<sc6502>
Hey Xogium , long time (too long) no speak.
<Xogium>
hey ! How are you doing ?
<sc6502>
I seem to be perpetually busy with either work, other projects or life in general.
<Xogium>
oh yeah… I know how that is
<sc6502>
Peter tells me OLED is spamming the logs when there is no OLED.
<Xogium>
yeah
<Xogium>
raised an issue on the navit github about it
<Xogium>
er navit repo
<Xogium>
our navit that is
<sc6502>
Ah, just found it :)
<sc6502>
I was looking in speedsaver, not navit
<Xogium>
but today I gave the last push so to speak and made a (in theory) working image for the mangopi, that gets flashed on the nand flash. No more micro sd, yay !
<Xogium>
and with filesystem of course, wear leveling and bad block aware
<sc6502>
Sounds good.
<sc6502>
So navit should just exit if there is no display.
<sc6502>
But it seems the OLED is lying about success
<sc6502>
Do you want navit to carry on, but without a display?
<Xogium>
yeah that could be good like make it wait until there's one maybe ? That way the service will not crash
<Xogium>
just like standby until there is a display
<sc6502>
Was there a "I2C ioctl error " at the start of the spam?
<Xogium>
I'm not sure
<Xogium>
it really went so fast I couldn't see anything but that error I gave you
<Xogium>
I'll try to see
<Xogium>
yeah nah I really just get spammed by i2c write error
<Xogium>
but it seems like navit simply exits on its own after
<Xogium>
but it is still huge spam
<Xogium>
there's like 8650 lines in the log file I made and it wasn't yet finished writing
<Xogium>
but I know why it slowed down the system so much, it was wring to the log daemon
<Xogium>
er writing
<Xogium>
the logs aren't stored anywhere in the flash or in ram, but the daemon still received the messages
<Xogium>
seems like the arduipi oled lib might lie to navit and make it think there's something connected when there's nothing?<
<Xogium>
not sure
<sc6502>
Yes it does.
<sc6502>
If it was __NetBSD__, it would have ended properly
<sc6502>
:)
<Xogium>
er why's that
<sc6502>
Anyway, new push to To github.com:Speedsaver/ArduiPi_OLED.git
<sc6502>
Just the way the conditional compilation code was written
<sc6502>
#elif defined(__NetBSD__)
<sc6502>
printf("I2C open error\r\n");
<sc6502>
if(i2c_fd < 0) {
<sc6502>
}
<sc6502>
return 0;
<sc6502>
#endif
<sc6502>
The error check is there, but only for one platform - doh!
<Xogium>
do you think it would be possible to have navit standing by while there isn't a display though ? In my case here I don't have one, but it would be useful so I could verify that navit at least runs okay, minus the display that is
<sc6502>
Yes, that's what it should do now
<Xogium>
darn that's a bit dumb
<Xogium>
ah good
<sc6502>
You should get a 1-time message that there is no display
<Xogium>
good, will try
<sc6502>
After that, silence (from the display at least)
<Xogium>
aye
<Xogium>
sounds perfect
<Xogium>
god, don't know if you ever worked with raw nand flash, but sure was a tricky one
<Xogium>
I mean I worked plenty with micro sd, with eMMC, with ssd even, but working with raw flash was my first time on this board
<Xogium>
not to mention that right from the start it was a real mess with this board, the vendor's provided instructions didn't even work !
<Xogium>
but I mastered the beast ;)
<sc6502>
Yeah I know how that feels when line 1 of the instructions don't even work.
<Xogium>
yeah honestly the 2 boards Peter sent me, one came dangerously close to encountering the wall made of concrete to my left
<Xogium>
oh jesus I also just figured out why our git commit bot exploded
<sc6502>
It wasn't because of my commit message was it?
<Xogium>
no
<Xogium>
we used the git.io url shortening service, which has now been discontinued by github
<Xogium>
which… explains why the bot never said anything when we pushed commits
Speedsaver has quit [Read error: Connection reset by peer]