Read as: things I want to use, build, integrate, create, etc
I'm bad at remembering things. I tend to come across blog posts, Github projects, presentations and the like which makes me think "damn, that'd be cool to play around with!". I usually keep it in mind long enough to bounce the idea off of some friends at work, but then completely forget it until I get home. After doing this countless times, I've decided I really need to make an official list of things I want to experiment with and track the ones I create something with.
Distributed Computing - Hadoop and Accumulo focused
- Quantcast File System (QFS)
A Hadoop-compatible file system written in C++, touted at being more performant than HDFS. Install it locally, see how it performs/works. Github
- Node Accumulo
A small project to "ingest" records into Accumulo using Node.js and RabbitMQ. I need to make a blog writeup about it.
- Barebones QEMU Linux install
I was playing around with getting a barebones Gentoo install running via QEMU, theoretically so I could then have a single image someone could load in VirtualBox, VMware, or on bare-metal. It might also be worthwhile to look into Vagrant
- Integrate Apache Tika
Tika could make a neat extraction engine for use in a content-indexing Accumulo solution
- Create an Apache Solr-compatible API for the Accumulo wikisearch example
When it comes to content indexing, Lucene/Solr pave the way. I've already created some code which ingests data into the wikisearch table structure, but I haven't taken the plunge to find what is missing in the wikisearch query functionality, nor what it would take to make an API bridge.
Personal Hosting (this site)
- Transition this site from Django to Rails
I think this is what I want to do. I don't use Rails frequently enough to keep all of its tricks in mind. Forcing myself to learn the same tricks for Django will just keep me from updating it
- Offsite backups
Needed for this, as well as personal computers. Should've been doing this for a long time already.
- Better monitoring
Luckily, I've had next to no unplanned downtime for this instance. Obviously, I should take the extra time to prepare for a time when things do go south. It would be nice to find something like PhpSysInfo that can also report on things like Nginx, Gunicorn, Thin, etc.
Other things, generally speaking
- Storm, zeromq, rabbitmq
I really liked the simplicity of setting up RabbitMQ when I played with it for Node Accumulo. I've heard lots of great things about Storm and zeromq.
- Electric Imp
Because who doesn't want to connect their appliances to the web??
- Build a GPS tracker in my car
Ignoring any "my car gets stolen" applications: wouldn't it be cool if my car tracks where it's been? I could actually collect data on how many times it's traveled the same stretch of road to/from work, etc. I could also probably very easily splice into the power already at the dash for my stereo. I could then bring it inside, have it auto-connect to my wifi, and upload the stats to some remote server. Seeed studio has some neat stuff, along with the typical Arduino, Beagleboard, and others.