As an example in mucomp, there is a certain part of the code that deals with the audio player. This is inherently a resource that should only be handled by one thread at a time. Clojure comes with a very good abstraction for exactly this problem, agents. An agent is simply some state, that is manipulated by only thread. Using an agent is done by through sending a function to the agent that will take the old state and return a new state. So with that one gets everything that is needed to write an audio player: serialized access and safely mutable state.

The only bad thing about agents is that if one forgets to return something from a function that run on the agent, then the new state of the agent will be nil. After being bit by this two times I decided that enough was enough. One of the very nice things about languages in the lisp family, is that one can mold your own abstractions to make code better (easier to read and with less bugs in this case).

The following macro creates a new way to define functions. Functions defined in this way will check for nil on return, and return the old state instead. The only change that is needed in code is to use defa instead of defn

The new album is written and just needs the funding to get recording. One great thing about Lord K, the main man behind the band, is that he cares a lot about the sound quality. The music sounds so much better in flac than in mp3 and proper speakers / headphones of course helps a lot *hint* Slayer and Metallica.

The band has been searching for a new record label, but has decided to try a donation experiment to see if they can get the money needed to do the recording through generous metal heads instead of a record label. I suggested that he tried kickstarter, but I guess he just wanted a low-tech email solution.

It is going to be extremely exciting how this works out. The internet and p2p networks has to a large degree in the mainstream media been associated with destroying the music industry. This is our chance to show that it can also be used to create music.

When you look at kickstarter and flattr, they are attacking the same problem, funding, at different angles. Kickstarter tries to get all the funding up front, while flattr is more of a tip jar model for something already produced. So in a way, they are complementary. I would argue that they both work best if the content is placed into the public one way or another. And that is where I think there is a huge potential.

Something like flattr creates an alternative to paywalls and an alternative to ads. And that is something I would very much like to see.

Failure 1 (I’m feeling lucky):

Google has a very nice feature called “I’m feeling lucky” that will direct you to the first result. If I could specify my queries good enough, I could rely on that, and not have to parse google to get the url. It’s very simple, you just add &btnI at the end of your query and google will redirect. Sadly it works fine most of the time, but sometimes it just fails to redirect you. I couldn’t find any patterns to this randomness and a “works sometimes solution” is not a good one

Failure 2 (google ajax):

I then found out that google has a seemingly very nice api that lets you do queries and get JSON back. JSON is easy to work with and it also allows one to go through several results, in case google doesn’t return the right one as the first result. After a bit poking around I found out that google ajax randomly returns different results from the normal google. It’s like using Yahoo instead of google. A bit of poking around returns the following 2 year old bug report. Furthermore the TOS directly forbids using the API for this kind of activity. Oh well, it didn’t work anyway.

Failure 3 (parsing google results directly):

After two bitter defeats I thought screw it, I’ll just parse the damn google result pages, how hard can it be? At least I know that it gives me the right results. So I did that, coded everything up and checked that things was working. Then let it loose on my collection (2×275 requests) and around the middle it stopped working. I poked a bit around, and found out that google has identified my program as a bad boy and decided to spank it by returning a “Please identify yourself as a human” page back instead of the normal google result page.

As a side note, after 3 bitter defeats I was ready to jump ship and try bing or Yahoo. That was a quick detour though, as none of them where up for the challenge of returning good results.

A quick google revealed that MPlayer has several tricks (audio filters) that might potentially work: volume, volnorm, pan, hrtf. I quickly discarded volume and volnorm since I don’t want to just boost the sound, I want it to distribute the channels properly. hrtf seemed like a good simple choice, since pan looked very complex. Sadly in the middle of Harry Potter I had to turn it off because it was making lots of clipping of the sound. So I was left with pan. It took a while to get a good default, but a bit of googling around revealed one with a decent default. I first just tried turning sub + center up to one but in one or the other movie introduced the dreaded clipping. So I had to keep it down a bit while still retaining decent boost of center and sub. After an afternoon of testing I came to the following “magic” formula:

-af pan=2:0.5:0:0:0.5:0.33:0:0:0.33:0.5:0.5:0.5:0.5

Please do note that one needs to add a -channels 6 in order for mplayer to decode all 6 channels so that it can mix it down to two. One can read more about the pan filter here.

• 40 years after the invention of relational databases we are still manually defining indexes
• 40 years after the invention of Unix, the scheduler in Android (= Linux) still does a terrible job at scheduling the tasks that really depend on it (games and audio)

We need more of this in software. It’s so easy, and has become so accepted, to ship a half-baked product and just slap a beta sign on it nowadays. Release early, release often is great, but there is also a time for making fixing all the small annoying bugs. If they plan on fixing this bug, then this is a step in the right direction.

deb http://dl.google.com/linux/deb/ stable main

Then you can install Chrome using the package manager and it will keep itself updated. Sadly the browser is quite alpha (no flash, random errors, no tabs saving on exit) but they update it every week and it runs insanely fast compared to firefox.

sudo vol_id -u /dev/sda4

Next step is to fix up the location of the resume partition. Open /etc/initramfs-tools/conf.d/resume and change the UUID to the same as you just put into fstab. Then run:

sudo update-initramfs -u

And you should be running again