The beginning...
It was December 18th and I was sitting at my computer, tinkering with the small OLED screens that had been sitting around unused for a month when I started to see red...Bitcoin took its heaviest hit in months and the charts visible out of the corner of my eye started looking more and more grim...Updates, rumors, and panic about China's new policy on Bitcoin were fly fast...I needed a distraction...
Over the next couple hours, I tried my best to ignore what was going on and came up with an idea...I was going to make myself a cold storage wallet with an LCD to read out QR codes...THAT should distract me for a while...A couple hours later I had this:
Nothing too special, but a success on the distraction front.
So, shortly after that I made this post on Reddit:
http://www.reddit.com/r/Bitcoin/comments/1t5xa9/knocked_out_a_quick_prototype_while_watching_the/
The number of responses and level of interest completely blew me away...I received a massive amount of feed back and suggestions about hardware, encryption, features, etc...
One week later...
Copied from the most recent Reddit post:
##############
Hello again guys, I was asked to keep giving updates on the hardware cold storage wallet I started working on about a week ago, so here is the first one. (I plan to do them weekly if I can.)
From now on I am calling this the "Crashwallet" since I first started working on it while watching the Dec 18th crash unfold. Since then I have been hard at work sifting through half a dozen different libraries I have found to handle QR code generation, reading up on encryption and storage options, and ordering some other components.
- QR codes - Despite trimming as much fat as I can from the libraries I have found that support version 3 codes (the minimum needed for bitcoin addresses) I have been unable to shrink one down to a level the Arduino Pro Mini I used for the original post can handle. (For the record, I am modifying the same library that is used on the Trezor.) This is actually good and bad...Good because it means I will need to upgrade the microcontroller I was using, allowing room for more features, but bad because I don't have any on hand at the moment.
- Encryption - Honestly, most of the time I have spent in the last week working on things has been try to tackle the QR libraries, so I have not gone to far in depth here. I have decided to do things a bit differently than in my first post however. Either I am going to use two SD card slots, one (easy to remove) that contains the encryption keys and the method for loading new unencrypted private keys, and a second (device must be disassembled to remove) on which the device will store the encrypted private keys...The other option will be a single SD card to load the private keys and encryption keys and use the microcontroller's internal storage for the encrypted private keys. This of course depends on reliability, and if I am not satisfied with the NVRAM's ability to retain data I will explore other methods.
- I have also ordered a few new components to experiment with. I mentioned in another post that donations I received paid for a pair of microSD breakout boards and a LiPo battery/charger from Adafruit (the only component source I know of accepting Bitcoin so far). Also from Adafruit, I ordered a couple different LCD displays. One, is a smaller monochrome dot-matrix LCD used in older Nokia phones which I am hoping will provide an extremely low power consumption option, but will probably not be in my final version...The other, which I am more likely to use, is the 1.8" color LCD that comes paired with a microSD board. This screen is rectangular, leaving room for a simple menu to still be visible while still displaying a QR code and since it includes a card slot, will simplify things considerably.
- Upgraded microcontroller...As I mentioned above, the Arduino just could not hack it (surprise surprise) so I also have a couple Teensy 3.1's on the way which are considerably more powerful and will allow for several other options to be added, namely an onscreen menu, onboard encryption, a prettier user interface, and much faster response time among other things.
The downside is, now I have to wait for things to get here in the mail...Unfortunately, I am on a US Army base in South Korea for another couple months and it is Christmas time, which means the mail runs PAINFULLY slow, so most of what I ordered probably will not be here for another week or two...Thankfully, any day now I should be receiving my Spark Cores which are very similar and should let me get started on the coding while I wait. Until then, I am going to keep streamlining the QR library, reading through some very promising encryption libraries I have found, and working on the code for the user interface.
#############
One day later (today)...
Today, I am looking into getting some graphic design done for me that will eventually become the loading screen for the LCD as well as a logo for the project for me to plaster all over the web and getting this blog setup so anyone interested can better follow my progress.
Apart from that, I have the user interface mostly planned out and just have to wait for the parts I ordered to arrive.