Topic: Mycelium Entropy update - page 2. (Read 8987 times)
Could be a good time to give another update on ETA of this Entropy device.
looks good. hopefully its cheaper than trezor 
The crowd funding buy links should still be open, 40 USD:
http://mycelium.com/entropy
looks good. hopefully its cheaper than trezor
Nice to include Shamir's Secret Sharing! How do these print-outs look like?
Somewhat like this:
https://i.imgur.com/CGHWIcS.jpg
Looking good!
Nice to include Shamir's Secret Sharing! How do these print-outs look like?
Hi guys. Here is the update on the Entropy progress. The software is done, the hardware is finalized, the CE compliance is done (shown on the USB connectors in the image below), and the circuit boards are finished and ready for assembly. We were expecting them earlier, but because we changed the cases from plastic to metal, we ended up having to move chips around on the board to make sure they fit into the slimmer cases. As a result, the devices are much more durable, the metal case shielding makes them resistant to any EM interference, and it spares us from needing to go through extensive FCC and CE testing. (You can see the comparison of the old design to the new cases in the image links below).
Because we were not completely sure that the slightly redesigned boards would fit and align with the cases correctly, we had to only order a few sample cases so we could test them to see if they fit right. Since the circuit boards are at the assembly facility, we created an exact 3D printed replica of the board to test the fit (picture below). Now that we know that they fit ok, the cases are being ordered, and as soon as they get here, everything will get assembled, flashed, packaged, and shipped out. Sadly, and unexpectedly, the company making the metal cases has given us a 6 week estimate as their expected production and assembly time. So although we have everything else ready, we apparently miscalculated on the estimate for how long it would actually take to put the things together. In short, when we guessed way back in June how long this would take, we ended up guessing wrong, and we are extremely sorry about that. Yes, we are aware of the ridiculous trend of delays in this market, and we were really hoping that we wouldn't be another example (for a while we actually thought that maybe we could get them out even before October), but we are practically finished at this point, and just got surprised by the final step.
On the bright side, this does give us extra time to tighten up the software security even more. Actually, in a way, all our delays - from moving expected shipping to late October instead of earlier (after initial feedback led us to add many more entropy sources to protect against compromised hardware), to the metal cases slowing us down now (which we decided on after we received concerns about interference from nearby high voltage power lines) - have been due to the feedback and suggestions in security improvements from our community. In a way, this project became an open collaborative process, with a lot of discussion and feedback, and although it slowed things down a bit, it did increase the security of the devices considerably. We have increased the number of entropy sources, and added even more internal tests against outside interferences (such as fluctuations in temperatures, magnetic interference, timing, etc) to make sure the entropy sources are not compromised. We also added custom settings, such as switching between Bitcoin, Litecoin, and testnet, switching between compressed and uncompressed keys, and software key signatures for those who want to build their own firmware. With the delay, we will use this extra time to add even more customizable security settings, such as the ability for you to add your own random salt (such as Diceware) so you can generate verifiably random keys even without having to trust us and our hardware suppliers, as well as tools for users to obtain the raw random data which is generated by Mycelium Entropy every time it is plugged in, and to analyze it to verify that this seed random data contains more than enough entropy on the random seed to properly generate a unique private key, that the random seed data is unique, and there are no deterministic techniques used to generate it (that it is really an entropy-based white noise collected in the silicon of the Entropy device microcontroller).
When we are done with these things, they will probably be the most secure private key generators on the market (not counting your printer security...). To help with that, we are also now releasing the source code for the Entropy devices, which you can find here:
https://github.com/mycelium-com/entropy
Please feel free to review the code, and let us know if you have any concerns about anything, or if you want us to add something or even contribute your own code. For those looking to add altcoin support, now is your chance to figure out how as well.
* [Old device design](https://i.imgur.com/IVmisRt.jpg)
* [New device design](https://i.imgur.com/aLq7nbN.jpg)
* [New device in case](https://i.imgur.com/lKU8OM0.jpg)
* [3D printed circuit board to test fit](https://i.imgur.com/gwNqpso.jpg)
* [USB connectors w/ CE certification](https://i.imgur.com/UHlvZlN.jpg)
* [Engraved cases](https://i.imgur.com/6ySzZMz.jpg)
Because we were not completely sure that the slightly redesigned boards would fit and align with the cases correctly, we had to only order a few sample cases so we could test them to see if they fit right. Since the circuit boards are at the assembly facility, we created an exact 3D printed replica of the board to test the fit (picture below). Now that we know that they fit ok, the cases are being ordered, and as soon as they get here, everything will get assembled, flashed, packaged, and shipped out. Sadly, and unexpectedly, the company making the metal cases has given us a 6 week estimate as their expected production and assembly time. So although we have everything else ready, we apparently miscalculated on the estimate for how long it would actually take to put the things together. In short, when we guessed way back in June how long this would take, we ended up guessing wrong, and we are extremely sorry about that. Yes, we are aware of the ridiculous trend of delays in this market, and we were really hoping that we wouldn't be another example (for a while we actually thought that maybe we could get them out even before October), but we are practically finished at this point, and just got surprised by the final step.
On the bright side, this does give us extra time to tighten up the software security even more. Actually, in a way, all our delays - from moving expected shipping to late October instead of earlier (after initial feedback led us to add many more entropy sources to protect against compromised hardware), to the metal cases slowing us down now (which we decided on after we received concerns about interference from nearby high voltage power lines) - have been due to the feedback and suggestions in security improvements from our community. In a way, this project became an open collaborative process, with a lot of discussion and feedback, and although it slowed things down a bit, it did increase the security of the devices considerably. We have increased the number of entropy sources, and added even more internal tests against outside interferences (such as fluctuations in temperatures, magnetic interference, timing, etc) to make sure the entropy sources are not compromised. We also added custom settings, such as switching between Bitcoin, Litecoin, and testnet, switching between compressed and uncompressed keys, and software key signatures for those who want to build their own firmware. With the delay, we will use this extra time to add even more customizable security settings, such as the ability for you to add your own random salt (such as Diceware) so you can generate verifiably random keys even without having to trust us and our hardware suppliers, as well as tools for users to obtain the raw random data which is generated by Mycelium Entropy every time it is plugged in, and to analyze it to verify that this seed random data contains more than enough entropy on the random seed to properly generate a unique private key, that the random seed data is unique, and there are no deterministic techniques used to generate it (that it is really an entropy-based white noise collected in the silicon of the Entropy device microcontroller).
When we are done with these things, they will probably be the most secure private key generators on the market (not counting your printer security...). To help with that, we are also now releasing the source code for the Entropy devices, which you can find here:
https://github.com/mycelium-com/entropy
Please feel free to review the code, and let us know if you have any concerns about anything, or if you want us to add something or even contribute your own code. For those looking to add altcoin support, now is your chance to figure out how as well.
* [Old device design](https://i.imgur.com/IVmisRt.jpg)
* [New device design](https://i.imgur.com/aLq7nbN.jpg)
* [New device in case](https://i.imgur.com/lKU8OM0.jpg)
* [3D printed circuit board to test fit](https://i.imgur.com/gwNqpso.jpg)
* [USB connectors w/ CE certification](https://i.imgur.com/UHlvZlN.jpg)
* [Engraved cases](https://i.imgur.com/6ySzZMz.jpg)
Jump to: