Return Styles: Pseud0ch, Terminal, Valhalla, NES, Geocities, Blue Moon.

Pages: 1-

Practicality beats purity: PHP vs Lisp

Name: Anonymous 2008-09-22 18:39

Name: Anonymous 2008-09-22 19:07

He has a point, even though he's only using it as an excuse to whine. STFU and write libraries.

Name: Anonymous 2008-09-23 20:10

>>2
Everybody writes libraries. Problem is, that's about all they do. No documentation, no maintenance, no coordination, no goal for complete, consistent frameworks, no standarization, and no unified, simple distribution.

I wish so hard we had something as cool as Python's standard library for Scheme and/or Common Lisp. A simple, consistent, complete set of tools for everything, with (preferrably) a single tool for everything. I don't want to have to lurk obscure, long abandoned, half-working servers in search for a particular library that does what another is missing, then discover it's completely undocumented and kinda broken on my Lisp. (And don't get me started on the huge set of non-standard libraries and frameworks Python has; a few are shit but many of them are awesome, simple, powerful and nicely integrated.)

The problem is, it takes a huge effort and lots of coordination and centralization to achieve this, and the Lisp community is just too small, too experimentally-oriented, too playful and too goal- and leader-less to do this.

So alternatively, I propose this: if we can't get Python's library in Lisp, we need to get Lisp in Python. So we do the following:

1. Write a Lisp dialect compiler (possibly in Python) that generates Python functions, classes, code objects and/or .pyc/.pyo files and uses Python's type system (so it'll be a lot like Scheme, with objects). (Something like Lython.)
2. Add "." syntax for attribute access (which is going to be godsend for dealing with Python objects and the standard library).
3. Add a macro facility and some Lispy utilities to it.
4. ???
5. PROFIT
6. Lisp is popular now.

Name: Anonymous 2008-09-23 20:13

>>3
Gee, that sounds like Java. But ZOMG teh jaev libz r blooted.

Name: Anonymous 2008-09-23 20:20

>>4
Are you retarded? The Java standard library is ugly and overengineered. One of the things I said I like from Python's is that it's simple and powerful at the same time. Besides, Java's is complicated partly due to its shitty type system, its toy object model and its lack of functions, let alone first-class functions, let alone closures, let alone anonymous closures, let alone- oh wait, we all Java is a bullshit language for stupid programmers and "industry".

The point is Python has an awesome library, an awesome object model, dynamic typing, and semantics which are so close to Lisp. It should be relatively easy to write a Lisp compiler for CPython's VM (or just a translator into Python code) that can use Python's standard library and interact with existing Python code.

Name: Anonymous 2008-09-23 21:04

Sounds half-assed to me. You might as well use newLISP.

Name: Anonymous 2008-09-24 6:46

>>6
... which is half-assed.  It doesn't implement many standard functions and it isn't compiled.  It is not LISP!

Name: Anonymous 2008-09-24 8:22

What about Clojure? I know you don't like Java (I also hate it), but the JVM isn't bad.

Name: Anonymous 2008-09-24 8:22

>>4
Clojure

Name: Anonymous 2008-09-24 11:36

>>8-9
HiveMind hm = MindAdder.createHiveMindFromHive( HiveFactory.getInstance().createNewHive("MindAdder").getHive(), MindFactory.createNewMind(2).getMind() )

Name: Anonymous 2008-09-24 14:29

>>7
...yes. That's what I meant.

Name: Anonymous 2008-09-24 14:38

lisp isn't pure you guys are dumb

Name: Anonymous 2008-09-24 14:45

No one claimed lisp was pure except possibly the OP - you are dumb.

Name: Anonymous 2008-09-24 15:02

php isn't practical you guys are dumb

Name: Anonymous 2008-09-24 18:30

>>7
newLISP *might* be half-assed, but if it is, it's not for any of those two stupid reasons. There's no such thing as Lisp standard functions (there are Common Lisp and Scheme standards, but newLISP is neither), and compiled? Who gives a fuck outside OMG OPTIMIZED fags? You ain't using newLISP to write GNU HURD, agreed. (BTW, most other Lisps aren't compiled either.)

Name: Anonymous 2008-09-24 19:27

Why are we comparing a programming language to PHP?

Name: Anonymous 2008-09-24 19:40

>>15
Most other Lisps certainly are compiled.  They have been getting compiled since before most other languages existed.

Name: Anonymous 2008-09-24 22:52

>>15
It apparently doesn't even have CAR or CDR. Aren't those standard functions? And its implementation of cons cells is wrong. And not only does it fail to be compiled, it uses reference counting. :(

Name: Anonymous 2008-09-25 2:13

Wait wait wait...  newLisp doesn't have CAR and CDR?  BLASPHEMY!#!$

Name: Anonymous 2008-09-25 4:25

Haskell

Name: Anonymous 2008-09-25 13:32

>>20
My other head is a tail.

Name: Anonymous 2008-09-25 15:07

php is all nice, but i would choose lisp everyday over it if it is installed on a webserver and well supported/has all needed libraries.

Name: Anonymous 2008-09-25 15:09

#define car head
#define cdr tail

PROBLEM SOLVED USING LISP MACROS

Name: Anonymous 2008-09-25 15:11

>>23
lisp uses the c preprocessor?

Name: Anonymous 2008-09-25 15:14

#defmacro car head

Name: Anonymous 2008-09-25 15:20

I meant

(define car cdr)
(define cdr car)


(Is that an _|_?).

Name: Sageing fail since 1463 2008-09-25 15:29

>>26
fail. GTFO

Name: Anonymous 2008-09-25 15:29

>>26
I think it is, but I don't know if the interpreter will successfully detect it and return _|_.

Name: Anonymous 2008-09-25 18:46

>>24
Sorry, fixed:

(give head on car)

Name: Anonymous 2009-09-19 1:02

Lain.

Name: Anonymous 2009-09-19 1:02

Lain.

Name: Anonymous 2009-09-19 1:03

Lain.

Name: Anonymous 2009-09-19 1:03

Lain.

Name: Anonymous 2009-09-19 1:03

Lain.

Name: Anonymous 2009-09-19 1:03

Lain.

Name: Anonymous 2009-09-19 1:04

Lain.

Name: Anonymous 2009-09-19 4:13


Don't change these.
Name: Email:
Entire Thread Thread List