Sat Nov 17 2018 15:55:50 GMT-0800 (Pacific Standard Time)
TL;DR - I archived 78 episodes (~150 hours) of a radio show and you can listen to them here.
It was probably 19 years ago that I discovered the FM station WCSB streaming on the internet. It was a special time, those Napster days when the internet still felt like something new...and most people were still on dialup or at best a home DSL connection. 1999 held all kinds of new opportunity to discover exciting things in a newly connected world...like finding an mp3 of a rare/hard-to-find b-side or having your head melted by stumbling into entirely new genres of music.
For me, listening to audio streams on shoutcast/icecast and direct from a few misc sources was a crucial part of that exploration/growth/discovery.
The "Noise Rotation" (sometimes fondly called the "noise rot") on WCSB was a 2-hour block of specialty programming that rotated its DJ/host each week. I have no idea when it started or ended, but I can say that it was alive at least from May 2000 through May 2002. I enjoyed listening to it because I could catch bands that I was familiar with (like Negativland and Foetus, for example) followed by a 40 minute wall of droney noise and shit that I'd never heard of. The freeform nature of the programming was really inspiring, and completely different from the college station I volunteered with. It seemed like they could get away with anything!
These days, many radio stations keep archives of their shows, and you can listen to them on your own time. Hell, some shows are just podcasts now that happen to also be broadcast on the FM band and the internet at certain times. But back then, storing and indexing and making a nonstop rotating archive of material was rarely feasible. Also, storage was expensive (a 20GB hard drive might have set you back about $120). For me, I just wanted to listen to shows on my own schedule...so I began recording.
I'm sure that I had a cron job set up to just begin recording the cbr stream and dump the output to a file (probably using wget or curl). Due to the somewhat chaotic nature of college/community radio and computer time drift, I started recording a few minutes before and let it run a few minutes late. Even then, I'm sure that I missed content.
I blogged about the Noise Rotation back in early 2001 and hadn't thought much about it since. At that time, I was excited to have edited down 24 hours of material by trimming and removing ads/commercials. I'd listen to a show every few years, but mostly since access to everything is so plentiful and the quality is greater, I didn't revisit it much.
I'm pretty sure that I stopped recording the show in May 2002 when we moved back to Oregon and just never set it back up.
This year I found the recordings again and decided that enough time had passed that others might now find them useful/fun/important/historical. I spent many many nights trimming the rest of the episodes and normalizing/preparing them for upload. I decided that in the interest of historical preservation and to expedite editing, that I would not remove any host chatter and I would leave all the station IDs and event announcements and disclaimers in place. I removed dead air a few times though. :)
In the process of working the recordings, there were a few interesting times...like the time where the prior (blues?) show just continued because the Noise Rot person didn't ever show up. I'm pretty sure daylight savings bit me on a couple of the recordings, and one of the hosts just talked more than he played music. I think the station went off-air at least one time as well.
The final result wasn't that much data by today's standards (only about 1.6GB), but I couldn't imagine creating each item by hand using the web interface on archive.org. Luckily, tools exist to help with this stuff, so I hacked up a script to help automate things. The first attempt at upload/archive failed after about 3 or 4 files due to automated spam throtting...but a quick email to a support admin at archive.org fixed things up and I was able to complete the process the next day.
I ended up having 78 usable recordings, which comprise about 157 hours of noise. Click here to enjoy or just click the logo above.
Tue Oct 23 2018 18:56:44 GMT-0700 (Pacific Daylight Time)
Fri Jul 13 2018 21:30:30 GMT-0700 (Pacific Daylight Time)
We've been back in pdx for about 16 years and I had not yet made it out to the Witch's Castle in Forest Park.
So I decided to do that on Monday.
I can't really say that I recommend it on a road bike, but it was mostly fine. I couldn't ride about half the trail, due to the rocky/steep/slippery/muddy nature of a trail in a park...but it was quite ok early on a Monday morning. Only passed about two other people on the trail (one hiker, one jogger).
The trail dumps you out on Cornell, and then you get to take the lane and bomb maybe 2 miles downhill, including 2 tunnels. I didn't know the area that well, so I was fully clenched and expecting to fall/die.
My device said I hit 40MPH. It was insane and also fun as hell...
Tue Jan 02 2018 09:45:00 GMT-0800 (Pacific Standard Time)
A new year, and a new incarnation of noisybox.net.
I've been wanting to do a rewrite for quite some time...probably over a year. The maintenance on the pyramid site was just crufty and I let it stagnate. I was also never happy with the custom css junk I hacked togehter, and the site was never very mobile friendly. Well, frankly, I think it looked terrible on mobile...and I wanted to change that.
After a couple of weeks of evaluating several popular static site generators, including jekyll, hexo, and hugo, I decided that none of them would work well for my needs. Granted, they are all fine projects and people have done and continue to do a stellar job of supporting and enhancing them...but in most cases they quickly began to feel limiting. In some cases, they feel frameworky and require the user to learn their internals in order to get nonstandard things accomplished. In other cases, the tools assume (or mandate) a specific way of working, likely based on assumptions.
In the case of jekyll, I think I added a paging blogs plugin and performance just absolutely cratered. Like by a factor of 10x or something. I dunno about you, but waiting 20+ seconds to verify a small content change sure isn't fun -- it's debilitating.
I had some high hopes for hugo. It was enjoyable to get started with and the performance definitely has that Go wow factor! I was a very bummed at how difficult it was to make small layout/theme changes...they have a ways to go there!
I really should have done a better job of documenting the pros and cons of each one, since I really did do a pretty in-depth evaluation of them...but alas, weeks have passed now and I didn't write things down and it's no longer fresh...and I no longer feel like revisiting them. So many other things to do.
As a result, I decided to write my own static site generator: prepply.
Maybe it's stupid to do such a thing, and perhaps nobody else will ever see value in it, since there are thriving communities around the existing tools. And I'm fine with that! But I also host it on github in case somebody else can learn from it or help to improve it or whatever.
Like so many personal projects, it's merely a first effort and is hacky and without tests. Shameful, I know. But I'm still pretty happy with how it turned out and remain excited about the idea of static site generation. You know...because everything old is new again. Or something. Maybe we didn't learn from the past. :)
Tue Jun 27 2017 20:23:20 GMT-0700 (Pacific Daylight Time)
I wanted to build out my own sample pack by extracting audio from a set of video files, and built a tool to help do that. Maybe you're into this stuff and want to check it out too? Here's the source on github: