infinity0

Contents

• Kid from a poor family goes to a famous university: what happened afterwards | 寒门子弟上名校之后

infinity0

Haskell

This is a quick "zero to hero" guide of various Haskell and FP-related topics. The intended audience (you) is self-directed people that want to quickly become a world-class Haskell programmer. Actually many of the topics are not specific to Haskell, so it can probably also help you become a world-class programmer in a similar language too.

infinity0

#define _GNU_SOURCE
#include <stdlib.h>
#include <unistd.h>
#include <limits.h>
#include <stdio.h>
#include <linux/types.h>

size_t max = (size_t)LONG_MAX +1; // (size_t)LONG_MAX+1
loff_t off = 0;

infinity0

import qualified Data.Sequence as S
import qualified Data.ByteString as BS

import Control.DeepSeq
import Data.Foldable
import Data.Traversable
import Data.IORef
import Data.Word
import GHC.Clock
import System.IO

infinity0

Tested 2020-07-19 chromium 83

You need to do:

• chrome://flags and enable load-media-router-component-extension, might have to restart 1-2 times to take effect
• download libwidevinecdm.so, see README.debian for details, restart to take effect
• ublock filters (not whitelist, those are something else):
  • @@||www.gstatic.com/cast/*$script,domain=open.spotify.com
  • @@||www.gstatic.com/cv/js/sender/v1/cast_sender.js$script,domain=open.spotify.com

infinity0

\section{Background}
\paragraph{Gossip} In a peer-to-peer gossip network we have to figure out how the nodes connect to each other. Two possibilities are a r-regular random graph and the SlimFly topology.

For the current assignment you can assume that the full list of peers is known, and available to each peer. (Networks where this is not the case, are vulnerable to the Sybil attack, which is out-of-scope for the time being.)

The topology should work for any sized network. Some topologies are optimised for particular network sizes (e.g. n = k ^ 2 for some k) and degrade for other sizes. This is bad.

There should be an efficient algorithm that each node can run to figure out its own neighbours. For example, for r-regular random graphs the only known way of doing this[1] is for everyone to generate the whole graph, whose computation costs O(n^3). This is not ideal. For our purposes, anything below O(n^2) can be considered reasonably efficient.

[1] (The naive method of selecting r random outgoing neighbours 

infinity0

{-# LANGUAGE
    AllowAmbiguousTypes
  , DataKinds
  , FlexibleContexts
  , FlexibleInstances
  , GADTs
  , MultiParamTypeClasses
  , PolyKinds
  , RankNTypes
  , ScopedTypeVariables

infinity0

Propaganda

"Official" (played @ actual campaign rallies):

• https://www.youtube.com/watch?v=X0dg7dUprmU #8MileYang - MC Jin
• https://www.youtube.com/watch?v=WhIv_iWetGo Drew Yang Gang That's Who I Hooked Up Wit - MC Jin

Memes & shit talk:

infinity0

[merge]
	conflictstyle = diff3
	tool = kdiff3
[mergetool]
	keepBackup = false
	keepTemporaries = false
	prompt = false

## then when you get a merge conflict, run "git mergetool" and it will run kdiff3 resolving as many conflicts automatiacyll as possible

infinity0

$ apt-file search -x ^/usr/bin/ | cut -f2 -d: | sort -u
 /usr/bin/[
 /usr/bin/0alias
 /usr/bin/0desktop
 /usr/bin/0install
 /usr/bin/0launch
 /usr/bin/0store
 /usr/bin/0store-secure-add
 /usr/bin/0xFFFF
 /usr/bin/2csv