There's a meme in the JS ecosystem that CSS is abominable and all its troubles would be fixed by making it a real boy. It's a strong meme.
-
-
Replying to @wycats @chriseppstein
I'm glad you're pushing back against it, but omg the number of times ppl have cargo culted the claim that CSS would be better as a real PL.
1 reply 2 retweets 23 likes -
Replying to @wycats
I talked about this in my last conference talk. My feelings on it are nuanced ;) A Turing Complete css runtime would be bad for browsers.
3 replies 0 retweets 12 likes -
Replying to @chriseppstein
I think it comes down to the Rule of Least Power. We need some more abstraction tools, but that doesn't imply a general purpose PL.
2 replies 0 retweets 7 likes -
Replying to @wycats @chriseppstein
Given a sufficiently good type system, is the rule of least power still applicable?
1 reply 0 retweets 0 likes -
Replying to @spion @chriseppstein
It is, but Rule of Least Power becomes an API design guideline rather than a PL guideline.
1 reply 0 retweets 0 likes -
Haskellers would say don't pass the state or io Monad around gratuitously. RoLP is saying if a PL makes that hard, don't always use it.
2 replies 0 retweets 0 likes -
Replying to @wycats @chriseppstein
the problem of ROLP is that you can't do anything in that nice little edge case where you do need more power to save you days of work
1 reply 0 retweets 0 likes -
Replying to @spion @chriseppstein
But if you model it using types, you'll see that the Io Monad has now leaked into the rest of your program ;)
1 reply 0 retweets 0 likes -
Which will take days to refactor.
2 replies 0 retweets 0 likes
I feel like it's a pipe dream to have an ergonomic general purpose PL that satisfies the spectrum of PoLP but won't dismiss it out of hand.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.