Sometimes working in middle management feels like being trapped purgatory between making policy and making software. Tom DeMarco's Slack paints a much rosier picture.
Digg's fourth version isn't running anymore, but was an interesting system to work on. This article describes the system architecture, as well as the context behind those choices, and will hopefully be an interesting read for those scaling engineering teams and systems.
In this article I'm releasing real-time analytics for this site, as well as analyzing historical data thus far.
If you're doing analytics, reports or dealing with memory constraints in Redis, you're probably dealing with keeping two sorted-sets mutually consistent. This article also takes a look at using multi/exec to keep it fresh.
Last year Digg released a Streaming API which exposes real-time activity on its site. It's available via JSONP, so this tutorial takes a quick look at dynamically populating a Flot.js graph using the API's data.
Designing a system which scales to a high number of requests isn't critical for most applications, but you'll never know that it was important until after the fact, so it's worth putting some thought into ensuring your architecture can grow with your success.
If you've ever built a homebrew analytics system and tried to get actionable insight from it, then you know the pain of filtering out automated and suspicious requests. Here are a few notes on what I've learned on filtering bots from analytics data.
Recently I grabbed Steve Krug's "Don't Make Me Think" and found it a reasonable discussion to the topic. My one enduring frustration is that it is damningly faint on evidence.
In this post I briefly discuss the (questionably) novel aspects of the new Sisyphus blog engine which hosts this blog.
All Rights Reserved, Will Larson 2007 - 2013.