As far as I know, large part of Code_Aster Finite Element solver is written in Fortran. And that’s really impressive software.
1 Like
Right. It’s more used in scientific, stats and math oriented jobs, not software development as such. I know a stock trader and cancer researcher who both use it, but 0 career devs.
1 Like
I weep when I think that, had Codd got his way, we could have had mainstream adoption of relational algebra.
Maybe with fewer Greek letters, though.
3 Likes
It’s Turing-complete. That said, for anything but the most basic commands, it’s a lot easier to write it in Python.
Unless you are doing very advanced, arcane, and obscure statistics, and what you want to do has already been implemented in R, but not Python, Python is better. I know it still doesn’t have a replacement for ggplot, but with packages like Seaborn, it is getting there.
It’s good for figuring out what library has already implemented your functionality, letting you do it in three lines instead of spending days on it.
2 Likes
I tried to learn R once. That quote pretty much suns up my impressions.
3 Likes
So’s PostScript.
7 Likes
Don’t be whacking on AWK too hard. It was good for its time until better string processing tools arrived.
Aho, Weinberger, and Kernighan got so thoroughly sick of SNOBOL4 that they did something about it. AWK has a relatively regular grammar and lacks weirdnesses like using whitespace as a string concatenation operator.
1 Like
I suspect that this is why the most-hated side of the list seems fairly heavy on vocationally relevant languages. Hating whatever the tribal enemy of your favorite hobby language is is kind of deviant unless done lightly and tongue in cheek.
Hating something that is common enough to both be loaded with awful legacy code and hacks; and have a constant demand for people to trudge into the data mines and hammer away at it is much more likely. The don’t call the money you get for that ‘compensation’ for nothing.
7 Likes
It’s OK, I was going to kill off the financial services industry anyway.
8 Likes
I always thought of Perl as APL for the twenty-first century.
As for hating it, I kind of understand why. Any time I’ve had to use it, I’ve felt like I was strapping an unlicensed nuclear accelerator to my back, and I was in dire danger of crossing the beams…
2 Likes
Whatever works for you, man.
8 Likes
A few notes:
Most pre-ANSI Fortran is barely comprehensible cruft whose maintenance is used as a form of torture. F90 and later looks like a relatively modern procedural language. Fortran 2008 is an impressive tool, as in, “Holy shit! F2k8 syntax supports a Pythonic list comprehension for array indices.” Serious stuff.
Clojure isn’t broadly hated enough because so few even care about a recrudescence of Lisp.
Forth may happily be assumed dead and forgotten. Religion and software engineering should not mix.
Java should be more broadly hated for its enforced needless verbosity.
C++ should be more broadly hated for dragging along C compatibility thus obligating needless complexity. It’s the PL-1 of the new millenium. Give it up already.
I want ease of use, readability, ease of maintenance without the language requiring itself to be a Way Of Life (cf. x86 assembly). Python fits the bill well enough but that’s only me because I haven’t written system code for a long time.
1 Like
Not so. Most Quantum Chem/Phys apps are written and maintained in dialects of Fortran. If you’re running O(n^7) calculations, Matlab isn’t even a remote consideration.
2 Likes
Ruby still has fanbois. If you like it, you probably like it a lot.
Yes. It’s really the only truly Object Oriented scripting language out there. Much like Smalltalk, everything is an object in Ruby – even numbers! Other scripting languages like Python mix objects with non-objects.
1 Like
i haven’t successfully written anything large in python. i know people do - and i seriously don’t understand how. small to medium tools are great in python. things beyond a certain size seem to fall apart due to what seems to be the near impossibility to refactor.
English dominance is because that’s were the money was for the longest time.
The next gen may be learning pidgin Chinese.
1 Like
It’s fairly idiomatic, too. SciPy fits the def of “large” well but I can’t think of any others offhand.
This statistic is garbage.
First, the statistic is not even “liked/disliked” - the actual language on the form is “tech you prefer to work with” vs. “tech you prefer not to work with”, which has an extremely different connotation from “liked/disliked”.
Second, there are all sorts of reasons to put something in “don’t want to program in this” other than that you hate the language. Some possibilities are: the kinds of jobs available, your prior experience, your desire to move on from a particular programming language. Objective-C and Ruby, for example, seem like good candidates for the latter - many people come to programming through boot camps where they learn Ruby or Objective-C and might want to boost their skills in other areas.
Finally, fraction disliked / liked is obviously a function of familiarity. People who know languages like Kotlin are clearly self-selected for people who are likely to enjoy Kotlin.
Anyway, this post is a good lesson in how easy it is to interpret data however you like and draw spurious conclusions which other people will then blithely accept as valid because you made a graph, so there is that.
6 Likes
