Is C++ Gaining in Popularity? (i-programmer.info) 106
An anonymous reader shares this report from Dice.com:
C++ is enjoying a surge in popularity, according to the latest update to the TIOBE Index, which tracks programming languages' "buzz."
C++ currently sits right behind C and Python on TIOBE's list. "A few months ago, the programming C++ language claimed position 3 of the TIOBE index (at the expense of Java). But C++ has not finished its rise. C seems to be its next victim," added the note accompanying the data... ["At the moment, the gap between the two is only 0.76%."]
Matlab, Scratch and Rust also match their all time high records at respectively positions #10, #12 and #17.
So here, according to TIOBE, are the 10 most popular programmings languages:
1. Python
2. C
3. C++
4. Java
5. C#
6. JavaScript
7. Visual Basic
8. SQL
9. PHP
10. MATLAB
The site I Programmer digs deeper: C++ was the only one of the top four languages to see a positive year-on-year change in its percentage rating — adding 0.79% to stand at 10.8%. Python had the smallest loss of the entire Top 20, -0.01% leaving it with a share of 13,42% while Visual Basic had the greatest loss at -2.07%. This, combined with JavaScript gaining 1.34%, led to JavaScript overtaking it to occupy #6, its highest ever ranking in the TIOBE Index.
They also note that COBOL "had a 3-month rise going from a share of 0.41% in April to 0.86% in July which moved it into #20 on the index."
C++ currently sits right behind C and Python on TIOBE's list. "A few months ago, the programming C++ language claimed position 3 of the TIOBE index (at the expense of Java). But C++ has not finished its rise. C seems to be its next victim," added the note accompanying the data... ["At the moment, the gap between the two is only 0.76%."]
Matlab, Scratch and Rust also match their all time high records at respectively positions #10, #12 and #17.
So here, according to TIOBE, are the 10 most popular programmings languages:
1. Python
2. C
3. C++
4. Java
5. C#
6. JavaScript
7. Visual Basic
8. SQL
9. PHP
10. MATLAB
The site I Programmer digs deeper: C++ was the only one of the top four languages to see a positive year-on-year change in its percentage rating — adding 0.79% to stand at 10.8%. Python had the smallest loss of the entire Top 20, -0.01% leaving it with a share of 13,42% while Visual Basic had the greatest loss at -2.07%. This, combined with JavaScript gaining 1.34%, led to JavaScript overtaking it to occupy #6, its highest ever ranking in the TIOBE Index.
They also note that COBOL "had a 3-month rise going from a share of 0.41% in April to 0.86% in July which moved it into #20 on the index."
C# (Score:1, Flamebait)
C# is ranked above Javascript? Seriously? Is it all a bunch of Unity people? I can't believe it. No wonder many games suck.
Re: (Score:3)
You might be surprised how many businesses have in-house software written in C#.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Look at the comment I replied to:
The GP backslashdot seems to be under the impression that Unity games are the only widespread application of C# or something. It shouldn't be surprising to anyone with much industry experience, whether they're a C# developer themselves or not.
Re:C# (Score:5, Interesting)
It doesn't really mean anything. Anybody who has used both C# and Java will tell you how much a pile of shit Java is compared to C# in every way imaginable, yet Java outranks C# in this. PHP isn't even a real programming language and somehow it is at #9, and SQL has so many differences in its variants that it's hard to justify calling them all the same language.
And all of this is enough to tell you that for C++ to end up where it is doesn't really mean anything, particularly given how ugly and incredibly inconsistent the language is without having any redeeming qualities whatsoever other than that it gets to claim that it's almost as fast as C while offering you a lot more convenient ways to shoot yourself in the foot as if that were a virtue.
Java pays more than C# (Score:5, Interesting)
It doesn't really mean anything. Anybody who has used both C# and Java will tell you how much a pile of shit Java is compared to C# in every way imaginable, yet Java outranks C# in this.
I've used both and found Java to be quite nice an C# to be limiting and overhyped. Not to mention the most important part...there are like 20x the number of Java jobs over C# in my area and the pay is higher. So C# can hold whatever masturbatory title in your head, but the people paying my mortgage prefer Java and I would rather miss out on being a smug C# supremacist than homeless.
Re: (Score:2)
How exactly is c# limiting?
Re: (Score:2)
Compared to Java or compared to C++? Compared to C++, C# (or Java) generics are far more limited than C++ templates which have compile-time duck typing. That alone makes it feel like your hands are tied.
Compared to Java, C# has light-weight "structure" types with less overhead than full objects, non-virtual member functions (faster when virtual dispatch isn't needed), and generics are faster and more flexible because it has runtime support and doesn't rely on type erasure (see here [dev.to] and here [jprl.com] - scroll down)
Re: (Score:1)
Java also has "non virtual member functions", lol.
And Java also has "light weight structs", they are called "records".
While C# generics are "better" and Java has "type erasure", they are only faster on primitve types, and also only in rare cases.
In real world there is speed wise no real difference in Java or C#.
Having unsigned types, makes C# more handy if you have to do bit fiddeling.
C# / .Net is pretty catching up in the library department though.
Re: (Score:2)
Java also has "non virtual member functions", lol.
As does C#
And Java also has "light weight structs", they are called "records".
And C# has...structs...
While C# generics are "better" and Java has "type erasure", they are only faster on primitve types, and also only in rare cases.
You'd need to provide numbers on that one. I always get much better performance out of C# in general. And, you NEVER have to worry about stupid type erasure. Or Java's absolutely horrid dependency management (even with good tooling like gradle, it's still really bad. Why? Dependency shading is why.)
In real world there is speed wise no real difference in Java or C#.
That's definitely not true:
https://benchmarksgame-team.pa... [debian.net]
TL;DR: C# beats Java on damn near everything.
C# / .Net is pretty catching up in the library department though.
I'd argue it's well ahead, and has been.
Re: (Score:1)
You'd need to provide numbers on that one. I always get much better performance out of C# in general.
Unlikely. As you actually never tested "the other language": impossible to know.
The benchmarks on your link are meaningless. The longest run I saw when I glanced over it was 10 seconds. So: the Java code was not even jit compiled while the C# code most likely was AOT compiled. Seriously? That is like comparing a race car with running engine versus a soap box car the owners need to assemble first while you a
Re: (Score:2)
Unlikely. As you actually never tested "the other language": impossible to know.
Java? I certainly have. It's a pile of crap.
So: the Java code was not even jit compiled while the C# code most likely was AOT compiled. Seriously? That is like comparing a race car with running engine versus a soap box car the owners need to assemble first while you are already on the track.
You should try to understand how that benchmark works before making assumptions. Bad assumptions are why you produce bad code, and why your students are always complaining that you ripped them off.
Re: (Score:2)
No. There is a separate comparison between C# .NET and C# aot
Still don't get the point.
So you find hand selected "benchmarks" where Java looks slow? Your problem, not mine.
Re: (Score:2)
I've done both (but much more C#, especially recently).
While they're both fine languages, I would almost always choose C# over Java, because:
Not a slam on Java or its many happy
Re: (Score:1)
Reflection is not affected by type erasures. Unless you force the compiler to remove it from the byte code.
A Thing<RealClass> still shows it is a "RealClass" when you reflect over it.
I never saw a factory of factories. And it is most certainly not a language thing, but: bad programming.
Re: (Score:2)
Factories suck period, and that IS a language thing.
Re: (Score:2)
A "Factory" ia. design pattern.
You can use it in what ever language you want.
Or don't use it.
Dumbass very much?
Hint, you can always do:
Type var = new Type(random, amount, of, arguments);
No one requires you to use a "factory".
I suggest to read up what. a factory actually is.
And: factories of factories, they simply do not exist. Would not make any sense at all: dumbass.
Re: (Score:2)
which have compile-time duck typing.
Personally I'm really not a fan of duck typing; I prefer very strict typing in all programming. A lot of reasons really, among them being more readable code, less bugs, etc. But I'm also the kind of guy who uses type annotation at every initialization both for myself and for whoever has to see my code later. Even in python I do that, and even when I call a function that can output more than one type. The only language I don't do that in is rust, which not only has very strict typing anyways, but rust-analyz
For starters: Microsoft's ecosystem (Score:2)
As superior as the language may be, I spend VERY little time writing code in the
Re:C# (Score:4, Informative)
> without having any redeeming qualities whatsoever other than that it gets to claim that it's almost as fast as C
How about three to eleven *times* faster than C ?
http://theory.stanford.edu/~am... [stanford.edu]
Re: (Score:2)
without having any redeeming qualities whatsoever other than that it gets to claim that it's almost as fast as C
How about three to eleven *times* faster than C ?
What's the point of this kind of cherry picking?
Re: (Score:2)
It isn't cherry-picking. It's pointing out that C++ template algorithms are faster and more flexible than generic algorithm functions in C because the compiler instantiates them for specific use cases. For sorting, that means, among other things, element size becomes a compile-time constant and there's no need for indirect calls for doing the comparisons. This gives a lot more opportunities for optimisation as well as reducing overhead.
To get the same performance in C, you need to copy/paste the code for
Re: (Score:2)
Re: (Score:2)
You wouldn't say that if you actually understood C++ templates. Trying to achieve the same thing with templates creates an unmaintainable mess.
Re: (Score:2)
It isn't cherry-picking.
Then what is it because the claim C++ is three to eleven times faster than C is absurd on its face.
It's pointing out that C++ template algorithms are faster and more flexible than generic algorithm functions in C because the compiler instantiates them for specific use cases. For sorting, that means, among other things, element size becomes a compile-time constant and there's no need for indirect calls for doing the comparisons. This gives a lot more opportunities for optimisation as well as reducing overhead.
This is equally absurd. The link you sent compares an implementation of quick sort with STL's sort which uses entirely different algorithms. You are comparing apples and oranges.
Re: (Score:2)
How about three to eleven *times* faster than C ?
Under certain conditions, C is faster than ASM. *shrug*
Under certain conditions, ASM is faster than machine code.
Do you see how absurd this gets? The reality is that C is generally faster and more efficient than ASM and machine code is faster and more efficient than ASM.
Increasing complexity, can, under certain circumstances, make some things faster; however, nothing can beat a master poking data directly into a computer's memory. Of course, it is 'faster' for me to write a web server in Python than it is t
Re: (Score:2)
>And all of this is enough to tell you that for C++ to end up where it is doesn't really mean anything, particularly given how ugly and incredibly inconsistent the language is without having any redeeming qualities whatsoever
Eh, C++ is really nice to program in. While it doesn't mandate certain style guidelines, the language itself is quite consistent in how the standard libraries work, and serve as a good basis for your own projects.
Re: (Score:3)
You've said a lot of the same silly things people often mindlessly repeat. Just so you know, there's a reason why we call posts like that 'karma whoring'.
Anybody who has used both C# and Java will tell you [...]
I'm going to guess that you tried out Java after you learned C# and struggled because you thought they were supposed to be similar and so didn't take the time to learn the language. Java and C# were similar ... 20+ years ago. I've been annoyed with C# since they changed how foreach loops worked (but not other kinds of loops for some reason) because they
Re: (Score:3)
People bitch about "bloat" in C++. I call them "features". And what's nice is that C++ doesn't really force you to use any of those features in particular. You can go right on creating C-style procedural code if you like, all the way to utilizing all the newest whiz-bang features. C++ is a multi-paradigm languages, so you can pick and choose among both features and styles.
Personally, I greatly appreciate the improvements that have been made in the past twelve years. They've made the language a lot safe
C++ (Score:2, Informative)
Dutch Gun C++'s. :)
You are is exactly right. I use c++ in a very specific way that gives me the tools I really want, without burying me in features I really don't care about or need.
I appreciate that it gives others the same opportunity to pick and choose. They can use it their way, I'll use it mine.
It's a great language. If you're a good programmer. If not, you really should stick to languages with life jackets, safety harnesses, lifeguards and proctors. And thanks for doing so.
Re: (Score:2)
You've said a lot of the same silly things people often mindlessly repeat.
There's a kind of logical fallacy in there: smart technical people have strong opinions on technical topics, therefore if I have strong opinions on technical topics, I must be smart. It pisses me off, and I've got really tired of marginally competent programmers presenting strong opinions as fact especially when those opinions involve implicitly shitting on someone.
There is a lot of value in a stable, consistent, language.
Indeed. My a
Re: (Score:1)
I used C# back when I had to code for a PDA, yeah that long ago. I rather liked it, compared to JAVA it's on equal terms, but since I only done JAVA for android, and since they keep changing the api calls for android with every iteration I am not liking JAVA as much. JAVA has some features I suppose I'd miss if I tried C# today, but since I haven't coded in C# for a very long time, maybe those functions are available there as well by now.
Re: (Score:2)
I've used both. There's really no big difference at all between the languages, which isn't a surprise since C# was originally a clean-room implementation of Java (Microsoft's "project cool" at the time). I've ported code from C# to Java, and from Java to C#. Each has some advantages. For people to get worked up about one being way better than the other at this point is just creepy.
Re: (Score:1)
"Anybody" who has used both will tell you how much a pile of shit Java is compared to C#?
Well, you sure don't speak for me, so make that "anybody except for one".
The syntax of the languages is very similar.
In 2008-2010 C# had some syntax bits I wished Java could have had... but now has (var (from Java 11) and try-with-resource (which it has had since at least Java 8, I think...?)).
When porting a plugin-based Java application (using OSGi) from Java to C++, in 2008-2010, we ran into assembly loader is
Re: (Score:1)
Calm down. This only programming languages popularity list. No need to get angry. Angry people live shorter.
Re: (Score:1)
Anybody who has used both C# and Java will tell you how much a pile of shit Java is compared to C# in every way imaginable,
That is just a lol statement.
Only 50% of "anybody" will tell you that.
The other 50% will tell you the opposite.
I personally know no one who thinks C# is in anything better than Java. Bottom line the libraries and the ecosystem is what counts.
The shops I know which use C# develop on windows and deploy on Linux and have nothing to do with Human Machine Interfaces ... and never tried Java,
Re: (Score:2)
PHP isn't even a real programming language
If it is Turing complete, then it is a real programming language. Are you claiming that PHP is not Turing complete? Honestly, your opinion sounds like gatekeeping.
Re: (Score:2)
You mean like this?
https://learn.microsoft.com/en... [microsoft.com]
I'm not sure what the point of that comment is anyways because java has a similar limitation. Besides, the .net runtime is a LOT less of a headache to deal with than JRE.
Re: (Score:1)
Besides, the .net runtime is a LOT less of a headache to deal with than JRE. .net runtime?
A sentence like that implies:
a) you have headaches (sometimes?) dealing with the
b) you have headaches (sometimes?) dealing with the JRE?
Not really sure what to say about that.
I rarely use .Net. Never had any headaches/problems what so ever with its runtime, since roughly 20 years. What exactly does one need to do wrong to have any headaches?
Same for Java, I use it since roughly 1995. How the funk would anyone have "pro
Re: (Score:2)
Not all JRE's are created equal. Many Java apps out there require old versions of Java and cannot run on newer JRE's.
Not sure about .NET in the cross-platform environment. I think .NET has only been cross-platform since 4.x came out, so it might all be backwards compatible for all I know. .NET 2.0, 3.0 and 3.5 were definitely not compatible with each other or 4.x
Re: (Score:2)
Not all JRE's are created equal. Many Java apps out there require old versions of Java and cannot run on newer JRE's.
In general, that is not true. Modern JVMs usually support old code.
And: no one prevents you from having an old perfectly well working VM on your machine.
Re: (Score:2)
You do know .net has been multiplatform for about seven years now right? It's also even more open than Java is.
Re: (Score:2, Informative)
Statically-compiled C# (using Microsoft's Roslyn compiler) is basically "C++ with nice syntax".
The catch is... a lot of C#'s creature comforts and niceness comes not from the language, but dotnet. Strip away dotnet so the bare, naked C# language is all you have left, and just about the only major advantage you have left is the absence of "vexing parses" (like, where a sequence of source could be interpreted in multiple ways unless you explicitly rewrite it to help the compiler find the semantic boundaries).
Re: C# (Score:2)
Oh, please. How many "vexing parse" situations are there in C# and Java?
Or, another gem... In c++, you can't simultaneously declare and initialize a static variable, like:
class Foo { // boom!
public:
static int fooValue = 0;
}
Or even
class Foo { // boom!
public:
static int fooValue;
}
Foo::fooValue=0;
No, in c++, if you declare a static int named fooCounter in Foo.h, you have to remember to explicitly initialize and assign a value to it somewhere that is NOT
Re: (Score:3)
I'll take that over lack of typedef any day of the week.
PS:
class Foo { // boom!
public:
static int& fooValue() {
static int fv = 0;
return fv;
}
}
Re: (Score:2)
Undefined behaviors are even worse. C++ is the only language I'm aware of where something can compile successfully, pass unit tests & produce the expected values, and STILL officially be incorrect, invalid and wrong.
Then you are not aware of many languages.
And since nobody... Not even Bjarne Stroustrup, flawlessly knows EVERY single nuance -- chapter and verse -- nobody can ever feel truly 100% confident that their nontrivial program is actually *correct*.
If you're using every feature of the language in
Re: (Score:3, Informative)
Yep.
Basically, these are being driven by Unreal (C++) or Unity (C#)
Python is a BAD language overall, impossible to package, and making a "self-contained" binary of some ML app ends up with a 20GB some odd EXE file that takes forever to launch, if it even becomes possible to launch, usually it chokes well before it gets to the actual model.
The only languages that are "likely worse" than Python in that regard are PHP and Perl, which share the same deficiencies and the only advantage PHP has is that it's a "we
Re:C# (Score:5, Insightful)
Re: (Score:3)
>Python is a BAD language overall, impossible to package
There are many different uses for Python. In engineering test and debug, your concerns are not relevant. So for many people, including me, it is absolutely fabulous. If I were rolling out an enterprise solution with it, I can't comment.
My test setup for the code of a science project on a soon to be launched satellite is written in Python.
Re: (Score:2)
There are no native threads in C
At some point, the abstractions stop and the heavy lifting must be done. And if you don't like C, there's always assembly...
Re: (Score:2)
Microsoft's .NET framework is centered around C# (though it supports other languages, that's its core) so you'll have a lot of MS .NET devs who use C#. There is a lot of C# development going on, regardless of if you hate MS etc.
And as much as I love working with Epic's Unreal Engine (C++ based), Unity is much more approachable with C# and the learning curve is not as steep due to (imo) they way they designed their dev tools. Unreal is amazing, but it takes a lot of time and effort to get comfortable with it
Possibly (Score:4, Insightful)
It's quite possible C++ is gaining popularity, but the TIOBE index isn't going to give you a meaningful answer to the question. It's lazy garbage that tells you more about search engine algorithms than anything else.
Re:Possibly (Score:4)
I don't believe that C++ is gaining any popularity. I don't see it losing it ether. A lot of languages come and go, some find a nitch and remain there for the entire lifespan of the language.
C and C++ have just always been there. They don't have a nitch, and never seem to go away. They are not the best language for everything, but they are not the worse language for anything. They are just there.
Re: Possibly (Score:3)
Do you mean niche?
"C++ is dead." (Score:2)
Re: (Score:3)
Re: (Score:2)
C++ is gaining in popularity with me, but I really hate the way they handle unicode, esp. utf8. I can understand why the legacy libraries can't handle it, but the newer ones don't seem to do any better. Even Java is better for that task, and it uses 16bit chars for gods sake.
Re: (Score:3)
C++ is gaining popularity in the same way that herpes is gaining popularity -- via people who never wanted it, but ended up having to deal with it anyway, and now it's become a part of them that they are stuck with for the rest of their lives. ;)
Re: (Score:2)
It's lazy garbage that tells you more about search engine algorithms than anything else.
Thank you. Was hoping someone else felt the same about the fucking point of all the code-measuring.
(Employer) "I'm sorry John. Business isn't good, we're gonna have to let you go."
(Python John) "But...but...my language is #1..."
Yeah, let me know when that shit works.
Eh, these indexes (Score:2, Interesting)
These indexes are like distrowatch. It's just a list of whatever whim people are thinking about at the moment, not what is actually being used by people that know what they are doing. Programmers, in particular noobs and unskilled (which are like 99%), are very prone to bandwagon hopping, hyping up fads and other garbage. They lack wisdom.
The problem is most people don't know this. They're idiots. Too many take these lists seriously so it snowballs.
Re: (Score:2)
Re:Eh, these indexes (Score:5, Informative)
TIOBE isn't a poll. It's computed by counting results of specially constructed search engine queries. It's even more useless than a poll.
Re: (Score:2)
> TIOBE isn't a poll. It's computed by counting results of specially
> constructed search engine queries. It's even more useless than a poll.
You'd think that by now the monthly Tiobe numbers would be auto-posted to Slashdot,
and auto-approved. But maybe they are already.
However Ypu Feel About This (Score:4, Insightful)
Don't worry, in a few weeks there will be another chart showing something different.
Honestly, we get way too many articles of this nature.
Re: (Score:2)
Honestly, we have way too many people addicted to clickbait.
FTFY, in case you were still somehow wondering why this seems so repetitive. Hard to walk away from revenue, no matter how stupid the reason it is for being generated.
Why is this posed as a question? (Score:2)
why is slashdot pushing TIOBE still? (Score:2)
Is slashdot owned by the same people, or does it have a stake in TIOBE's continued existence?
You can get a better analysis of programming languages from any astrologist than TIOBE. For example, I just read the other day that RUST is cancer.
Re: why is slashdot pushing TIOBE still? (Score:2)
Re: why is slashdot pushing TIOBE still? (Score:5, Funny)
A rock. A wet noodle. A magic 8 ball. Pretty much anything is a better indicator about programming languages than TIOBE.
Re: (Score:3)
Nah. Tiobe is an indicator of what people in certain areas are talking about. And it's a reasonable guess that if they're asking questions about C++, then they're interested in it. Popular might be stretching it a bit, though.
Re: (Score:2)
So what is a better indicator than TIOBE?
Want Ads?
Re: why is slashdot pushing TIOBE still? (Score:2)
We use python. One of my first slides is the TIOBE index as a rationale for why we teach Python. I also referenced another index, but can't remember the name. I knew these indeces are under debate, but am surprised to read the many "TIOBE is garbage" remarks. So my question is sincere. What is a better measure?
Re: (Score:1)
Why do you need a measure?
Just look at language features and look where the languages are used. Put that in relation to what you need and/or what to teach.
Popularity of a language, regardless if perceived or real: is a damns tupid thing to base your teaching of a programming language on.
Or google something like "programming language by popularity", or check which languages are most used on github/gitlab.
Re: (Score:2)
Look, if you have nothing usefull to contribute, it may be better to keep your mouth. Don't worry, inhibition is a higher order form of thinking (executive functions of the brain). It is one of the last things to develop before we fully mature, you will get there, but you need to realise you are not there yet. That's why I am rubbing it in.
Re: (Score:2)
Re: (Score:2)
Sounds like a plan. :P
However my Dutch, Flaemic and French is rather lacking
Re: why is slashdot pushing TIOBE still? (Score:2)
Samsara (Score:3)
Queue all of the predictable comments about how this story isn't accurate, add in some bitching about some common languages, and heated defenses on how that isn't fair.
Re: (Score:1)
Yup, and your comment is just as predictable.
But we're all having fun, aren't we? Yes we are.
Interesting to see MATLAB on the list (Score:2)
I think two kinds of programming languages would get a lot of search engine hits:
1. The one's that are really in vogue and being used extensively
2. The ones that are absolutely not in vogue and folks having to maintain old code are searching hard to find some help
Python, C++, C#, JAVA, Javascript, MATLAB etc. I think fall into category 1. FORTRAN, Assembly, Pascal, COBOL etc. fall into category 2.
But strange that Julia is not on the list.
Re: (Score:2)
I think C++ for me is category 2. (I said for me; YMMV.)
I would never choose, in my problem space, to use C++. I do need lower- and higher-level languages, but not typically for the same layer of code.
But my employer has bought several large code bases that do use C++ and that it falls partly on me to maintain.
I promise that I'm not looking up anything on Google because I chose to use C++. I'm doing it because that choice was made for me, and, since it mostly drives either SDL2 or Qt and there are good b
popular != good (Score:5, Insightful)
Re:popular != good (Score:4, Funny)
However, I disagree vehemently about the Kardashians: I really love to hate Gul Dukat, Gul Lamar had a most interesting redemption arc (just too late for himself, just in time for the rest of the galaxy), and Garak was just plainly (and creepily) awesome.
Re:popular != good (Score:5, Funny)
You have made the common mistake of confusing Kardashians with Cardassians.
One group is a bunch of amoral, leather-clad, vaguely reptilian beings who will stop at nothing to further their aims. The other group, of course, occupied Bajor.
Re: (Score:1)
I wish I hadn't commented already, because I would have modded the above "funny".
"It's funny because it's true."
-Family guy
"Simpsons (already) did it!" (Score:2)
-Homer Simpson
Re: (Score:2)
Big News Would Be Cobol Doubling in Three Months (Score:2)
They also note that COBOL "had a 3-month rise going from a share of 0.41% in April to 0.86% in July which moved it into #20 on the index."
But TIOBE is such a poor index that it is not believable.
How is this a popularity contest? (Score:2)
Quite seriously, I don't choose my language by how "cool" it is perceived, I choose it by the requirements of my project.
Frankly, if project managers do that differently, I can see why software is as crappy as it is.
C++ is unsafe programming language (Score:1)
Re: (Score:2)
We *know* companies skimp on QA and documentation, outsource coding jobs or just hire cheaper and inexperienced devs. This resulted in DECADES of shitty code that spawns all kinds of exploits.
GIGO. You can do that in any language, not just C / C++. No language is truly safe from GIGO, as the second you think you've made the perfect language impervious to garbage, the world creates better garbage.
SQL? (Score:2)
That's sorta like listing json or ini there... Who "programs" in SQL? Could we not restrict this list to Turing-complete languages?
Stored procedures (Score:3)
SQL with stored procedures probably is Turing-complete. Even MySQL/MariaDB got stored procedures eventually, albeit with really ugly syntax.
Re: (Score:2)
Relational algebra isn't Turing complete, though all SQL needed was the right kind of loop, which it eventually got with recursive CTEs.
A rule-of-thumb: you need recursion and arbitrary memory. Lookup tables have neither, finite state machines lack memory, relational algebra lacks recursion. Turing machines have both, as does Conway's Game of Life and some single instruction computers. It's amazing how little it takes.
Just for fun, relational algebra is Godel complete.
Re: (Score:2)
Oh, God (Score:1)
Re: (Score:1)
Not too stupid, just to set in my ways to switch from plain old c.
Less useful than people think (Score:2)
People that favor one language over another in their companies based on this index are morons.
but yeah, COBOL?!? (Score:2)