Any Ethernet solution for the Arduino will use the hardware SPI pins. This causes a problem with the standard ShiftBrite shields, as they expect to use those pins as well. You can use an Arduino Ethernet, or an Ethernet Shield, and use software SPI on some other pins for the ShiftBrites.
Alternately, you could put a much cheaper/smaller Arduino in the lamp post, like an Arduweeny or Arduino Pro Mini. This is a more soldering-based approach. I'd use the CAT5 to carry RS485 on one pair, using SN65176 chips on both ends. You can use an Arduino in the house, or a PC with a serial/usb/RS485 converter to control the remote Arduino. This is more work, but potentially cheaper than an Arduino Ethernet with PoE.
There is no hidden secret or trick to writing successful code from scratch. It is all hard work and research. You don't learn how to build a house by walking up to a construction crew, saying "Hey, I'm building a house too!" and asking them how to hold a hammer while they're working, and you don't own a hammer.
That's correct. You can string the shields together with a few wires, connect the SDA and SCL in parallel. I don't have a Teensy to test, so there is no Centipede Library for it...but you should be able to get it working quickly, the library is merely convenient.
A chess board is also the perfect layout for a matrix arrangement...you may not really need that many channels. The reed switches could easily be read with 16 pins.
An external, battery-backed RTC of some sort would be a good idea, though extreme accuracy isn't needed. The main reason is that if the power drops out for a couple minutes or someone trips over a power cord, the Arduino will lose track of whatever time has passed. 21 days is a long time to assume that something will have 100% reliable power, especially when so many lives are on the line
Forgive me if there was significant discussion about this last year, or if there was another thread I should be reading. I wasn't able to locate discussion about this via search.
I know this was discussed in reference to the new USB interface on the UNO as well. The upshot of that discussion was that no one could make an Arduino UNO clone, put the same ATMega8u2 firmware on the USB interface chip, and use Arduino's VID/PID. They would need to use their own VID/PID, lose seamless integration (make user install several files specific to the device, rather than included with Arduino download).
I'm just checking to make sure this holds true with the Leonardo as well. It would be nice if the Leonardo bootloading code was perceived as a virtual Arduino device that resides within the single ATMega32u4 chip, and could be used by anyone. However, this seems unlikely based on the old Arduino UNO controversy.
If we can't make our own devices appear as an Arduino Leonardo, I believe I'll stick with a two-chip solution with the new cheaper FT230X device, which I have verified works well with a Diecimila type setup, especially at 3.3V.
I think they've had that deal since forever. It's competitive if you actually need a large PCB, but they'll charge you the same if you only need a 1x2 inch board. For smaller boards, Seeedstudio and Iteadstudio are still unapproachable in value.
Surprised how many people don't know about nibblers.
I would carefully mark the cutout I need, drill a hole somewhere inside the area, then nibble out to the edge and around the perimeter. Very minimal filing or sanding after that, if any. You can make a nice square hole this way if you need it.
I made a network-connected, dimmable neon sign controller for company once, using an Arduino and Ethernet Shield. You should be able to find neon ballasts that have an extra set of inputs for dimming; just make sure the signal is compatible with the Arduino. Also, the standard PWM frequency for the Arduino won't work correctly. You need to lower the PWM frequency to something closer to 100Hz; then the neon dims nice and smooth with no humming (at least with the ballasts I was using, and I don't remember the name of them).