I suppose pi could eat into some of the arduino projects currently being done. Though arduino is much simpler to interface with various sensors - thanks to shields, once a few 'shields' for pi show up, the lure of powerful hardware and full blown OS will move many to pi.
I agree; however a particularly alluring component to an arduino is driving it with a 9V battery and throwing it up into the false ceiling and letting it do its thing. (Or whatever your use case may be.)
I think in the short term there will be a lot of work done very quickly to make Pi & Arduino talk to each other - the Pi's power and the Arduino compatibility are a great combination.
I agree that in the longer term the Arduino processor board isn't necessary. It shouldn't be too hard to make a cheap dongle to connect Arduino shields to the GPIO pins on the Pi, with some attendant software to drive them.
If I could make one weatherproof, it would be neat to use one outside in my back yard to take some weather information from the internet over wifi and then control some relays that turn some water spigots and save a ton of water for my garden.
In my head whenever I think about a 60 year old version of me in the year 2040, I've sort of burned out on writing code but still bust some out whenever I can use it to save water. And the kids will think I am lame for writing code that saves a few gallons, but we'll be right on the brink of that starting to matter a lot.
Also it would be cool to use the same machinery to drive those little windows and fans to make a greenhouse stay perfectly the same temperature during the day.
This is probably total overkill for one of these devices. But I could code the perl up in an hour or less and it would just work, and having something exist and just work is so much more satisfying than "this thing I keep saying I will do some year".
Another more romantic, less useful idea is that a raspberry pi mounted outside sampling the sky quality and scraping a few astronomical forecasts could come up with a composite of "this really is a night you should go outside for an hour before bed" - it would send a txt message to me at a certain time (or even be allowed to wake me up! [for a meteor shower or solar flare report). It seems I can write code and work on house chores 7 days a week and never remember nary a bit of it, but if I spend just an hour gazing at constellations and trying to figure it all out on my own, I remember that slice for a good long while and I dont feel like I missed out on something as much; this would be like a watchdog timer for my distracted life.
I think using XBMC will be huge. Cheap, low power, and simply point it at a computer/external HDD with a media collection to play all available media.
Also, using it as a gaming machine (see the Quake III video on Raspberry Pi's site) and game emulators for older console games.
I would love to see different sensors being attached for cheap, open-source home automation projects (home heating, turning off lights when no-one is present, monitoring home energy consumption, talking to mobile devices etc.)
Stand alone network backup device. I plan on hooking a low power external USB drive, and using something like rsync over the network so that backups are isolated from the computer I'm backing up. There's already several techniques documented for using rsync combined with "cp -al" to create daily/weekly/monthly snapshots using incremental space; combine this with an appropriate "command=" parameter in ssh's authorized_keys file, and you can have a backup that can run automatically, and that can't be destroyed by malware on the host you are backing up.
I no longer allow myself to use backup solutions I invent. And that's just a calling to my own confidence in my code, but I have lost data that I liked because I did a home brew backup. It's all that logic that makes sure it's still working correctly over time. There are so many devices available for very low cost that provide this exact solution. For me, it's one of those things where it's dangerous to enjoy inventing because I am not a master tool maker.
Hope you're thinking of a self-contained LCD/touchscreen off the shelf.
I've been trying to ascertain how to hook up a small LCD and capacitive controller over the LVDS/DSI port and I'm not getting a lot of information. Asking on the forum isn't getting much data.
Is there SPI or I2C on the flex connector? Nobody knows. How about power for the LCD backlight? Oh, the 5V rail is limited to 1 amp. How about the DSI driver? Broadcom won't talk.
You start pressing for details about this and then the stance swivels to "Hey! This is an educational computer! You shouldn't be working on wacky applications like this!" I don't see a lot of people proposing educational applications for the Rpi so far. Just people that want a cheaper Pandaboard.
From what I understand, the idea is that the first few batches will go to hackers so they (we) can get to work stress-testing and creating with these. After we're done, they'll sell the next batch to the education market hopefully with the new tools/accessories that the hacker community had come up with.
Can't create much when we can't see schematics and ascertain if the I/O ports carry what we need. And if there was a deficiency, they've publicly said they won't respin boards.
They do plan (and will have) IO extensions ala Arduino. I can't remember what they call the daughterboard they're releasing.
Mainly what they want is software and cases. It's meant to teach kids computer programming and software basics outside of the Windows and OSX "don't worry about it" mindset they're currently growing up in. Robotics and extraneous hardware design is outside of their current scope.
If you're thinking of the "gertboard" that was announced, that's a breakout board that is connected to the Rpi over a SPI line. So it's not turning GPIO lines off the processor into usable connections, it's just an I/O extender. You could be using that exact same board with an Arduino today and bit-bang your SPI commands.
Oh, and a pocket voip p2p wifi mesh phone would be nice.
Though the latency might be crap over too many hops and I am not entirely sure how well it would fare in really huge city-wide networks of the things. :)
