andyvanee

Peter Norvig is a big name in computer programming. I stumbled across some of his code and was struck by how he could often accomplish in 4 lines of code things that might take me pages.

A bit more searching led me to a course taught by Peter on Udacity Design of Computer Programs

There were a number of places where I was absolutely blown away. Almost always it was a case of me over-engineering the problem (usually with unneccesary OOP) where his simple data types and list comprehensions nailed it.

As I’ve given this blog a little refresh, it’s caused me to look back on what I’ve written so far and think about what I want to write in the future. My first reaction was embarrassment at some of...

Read More

OpenResty is a bit of an enigma. I first heard about the project a few months back when reading about some benchmarks of various web application platforms. OpenResty was WAY ahead of all the others...

Read More

This blog is now compiled with Middleman and deployed with Wercker (still deploying to S3). I’ve updated the look/feel as well, mostly taking stuff away to make it really clean & simple.

I think this will be a nice setup :)

I haven’t been doing a lot of JavaScript lately, but every time I do i try to explore the idea of doing everything event driven. “I want the world to be made of hooks” -Andrew Plotkin This quote...

Read More

When Google announced that Reader was to be no more, I, like many other geeks took things into my own hands. I was, until recently, a daily user of Reader. There are only a handful of RSS feeds that...

Read More

…for a number of reasons. One reason, which is often addressed, is the challenge of effectively communicating your ideas. This is partly technical, but has more to do with how effectively you communicate...

Read More

In an effort to make the static site workflow a little less manual, I’ve created a little web-service to automate things a bit. Here’s how it works:

1. Edit posts/pages/layouts etc with prose.io
2. Github web-hook hits my service, triggering a build.

The build stage looks like this:

1. git pull (or clone if need be)
2. make
3. s3cmd sync

Setting up the server was a bit difficult for a couple of reasons. My blog is pretty straightforward, but it did need a number of things installed on the server for a successful build: make, jekyll, uglifyjs, recess, pygments and s3cmd. Also, this all had to be run as the www-data user, which had a pretty minimal environment.

So far, it’s awesome. Editing and publishing my blog as a static site on S3 is now completely automated. Code coming soon on github…

This post caught my eye: Why you need a home lab to keep your job. Your boss won’t pay for training, so your partner has to put up with servers at home What intrigued me, as a web developer, was not...

Read More