In this post, I’ll document my actual requirements for the project. I’m sure they’ll evolve as I go - requirements always do - but this should provide a good starting point. In no particular order:

1. The player will provide several buttons or keys - and each one will play a specific mp3 from the filesystem.
2. Pressing the same button a second time will stop the playback.
3. Pressing a different button while a track is playing will stop playback, and start playback of the new mp3 from the filesystem.
4. (Maybe) Some kind of switch will allow the user to change music sets associated with the buttons. This implies that it must be able to query the state of that switch after a power off…
5. The volume should be a controllable with a slider/switch that allows 2 or 3 preset volume levels. As with the previous item, this implies that the current state/position of the switch must be queryable after a power loss.
6. Turn on relatively quickly: a 2-year-old doesn’t understand turning something on, walking away for a minute or two, then returning to use it. So either extreme power management, a standby mode (instead of “off”), or very fast boot time. I see that with a class 10 SD card, boot time may be as low as 6-7 seconds - but with no feedback, this will still be long enough to confuse/frustrate a youngster?
7. Keys/buttons should light up when pressed? (Perhaps rev 2…)
8. Bright colors!
9. All controls must be suitable to a toddler’s hands - that is, larger than normal.
10. Durable - need to be able to be dragged around the room, dropped, kicked, sat/stood upon, and in general treated like any of his other toys.
11. Ideally battery-powered and portable, with a decent battery life. May not be practical given the requirement for fast boot.
12. I need to be able to update/swap out the music selections available without a lot of pain. Over WiFi ideally, or perhaps simply via a USB drive.
13. Internals must be securely protected from little fingers and incidental spills (he’s moving on to open cups these days…)
14. Usage stats/play counts! I’d love to know which music he prefers and which he doesn’t ever listen to.

Already I see this is going to be more complex than I thought, though not unmanageably so. I’m most concerned about quick startup time and battery life, since RPi doesn’t seem to have much in terms of suspend/sleep modes. I’m not quite sure about materials - perhaps wood? I have no 3d printer and am not interested in paying for custom molded parts.

Next post will be a bit of brainstorming around implementation, and posting results of any research I do between now and then.

Tonight my wife and I were talking about how our son would cycle through different songs on one of his toys until he reached the one he wanted. Unfortunately, each “song” is just a couple of measures long, so just as he gets into the groove the music ends.

This got us to wondering if there was a music player built for kids. Turns out there are a couple, but they’re overpriced and for the most part intended for someone a little older than our son (he’s 2.5). They’re basically oversized, limited-capacity MP3 players - not really what I want.

The more I looked, the more I realized what I wanted didn’t exist - and that perhaps here was the RPi project I was looking for. However, I’ve not touched a soldering iron for 20 years or more - and when I did last use one, it was at the most basic level.

A quick google search showed that nobody has attempted what I have in mind (or if they have, they’ve not written about it), so I popped into #raspberrypi on freenode and asked folks there if this would be a good beginner project. I pointed out that I could code anything needed, but was starting from zero when it came to knowing how to put the hardware together. I received a positive response and a couple of helpful hints - so I’m going to go for it.

I’ll be documenting here as I go, and will make available schematics. Related posts will be in category TMP for those who are interested in following.

Next post will describe my requirements.

My well of ideas seems to have run dry as well. Early on I made a list, but looking back over it - none of the items there are very interesting.

Anyone have thoughts for a good RPi project?

Our choices aren’t perfect, but the only message you actually send when you choose to abstain is that you don’t care about the future of the country you live in. The best thing you can do is vote for the candidate closest to what you want.

Consider also: this is not just an election for the presidency. Other seats are up as well, including your local government.

Go forth and vote. It’s the loudest voice you have if you want things to get better.

So I started looking around for alternatives, and came across Octopress. Now this, I thought to myself, looks promising - I can actually set up a git-based flow for writing and publishing, without having to muck around with Wordpress’s controls. Being able to use markdown in vim is also a huge plus - web-based editors annoy me, and I’m far less likely to accidentally lose a post by mis-clicking somewhere on the screen.

It doesn’t hurt that I like the default theme - clean and simple. I may tweak it in the future, but I’m comfortable leaving it as is for now.

With some minor tweaks I should be able to deploy using rsync. I may explore using the Chef SCM resource and a custom LWRP to regenerate the site (build artifacts) on the server itself, but for now rsync will be fine.

* That’s a referrer link, BTW - if you follow it and sign up for Linode services, I get discounts. And if you’re looking for a reliable VPS host, Linode offers some of the best service and most reasonable prices out there.