[Bug] Flashing firmware (from build #760) on movehub does not complete

[BertLindeman]

Describe the bug
Flashing this firmware onto the movehub does not complete

To reproduce
Steps to reproduce the behavior:

1. Download movehub firmware from build 760
2. Drop the zip file onto the "Flash the Firmware" button
3. Press and hold the green button on the movehub
4. pair that hub
5. wait until the progress bar is full
6. wait five minutes [edit] "or twenty minutes" more and drop your patience
7. press the green button and hold until the hub switches off.
8. pair the hub and look at the current firmware:

from pybricks import version
print(version)

('movehub', '3.0.0a12', 'v3.0.0a12-11-g11a7ecbc on 2020-12-18')

So the firmware is flashed. And reports the same as the technichub:

('movehub', '3.0.0a12', 'v3.0.0a12-11-g11a7ecbc on 2020-12-18')
----------
('technichub', '3.0.0a12', 'v3.0.0a12-11-g11a7ecbc on 2020-12-18')

Expected behavior
Would expect the hub to switch to pairing mode after the firmware flash is complete

Screenshots
None sofar

[BertLindeman]

Firmware in Build #756 does complete normaly.

Reports:

('movehub', '3.0.0a12', 'v3.0.0a12-8-gd590a50b on 2020-12-17')

[dlech]

I think sometimes the move hub just doesn't reboot correctly after flashing. I don't think there is anything we can do about this since the bootloader can't be changed.

[BertLindeman]

I think sometimes the move hub just doesn't reboot correctly after flashing. I don't think there is anything we can do about this since the bootloader can't be changed.

Thanks @dlech Can we tell users that after (say) 30 seconds after the "flashing firmware progress-bar" is at 100% it would be safe to hardstop the movehub and re-boot?

[dlech]

Yes, we plan on improving the firmware update process.

[laurensvalk]

I think sometimes the move hub just doesn't reboot correctly after flashing. I don't think there is anything we can do about this since the bootloader can't be changed.

I wonder what the statistics are.

I always flash with pybricksdev during development, and it never fails at 100% for me. It does fail occasionally, but only at a point somewhere halfway, but this is detectable at least.

So maybe we can find the best of both worlds somehow.

[BertLindeman]

On the other hand before this build 760 I never had this behaviour and with this firmware it consistently is"going-on" after 100% here. And the previous firmware I have does reboot at 100% firmware flash.

[dlech]

Maybe it was just a bad CI build. There is a newer one to try now.

[BertLindeman]

Maybe it was just a bad CI build. There is a newer one to try now.

Thanks, will do.

[BertLindeman]

Took movehub firmware from Build #763
No success, to be sure: twice.

Then flashed previous one and went back to
'movehub', '3.0.0a12', 'v3.0.0a12-8-gd590a50b on 2020-12-17')
without the problem.

[dlech]

OK, I'm seeing the same thing.

Console log shows:

io-6de156f3.js:111 Error: failed to get final response: ,,true
    at flash-firmware.ts:321
    at u (runtime.js:63)
    at Generator._invoke (runtime.js:293)
    at Generator.next (runtime.js:118)
    at p (redux-saga-core.esm.js:1157)
    at c (redux-saga-core.esm.js:1251)
    at t (redux-saga-core.esm.js:698)
    at c (redux-saga-core.esm.js:1251)
b @ io-6de156f3.js:111
b @ redux-saga-core.esm.js:1003
s @ redux-saga-core.esm.js:813
t.cont @ redux-saga-core.esm.js:828
p @ redux-saga-core.esm.js:1178
c @ redux-saga-core.esm.js:1251
n @ io-6de156f3.js:177
c @ redux-saga-core.esm.js:1251
b @ redux-saga-core.esm.js:1012
s @ redux-saga-core.esm.js:813
t.cont @ redux-saga-core.esm.js:828
b @ redux-saga-core.esm.js:1012
s @ redux-saga-core.esm.js:813
t.cont @ redux-saga-core.esm.js:828
b @ redux-saga-core.esm.js:1012
s @ redux-saga-core.esm.js:813
t.cont @ redux-saga-core.esm.js:828
p @ redux-saga-core.esm.js:1178
c @ redux-saga-core.esm.js:1251
t @ redux-saga-core.esm.js:698
c @ redux-saga-core.esm.js:1251
Promise.then (async)
T @ redux-saga-core.esm.js:395
O.<computed> @ redux-saga-core.esm.js:527
(anonymous) @ redux-saga-core.esm.js:1203
m @ redux-saga-core.esm.js:1271
(anonymous) @ redux-saga-core.esm.js:721
O.<computed> @ redux-saga-core.esm.js:716
(anonymous) @ redux-saga-core.esm.js:1203
m @ redux-saga-core.esm.js:1271
p @ redux-saga-core.esm.js:1161
c @ redux-saga-core.esm.js:1251
n @ io-6de156f3.js:165
c @ redux-saga-core.esm.js:1251
t @ redux-saga-core.esm.js:698
c @ redux-saga-core.esm.js:1251
l @ redux-saga-core.esm.js:503
e.put @ redux-saga-core.esm.js:339
L.e.put @ redux-saga-core.esm.js:371
(anonymous) @ redux-saga-core.esm.js:1412
dispatch @ redux.js:638
(anonymous) @ io-6de156f3.js:133
(anonymous) @ redux-saga-core.esm.js:472
d @ redux-saga-core.esm.js:31
b @ redux-saga-core.esm.js:87
m @ redux-saga-core.esm.js:58
O.<computed> @ redux-saga-core.esm.js:584
(anonymous) @ redux-saga-core.esm.js:1203
m @ redux-saga-core.esm.js:1271
p @ redux-saga-core.esm.js:1161
c @ redux-saga-core.esm.js:1251
l @ redux-saga-core.esm.js:503
put @ redux-saga-core.esm.js:173
(anonymous) @ redux-saga-core.esm.js:264
t @ lwp3-bootloader-ble.ts:133
Show 24 more frames
io-6de156f3.js:112 The above error occurred in task vr
    created by takeEvery(flashFirmware.action.flashFirmware, vr)
    created by bo
    created by bo
Tasks cancelled due to error:
takeEvery(flashFirmware.action.flashFirmware, vr)

So apparently it is timing out waiting for the final response.