newRPL - can't restore to 39gs

[Claudio L.]

Just a quick followup: I was able to reproduce some timeout problems that happen randomly (unfortunately not often enough on my hardware to diagnose consistently). So I'm working on an improved communications protocol that can retry packets so we can overcome this issue. The original design assumed these failures were so rare that it was OK to just hang up and have the user retry. Unfortunately that won't work if you are trying to update the firmware, it's actually catastrophic. I just need time... wish I had more of it.

[franco51]

Hi, it's a pity your hardware is not reproducing the issue consistently.

Any help I can offer to resolve it is pretty limited but as my hardware fails every time, I would gladly undertake some testing when you are in a position to release something.

BR
Frank

[Claudio L.]

(04-05-2019 08:56 PM)franco51 Wrote:  Hi, it's a pity your hardware is not reproducing the issue consistently.

Any help I can offer to resolve it is pretty limited but as my hardware fails every time, I would gladly undertake some testing when you are in a position to release something.

BR
Frank

No need, I know why your hardware fails. Actually is not your hardware but a software issue.
Windows enforces a fixed packet size for HID USB transmissions. Other OSes don't, and I guess at some point, I got fed up with the slow speed and increased the packet size. Works well in Linux but completely breaks in Windows, as Windows plainly rejects the packets.
I could just change one line of code, recompile and publish to solve your problem... except now nothing works until I finish the new protocol. But rest assured it will work on Windows, just slower than it could be.

[franco51]

Understood. Many thanks for including the Windows OS!
BR
Frank