New development firmware updater 1.15.2.b1

[Xykon]

I have released a new development firmware updater 1.15.2.b1

Windows download
MacOS download
Linux downloads

This version fixes a number of issues discovered in the previous release:

CORE:

• Fixes a bug that would cause the firmware updater to get stuck at a lower baudrate than requested

CLI:

• Improved progress output when using flash, copy and erase_fs commands

GUI:

• The firmware updater should now remember the last file flashed (if it still exists)

• The Pybytes window should (hopefully) no longer appear unexpectedly

• When using an Expansion Board 3.0 or Shield (Pyscan etc.), the firmware updater will now make 3 attempts to flash the board before throwing an error. This should mitigate some USB issues.

As usual please let me know if you encounter any issues.

[Xykon]

@sympatron Can you try this link please? https://xykon.pycom.nl/s/M8braq020b9gInb

Internet and China don't work well together...

There isn't really documentation for the new format yet because it's not yet been officially released. If you do manage to download the firmware updater you'd be the first person outside Pycom to actually use it. And as I said before it's still experimental and I'm aware of some issues with it such as the otadata partition definition missing, so please don't try to modify the script2 file in a firmware package.

I'm still in China this week and things tend to get more and more hectic towards the end of the trip so I likely won't be able to do any work on the firmware updater until I'm back in Eindhoven.

[Sympatron]

@xykon
I cannot access the files. OwnCloud wants me to sign in...

Is the new script format documented somewhere? I just found this documentation of the old format.

[Xykon]

@Sympatron Sorry it's been quite a while since I last worked on the firmware updater. I've been taking on new responsibilities at Pycom and we're still recruiting new developers.

I built a Windows version from the latest sources that you can download here but this has not yet been properly tested. Specifically, this version added support for a new format of the script file to support the changes we made to the partition layout in the development / release candidate firmware. However this is not yet 100% complete and may lead to unexpected results in certain circumstances.

This version does remember the last COM port being used so maybe you'll still find it more useful than the official development release.

I'm currently in China visiting our manufacturer but I'll try to look at the "indefinite" hanging issue when the COM port is blocked after I return to the Netherlands at the end of next week. This seems to be a "Windows only" issue and since most of the development and testing is done on MacOS I didn't notice that before. I've also noticed that there are some issues when you have a really slow internet connection.

[Paul Thornton]

@sympatron Thanks for the clarification.

[Sympatron]

@paul-thornton I don't think it does hang indefinietly (I never waited long enough), but very long considering that it could detect immediately that the port cannot be opened and abort with an error message.
Or at least retry very often while waiting so that it can begin when the port is free again, which it does not as far as I can see.
The timeout is simply far too long. Either the port can be opened in 1-2s or it has to be considered a failed attempt and either aborted or retried.

[Paul Thornton]

@sympatron said in New development firmware updater 1.15.2.b1:

@xykon
And the handling of failed connection attempts is not very good. If the serial port cannot be opened the program hangs without an error message or an option to abort or retry. This is annoying when you often reflash the firmware and forgot to close the port in another application.

I tried this and after a while it timed out. I've not been able to replicate it just hanging there indefinitely on windows 10 at least could you share some more info about your setup?

[Sympatron]

@xykon
Also the updater does not release the serial port on the result page. Only after you click on "Restart".
And the handling of failed connection attempts is not very good. If the serial port cannot be opened the program hangs without an error message or an option to abort or retry. This is annoying when you often reflash the firmware and forgot to close the port in another application.

[Sympatron]

@xykon
When will there be a new version? I assume the change would be rather trivial...

[Xykon]

@sympatron Sorry I misunderstood this previously. I interpreted it as the serial port being reset to the default speed value (which was actually true for the PIC based boards).

I'll add this in the next release.

[Sympatron]

The serial port is still not saved.