3 ms·
Don't get me wrong, I respect these try-X efforts. But they have glaring annoyances (usually popping up in the name of Security) that are just begging me to ope
by uros643 16y ago
Don't get me wrong, I respect these try-X efforts. But they have glaring annoyances (usually popping up in the name of Security) that are just begging me to open a good old-fashioned command-line REPL (but that's if I'm at my own computer).
For instance, in Try Erlang, io:format is out of reach, not to mention spawn. Same goes for Try Clojure: there doesn't appear to be a way to create threads.
Don't you think, given the languages we're talking about here, that these projects are self-handicapping from the start? (the seriousness of those security measures notwithstanding)
I'm getting the feeling that these online REPLs were created just to show that it's possible to do, not to address any practical concern or to highlight a strength of these languages.
Again, I don't mean to disparage the efforts of the authors. Just my $0.02
- andrewtj 16y agoI think these sites are trying to cater more toward FUD-encumbered newbies than those for whom installing and having a play with a new language is second nature.
- andrewtj 16y agoWould the down-voter care to explain their rationale?
- chrisdone 16y agoI made Try Haskell and purity and types are Haskell's whole thing, so technically I can teach the good stuff about Haskell without needing IO and threads and such. The security measure for disabling IO was to merely not run in the IO monad. Interestingly, there is a completely pure filesystem library for Haskell http://lambda-the-ultimate.org/node/1036 http://lambda-the-ultimate.org/node/1036 that I'm thinking of using, to demonstrate the difference between side-effectual and pure functions. I agree regarding the other two, security is a big problem for them. I'm trying to get a trylisp setup but many of the Lisp guys think it's fairly impossible to do a safe sandbox.