3 ms·
Most people are pragmatic. Simple, axiomatic principles can be very intellectually appealing (probably more so amongst the STEM-crowd than the general populatio
by martythemaniak 6y ago
Most people are pragmatic. Simple, axiomatic principles can be very intellectually appealing (probably more so amongst the STEM-crowd than the general population), if they don't produce good outcomes, they will and should be discarded.
In our field we constantly see languages with stronger intellectual foundations (Haskell, Lisp, Scala, Perl6) constantly lose out market share to pragmatic, practical, messy and oftentimes ugly languages/libraries/tookits etc. We often knowingly choose the latter set knowing full well what their downsides are, because we still come out ahead.
If you're wondering why someone would support Facebook/Twitter/YouTube's recent moves, it really is that simple. Their foundations don't have to be perfect, or even particularly good or fair. They just have to be better the deluge of toxic sludge that's currently overwhelming these systems. That is to say, there is no answer to the Paradox of Tolerance and you don't need an answer. If you feel people are taking advantage of you, you will (and should) crack down, and if you crack down too much eventually people will (and should) rebel.
- tomp 6y agoHaskell loses because it takes 2000 lines just to write a logging statement. Scala loses because you can’t even write a fast for loop. Lisp loses because it disrespects 100 years of math. I’m exaggerating a bit but not by much.
- lanstin 6y agoI don’t get the lisp thing? I thought lisp lost because of lisp weenies and worse (simple) is better as far as writing maintainable code bases?
- tomp 6y agoInstead of writing 1 + 2 as in mathematics, in Lisp you write (+ 1 2). I'm not sure I get the second statement of yours, but personally I find Clojure in particular very inspiring and amazing, if only it had a better syntax...