4 ms·
You're touching the reason yourself. When 90% of the code you find is crap, it simply means that the language has a low barrier of entry. When languages like ML
by hvidgaard 10y ago
You're touching the reason yourself. When 90% of the code you find is crap, it simply means that the language has a low barrier of entry. When languages like ML/Scheme/LISP variants, Haskell, ect. don't have that much crap, it's because the barrier of entry is higher.
And this goes not just for languages. Frameworks like Akka, is the same. The idea of actors to form the system is simple and elegant, but it's far from simple to get started with.
- bigiain 10y agoFor Perl and PHP - as well as the low barrier to entry, there was also the almost ubiquitous availability - at least across web hosting platforms in the late '90 and early 2000s. I was a Perl guy through and through back then, but even I wrote some PHP because getting web hosting companies to allow/enablePerl cgi (or even worse fastcgi or mod-perl) was expensive or impossible, where PHP was "just there" and "just worked". (But yeah, you don't get frustrated pixel-pushers from the coloring in department getting over ambitious and writing up a site backend in Rust or Go... Your typical weekend Haskell hacker is about 100% likely to be able to spell "algorithm" and at least handwave their way through a Big O discussion... No big criticism of talented designer types, but in general they've got as many holes in their understanding of the "S" bit of "CS" as most developers have in their aesthetic abilities...)
- collyw 10y agoPython hasn't got a high barrier to entry (unless you are talking about deploying it to a web server). The code is generally better than PHP or JS code (though I have seen my share of crappy Python code).
- Klathmon 10y agoIMO python has a much higher barrier to entry. Hell just getting PIP installed can be a bear on windows, remembering the damn --upgrade flag, and the whole concept of a virtualenv means that often the casual python hacker doesn't have access to packages in python until they are much further along. Contrast this with Node where npm is always right there, and installs are local by default means the kid that just sits down with js has access to every package out there and can even publish things his/her first few days. I'm obviously more comfortable with the NPM ecosystem, and I might just be biased, but it does seem much easier to work with and use.