|
current scala status: still pretty neato haven't done anything practical with yet of course, that's when programming stops being fun
|
# ? Apr 9, 2013 21:33 |
|
|
# ? May 10, 2024 14:17 |
|
Cybernetic Vermin posted:wow, i think that is the first language in the thread i have genuinely never heard of, and it looks entirely uninteresting are you one of those hipsters who resents c syntax for no reason
|
# ? Apr 9, 2013 21:37 |
|
Werthog 95 posted:anything practical, that's when programming stops being fun agreed, werthog 95
|
# ? Apr 9, 2013 21:38 |
|
squirrel is essentially lua with built-in classes, sane boolean behavior, errors for undefined variables/table element access, and automatic reference counting
|
# ? Apr 9, 2013 21:41 |
|
Win8 Hetro Experie posted:so you're just fine with WoW players making their first lua addon advocating lua and writing addons in lua as how they should be developing software in their day jobs? that's not what i was talking about at all
|
# ? Apr 10, 2013 03:40 |
|
wrote a lexer for handlebars syntax highlighting in pygments still baffled that we all use regular expressions in 2013. you'd think people would at least start habitually commenting them, or at least stop leaving them as weird undocumented strings all over the place. i mean no one's made anything better but it's just one of the least intuitive things in coding today
|
# ? Apr 10, 2013 04:14 |
|
abraham linksys posted:wrote a lexer for handlebars syntax highlighting in pygments emacs has some lispy dsl for writing them but honestly i just end up using weird undocumented strings instead since they take up like 10x less space and who doesn't know how to read a regex in this the thirteenth year of the twenty-first century of our lord and saviour jesus "big jc" christ.
|
# ? Apr 10, 2013 04:25 |
|
Police Academy III posted:emacs has some lispy dsl for writing them but honestly i just end up using weird undocumented strings instead since they take up like 10x less space and who doesn't know how to read a regex in this the thirteenth year of the twenty-first century of our lord and saviour jesus "big jc" christ. you are so amazingly stupid that i'm not surprised at all that you'd write something like this
|
# ? Apr 10, 2013 04:31 |
|
vapid cutlery posted:you are so amazingly stupid that i'm not surprised at all that you'd write something like this uh ok gently caress you too buddy
|
# ? Apr 10, 2013 04:38 |
|
the only good thing about regexes is that there's stuff like http://regexpal.com/ that makes it way less tedious to write i mean they work they're just one of the few useful things that everyone just gave up making any easier to write
|
# ? Apr 10, 2013 05:04 |
|
abraham linksys posted:the only good thing about regexes is that there's stuff like http://regexpal.com/ that makes it way less tedious to write writing them isn't really that hard because typically you're just trying to generalize from an example. reading them is loving stupid though.
|
# ? Apr 10, 2013 05:33 |
|
vapid cutlery posted:writing them isn't really that hard because typically you're just trying to generalize from an example. reading them is loving stupid though. reading them isn't bad. it's finding them in the wild, all alone, without a comment describing their intended input and output. peak regex
|
# ? Apr 10, 2013 05:38 |
|
ultramiraculous posted:reading them isn't bad. it's finding them in the wild, all alone, without a comment describing their intended input and output. ok
|
# ? Apr 10, 2013 05:38 |
|
so have a thingcode:
|
# ? Apr 10, 2013 07:00 |
|
use sublime text learn regex stop wrapping wrappers
|
# ? Apr 10, 2013 07:14 |
|
Police Academy III posted:so have a thing gross
|
# ? Apr 10, 2013 07:25 |
|
(aa)*+a matches {a,aaa,aaaaa,...}. ((aa)*+a)* matches {a,aaa,aaaaa,...}. notice that taking the * of a language containing a does not result in a language matching aa. ((aa)*+a)*a matches.... {a}. perl regular expressions are pretty broken in their compositional semantics. presenting a paper about it at a conference in june in fact. Cybernetic Vermin fucked around with this message at 09:19 on Apr 10, 2013 |
# ? Apr 10, 2013 09:16 |
|
wait really that's weird
|
# ? Apr 10, 2013 09:23 |
|
to not confuse anyone about the severity, *+ is the 'possessive quantifier', which matches as much as it can and refuses to backtrack. so a*+a matches nothing, the a*+ will always eat all a's so the final a can never match. fact remains that the semantics are largely undocumented, very weird in practice, and the only known implementation technique is a hack in backtracking engines, giving perl and pcre different semantics. working out proper (compositional) semantics made for a neat article
|
# ? Apr 10, 2013 09:57 |
|
ultramiraculous posted:reading them isn't bad. it's finding them in the wild, all alone, without a comment describing their intended input and output. lol if you use a language which doesn't allow whitespace and comments in regexes
|
# ? Apr 10, 2013 11:26 |
|
vapid cutlery you seem to be mad at a lot of people and i hope you aren't under too much stress in your life ok god bless
|
# ? Apr 10, 2013 11:52 |
|
People should check out whiley a cool language with flow typing allowing you to model assertions in a natural way and have the compiler check them Intersection types are cool and more languages should have them now that people have figured out that sum/union types are a thing Also the verification is decidable which is neat.
|
# ? Apr 10, 2013 15:21 |
|
http://whiley.org inshallah
|
# ? Apr 10, 2013 15:23 |
|
Malcolm XML posted:http://whiley.org inshallah so i saw this, said "wow neat" then i saw this : http://whiley.org/2013/01/14/whiley-puzzler-1/ and closed that tab forever
|
# ? Apr 10, 2013 18:09 |
|
whitey, the most static plang
|
# ? Apr 10, 2013 18:19 |
|
operational transformation is cool as hell: http://www.codecommit.com/blog/java/understanding-and-applying-operational-transformation (the bottom part of the article where the graphs get all crazy is still hard for me to grok tho)
|
# ? Apr 10, 2013 18:34 |
|
im just gonna leave this here: http://bluehackers.org
|
# ? Apr 10, 2013 18:38 |
|
trex eaterofcadrs posted:so i saw this, said "wow neat" then i saw this : http://whiley.org/2013/01/14/whiley-puzzler-1/ and closed that tab forever wait are you saying there's something wrong with this behavior or what.
|
# ? Apr 10, 2013 18:38 |
gas
|
|
# ? Apr 10, 2013 18:38 |
|
ultramiraculous posted:wait are you saying there's something wrong with this behavior or what. not to get all tbc here but permanently retyping a variable due to a query on its type is some academic masturbation of the highest order
|
# ? Apr 10, 2013 18:48 |
|
ultramiraculous posted:wait are you saying there's something wrong with this behavior or what. Don't know if I'm woooshing myself here or what, but it sounds like a bad idea that a purely functional g(x) called twice with the same argument can give different results based on the context. If you can't figure out the type, I'd say you should either disallow the use (say like what Hindley-Milner type systems do) or go dynamic for that call, and maybe warn. Here it just deals with ambiguity by typing with 'any' so that type inference can change how g(x) will be executed based on context rather than the function itself. Kind of like dynamic scoping, but for type systems. Or maybe there's something I don't get about that one, but that's what it sounds like.
|
# ? Apr 10, 2013 18:51 |
|
it seems that the type of the object isn't used in dispatch, but the type of the variable holding the object
|
# ? Apr 10, 2013 19:02 |
|
maybe this is a joke designed to show why most naive static analysis techniques fail
|
# ? Apr 10, 2013 19:11 |
|
trex eaterofcadrs posted:not to get all tbc here but permanently retyping a variable due to a query on its type is some academic masturbation of the highest order i guess i'm just used to scala, because it's roughly equivalent to doing a match/case, which is basically a fancy switch that types the output for you
|
# ? Apr 10, 2013 19:16 |
|
the dispatch this is also how scala, and i maybe java too, handles overloaded methods that have covariant parameters. if you have an Object that happens to be a String, you call the method with the Object parameter because the defined behavior is not to do dynamic dispatch. from there you'd figure out its a string and call g(x) with a string type. it's scala regardless so trex eaterofcadrs posted:academic masturbation of the highest order ultramiraculous fucked around with this message at 19:25 on Apr 10, 2013 |
# ? Apr 10, 2013 19:22 |
|
i might have misread but it looked like a bad interaction between what was just an implicit unpacking of an object from a union type (which, done explicitly, is bog-standard in languages with union types, right?) and an overloaded g(x)
|
# ? Apr 10, 2013 19:43 |
|
abraham linksys posted:operational transformation is cool as hell: http://www.codecommit.com/blog/java/understanding-and-applying-operational-transformation (the bottom part of the article where the graphs get all crazy is still hard for me to grok tho) looks like darcs' patch theory
|
# ? Apr 10, 2013 22:30 |
|
patches can be composed except when they can't be composed yo
|
# ? Apr 10, 2013 22:57 |
|
rebinding variables is a bad idea silently changing the type of a variable is utterly horrible. its worse than php a fun esolang would be one that employed the principle of most astonishment tho
|
# ? Apr 11, 2013 01:45 |
|
|
# ? May 10, 2024 14:17 |
|
"use strict";
|
# ? Apr 11, 2013 18:35 |