These are chat archives for esp8266/Arduino

17th
Jan 2019
Clemens Kirchgatterer
@everslick
Jan 17 11:49
libraries/ESP8266mDNS/src/LEAmDNS_Control.cpp:247:102: error: 'ip4_addr_netcmp' was not declared in this scope current master borked?
Develo
@devyte
Jan 17 11:59
No
CI runs fine, no issues
What are you trying to do?
david gauchard
@d-a-v
Jan 17 12:35
@everslick are you using lwip-1.4 ?
Clemens Kirchgatterer
@everslick
Jan 17 14:07
@d-a-v yes.
i just recompiled with LWIP_VARIANT=lm2f
compiles now fine
david gauchard
@d-a-v
Jan 17 14:08
Good, but maybe we should fix new mDNS with lwIP-1.4...
Clemens Kirchgatterer
@everslick
Jan 17 14:10
what is the difference between lwip with features and without? :-D
ip6 is clear, low mem and high bandwidth likewise, but features?
david gauchard
@d-a-v
Jan 17 14:13
Features are TCP's SelectiveACK-OUT, IP fragmentation&Reassembly, AutoIP(169.254)
Clemens Kirchgatterer
@everslick
Jan 17 14:13
THX!
david gauchard
@d-a-v
Jan 17 14:14
SACK is (very very) nice for big transfers or poor quality links
Clemens Kirchgatterer
@everslick
Jan 17 14:14
ic
Clemens Kirchgatterer
@everslick
Jan 17 14:27
compiling without features only saves a couple of KB flash, i suppose. does it save ram at runtime?
Develo
@devyte
Jan 17 14:27
Yes
Clemens Kirchgatterer
@everslick
Jan 17 14:30
last question, ping is only available when LWIP_VARIANT=src ?
david gauchard
@d-a-v
Jan 17 14:32
No, it is there
Recently readded. Maybe in 2.5.0-beta2, surely on git head
Clemens Kirchgatterer
@everslick
Jan 17 14:35
great, thx
Clemens Kirchgatterer
@everslick
Jan 17 15:26
latest master runs EXTREMELY well here!!! great work guys @d-a-v , @earlephilhower , @devyte
thank you so much, for your work!!!
not much left to be desired! :-D (besides unified FS API, and a better webserver w/ websockets) ;-P
david gauchard
@d-a-v
Jan 17 15:55
Unified FS API is coming soon thanks to @earlephilhower. External websocket libraries work well too.

latest master runs EXTREMELY well here!!!
thank you so much, for your work!!!

Thanks :) that's appreciated

Ah you mean a webserver + websocket running on the same port
Develo
@devyte
Jan 17 16:01
We'll get to a webserver redesign eventually (unless someone wants to pick up the challenge?), but there are still higher priority things to handle before that
Clemens Kirchgatterer
@everslick
Jan 17 16:27
@d-a-v exactly. the current solution goes not well when sitting behind NAT. besides that the webserver fragments the heap a lot.