Report Bugs Here

Moderator: Mmiscool

User avatar
By Electroguard
#55661 I have a very predictable and repeatable bug with V3.
The first time of trying to 'save' after wifi connection will invariably succeed.
But the second time of trying to 'save' will always cause an esp crash and reboot.
Retry the 'save' after allowing time for reconnection again and it will succeed, but edit and try to save changes again will cause another crash and reboot.
It happens every other time after connection!

I should add that this may be something to do with the fact that I have never been successful in doing a V3 Format Flash - which always pops up a window saying...

'esptool.exe has stopped working.'
'A problem caused the program to stop working correctly. Windows will close the program and notify you if a solution is available.'
The only option then is to click the big 'Close program' button and close everything down.


If I reboot the esp device back into reflash mode, then just go for a Firmware Flash without a Format, it succeeds ok. Everything then appears to work, except as mentioned above, where every other 'save' causes an esp disconnect and reconnect.

I should point out that this only applies to the V3 Flasher (and all versions which I've tried), whereas if I use the V2 A24 Flasher afterwards it Formats ok every time.

I should also point out that the V3 Format failure occurs with every device I've tried, including nodeMCU, but they can all be V3 Flashed ok.

Are these 2 problems related?
Is there any solution for them?
User avatar
By Mmiscool
#55666 I have not experienced this. How many KB is your program stored as a text file on your PC and what size models are you using?
User avatar
By Oldbod
#55668 Not at home to document this but i have a similar problem when i try to format black nodemcu 4Mb flash. Save works fine for me though, just format doesnt. Flashing a new version works fine (apart from buffering the write display to the end). Running on vista home premium.
User avatar
By Electroguard
#55669 The script file is a 6K .txt on computer which gets pasted into the edit window rather than saved on the device as a .bas file.

Mostly I've been developing on 4Mb ESP devices loaded with the 1Mb V3 firmware core for device compatability, but I get exactly the same symptoms when using 2Mb or 4Mb cores.

The voice script I've been working on recently does usually run ok if everything is absolutely to its liking, but it is a bit temperamental cos it's reading an interrupt from the busy signal of a voice chip to output words from a FIFO buffer (array) one after the other (just to speak the local IP address at the moment).

The thing is though, I don't think the problem is script specific... I've never been able to Format any device of any size with V3 - so that's before any script is even loaded. And all I've said applies equally to any previously unused virgin ESP devices, not just the various ones I've already been using.

It's not just the latest A49 Flasher either, I've had the same 'unable to Format' error with any previous V3 versions I've tried to use, but as I already said, V2 still works fine.

I'm wondering if it's because I'm using a virtual machine running on VMware, which is unbelievable transparent and stable, but presumably has to interface real-world device drivers to virtual com ports. So perhaps it's because of the V3 flashers initial checks to discover available com ports etc which is somehow giving it indigestion which V2 doesn't suffer from because it's not even looking.

The only way of proving it one way or the other would be with an option to bypass the automatic initial com port tests, but it's not worth the hassle if I'm the only one experiencing such problems. So perhaps I'm the only one using a virtual machine (and just seen his post, so pehaps Oldbod too).