3 ms·

I'm so phenomenally excited by this. I bought a ReMarkable 2 excited about the screen, without fully understanding the degree to which the general-purpose-devi

I'm so phenomenally excited by this. I bought a ReMarkable 2 excited about the screen, without fully understanding the degree to which the general-purpose-device functionality had been hobbled. (No amount of we-let-you-ssh-into-it will get you bluetooth keyboard support when the wifi+bt chip has bluetooth hardware-disabled in solder. Caveat emptor.) Of course it's early days yet for the PineNote, but it's moving in the direction of more general-purpose functionality, not less. And that can only be a good thing.

csdvrx 5y ago
> the wifi+bt chip has bluetooth hardware-disabled in solder Just... why??
- josephcsible 5y ago
  Attack surface reduction, maybe? Bluetooth vulnerabilities at really low levels in the stack aren't that uncommon. And if it's just disabled in solder and not on the silicon die, wouldn't it be practical to re-enable it by taking it apart if you really wanted?
  - cge 5y ago
    Attack surface reduction? On a device with unencrypted storage, the root password displayed in the about screens, no verified boot, no password by default (with password protection that amounts to a lock screen), and unprotected USB access to all user files by default? Where there are four exposed pins on the exterior that allow mounting of the root filesystem as USB mass storage regardless of password protection (https://github.com/ddvk/remarkable2-recovery https://github.com/ddvk/remarkable2-recovery), and the root password is stored in plain text?
    - josephcsible 5y ago
      Doesn't everything you listed require physical possession of the device to do anything with, whereas Bluetooth can be exploited just by walking near a malicious device?
      - tata71 5y ago
        RCEs in your kernel mean verified boot is a baseline desire.
- detaro 5y ago
  Potentially certification. Or it's just missing support components that would be needed for BT to work, but weren't put in because BT wasn't part of the design, and thus disabled since it wouldn't work either way.
- MartijnBraam 5y ago
  I guess they just didn't hook up the uart connection for the bluetooth side of wifi/bt chip since they were not using it?
paulcole 5y ago
Is the Remarkable2 ever advertised as having Bluetooth keyboard support? I bought mine based on the advertising which pretty clearly portrayed it as an e-ink version of a paper notebook and nothing more.
cge 5y ago
Much of the Remarkable 2's design shows this level of suspicious hardware design choices that cripple extra functionality, functionality that wasn't advertised, but would have been possible had work not been done to prevent it. Beyond just bluetooth keyboards, while USB keyboards work on the Remarkable 1, support was apparently intentionally broken on the Remarkable 2. While otherwise completely supporting USB OTG, the USB port has been modified, apparently at a hardware level, to not provide power to devices. This appears to be related to the addition on the Remarkable 2 of an additional, physically-adjacent USB-protocol interface with a proprietary connector, for potential future accessories, which does provide power. For those who have made adapters, this proprietary interface supports USB keyboards quite readily: apparently, the possibility of developing a keyboard accessory in the future was sufficient motivation for the company to break standard keyboard support at a hardware level. While I have found the Remarkable enormously useful, that has largely been in spite of the company. Had I researched it more closely before purchasing it, I would have realized that the seeming openness (SSH access, Linux, a development community and unofficial third-party software repository) appears to have been forced on an otherwise hostile company by the GPLv3. So it is promising to see the same style of device with a manufacturer that actually supports it being open and adaptable.
- squarefoot 5y ago
  > functionality that wasn't advertised, but would have been possible had work not been done to prevent it Not to defend them (I would choose the PineNote anyway) but I think the practice is common: crippling something that is very easy to get working in a current version of a product is how they can later use it as a bullet point to advertise the sale of the next version that contains the "latest development", "much wanted feature", etc. If we don't see the practice employed often, it's likely because most products are tight closed and the turned off features aren't easy to spot.

csdvrx 5y ago

> the wifi+bt chip has bluetooth hardware-disabled in solder Just... why??

josephcsible 5y ago
Attack surface reduction, maybe? Bluetooth vulnerabilities at really low levels in the stack aren't that uncommon. And if it's just disabled in solder and not on the silicon die, wouldn't it be practical to re-enable it by taking it apart if you really wanted?
- cge 5y ago
  Attack surface reduction? On a device with unencrypted storage, the root password displayed in the about screens, no verified boot, no password by default (with password protection that amounts to a lock screen), and unprotected USB access to all user files by default? Where there are four exposed pins on the exterior that allow mounting of the root filesystem as USB mass storage regardless of password protection (https://github.com/ddvk/remarkable2-recovery https://github.com/ddvk/remarkable2-recovery), and the root password is stored in plain text?
  - josephcsible 5y ago
    Doesn't everything you listed require physical possession of the device to do anything with, whereas Bluetooth can be exploited just by walking near a malicious device?
    - tata71 5y ago
      RCEs in your kernel mean verified boot is a baseline desire.
detaro 5y ago
Potentially certification. Or it's just missing support components that would be needed for BT to work, but weren't put in because BT wasn't part of the design, and thus disabled since it wouldn't work either way.
MartijnBraam 5y ago
I guess they just didn't hook up the uart connection for the bluetooth side of wifi/bt chip since they were not using it?

paulcole 5y ago

Is the Remarkable2 ever advertised as having Bluetooth keyboard support? I bought mine based on the advertising which pretty clearly portrayed it as an e-ink version of a paper notebook and nothing more.

cge 5y ago

Much of the Remarkable 2's design shows this level of suspicious hardware design choices that cripple extra functionality, functionality that wasn't advertised, but would have been possible had work not been done to prevent it. Beyond just bluetooth keyboards, while USB keyboards work on the Remarkable 1, support was apparently intentionally broken on the Remarkable 2. While otherwise completely supporting USB OTG, the USB port has been modified, apparently at a hardware level, to not provide power to devices. This appears to be related to the addition on the Remarkable 2 of an additional, physically-adjacent USB-protocol interface with a proprietary connector, for potential future accessories, which does provide power. For those who have made adapters, this proprietary interface supports USB keyboards quite readily: apparently, the possibility of developing a keyboard accessory in the future was sufficient motivation for the company to break standard keyboard support at a hardware level. While I have found the Remarkable enormously useful, that has largely been in spite of the company. Had I researched it more closely before purchasing it, I would have realized that the seeming openness (SSH access, Linux, a development community and unofficial third-party software repository) appears to have been forced on an otherwise hostile company by the GPLv3. So it is promising to see the same style of device with a manufacturer that actually supports it being open and adaptable.

squarefoot 5y ago
> functionality that wasn't advertised, but would have been possible had work not been done to prevent it Not to defend them (I would choose the PineNote anyway) but I think the practice is common: crippling something that is very easy to get working in a current version of a product is how they can later use it as a bullet point to advertise the sale of the next version that contains the "latest development", "much wanted feature", etc. If we don't see the practice employed often, it's likely because most products are tight closed and the turned off features aren't easy to spot.