Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
Dmitry Kireev
@AutomationD
@igrr thank you
Ivan Grokhotkov
@igrr
basically when someone builds an executable on OS X, by default compiler will target the same OS X version. the option i mentioned is to make sure the binary will run on older versions as well
it's used by clang/gcc
Dmitry Kireev
@AutomationD
ah, that’s for gcc?
got it
thank you. mac specific.
Ivan Grokhotkov
@igrr
there's also an equivalent GCC/clang command line option, -mmacosx-version-min, but i find it easier to set MACOSX_DEPLOYMENT_TARGET once in the environment rather than making sure -mmacosx-version-min gets passed into each configure script correctly.
just my 2 cents, i remember people having issues when I built toolchain without this option
Dmitry Kireev
@AutomationD
@igrr appreciate it!
I like esp-open-sdk because it’s open… but it doesnt’ work on windows yet :(
doesn’t build..
Ivan Grokhotkov
@igrr
yeah, i saw pfalcon is reluctant to fix that, despite the fact that windows has about 80% user share in the esp8266 community
Dmitry Kireev
@AutomationD
yep :(
I’m trying to arrange build environment here https://github.com/kireevco/packages/tree/master/esp-open-sdk/windows
tried cygwin, mingw… there are issues… and testing cycle is pretty long .
(just complaining) :-D lol.
Ivan Grokhotkov
@igrr
yeah, i remember it took me a few nights to troubleshoot everything and build windows toolchain on my old laptop
last time i had to do this, i got myself a D4 instance from azure cloud
Dmitry Kireev
@AutomationD
I got myself a host on hetzner, and built a proxmox host
you did a great job, thanks for your help too :)
Ivan Grokhotkov
@igrr
ok, actually windows is down to 67% 
esptool-0.4.5-linux32.tar.gz: 1103 downloads
esptool-0.4.5-linux64.tar.gz: 5845 downloads
esptool-0.4.5-osx.tar.gz: 5592 downloads
esptool-0.4.5-win32.zip: 27600 downloads
Dmitry Kireev
@AutomationD
nice data
still, lots of windows users
Ivan Grokhotkov
@igrr
@luc-github what does the serial log show after (and during) restart?
Luc
@luc-github
@igrr I use swap to avoid serial garbage when I send a retart command and I do not have serial enable few seconds after restart for same reason, so If there is no obvious mistake in procedure I describe to you - now I will dig using simple sample and report the output - thanks for your help
Luc
@luc-github
I used BasicOTA.ino just add an output at each loop and a delay(2000) and same output - upload ok - ERROR[0]: No Error but after sample continue and do not reset - here the output
OTA server at: ESP8266-OTA-22e89:8266
Ready
First version
First version
First version
Update Start: ip:192.168.0.101, port:48266, size:346816
ERROR[0]: No Error
First version
First version
First version
if I restart it is like no upload happened :
OTA server at: ESP8266-OTA-22e89:8266
Ready
First version
First version
First version
the uploaded is same code but with output "second version"
Luc
@luc-github
from another hand, if I use WebUpdater.ino, it works!
after upload the reset happen and new sketch is used - weird
Ivan Grokhotkov
@igrr
since this line doesn't run, apparently there is some error:
Serial.printf("Update Success\nRebooting...\n");
can you please uncomment this line in Updater.cpp and try again?
//#define DEBUG_UPDATER Serial
Updater.cpp is in cores/esp8266 directory
Luc
@luc-github
I have compared code and found difference between ESP8266HTTPUpdateServer.cpp and ArduinoOTA.cpp, first one use if(Update.end(true)) second use if(Update.end())
by changing if(Update.end(true)) in ArduinoOTA.cpp first time update succeed but go into dead loop second try it worked
ok will do #define DEBUG_UPDATER Serial
ok result is :
First version
Update Start: ip:192.168.0.101, port:48266, size:346928
premature end: res:0, pos:339968/346928
ERROR[0]: No Error
First version
First version
Luc
@luc-github
looks like some part is missing
missing 6960 = 4096 + 2864
sending 4096
sending 2864
done!
could be first and last packet ? or 2 last packets ?
Luc
@luc-github

the web update - works without problem:

Booting Sketch...
HTTPUpdateServer ready! Open http://esp8266-webupdate.local/update in your browser
Staged: address:0x00059000, size:0x00054B30

ets Jan 8 2013,rst cause:2, boot mode:(3,7)

load 0x4010f000, len 1264, room 16
tail 0
chksum 0x42
csum 0x42
@cp:0
ld
Booting
OTA server at: ESP8266-OTA-22e89:8266
Ready
Second version
Second version

could be a bug in ota_server.py ?
I tried espota.py but it is not working
Luc
@luc-github

somehow ota_server.py connection.close() is taken in account too fast - I have added a sleep and now the number of packet is correct

adding : 

finally:
  time.sleep(2)
  connection.close()

First version
First version
Update Start: ip:192.168.0.101, port:48266, size:346928
Staged: address:0x002AB000, size:0x00054B30
Update Success
Rebooting...

ets Jan 8 2013,rst cause:2, boot mode:(3,6)

load 0x4010f000, len 1264, room 16
tail 0
chksum 0x42
csum 0x42
@cp:0
ld
Booting
OTA server at: ESP8266-OTA-22e89:8266
Ready
Second version
Second version
Second version

so problem was in python script
Luc
@luc-github
I also was able to use python espota.py if I do not use the port parameter
it says it use port 48266. if I use:
python espota.py -i 192.168.0.120 -p 48266 -f BasicOTA.cpp.bin I got no device response error
if I use :
python espota.py -i 192.168.0.120 -f BasicOTA.cpp.bin it is fine
Me No Dev
@me-no-dev
-p is the port number
in your case it's different than 48266
be default OTA should be at port 8266, unless somebody decided otherwise and I have not seen the commit
mDNS service should also advertise that port