Going it Solo

I’ve been seeing the original billboards of this ad, with “Undies On. Game On. Solo” on them. I think mine’s better:

Undies Off. Game On. Solo.
Undies Off. Game On. Solo.

Undies Off. Game On. Solo.

Going it Solo

BlahBlahBleh – The Simplest Youtube Radio

BlahBlahBlehI’ve written before about how cool a simple youtube based “radio” site would be. What I mean by that is something which does the music programming for you, so you don’t have to choose each song, but instead can just kick it off and leave it running.

Well, I’ve built it and it works wonderfully! It’s called BlahBlahBleh,and it’s at http://www.blahblahbleh.com.

I’ve been threatening to build this for ages, so my family and friends might be glad that this is one less idea that I’ll bog on about at great length. I personally much prefer a realised idea to idea crack.

Just in case it’s not immediately obvious, it works as so: you go to the site, type in a search (a band or song name or combination works best), press “Go!”, and a clip from Youtube matching the search will play. When that clip is done, another clip will play automatic, and so on indefinitely. If you meet any songs you don’t want to hear, just press “skip” and you’ll go to the next one. At any time you can see links to a list of next songs, which is not the next few songs, but is the list from which the next song will be randomly chosen. You also have links to up to the last 3 songs you already heard. Of course, you can also put in a new search to restart the process.

How does this work? Well, the search is sent to the youtube API, and youtube sends back a list of matching clips. I filter this to keep only clips which have the category Music, which have more than a minimum number of views (iirc it’s 10,000 at the time of this writing), and order them by rating (like/dislike score). From this list I choose randomly from the top 5, and play that.,

Then, when it’s time to choose the next song, I ask the Youtube API for the list of related videos for the video just watched, then I filter it again as above (Music, minimum views, sort by rating), and again choose randomly from the top 5. Actually getting the list of top 5 to choose from is done when the previous video starts playing (in the GET handler), and the random choice is made once that video finishes (processed in a POST handler). That give me the chance to present the next videos list to the user while the previous video is playing.

All next songs are chosen as just described. So, BlahBlahBleh will eventually wander away from what you chose initially, sometimes into something annoying (so just search again), sometimes into something new and great, which is always fun!

BlahBlahBleh will be there indefinitely. It’s hosted on Google AppEngine, and currently costs me nothing. If the load ever builds to where I have to start paying for it, I might consider carrying some google ads to defray the cost, or solicit donations, or just wear the cost, depending on how I feel and how much money we’re talking about! Even when Google AppEngine apps start costing money, they’re very cheap, so I’m not too worried.

But it’s not about money. It’s something I’ve built to use myself, and everyone else is also welcome to use it. Also, it’s based in my philosophy that we shouldn’t hoard information. Instead, make it all free, and draw from the commons. BlahBlahBleh is a great example of something which couldn’t exist in a world of purely private silos of music.

So give it a try, and if you like it, then feel free to use it as much as you like. Have fun! Feel free to comment here if you have suggestions, bugs, whatever.

BlahBlahBleh – The Simplest Youtube Radio


I’ve started the earliest version of a digital commonplace book. It’s simply a blog with google’s blogger, into which I post a copy + link of anything I read of significance. To make the posting experience smooth, I use the Google Toolbar in Firefox, which has a button for posting to blogger.

I chose blogger because the terms of service say you can have as many posts as you like, of whatever length you like, and there’s also a comprehensive API for any future custom code of mine to use.

Is it working out? Well, my idea was that google would index the blog entirely and in realtime (as it seems to do with this blog). However, that isn’t the case so far. A few random articles are indexed, the rest not. Perhaps I need to wait longer?

Or, perhaps because nothing except this blog links to the digital commonplace blog, the Googlebot all but ignores it? This could continue to be the case, in which case I could “fix” it by encouraging links from other sites. Really though, this isn’t about being linked to; it’s for my consumption, it doesn’t need to interact publicly with other information. So, it might be that using a public blog + google search is a bust.

In that case, plan B is to create my own knowledge base on a machine at home, and point an open source search engine at it. That’ll require custom code, but not all that much (a bit of python likely). It’ll have the benefit that I can make it private, and thus be able to put anything in, whole commercial books, whatever. I probably have to build this anyway with plan A, purely to hold closed commercial sources that I can’t publish.

But I’ll give the blog approach a little longer.

Also, in parallel I’ve turned on Google Web History (google.com/history). If you use Google Toolbar, you can get it to log all your web activity to google, which they make available to you in the Web History tool, privately. It’s fully searchable, a great service actually. Anyone prepared to trust google with their surfing data will benefit from this tool. Check it out.

However, for this project it’s not enough. I can’t add other, non website info to it (except perhaps by publishing that info on a blog then navigating to it in firefox). Also, there’s no API; I can’t do anything programmatic with it. So that’s a bit of a non starter. A shame.

As to auto-memeticism, I’ve noticed that I’m automatically censoring what I add to the commonplace book. In particular, I’m leaving out sites where I think the information is suspect. I just noticed myself doing this a little while ago, and the thought was “I don’t really want to pollute my memory with crap like that”.

Without really thinking about it, I’ve been operating in the knowledge that my present self is managing a mental environment for my future self, and that my present self can make decisions knowing that my future self will be unaware of them; present self is in control of what is remembered.

I’m manipulating the future me!

This is a strange rabbit hole I’m going down. I’m a huge fan of the movie Memento, but do I want to live like that? Come to think of it, that might actually be the genesis of this idea. #freakingmyselfout .