ELLIOTTCABLE changed the topic of #elliottcable to: #ELLIOTTCABLE — “do something cool, shove it into throats, everyone thinks it's crap, then it's all amazing.” “everything else is just details.”
Sgeo has joined #elliottcable
yorick has quit [Read error: Connection reset by peer]
<purr>
<micahjohnston> what was that one webcomic with lots of dicks in it
<whitequark>
oglaf
<devyn>
so, I realize I am late to the party
<devyn>
but how many of you have seen urbit?
<whitequark>
I did
<whitequark>
I think it's an incredibly elaborate trolling attempt
<Nuck>
I thought the same thing about this channel
<devyn>
whitequark: it's very bizarre and definitely impractical but it has some neat aspects
<joelteon>
it's a trolling attempt
<whitequark>
devyn: that's fine actually
<devyn>
honestly I'm treating it as a work of artistic expression
<devyn>
and it's pretty neat as one
<whitequark>
what's not fine is that the whole thing is written with zero regard to the reader, all while pretending right the opposite
<devyn>
explain. I haven't gotten very far lol
<purr>
lol
<whitequark>
well, for example it claims to be simpler [to understand] than contemporary functional languages
<devyn>
yeah, while the author claims that it would have been impossible to write had he learned Haskell or Lisp
<devyn>
o.o
<whitequark>
yet it's in practice much more complex, 20% because it is turing tarpit-ish, and 80% because the terminology is deliberately chosen to be confusing
<whitequark>
he obviously is familiar with haskell, lisp, etc
<devyn>
yes, I found that to be a weird claim
<devyn>
I don't very much like the idea of making every single component a four letter nonsense word
<whitequark>
so what i predict is:
<devyn>
lol
<purr>
lol
<whitequark>
the majority of people is going to dismiss this as an insane weird thing
<whitequark>
and there'll be a minority of incredibly arrogant nerds running around and claiming it's the best thing since sliced bread
<devyn>
well you can predict that one of two things will happen with every thing released in the software world
<devyn>
1) it becomes a gigantic marketing thing and everyone must learn it (ruby, node.js, etc.)
<devyn>
2) it becomes a fringe and everyone who knows what's up must learn it
<whitequark>
disagree
<whitequark>
categorize: linux, libxml2, qt, ...
<devyn>
okay. never mind. that was a bad claim. it just seems like anything hyped these days is like that
<whitequark>
it's some kind of cognitive bias. no one is going to be butthurt because qt is too fringe/not fringe enough
<whitequark>
so you don't notice it
<devyn>
yes
<whitequark>
oh, hm
<whitequark>
I think that things which are deliberately marketed by someone tend to polarize people
<whitequark>
and infrastructure stuff which just takes over by default does not, at all
<devyn>
it just seems like a bunch of random words that don't mean anything
<whitequark>
never seen any forth? or apl?
<devyn>
no, no
<devyn>
I have
<devyn>
but... using nonsense words instead?
<devyn>
lol
<purr>
lol
<whitequark>
well
<whitequark>
there's ruby gem names
<whitequark>
but that is a creative way of avoiding namespace collisions
<whitequark>
it's not like you can have collisions when you directly control the namespace.
* whitequark
shrugs
<purr>
¯\(º_o)/¯
<whitequark>
he just wants to fuck with you.
<devyn>
haha
<whitequark>
on the more serious side, there's primitives
<whitequark>
which are implemented in urbit "for correctness"
<whitequark>
and then pattern-matched by the C runtime "for speed"
<whitequark>
so he claims to have the advantages of both and disadvantages of none
<whitequark>
but it's actually the opposite lol
<devyn>
hahaha
<whitequark>
1) how do you verify C code which is actually used? you don't
<devyn>
yeah, so there goes the correctness claim
<whitequark>
2) how do you ensure sensible portability of primitives? you tie yourself to a fucking implementation
<devyn>
but then he would just dismiss interpreters that did things wrong as, well, wrong
<whitequark>
because if you don't and it falls back to urbit, it is *not* usable
<devyn>
is the implementation not portable?
<devyn>
I'm not certain
<whitequark>
it doesn't matter
<devyn>
but I get the point
<whitequark>
he claims that the implementation is not a part of the language, but it totally is
<whitequark>
so no, it's not those 300 lines
<devyn>
this kind of reminds me of toki pona somehow
<whitequark>
I see zero difference with using, say, lambda calculus, which is a turing tarpit
<whitequark>
and extending it with stuff when you need it
<whitequark>
it's literally the same approach, except with more weird names
<whitequark>
guess he should have read that lisp manual :]
<devyn>
the weird names really are stupid
<devyn>
every single piece of code looks like some weird piece of Vogon poetry
<whitequark>
I would totally be for them, *if* he actually invented some really new stuff, calling which "function" or "closure" or whatever would actually be confusing
<whitequark>
but it's, again, just the opposite
<whitequark>
vogon poetry lol.
<purr>
lol
<whitequark>
this is probably the best description i heard
<devyn>
you know I feel like the names are actually taxonomic somehow
<devyn>
right? I feel like he's trying to build some kind of MUD...networked machine thing
<devyn>
; ~podlyr-ralsev-picben-laglus--silsep-holnup-rigmud-sitfet: cgyarvin said on IRC that %mint-coke is the most obscure error
<devyn>
what even
<devyn>
an error named mint-coke
<devyn>
.
<devyn>
.
<devyn>
.
<devyn>
hah
<devyn>
it really seems like he's trying to build something similar to, well, a hack lab, but over the internet... I get the feeling that this is all meant to be considered one uniform machine, basically
<whitequark>
to me it looks more like an OS where distributed computing is builtin
<whitequark>
sort of like plan9
<devyn>
but I think it's more than that - yes, that's a feature but I think it's a side effect of something bigger
<devyn>
it's basically got community interaction features built into the core OS
<whitequark>
hm
<devyn>
hell as I'm on IRC here, in the window where I'm running urbit
<devyn>
I can see people talking
<devyn>
he is fucking obsessed with four letter names
<devyn>
weev = 'web event'
<devyn>
cron = 'wakeup' (heh)
<devyn>
but the whole four letter thing like
<devyn>
this is all so OCD
<devyn>
and then less important things, temporary variables or something
<devyn>
seem to be three letters
<devyn>
fuck it, I'm looking at his C code. hold on.
<joelteon>
was that you talking
<devyn>
joelteon: ?
<joelteon>
oh, the guy talking about mint-coke wasn't you
<devyn>
nope
<devyn>
that was podlyr-ralsev-picben-laglus--silsep-holnup-rigmud-sitfet
<devyn>
what even
<devyn>
commit message: "Revise some bowl handling."
<devyn>
like
<devyn>
what
<devyn>
holy shit
<devyn>
his C code is just as ridiculously structured