<alexgordon>
aha so coffeescript has all its codegen in nodes.coffee, and you just build up a tree then call compileToFragments
<alexgordon>
couldn't be easier
<alexgordon>
coffeescript is 70% good, 30% bad
<alva>
It seems to fix the least bad parts of JS: what it looks like
<alva>
Or am I confused?
<alva>
I mean, I am, in general, but about CoffeeScript.
<alexgordon>
yeah it fixes the bad parts, but adds some awkwardness
<alexgordon>
like the way you're supposed to use callbacks is to remove the parens from the function you're calling. which is just weird
<jfhbrook>
alternately, it hides some of the good parts and creates a lot of problems of its own ;D
<jfhbrook>
also it wildly predates es6
<jfhbrook>
I'll give it points for -> and => though
<jfhbrook>
(es6 stole =>)
<alva>
Yeah was just about to say they need to update those examples, but that would make it less appealing I suppose
<jfhbrook>
nah I'm pretty sure nobody's actively developing cs right now
<alexgordon>
they are
<jfhbrook>
oh?
<jfhbrook>
pining for the fjords? ;D
<alexgordon>
last release was september 2015
<jfhbrook>
seriously though, I'm impressed
<jfhbrook>
I really did think it was an ex-parrot
* alva
doesn't get the compile-to-JS thing. Seems like a terrible target.
<jfhbrook>
well
<jfhbrook>
if you want it to run in the browser
<jfhbrook>
even if you do hate javascript, what choice do you have
<alexgordon>
alva: it's because es5 kinda sucks, so people don't want to write it
<jfhbrook>
also coffeescript is in a lot of ways Just (a subset of) Javascript
<alva>
Well that would be a nice bonus, but I don't get it as the sole target.
<alexgordon>
that's why we have babel and coffeescript and that haskell one
<alexgordon>
elm
<alva>
Like, if I could run my game albeit slowly in a browser, but also spit out a native executable that is not slow.
<alexgordon>
yeah that would be good
<alva>
If I can only do the former.. I don't know.
<jfhbrook>
keep in mind a lot of these tools are built by javascripters that are pissed about the browser, not PL nerds
<jfhbrook>
or, if PL nerds, heavily tempered by the former
<alexgordon>
alva: there is probably some way to compile js to jvm
<alva>
jvm is another thing I don't get. I guess I don't get a lot of things. Why do I need a virtual machine, my physical one is pretty good at computing.
<jfhbrook>
also the thing where you're just goofing off and already think in javascript
<alva>
Portability was and remains solved before jvm came about.
<alexgordon>
alva: it's because llvm didn't exist
<alva>
You just have to compile it once up front instead of every time it runs for everyone.
<jfhbrook>
that's the whole point though
<alexgordon>
before llvm there were two options: compile to C, or write GIMPLE
<alva>
But your source code remains the same.
<jfhbrook>
they want to ship artifacts that can be ran on any machine regardless of architecture
<alexgordon>
neither of these are very nice options
<jfhbrook>
without shipping sauce
<alva>
Oh, well if that's the purpose, then I get it.
<jfhbrook>
I mean, that's the pipe dream anyway
<jfhbrook>
we all know how well that works in practice (*okay*)
<alva>
But still. You can cross compile for all your platforms (which is probably 1 anyway), then ship those.
<alexgordon>
also in the past there were a lot more architectures
<alexgordon>
like SPARC!
<alexgordon>
nowadays it's just ARM and Intel
<alva>
I might have an ARM problem. All my things are ARM.
<alexgordon>
and then all the embedded stuff, which is gradually moving towards ARM too, because it's easier to throw a SoC in your toaster than to use some weird embedded thing
<alva>
Hehehe yeah.
<alva>
So I've noticed at work.
<alexgordon>
I remember laughing when Go was introduced, at their decision to "not use LLVM because it's too slow"
<alexgordon>
but now it seems pretty sensible, all they need are x86, x86-64, ARM and 64-bit ARM
<alexgordon>
not so much work
<alexgordon>
(I still don't get how LLVM is too slow though)
<alva>
What's their rationale to do the whole void * thing again and not support type params?
<alexgordon>
drugs
<alva>
"If we call it interface{} no one will notice"
<alexgordon>
go seems to attract a lot of people coming from dynamic languages
<alexgordon>
it's like the stepping stone from js/python to the world of static typing
<ELLIOTTCABLE>
replace the parens with brackets and, idk
<eligrey>
thanks
<ELLIOTTCABLE>
probably not, though
<eligrey>
i wanted to show a friend what paws looks like
<ELLIOTTCABLE>
because the current effort since early 2015 is a very subtle semantic change that will *completely*, unfortuantely, change how code is actually written
<purr\ec>
System/Master 5b97ea9 ELLIOTTCABLE: (- new sub py) Submoduling IPython-notebook-extensions
<purr\ec>
System/Master 178b01a ELLIOTTCABLE: (new sh script) Wrote a git-cd script. What a hot mess.
pelegreno has quit [Read error: Connection reset by peer]
pelegreno has joined #elliottcable
pelegreno has quit [Remote host closed the connection]
pelegreno has joined #elliottcable
ohhmaar_ has joined #elliottcable
brr has quit [Ping timeout: 268 seconds]
ljharb has quit [Ping timeout: 268 seconds]
incomprehensibly has quit [Ping timeout: 268 seconds]
Sorella has quit [Ping timeout: 268 seconds]
manveru has quit [Ping timeout: 268 seconds]
ohhmaar has quit [Ping timeout: 268 seconds]
ohhmaar_ is now known as ohhmaar
brry has joined #elliottcable
brry is now known as brr
Sorella has joined #elliottcable
manveru has joined #elliottcable
incomprehensibly has joined #elliottcable
ljharb has joined #elliottcable
pikajude has quit [Quit: Quit]
pikajude has joined #elliottcable
pikajude has joined #elliottcable
* pikajude
profiles node app
<pikajude>
"[Unknown] 64.4%"
<pikajude>
This is starting out well!!!!!!!!!
<pikajude>
hey ljharb, how do I get profiling info from node 0.10
<ljharb>
um
<ljharb>
i have no idea
<pikajude>
nice
<ljharb>
ask in #node-dev
<ljharb>
(also upgrade from 0.10 :-p)
<pikajude>
thanks for the rec ljharb but every time i have to investigate something node related i just get this sinking feeling in my stomach
<ljharb>
aw
<ljharb>
node-dev is the core team
<ljharb>
so give it a day and you'll probably get the right answer
eligrey has joined #elliottcable
Rorik has joined #elliottcable
Rarik has quit [Ping timeout: 244 seconds]
<pikajude>
hey, do any of you guys know why outlook's web app removes links from emails, but wraps the text that used to be a link in a span that changes it to the visited link color?
<pikajude>
that seems kind of mean spirited
<jfhbrook>
spite
<jfhbrook>
outlook's webapp text fields are almost as janky as confluence so I just have to assume it's, well, incompetence and/or lack-of-polish
Sgeo has joined #elliottcable
Navarr has quit [Quit: Connection closed for inactivity]
<alva>
Bleh, gave up on LLVM JIT, using the interpreter instead so I can progress on the actual language.
<alva>
Will revisit once ORC is more stable/documented.
eligrey has quit [Ping timeout: 264 seconds]
eligrey has joined #elliottcable
<pikajude>
VPN configuration in linux is the worst possible experience i can imagine