lisp

http://twitter.com/id_aa_carmack

Bit surprised at negativity about my lisp comments -- I am infatuated with the elegance of Scheme out of all proportion to its relevance

He likes Haskell more.

Bit surprised

I'm not

I play around with Lisping on iPad while on plane flights; it usually leaves me considering what niche lisp could still usefully occupy.

Not understanding the vast capabilities and current uses of LISP is no ones fault but your own.

>>3
What can LITHP do that Template Haskell can't?

>>4
easy to write macros

isn't lisp just an incredibly good well designed language? and the only reason people don't use it is because of bad taste?

it's had all these fantastic ideas in it that have taken decades and decades for normal people to accept the utility of.

lisp just makes it easy to do a few things that most programmers don't consider doing.

>>6
LITHP is a language with a "could be done" curse embedded. LITHPers constantly talk about how this or that could be done with macros, but almost never actually do anything.

>>5
The only thing that Lisp saves when macrobuilding or language-making is lexer+parser.

>>9
What shitty language macro system makes you write a fucking lexer/parser?

>>8
LITHP could be haskell

>>9
Given how many tools need to be able to parse program source code, I'm not convinced that savings doesn't pay off. Look at all the duplicated effort that goes into writing parsers for all the tools that support the use of popular curly brace languages. It's a lot easier to keep a simple syntax than to build a popular parsing library for a complex one.

>>1

I am infatuated with the elegance of Scheme out of all proportion to its relevance

Does he know about Common Lisp?

>>1
oh please, stop thinking on the surface of the gentleman. Go deep within.

He is a graphics programmer, therefore he loves the model of Math, therefore he loves functional programming languages. Get over it. Big fucking deal.

Cudder's into shoving pixie wands up his butt, therefore he loves the model of the Intel Chipset. You can't blame the fucking guy, he just loves shoving pixxie wands up his fucking butt.

quit your fear of "the best" and go outside and throw a fucking ball or something

>>14
u Need to Stop beING Retarded.

go deep within my anus

LOL DO I FIT IN YET GUYS? EPIC MEME

>>14

He is a graphics programmer, therefore he loves the model of Math, therefore he loves functional programming languages. Get over it. Big fucking deal.

goddamn you are a stupid nigger
i bash you upon the head nigger
bash
bash
bash

>>14

he loves

C++

OK, let's have a look at this.

------------------------------------------------------------

<ID_AA_Carmack> Is there a formal lisp dialect with no undefined behavior?
<jb55> closest thing I could think of is Shen http://www.shenlanguage.org

The first recommendation is shen, this stupid hipster bullshit written by an egotist who doesn't understand type systems or freedom.

<ID_AA_Carmack> That is one of the things I don't care for about scheme. Elegant simplicity, but not as rigorous as I would like.
<quadricode> Scheme pioneered rigor with its mathematical underpinnings, namely its single page of formal denotational semantics.

Knights of the TOLD republic. Carmack is clueless.

some faggot hipster with a pixel art avatar chimes in:

<dysoco> I'd say take a look at Qi or Shen, maybe Arc. But I'm also interested in knowing if there's a secure Lisp dialect.

some dumbasses who doesn't understand godels theorem or know what a language is chimes in:

<infogulch> are you asking for a language that is both consistent and complete?
<tom_forsyth> If there is, Godel says it's not very interesting.

------------------------------------------------------------

<ID_AA_Carmack> I play around with Lisping on iPad while on plane flights; it usually leaves me considering what niche lisp could still usefully occupy.
<swanodette> one of the best optimizing compilers written in Scheme http://icfp06.cs.uchicago.edu/dybvig-talk.pdf , are high quality production compilers niche?
<cemeric> Gentle suggestion: maybe evaluating an entire language family based on iPad editors/impls isn't a good strategy?

HAH HA HA HA HA FUCKING GIGATOLD.

<darinmorrison> I think there’s room for a fancy typed lisp based on ideas from generalized multicategories and polynomial functors/opetopes

math... I doubt anyone knows wtf you're talking about. This guys hair shows he ate much haskell in the 60s.

<adam_hill> Niche #1: Space flight! http://www.flownet.com/gat/jpl-lisp.html #lisp #space

neat.

------------------------------------------------------------

<ID_AA_Carmack> Bit surprised at negativity about my lisp comments -- I am infatuated with the elegance of Scheme out of all proportion to its relevance.

LOL he keeps trolling, ever time he gets his ass handed to him it just makes him more vicious!

<nega> as someone we respect, you make glib clickbait comment about what we know you're not an expert in, and your're surprised?

you tell em buddy

<jrychter> well, there is a useful lisp, it’s called #clojure, and it’s growing. People use it for large systems, but quietly, no drama.

<Johnathan_Blowhard> I think Lisp is in a separate enough neighborhood from most other languages that it invites that weird tribalism.

What absolute rubbish.. he is just completely out of touch.

<ID_AA_Carmack> @nega "Niche" was positive--I want to find a way to use it. I stand by the env comment--functions depending on mutable global state is bad.

so in the end he just expressed himself in a slightly unclear way and got jumped on.

blowhard is just trying to give him a suckjob by going on about "tribalism."

carmack went on about it when people didn't like that fb bought oculus.

tfw m y lines didnt wrap
no one gonna appreciate my post fully

>>21
mfw I read most of those comments on twitter already

>>19

math... I doubt anyone knows wtf you're talking about

You are retarded and uneducated. Please remove thyself from this board forever and ever.

>>10
Um, every one of them. It's just that with Lisp you can use Lisp's lexer as the lexer and the macro arglists as the parser. That's the only work that Lisp saves you, and it's not much.

>>24
Macros are not even good parsers. If you want to implement any serious kind of syntax, you have to implement most of syntactic checking and error reporting yourself.

>>24,25
Symta macros aren't coming on so well, are they?

>>12
You've never heard about code reuse, or what? Haskellers, for example, reuse a part of GHC to parse Haskell code:

https://hackage.haskell.org/package/haskell-src-exts-1.13.5

>>27
If only GHC based implementations use that interface, there is no reuse. gcc also has a parser interface that nobody else uses - oh sure, it could potentially be reused, but it isn't. No effort is saved making a thing reusable if it is never actually reused.

>>24
Lisp is built to be extended. Most languages are not. If you want to add to another language, you have to fuck around with the grammar. In lisp, two language extensions can merge with minimal work. In other languages, it could take a complete rewrite to do this.

>>25
A macro checks its form upon expansion. Yeah you have to write this checking code yourself. What alternative do you propose? For the validation for your language extension to magically exist in the implementation's native parser?

>>26
kill yourself.

>>27
Tell me, how much work would it take you for another bullet point to appear on that list of syntax extensions? It takes me 10-20 lines of code in lisp.