p-lang thread: (now (have you (problems two)))

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > YOSPOS > p-lang thread: (now (have you (problems two)))

«‹›1784 »

Abandon: Nov 23, 2006

# ? Sep 9, 2013 07:33

Adbot: ADBOT LOVES YOU

# ? May 13, 2024 12:57

MononcQc: May 29, 2007

http://resthooks.org/

because polling socks, let's reimplement everything TCP or TLS, but over HTTP over TCP/TLS, and with an ad-hoc description rather than an actual spec.

The docs includes use cases for:

congestion control
retransmissions of lost messages
in-order delivery guarantees
authentication and handshakes

They even acknowledge the existence of things like websockets that could be better and lighter (given they're just a way to tunnel TCP-equivalent poo poo but over HTTP), but declare that (instead of writing a websocket->rest proxy, or whatever else) it's just better to have all these entire patterns to do poo poo and reimplement it all over again because CRUD.

:smithicide:

# ? Sep 10, 2013 18:14

Opinion Haver: Apr 9, 2007

quote:

Over a representative time period, Zapier polled for changes 30 million times but only took action on 460,000 of those polls (1.5% efficient).

Contrast this with REST Hooks, where 547,000 hooks were received over the same period and all of them were acted upon (100% efficient).

In other words, if everyone implemented REST Hooks, server load for both the sender and receiver could be reduced by 66x.

lol

# ? Sep 10, 2013 18:56

Cocoa Crispies: Jul 20, 2001; Vehicular Manslaughter!; Pillbug

MononcQc posted:

http://resthooks.org/

because polling socks, let's reimplement everything TCP or TLS, but over HTTP over TCP/TLS, and with an ad-hoc description rather than an actual spec.

The docs includes use cases for:

congestion control

retransmissions of lost messages

in-order delivery guarantees

authentication and handshakes

They even acknowledge the existence of things like websockets that could be better and lighter (given they're just a way to tunnel TCP-equivalent poo poo but over HTTP), but declare that (instead of writing a websocket->rest proxy, or whatever else) it's just better to have all these entire patterns to do poo poo and reimplement it all over again because CRUD.

looks like it's just a brand name and a smug web page about something zillions of services already do

fitbit, withings, etc. have had subscriptions for years at this point

# ? Sep 10, 2013 18:59

uG: Apr 23, 2003; by Ralp

what if i wish to use the superior SOAP

# ? Sep 10, 2013 19:30

Plastic Snake: Mar 2, 2005; For Halloween or scaring people.

SOAP isn't web scale!!

# ? Sep 11, 2013 21:20

Nomnom Cookie: Aug 30, 2009

has anyone ever used soap over a non http transport

# ? Sep 12, 2013 00:27

Shaggar: Apr 26, 2006

I don't think so

# ? Sep 12, 2013 00:34

Nomnom Cookie: Aug 30, 2009

soap was created to make corba look good

# ? Sep 12, 2013 00:36

Shaggar: Apr 26, 2006

soap is fine.

# ? Sep 12, 2013 00:36

Shaggar: Apr 26, 2006

its just a bummer theres no replacement or new work being done on messaging standards.

# ? Sep 12, 2013 00:37

MrMoo: Sep 14, 2000

Nomnom Cookie posted:

has anyone ever used soap over a non http transport

I think Microsoft use it over MSMQ for some of their other terrible software. I've certainly seen SOAP on MSMQ on PGM.

# ? Sep 12, 2013 00:48

Bloody: Mar 3, 2013

i used soap once to serialize a .net application's settings to disk

# ? Sep 12, 2013 01:01

Nomnom Cookie: Aug 30, 2009

work uses mysql as a message queue

# ? Sep 12, 2013 01:03

Shaggar: Apr 26, 2006

lots of message systems use volatile storage

# ? Sep 12, 2013 01:08

Notorious b.s.d.: Jan 25, 2003; by Reene

Nomnom Cookie posted:

work uses mysql as a message queue

i used to work on bunch of in-house software that used mysql for all ipc

but data was passed to/from customers using ibm mq and xcbl and soap

the original devs just couldn't make the conceptual leap to use that poo poo internally i guess

# ? Sep 12, 2013 01:10

Notorious b.s.d.: Jan 25, 2003; by Reene

oh i forgot the best part

it was mysql 4, so there was no meaningful way to lock rows or anything. jobs just had to be very precisely scheduled not to step on each other's toes. with a homebrew job scheduler. can't make this up

who needs transactions
data integrity is for enterprise-y firms

# ? Sep 12, 2013 01:32

Workaday Wizard: Oct 23, 2009; by Pragmatica

Hiya cpphiles

I have a question regarding cpp pointer-to-members.

Why?

# ? Sep 12, 2013 11:04

shrughes: Oct 11, 2008; (call/cc call/cc)

For member functions, the reason why is pretty obvious.

For member fields, to be consistent with member functions.

# ? Sep 12, 2013 12:06

Zlodo: Nov 25, 2006

pointer to members are v useful to do things like serialization, scripting language bindings, using methods as callbacks for listeners and such

# ? Sep 12, 2013 12:33

Workaday Wizard: Oct 23, 2009; by Pragmatica

Thanks for the answers. I haven't thought of those applications :blush:

# ? Sep 12, 2013 12:42

prefect: Sep 11, 2001; No one, Woodhouse.
No one.; Dead Man’s Band

Shinku ABOOKEN posted:

Thanks for the answers. I haven't thought of those applications

this thread is frequently a place where you wind up saying "oh, i hadn't thought of that" :shobon:

# ? Sep 12, 2013 12:50

qntm: Jun 17, 2009

Nomnom Cookie posted:

has anyone ever used soap over a non http transport

over JMS, yup :shepface:

# ? Sep 12, 2013 13:28

Nomnom Cookie: Aug 30, 2009

std::bind is really cool and i wish java wish java had anything like it

# ? Sep 12, 2013 13:39

Notorious b.s.d.: Jan 25, 2003; by Reene

Nomnom Cookie posted:

std::bind is really cool and i wish java wish java had anything like it

java 8 includes lambdas to make this not painful
(you could hack around it now but the syntax would be so nasty it wouldn't be worth it)

edit: it just dawned on me that the hacky way to do this in java today is basically what std::bind does -- abuse generics with a special class that exists only to compose functions

# ? Sep 12, 2013 15:42

Nomnom Cookie: Aug 30, 2009

i doubt that java generics would let you do java.lang.PartiallyAppliedFunctionFactory

# ? Sep 12, 2013 18:03

Max Facetime: Apr 18, 2009

Nomnom Cookie posted:

work uses mysql as a message queue

Shaggar posted:

lots of message systems use volatile storage

Nomnom Cookie posted:

i doubt that java generics would let you do java.lang.PartiallyAppliedFunctionFactory

java generics won't, but MethodHandles should, with the caveat that the code is really ugly

this short demonstration calls Throwable.printStackTrace from another thread the traditional way and by adapting a MethodHandle:

Java code:

import java.lang.invoke.MethodHandle;
import java.lang.invoke.MethodHandleProxies;
import java.lang.invoke.MethodHandles;
import java.lang.invoke.MethodType;

public class Snippet {
  private static Runnable version1(final Throwable throwable) {
    return new Runnable() {
      @Override public void run() {
        throwable.printStackTrace();
      }
    };
  }

  private static Runnable version2(Throwable throwable)
      throws NoSuchMethodException,
      IllegalAccessException {

    MethodHandle printStackTrace = findMethodThrowablePrintStackTrace();

    MethodHandle noArgPrintStackTrace =
      MethodHandles.insertArguments(printStackTrace, 0, throwable);

    return MethodHandleProxies.asInterfaceInstance(
      Runnable.class,
      noArgPrintStackTrace);
  }

  private static MethodHandle findMethodThrowablePrintStackTrace()
      throws NoSuchMethodException,
      IllegalAccessException {

    MethodType printStackTraceSignature =
      MethodType.methodType(void.class);

    return MethodHandles.lookup().findVirtual(
      Throwable.class,
      "printStackTrace",
      printStackTraceSignature);
  }

  public static void main(String[] args) {
    try {
      Thread thread1 = new Thread(version1(new Exception()));
      thread1.start();
      thread1.join();
      Thread thread2 = new Thread(version2(new Exception()));
      thread2.start();
      thread2.join();
    } catch(InterruptedException | NoSuchMethodException
      | IllegalAccessException e) {
      e.printStackTrace();
    }
  }
}

# ? Sep 12, 2013 21:32

Notorious b.s.d.: Jan 25, 2003; by Reene

Nomnom Cookie posted:

i doubt that java generics would let you do java.lang.PartiallyAppliedFunctionFactory

uhhh something like that

trigger warning: this is worse than c++

http://www.drdobbs.com/jvm/functional-programming-in-java/184406320

# ? Sep 12, 2013 21:47

Notorious b.s.d.: Jan 25, 2003; by Reene

Partial.java is too big to paste w/ comments so here's a github gist

it's really, really ugly. with lambdas this entire file would be like five lines. using generics, it's a clusterfuck

https://gist.github.com/anonymous/6543643

here is a sample

Java code:

	/**
	 * Makes a partial function object with 4 arguments remaining.
	 * @param <T0>		is the type of the previous function argument
	 * @param <T1>		is the type of the next function argument
	 * @param <T2>		is the type of the second remaining function
	 * 					argument
	 * @param <T3>		is the type of the third remaining function
	 * 					argument
	 * @param <T4>		is the type of the final function argument
	 * @param <R>		is the return type of the function object
	 * @param <A>		is the type of the current ArgList
	 * @param <E>		is the type of the Executioner
	 * @param arglist	is the current ArgList
	 * @param execute	is the executioner
	 * @param f			is the current function object,
	 * 					used to infer types needed by this function
	 * @return			the function object with 4 arguments remaining
	 */
	public static
	<T0, T1, T2, T3, T4, R, A, E extends
		Executioner<R, ArgList<T4, ArgList<T3, ArgList<T2, ArgList<T1, A>>>>>>
//	Fcn<T1, Fcn<T2, Fcn<T3, Fcn<T4, R>>>>
	Function.O4<T1, T2, T3, T4, R>
	make4(A arglist, E execute,
			Fcn<T0, Fcn<T1, Fcn<T2, Fcn<T3, Fcn<T4, R>>>>> f) {
		return new O4<T1, T2, T3, T4, R, A, E>(arglist, execute);
	}

# ? Sep 12, 2013 21:55

SavageMessiah: Jan 28, 2009; Emotionally drained and spookified; Toilet Rascal

# ? Sep 12, 2013 22:17

Shaggar: Apr 26, 2006

abusing generics is always ugly

# ? Sep 12, 2013 22:19

Soricidus: Oct 21, 2010; freedom-hating statist shill

someone somewhere is using that in production code

# ? Sep 12, 2013 23:02

Zlodo: Nov 25, 2006

Notorious b.s.d. posted:

trigger warning: this is worse than c++

well duh

# ? Sep 12, 2013 23:06

Opinion Haver: Apr 9, 2007

i think i found tbc's g+ account

quote:

Types were invented by compiler writers to make their job easier. They do not make programming easier.

quote:

I program a lot in Perl. It has only one simple type, scalar. You can stick anything in it, a number, text, or a reference.

quote:

For every programmer I worked with, every time they had a problem with types was in a strongly-type language.

# ? Sep 12, 2013 23:13

gonadic io: Feb 16, 2011; >>=

natural programming just got strongly typed!

code:

Feature: The Data.List module

    In order to be able to use lists
    As a programmer
    I want a module that defines list functions

    Scenario: Defining the function foldl
      Given I want do define foldl
      Which has the type (in brackets) a to b to a (end of brackets),
                         to a, to list of b, to a
      And my arguments are called f, acc, and l
      When l is empty
      Then the result better be acc
      Otherwise l is x cons xs
      Then the result should be foldl f (in brackets) f acc x
                                (end of brackets) xs

gonadic io fucked around with this message at 23:39 on Sep 12, 2013

# ? Sep 12, 2013 23:35

Nomnom Cookie: Aug 30, 2009

a compelling alternative to cucumber

# ? Sep 12, 2013 23:42

uG: Apr 23, 2003; by Ralp

quote:

I program a lot in Perl. It has only one simple type, scalar. You can stick anything in it, a number, text, or a reference.

Perl has three built-in data types: scalars, arrays of scalars, and associative arrays of scalars :goonsay:

# ? Sep 12, 2013 23:56

Nomnom Cookie: Aug 30, 2009

uG posted:

Perl has three built-in data types: hash, weed, and pound sign

# ? Sep 13, 2013 00:03

kitten emergency: Jan 13, 2008; get meow this wack-ass crystal prison

uG posted:

hash weed and pound

friday night motherfuckers

# ? Sep 13, 2013 00:47

Adbot: ADBOT LOVES YOU

# ? May 13, 2024 12:57

NOTinuyasha: Oct 17, 2006; The Great Twist

protobuf: if i add values to an existing enum will it break the protocol or will old versions just set the enum to default if an unknown value is encountered?

# ? Sep 13, 2013 20:14

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > YOSPOS > p-lang thread: (now (have you (problems two)))

«‹›1784 »