Programming Language popularity by activity


FBnil

There is 1 impostor among us.
Joined
Dec 14, 2012
Messages
3,503
Location
Yurp
Programming Language Popularity Chart
http://langpop.corger.nl/

As measured by github changes and stackoverflow tags....


Also:
jobtrents

If you add "C" as a new term, then that outshines all those :)
But C++ is aweful because nobody is looking for a job in it...
 

PowerGod

Hardcore Member
Joined
Jun 20, 2011
Messages
3,693
Cool, but a little messy... could it be better if subdivided by type of language... I mean, how can you compare SQL to C# ?

EDIT:
Just seen LOLCODE in the list xD
 

Kev2442

Still French
Joined
Nov 4, 2014
Messages
635
Location
42 (Loire) is the answer.
Pretty interesting to read, but I think it lacks some clarity. I mean, number of lines changed in GitHub, that could mean anything.
See SQL. Does that mean they don't need many lines to get things done, or that it isn't very popular on the platform ?
 

levi

Still fresh, damnit!
Joined
Oct 6, 2008
Messages
13,167
Location
Somewhere off the coast of the EU
I'd guess SQL is rarely seen in raw files, but more commonly embedded in projects in some other language that submits those strings to an SQL server. It may also be less popular on a platform that mainly contains open source projects. I suppose it also does take less SQL lines to open a table and iterate over its lines to select specific records, and certainly less to cross reference one table against another than it would in a general purpose language.
 

edgex004

Advanced Member
Joined
Jan 5, 2008
Messages
1,204
LabVIEW is on it... What does number of lines changed mean with it?
Well, if LabVIEW is anything like Matlab/Simulink, the functions and signals that are laid out in the editor get saved as an XML file, and a diff can get run on that when a change is committed. However, often times small changes like moving a single function to a new location will result in a cascade of changes showing up in that diff.
 

Linux-SWAT

Hardcore Member
Joined
Feb 13, 2010
Messages
8,574
I'd rely more on sourceforge than github for that kind of statistics.
 
Last edited:

FBnil

There is 1 impostor among us.
Joined
Dec 14, 2012
Messages
3,503
Location
Yurp
What about Ook?
I always found Ook to be missing the command "Ook? Ook?". It could be used as a command shifter, thus, the next Ook command is syntactically the same but functionally different to its unshifted version. For example: print and dump-to-file can be implemented in this shifted space.

Also, if we compress this, then "Ook? Ook!" can be written as "O? O!", ultimately as "?!", and that means that you can compress it 4.5 times (deflate to 0.22%), this then can be used to used wpeg compresson and compress it even further to 5 bytes. Unfortunately, by then it suffers from unreadability for orang-utans.

I also like piet. The program, or resulting image, can be used as a keyfile for keepass. As the colours are very well defined, you can use steganography on the image, without losing your code. It is versatile in that way. Also, Programming in Chef makes me hungry.
 

Amnykon

Still Fresh
Joined
Aug 27, 2012
Messages
72
Location
Austin, TX
Well, if LabVIEW is anything like Matlab/Simulink, the functions and signals that are laid out in the editor get saved as an XML file, and a diff can get run on that when a change is committed. However, often times small changes like moving a single function to a new location will result in a cascade of changes showing up in that diff.
oh if it was that easy. They are not saved as XML or any other markup language. The LabVIEW code is pure binary...
If it is per line, it must be every 0xD ('\n') that is randomly found in the binary file outputs, completely changing the amount of them based mainly on the size of the .vi file.

I think the axis of the chart does not really indicate much... If I write in assembly I am going to have more lines changed per edit and it will be ranked higher. Swift has more type checking then C, so I don't have to fix bugs as much, making the graph is shows usage(what they are trying to measure) * ease of language. Ease of language is a constant that is different per language. [ assembly<C<swift ] Unless this constant is found and divided out, I think this graph is useless, but still interesting.
 

levi

Still fresh, damnit!
Joined
Oct 6, 2008
Messages
13,167
Location
Somewhere off the coast of the EU
Yes, it's quite a limited graph, and there are plenty of situations where direct comparison of the numbers doesn't work, but for languages you know, where you know their relative power, and what 'a line of code' means in that context, it's interesting. I didn't see it make any claims as to its applicability, but as a conversation starter it's fine. I guess that means I shouldn't be trying to shut down that line of reasoning that way I just did ;)

Cool idea, but the term "compression" seems so strange in this case...
Well, the way I read it is to call that 'compression' is to be sarcastic. It's a joke algorithm :)
 

ible

professional vim user
Joined
Mar 24, 2014
Messages
2,421
Location
Seattle, WA
there are 10 kinds of people in the world. those who understand bad programming jokes, and those who don't...
 
Top