Multiprocessing TSV repair
A few days ago I wrote about optimizing a TSV repair script that took large TSV files with unquoted newline characters like this:
Recent Posts
Optimizing TSV Repair in Python
This TSV file has a problem:
“Don’t automate complexity”
You ever hear someone say “We can just automate it,” and feel almost certain it’s a bad idea? I’m always fighting with this. And I can’t always explain why I...
More consistent hashing, or learning math for the advanced in age
I wrote my last post about consistent hashing, which has really stuck with me. The problem statement is so simple but the solution seems at a glance to be co...
The niftiness and diversity of consistent hashing
I’ve been interested in decentralized data in distributed systems lately. There’s too much data for any one storage node to handle, so you have lots of stora...
Abuse-tolerant interfaces
A common approach in the industry for forming a performance oriented SLA is to describe it using average, median and expected variance. At Amazon we have ...
Swatch time
I saw this on HN this morning. Nearly 30 years ago Swatch created Swatch Internet time. The units are called .beats, and they’re a decimal timekeeping system...
My kind of place
What is it about big social media that makes me feel like I’m part of the conversation? Conversations are being had, and there are people here, so this must ...
Replicating lazy replication in python
I’ve been reading about distributed systems lately. I have a lot to catch up on. When I started making a reading list a couple months ago, I had heard about ...
A week on NTPSec
In which I find out with some certainty what time it is.
An automated checklist for computer setup
A little while ago my laptop died once, and then twice, and in between each failure I had to use a spare laptop. I ended up setting up my computer fully for ...
A plain text to-do list
I’ve tried lots of to-do lists, from ToDoist to fine-grained Jira tickets and all things in between. I still felt disorganized. But I found1 Jeff Huang’s My ...
NTP in 30 seconds
What time is it?