Timothy Stebbing tjstebbing

Atomic refreshing in memory DBs without mutexes

This is a simple way of having an in-memory DB which is safe to use across multiple goroutines without needing to use any kind of mutex for updates.

1. Create an interface

type FooDB interface {

🚐 MINI BUS

Mini Bus is a drastically simplified version of D-BUS, specifically for plain text communication between terminal UIs / shell scripts and terminal multiplexers.

minibusd

Minibus daemon monitors /tmp/minibus for newly created UNIX Domain sockets (SOCK_STREAM) and connects to them based on their location:

Very simple method for writing API tests for REST services using bash, httpie, jq and shunit2.

Write some tests:

#!/bin/sh

# Integration test suite, install: httpie, jq and shunit2

alias http='http --body --pretty=none'

Bard testing tool

Bard is a commandline utility for running code tests in multiple languages using Markdown files as the source. Bard has two basic concepts:

Bard extracts code blocks from markdown and executes them with the native language test tools.
Bard interleaves shell commands using a special syntax that compares command output with expected responses.

At CampJS this week there were a lot of talks about frontend web components, mostly focused on views. It struck me, as a mostly backend engineer that we could do with some better abstractions around client side/server side data sync to go with the new view model systems out there. Here's a few thoughts I had over the weekend..

A client side, in memory data store:

Which is atomic (transactional?)
With data that is strongly typed (validated)
That syncs to a server (which is authoritative)
That syncs uncommitted content to local storage transparently
That is resilient to connectivity trouble
That is bounded by the developer

Anvil SQL helper

Anvil is a golang lib and wrapper for database/sql that aims to provide:

Out of the box CRUD operations for user defined struct types
Extensible operations via user defined SQL templates
A structured representation of Tag defined SQL field types for use in SQL templates
Helpers for working with multiple DB sharding

Anvil isn't:

	#!/bin/bash

	#apt-get install html-xml-utils
	bridgeURL="http://howmanydayssincemontaguestreetbridgehasbeenhit.com"
	hookURL="http://example.com"

	days=$(echo "$bridgeURL" \| \
	wget -O- -i- 2>/dev/null \| hxnormalize -x \| hxselect -i '.day_box' \
	\| grep -o -E '[0-9+]')

	package main

	import (
	"bufio"
	"encoding/json"
	"fmt"
	"io/ioutil"
	"log"
	"net/http"
	"os"

	const reg = /\s([^\s\{,]+(?:\s+[^\s\{,]+))\|\s\{[^\}]\}/gm

	css = `
	.car, div.train, #bus
	div.jetski:nth-child(4n)!important {
	display: none;
	}

	.banana ,
	#fruit.ap_ple.pear {

	11300841 ['1775','10921','3583','5838','15519','1794']
	11300841 ['10761','9251','69928','78022','10459','5188','120504','120506','128530','94604']
	1345 ['10761','9251','69928','78022','10459','5188','120504','120506','128530','94604']