5 ms·
I would say there exist more Matlab code than Go code, however I think Go is much more popular among all developers.
by tutfbhuf 7y ago
I would say there exist more Matlab code than Go code, however I think Go is much more popular among all developers.
- seanmcdirmid 7y agoA lot of people using Matlab aren’t developers.
- kragen 7y agoDo you mean they run Matlab programs but do not edit them? I don't think those people are being counted, are they? If they develop programs in Matlab it seems to me that tautologically they are developers, even if, like Con Kolivas, that isn't their job title.
- seanmcdirmid 7y agoA lot of people who write programs aren’t doing software development, they aren’t developers. Unless your definition of developer is one who programs (which is already covered nicely by the term programmer, but to each their own).
- kragen 7y agoI'm not familiar with the distinction you're making; would you like to elaborate?
- seanmcdirmid 7y agoOne whose career is focuses on building software vs. one who programs sometimes as part of their job that is otherwise not focused on building software. Most of us on HN are in the former category, but there are tons of people in the latter category as well.
- kragen 7y agoI see. Well, it's true that lots of Matlab programmers have something other than programming as their job, maybe most of them. I don't think this makes the activity fundamentally different; a baby is still a baby, whether delivered by a doctor or by a midwife, and birthing is still birthing. But you are certainly free to call it by a different name if you like.
- seanmcdirmid 7y agoIt has a huge influence on the language being used. There are simply many kinds of “programming”, and thinking about it as a homogeneous activity is just going to yield ridiculous PL popularity lists that compare apples and oranges.
- kragen 7y agoIt certainly does have a huge influence on the language being used, as do things like living in Nigeria, writing code to run on on-orbit satellites, and working at Microsoft. But I wouldn't venture to say that Nigerians, Microsoft employees, and space cadets are therefore not doing "software development", just that they're doing it under atypical circumstances. (And, as that list suggests, in the end all of us have some atypical circumstances.) I do agree that people frequently overestimate the homogeneity of programming, and that this homogeneity makes such lists less useful. Still, I think it's a useful observation that in 1985 Paul Lansky was writing music in FORTRAN and electrical engineers were writing circuit simulators in FORTRAN, while today musicians tend to do their software development in PureData or something similar, while electrical engineers tend to use MATLAB, Octave, C++, or Python. And I think part of the nature of software development is that it's typically a highly collaborative activity --- a library written by an electrical engineer can be used by a musician, but usually only if they're using the same programming language.