is soaring. New such conferences are appearing frequently. Each ISO C++ Standard meeting has had higher attendance than any previous one, for the past few years.
Don't they? I've seen hypes come and go every 3-6 years for 3 decades now (and I'm sure before that). Let's take web programming:
1996+ - Perl (is where it's at)
2000+ - PHP for the masses and JSP/Servlets for the enterprise
2006+ - Rails (remember when every second job posting was for RoR?)
2013+ - Node
I think that is specific to web programming. I wonder what is the trend after Node? Since Node's hype seems to have finally died out.
Apart from Perl in the table above, PHP Laravel, Ruby Rails, JS Node seems to each hold on its own.
There are still many times more Delphi programmers than are ever counted. Just not much turnover, and no hype.
And personally I now really appreciate the persistence of C in contrast to all the changes and new features in other languages.
Furthermore most environments that seem to have only C are a Gcc target, and then you have access to all the other languages Gcc can compile, without a detour to C.
Transcribing leads to a horrible debug experience. Immature languages maintained by a single/few are a no-go. Your team already knows C and is good at it, finding new programmers is easy.
You don't want another language, even if it's supported by GCC. If you want anything from C, then it's more safety guarantees. And C is flexible enough to run a better type system on top of it, e.g. look at what the kernel community is doing with `sparse`.
But to insist there is no alternative is just untrue, and (after you have been corrected) dishonest.
Trends among all programmers don't matter. It's much more important to ask yourself what trends can be observed among prominent top-tier development teams, and on a timeframe much more than a single year, because those people do know what they're doing and take a year or two to evaluate languages. I suspect the list would look a lot different. I can't imagine many teams are starting new projects today and saying, "You know what went wrong on our last project? We used C++ instead of C."
My best guess at the fastest growing programming language? Some variant of RTL. Hardware acceleration is huge right now. It's either that or Rust.
Rust gets a lot of attention here on HN, and the percentage growth in number of (non-production) coders, relative to any previous period, is high, but the absolute numbers are still very, very tiny, whoever you count.
If we count new production coders, any growing, established language is adding enormously more. C++, Javascript, maybe C#. The most surprising of these is C++, because it should already have penetrated everywhere it was viable, but paid conference attendance doesn't lie.
And, yes, Verilog etc. are probably adding quite a lot more than a year ago, again relatively.
Fastest growth rates usually come from new languages with hype. A language which grows from 1 user to 10,000 has a growth rate that is impossible for C to match.