These are chat archives for nightscout/intend-to-bolus

6th
Mar 2019
Riko L
@Ricco555
Mar 06 00:07

Anyone else using token based authentication for NS?
I've just run the new dev from scratch. Installation went smoothly regardin npm - no errors, untill in setup script this:

Running nightscout autoconfigure-device-crud https://xxxxx.herokuapp.com token=xx-here-token-was-OK-xx
added process://ns/nightscout/ns NIGHTSCOUT_HOST API_SECRET
process://ns/nightscout/ns https://xxxxx.herokuapp.com API_SECRET
Reading package lists... Done
... some OK stuff
2019-03-06 00:40:01,350 ERROR API_SECRET in ./ns.ini should start with 'token='
Error checking Nightscout permissions

With token based authentication ns.ini got filled with string "API_SECRET" instead of token. but I cannot find the error in nightscout.sh .. somewhere around line 338 API_SECRET gets lost, and NIGHTSCOUT_HOST is filling up nicely.

Or second argument never gets into the nightscout script
Riko L
@Ricco555
Mar 06 00:31
Switching to "normal" API secret instead of token passed this point successfully and broke down on compiling openaps-menu, which I do not need for pi+RFM69HCW.
Riko L
@Ricco555
Mar 06 00:52
so after commenting out the relevant parts from oref0-setup.sh (pi buttons and openaps-menu), which is a bit brutal, aaaand it finally completed successfully. Hoorah!
Jon Cluck
@cluckj
Mar 06 01:07
@Stargazer32584 let me know when you're around to troubleshoot your x12?
davidvjr
@davidvjr
Mar 06 02:00
@acember I'm trying to setup openAPS with a pi zero w and an explorer hat. I'm running openaps-install.sh and got the same error you posted about in September: alt I am new to python, could you explain how you got the script to continue by rerunning curl -s? I'm not having any luck googling what means. Thanks!
davidvjr
@davidvjr
Mar 06 02:17
@acember Nevermind for now, I figured it out, its running again, I'll just have to see if it works this time.
davidvjr
@davidvjr
Mar 06 02:48
I seem to have gotten it to finish...
Martin Haeberli
@mhaeberli
Mar 06 05:03
please remind me how to reflash an Edison Explorer radio on dev - I’m getting this error:
OREF0_DEBUG=2 oref0-mmtune
+ echo '{}'
+ echo -n 'mmtune: '
mmtune: + mmtune_Go
+ set -o pipefail
+ grep WW pump.ini
+ tee monitor/mmtune.json
+ Go-mmtune
2019/03/05 21:01:28 cannot connect to CC111x radio on /dev/spidev5.1
2019/03/05 21:01:28 /dev/spidev5.1: device is in use
+ '[' -s monitor/mmtune.json ']'
+ mmtune_Go
+ set -o pipefail
+ grep WW pump.ini
+ tee monitor/mmtune.json
+ Go-mmtune
2019/03/05 21:01:28 cannot connect to CC111x radio on /dev/spidev5.1
2019/03/05 21:01:28 /dev/spidev5.1: device is in use
try to flash, get
root@etghopenaps14:~/src/ccprog# ./ccprog -p 19,7,36 erase
Using pins: DC=19, DD=7, RESET=36
Erasing chip.
This code is only tested on CC1110. Unsupported chip id = 0x00.
Chip erase failed.
straykatz
@straykatz
Mar 06 05:45
Hmm ... I have seen this before ... @mhaeberli
Do you want to reflash? It's been the cure-all for most my weird issues I have encountered? Reseat the Edison, reflash.
Martin Haeberli
@mhaeberli
Mar 06 05:50
reflash -Edison- or -Explorer-? I know how to reflash Edison, and I used to be able to run ccprog (in fact, I helped clean up the documentation for ccprog) to reflash the Edison CC111x radio, but somehow on dev some of the magic has changed ...
I think
and thanks! @straykatz for suggestions...
straykatz
@straykatz
Mar 06 05:51
:)
straykatz
@straykatz
Mar 06 05:57
Don't think this is related ... and I don't recall the error codes, but it might have been like yours. I had issues with 2 explorer boards that "broke". One of them, I broke the UL antenna connector off along with a 915 MHz wire antenna, when I took it off one too many times.
straykatz
@straykatz
Mar 06 06:05
IMG_4061.jpeg
TranceCake
@TranceCake
Mar 06 07:21
I am having the above issue with a rfm69 module. Still troubleshooting if it's my soldering or if it's broken: This is the one that was connected to the pi that I accidentally fried, it had a short so I would not be surprised if it took the rfm69 with it... :O
boc-the-git
@boc-the-git
Mar 06 07:26
From what I can gleam from the above (albeit not reading EVERY message), it sounds like I'm safe to move back to the dev branch? Will give it a go now!
boc-the-git
@boc-the-git
Mar 06 09:33
Steps for those like me lacking git knowledge, that are looking to get back to the latest dev:
cd ~/src/oref0
git fetch origin
git reset --hard origin/dev
Jon Cluck
@cluckj
Mar 06 14:22
thanks, I just delete the directory and clone it over again :laughing:
tuzoenduro
@tuzoenduro
Mar 06 14:25
is it safe to update?/how do I update? my pi/hat rig is about 10 commits late but seing the recent bug bonanza I was reticent to try to update again...
elodaille01
@elodaille01
Mar 06 14:48
Hi everybody. .#OpenAPSuser.I planned to upgrade my oref0, but, here is the message I got (see below) any thoughts, please ? Thank you so much ! ;-)
Using /home/.rootfs/usr/local/lib/python2.7/dist-packages
Finished processing dependencies for decocare==0.1.0-dev
/root/src/oref0/ already exists; pulling latest
error: object file .git/objects/66/1fe61766ad31d576137935381ccfb4574ecfbd is empty
error: object file .git/objects/66/1fe61766ad31d576137935381ccfb4574ecfbd is empty
fatal: loose object 661fe61766ad31d576137935381ccfb4574ecfbd (stored in .git/objects/66/1fe61766ad31d576137935381ccfb4574ecfbd) is corrupt
Couldn't pull latest oref0
TranceCake
@TranceCake
Mar 06 15:24
@tuzoenduro mine is 13 behind and I'm keeping it there for now since I have no working backup ;)
Martin Haeberli
@mhaeberli
Mar 06 15:25
See comment from “boc the git” above, OR just “cd ~/src ; rm -rf oref0 ; git clone https://github.com/openaps/oref0”
tuzoenduro
@tuzoenduro
Mar 06 15:34
@TranceCake that is what I was afraid of... I have no backup for the mo, so don't want to jynx it... I have to flash an edison (for which I have no root password for), which would let me have 2 working ones, but for now I will wait then..
Eric
@ecc1
Mar 06 15:43
@TranceCake what issue are you having with an RFM69 module? (and what kind of module, standalone breakout board or Adafruit bonnet?)
TranceCake
@TranceCake
Mar 06 16:03

@ecc1 I think I have the adafruit one. https://www.sossolutions.nl/3070-adafruit-rfm69hcw-transceiver-radio-breakout-868-or-915-mhz
I am getting the error from the docs:

2019/01/14 15:14:25 cannot connect to CC111x radio on /dev/spidev0.0
2019/01/14 15:14:25 cc111x: no response
Usage: grep [OPTION]... PATTERN [FILE]...
Try 'grep --help' for more information.

They say it's because either I have run the runagain script (i'm sure I didn't) or the soldering is bad
I am afraid that the rfm69 is broken since it was connected to a pi that shorted out. Is there any way to know for sure if the board is broken or not?

e.g. if it's a soldering error
Eric
@ecc1
Mar 06 16:07
You have something configured wrong, since it says CC111x radio
TranceCake
@TranceCake
Mar 06 16:08
so maybe do a wipe and try again? maybe something got messed up during all the debugging of the last couple of days
Eric
@ecc1
Mar 06 16:08
You can build the low-level "regtest" command in the rfm69/cmd/regtest directory (cd there and do "go build"), then run it ("./regtest"). If it prints a bunch of nonzero values, and no warnings about unexpected values, the hardware is probably fine
Jon Cluck
@cluckj
Mar 06 16:08
you probably need to re-do that go get thing and copy the binaries?
Eric
@ecc1
Mar 06 16:08
You can do this ^^^ even before you re-run the setup
(this = running regtest to verify the radio is OK)
Jon Cluck
@cluckj
Mar 06 16:09
yeah, check the regtest first :)
TranceCake
@TranceCake
Mar 06 16:09
okay will do thx!
sorry I'm a bit of a noob, I seem to not be able to locate the rfm69 folder
This message was deleted
Jon Cluck
@cluckj
Mar 06 16:15
~/go/src/github.com/ecc1/rfm69?
TranceCake
@TranceCake
Mar 06 16:16
yeah thx!
when I do 'go build' I get a lot of cannot find package output
/usr/local/go/src/fmt/scan.go:10:2: cannot find package "math" in any of:
        /usr/local/go/src/math (from $GOROOT)
        /root/go/src/math (from $GOPATH)
is one example
Eric
@ecc1
Mar 06 16:32
what does "go env" print?
TranceCake
@TranceCake
Mar 06 17:22
in that folder?
GOARCH="arm"
GOBIN=""
GOCACHE="/root/.cache/go-build"
GOEXE=""
GOFLAGS=""
GOHOSTARCH="arm"
GOHOSTOS="linux"
GOOS="linux"
GOPATH="/root/go"
GOPROXY=""
GORACE=""
GOROOT="/usr/local/go"
GOTMPDIR=""
GOTOOLDIR="/usr/local/go/pkg/tool/linux_arm"
GCCGO="gccgo"
GOARM="6"
CC="gcc"
CXX="g++"
CGO_ENABLED="1"
GOMOD=""
CGO_CFLAGS="-g -O2"
CGO_CPPFLAGS=""
CGO_CXXFLAGS="-g -O2"
CGO_FFLAGS="-g -O2"
CGO_LDFLAGS="-g -O2"
PKG_CONFIG="pkg-config"
GOGCCFLAGS="-fPIC -marm -pthread -fmessage-length=0 -fdebug-prefix-map=/tmp/go-build354390532=/tmp/go-build -gno-record-gcc-switches"
Eric
@ecc1
Mar 06 17:30
and what does "go version" print?
TranceCake
@TranceCake
Mar 06 17:32
go1.11 linux/arm
you were wondering if I was running pre 1.10 due to the math package being added in that version?
Eric
@ecc1
Mar 06 17:34
no, math package has been around forever, but some new features were added. and those "can't find package" errors happened when your current directory was the rfm69/cmd/regtest dir? (you had cd'ed to that directory first?)
TranceCake
@TranceCake
Mar 06 17:34
just to be sure I went to the ~/go/src/github.com/ecc1/rfm69/cmd/regtest# folder and ran 'go build'
the cmd dir also contains a demo and rxtest folder
Eric
@ecc1
Mar 06 17:39
Not sure what's going on, is it possible your installation of Go was interrupted? can you do apt-get install --reinstall golang
TranceCake
@TranceCake
Mar 06 17:39
yeah sure
Eric
@ecc1
Mar 06 17:40
that uses the system version, so also rm -r /usr/local/go afterwards
TranceCake
@TranceCake
Mar 06 17:40
okay
go build now gives: -bash: /usr/local/go/bin/go: No such file or directory
Eric
@ecc1
Mar 06 17:45
do hash -r and try it again
TranceCake
@TranceCake
Mar 06 17:45
/usr/lib/go-1.7/src/github.com/ecc1/rfm69 (from $GOROOT)
        ($GOPATH not set)
which go reveals /usr/bin/go
Eric
@ecc1
Mar 06 17:46
aargh, are you running Debian stretch? I run buster, which has go 1.11
TranceCake
@TranceCake
Mar 06 17:46
oh sorry yeah :O
didn't know about buster
now I know why the script pulled it from github instead of using the system version
Eric
@ecc1
Mar 06 17:48
no, my mistake. I guess we'll to install a /usr/local version after all. Please try go get golang.org/dl/go1.11.5
and if that works, go1.11.5 download
TranceCake
@TranceCake
Mar 06 17:49
package golang.org/dl/go1.11.5: cannot download, $GOPATH not set. For more details see: go help gopath
I'm guessing I should go to /usr/bin/go ?
Eric
@ecc1
Mar 06 17:52
OK, let's get rid of the old version of Go first: apt-get purge golang
TranceCake
@TranceCake
Mar 06 17:53
yeah the usr/bin thing didn't work
Eric
@ecc1
Mar 06 17:53
Then,
cd /tmp
wget https://dl.google.com/go/go1.11.5.linux-armv6l.tar.gz
tar -C /usr/local -xzf go1.11.5.linux-armv6l.tar.gz
TranceCake
@TranceCake
Mar 06 17:54
running now
btw thank you so much for running me through this!
okay that's done
Eric
@ecc1
Mar 06 17:59
OK, if you still have /usr/local/go/bin on your PATH, does "go version" work?
TranceCake
@TranceCake
Mar 06 18:00
huh what.. I went there and it tells me go1.7.4
Eric
@ecc1
Mar 06 18:01
did you purge the system golang package?
TranceCake
@TranceCake
Mar 06 18:01
yeah
Eric
@ecc1
Mar 06 18:01
and then hash -r again?
TranceCake
@TranceCake
Mar 06 18:01
oh no not that
Eric
@ecc1
Mar 06 18:02
you should be able to run go version anywhere
TranceCake
@TranceCake
Mar 06 18:02
hm, still on 1.7.4
okay hold on
Eric
@ecc1
Mar 06 18:02
what does which go print?
TranceCake
@TranceCake
Mar 06 18:02
wnet to /usr and it still says go1.7.4
usr/bin/go
pointing to the wrong go maybe?
Eric
@ecc1
Mar 06 18:03
what does dpkg --get-selections |grep golang print?
yeah, the old system golang package is still there
TranceCake
@TranceCake
Mar 06 18:03
golang-1.7                                      install
golang-1.7-doc                                  install
golang-1.7-go                                   install
golang-1.7-src                                  install
golang-doc                                      install
golang-go                                       install
golang-src                                      install
Eric
@ecc1
Mar 06 18:03
more purging needed :-)
TranceCake
@TranceCake
Mar 06 18:04
so just run it again?
Eric
@ecc1
Mar 06 18:04
apt-get purge golang-{go,1.7}
TranceCake
@TranceCake
Mar 06 18:05
it asks if I want to autoremove some packages with 1.7. in the name
Eric
@ecc1
Mar 06 18:05
say yes
TranceCake
@TranceCake
Mar 06 18:05
oops
ok done
Eric
@ecc1
Mar 06 18:06
try go version again?
from anywhere
TranceCake
@TranceCake
Mar 06 18:06
no such file or directory
The following packages will be REMOVED:
  golang-1.7-doc golang-1.7-go golang-1.7-src golang-src
Eric
@ecc1
Mar 06 18:07
try the hash -r again
^^^ yes those removals are all OK
TranceCake
@TranceCake
Mar 06 18:07
ok
still no such file
which go: /usr/local/go/bin/go
well then that's logical cause it's looking at the old location
-bash: /usr/bin/go: No such file or directory
Eric
@ecc1
Mar 06 18:09
we want the /usr/local location now, that's where the tarball got extracted. you did the hash -r?
TranceCake
@TranceCake
Mar 06 18:10
oh now it's ok
aparently something didn't go through
its go1.11.5 now
Eric
@ecc1
Mar 06 18:10
yay! now try cd'ing to the regtest directory and "go build" again
TranceCake
@TranceCake
Mar 06 18:11
yeah that is already taking longer than before yay!
what did you mean by running it with ("./regtest")?
as in go one folder up?
Eric
@ecc1
Mar 06 18:12
no parens or quotes, just type ./regtest
TranceCake
@TranceCake
Mar 06 18:12
okay will try when it's done
I'm going to run a lap around my flat if it turns out my rfm69 is dead xD
uh ooh: 2019/03/06 19:13:21 unexpected hardware version 0000 (should be 0204)
that means it could be the soldering or the rfm69 itself right?
Eric
@ecc1
Mar 06 18:15
oh wait, which instructions did you follow for wiring up the RFM69 pins? you might need to do go build -tags walrusto get the older pin config
TranceCake
@TranceCake
Mar 06 18:15
oooh, doing that now, cluckj told me to do those yeah
go get -u -v -tags "rfm69 walrus" github.com/ecc1/medtronic/...
and then
cp -pruv $HOME/go/bin/* /usr/local/bin/
and then
mv /usr/local/bin/mmtune /usr/local/bin/Go-mmtune?
Eric
@ecc1
Mar 06 18:16
no need for most of that now
TranceCake
@TranceCake
Mar 06 18:16
ok
Eric
@ecc1
Mar 06 18:16
just do go build -tags walrus in the regtest directory
TranceCake
@TranceCake
Mar 06 18:16
just what you typed
ahwww yes! whole bunch of config registers
Eric
@ecc1
Mar 06 18:18
no (or only a few) lines with a bunch of asterisks?
TranceCake
@TranceCake
Mar 06 18:19
no asterisks but some hexadecimal numbers
Eric
@ecc1
Mar 06 18:20
well, good new or bad news, the RFM69 is fine. And you do have an antenna soldered to that pin on the breakout board, right?
TranceCake
@TranceCake
Mar 06 18:20
yes I do
read it can damage itself if you don't
Eric
@ecc1
Mar 06 18:21
we can also pretty easily try a receive test if you have either another rig, or an x23 pump
TranceCake
@TranceCake
Mar 06 18:21
I have a rig
yes let's do that!
Eric
@ecc1
Mar 06 18:22
OK, then change to the rxtest folder (cd ../rxtest) and do another go build -tags walrus
TranceCake
@TranceCake
Mar 06 18:22
and do it on the other as well I suppose?
it's done
Eric
@ecc1
Mar 06 18:23
other? no, this program will just "sniff" packets sent by the other rig, to check that the RFM69 is receiving OK
TranceCake
@TranceCake
Mar 06 18:23
oh right
Eric
@ecc1
Mar 06 18:23
OK, run it: ./rxtest
and do something with your other rig to make it talk to your pump (nearby)
TranceCake
@TranceCake
Mar 06 18:24
oh I need to specify a frequency, I happen to know my other rig does best at 868.350
Eric
@ecc1
Mar 06 18:24
should see hex dumps of packets
TranceCake
@TranceCake
Mar 06 18:25
a whole lot
with dbm ranging from -28 to -94
Eric
@ecc1
Mar 06 18:26
OK, RFM69 and antenna are fine then. So the rest of your issue is "just" configuration of openaps, which I'm not too knowledgeable about
TranceCake
@TranceCake
Mar 06 18:26
okay thanks a lot!!
Eric
@ecc1
Mar 06 18:27
but if it's configured right, the messages from mmtune etc. will say "connected to rfm69 on spidev0.0"
TranceCake
@TranceCake
Mar 06 18:27
yeah just checked and the loop is still giving me the CC111x error
I can try to run mmtune manually
yeah weird, it gives me the CC111x connection error
Eric
@ecc1
Mar 06 18:28

yes, here's where you can try those commands @cluckj said:

cd
go get -u -v -tags "rfm69 walrus" github.com/ecc1/medtronic/...

etc. Note those "..." at the end should be there

TranceCake
@TranceCake
Mar 06 18:29
think there may be an error in the setup script since you do need to choose what kind of radio you have
Eric
@ecc1
Mar 06 18:29
I thought there was an "other" option or something like that?
TranceCake
@TranceCake
Mar 06 18:29
oh, ok, any specific folder to run them in?
Eric
@ecc1
Mar 06 18:29
home (that's what the initial "cd" puts you in)
TranceCake
@TranceCake
Mar 06 18:30
there is a 1 for some cc111x and then another one with a bunch of c's and 1's and then 3 for the rfm, I'm sure I chose 3
ok
cool new trick :)
thx again for you generous time!
Eric
@ecc1
Mar 06 18:32
you're very welcome
TranceCake
@TranceCake
Mar 06 18:33
it is redownloading everything so I think it might have been that
Martin Haeberli
@mhaeberli
Mar 06 18:33
@ecc1 - can you advise re current magic for re-flashing CC111x in Edison/Explorer (under dev branch)?
Eric
@ecc1
Mar 06 18:34
I've only ever flashed cc111x radios with a TI CC Debugger (TI sticks etc.). No experience with Explorer boards I'm afraid
Martin Haeberli
@mhaeberli
Mar 06 18:34
@ecc1 ok thx; anyone else ?
also, wondering if anyone can comment on this - running Explorer / Edison / dev branch - I notice some rigs take a long time to see the pump, despite proximity and an antenna; sometimes papertrail logs just show mmtune, sometimes mmtune and a signal strength. It’s almost like I’d like a command in the Edison’s shell that can asynchronously trigger an mmtune, or asynchronously report the signal strength, or maybe see a log entry showing signal strength if known for every successful pump communication.
are any of these worth an issue or do any exist and just need documentation? obviously, relevant also for PiHat explorer rigs
TranceCake
@TranceCake
Mar 06 18:38
my rig regularly ony get comms up with the 3rd retry
Martin Haeberli
@mhaeberli
Mar 06 18:38
@TranceCake right - thx; i often go through many sets of 3 retries on some rigs before comms works - what do you see that is an indication of success? (besides looping working…)
TranceCake
@TranceCake
Mar 06 18:39
that it get's pump data
after the 3rd retry if it didnt go right it log a failure
Eric
@ecc1
Mar 06 18:39
@mhaeberli mdt rssi will show the signal strength (but it will conflict with your loop if that's running, so need to time it right)
TranceCake
@TranceCake
Mar 06 18:39
so when I say it works I mean no failure
Martin Haeberli
@mhaeberli
Mar 06 18:41
@ecc1 thx very nice - if I wanted to run that on every pump comms loop try, what file would i add that to (if you know); I’ll give that a shot; @TranceCake thx
TranceCake
@TranceCake
Mar 06 18:41
I also regularly have an empty mmtune in my logs
Martin Haeberli
@mhaeberli
Mar 06 18:42
@TranceCake :+1: - never clear to me what that means.
TranceCake
@TranceCake
Mar 06 18:42
I think a pump comms timing mismatch
the pump only periodically listens for wakeup signals (I assume) so I find it only normal that it sometimes misses the rigs wakeup call
but that's all guesswork from me
Eric
@ecc1
Mar 06 18:43
@mhaeberli oref0-pump-loop.sh but not sure exactly where (maybe tuck it in "update_display")
Jon Cluck
@cluckj
Mar 06 18:43
yep, in update_display is best
re:cc-prog on dev, you'll need to install MRAA
Martin Haeberli
@mhaeberli
Mar 06 18:45
@cluckj thx - does update_display still get invoked on display-less Edison rigs?
TranceCake
@TranceCake
Mar 06 18:45
@cluckj when you first told me the walrus command, you mentioned not restarting the rig first, is that key because my rig just died due to battery mid walrus download?
Martin Haeberli
@mhaeberli
Mar 06 18:45
walrus command?
TranceCake
@TranceCake
Mar 06 18:46
go get -u -v -tags "rfm69 walrus" github.com/ecc1/medtronic/...
Martin Haeberli
@mhaeberli
Mar 06 18:46
k thx @cluckj
Jon Cluck
@cluckj
Mar 06 18:46
(on mobile right now :laughing: )
Martin Haeberli
@mhaeberli
Mar 06 18:47
@cluckj - I think it was -you- I met at HackNS SF last fall - with the cool PiZeroW case - can you re-link that here or hardware-dev ?
Eric
@ecc1
Mar 06 18:47
@TranceCake I don't think we discussed restarting at all, there was just no need to build the medtronic sw at that point until we knew the rfm69 was working. redoing it once you have power again should do the right thing
Jon Cluck
@cluckj
Mar 06 18:47
you want to redo the Go bits for the rfm69 before reboot because you don't want the active loop crashing into copying the new binaries
TranceCake
@TranceCake
Mar 06 18:47
yeah no I mean way back with my first rig.
so is it bad that it shut down?
Jon Cluck
@cluckj
Mar 06 18:49
@mhaeberli it was :) the case is super DIY, but I'll post stuff about it in hardware-dev. you may want to check out @loztblaz 's case too
TranceCake
@TranceCake
Mar 06 18:49
i'll just stop cron before copying
boy am I happy the rfm69 is fine, because this was not fun to get right!
Travel rig.jpeg
soldering wise
yeah! it is looping and everything!
Jon Cluck
@cluckj
Mar 06 19:00
yay
Martin Haeberli
@mhaeberli
Mar 06 19:12
@cluckj thx
Riko L
@Ricco555
Mar 06 19:21
@ecc1 > oh wait, which instructions did you follow for wiring up the RFM69 pins? you might need to do go build -tags walrusto get the older pin config
what is the new pin configuration that is supported with regular tag without "walrus" ?
Eric
@ecc1
Mar 06 19:40
@Ricco555 the default is now for the Adafruit RFM69 "radio bonnet", because I expect most new users of RFM69s will go that route (display! buttons!)
Jon Cluck
@cluckj
Mar 06 20:09
THREE buttons!!!
Riko L
@Ricco555
Mar 06 20:19
Yay .. cool looking.. and I can play snake on it probably?
Abigail Cember
@acember
Mar 06 20:29
@davidvjr Sorry I didn't see your messages right away, but I personally also just did this blindly based on instructions -- I'm not aware of there being any magic except for the no. 1 IT fix of "turn it off and then turn it on again", or something equivalent. With my limited programming experience, I myself am baffled as to how it could be that running the same script twice results in different outcomes...Maybe someone with more experience on here can address that.
@TranceCake Can that thing you just built run without being plugged in?
(that cylindrical thing is a battery, right?)
Eric
@ecc1
Mar 06 20:59
@acember they're powering it via the micro-USB plugged in at the bottom. I think that AAA cell is playing the role of a banana (for size)
TranceCake
@TranceCake
Mar 06 22:23
@acember the cylindrical thing is an antenna, it needs a powerbank
amspoke
@amspoke
Mar 06 23:24
Hi, i have a question. My Rig (Raspberry) runs an openaps command each minute which seems to parse something from the logs, but the logs grow everyday, openaps.log is something like 7Gb now. The questions is about log rotating, what can happen if i rotate then daily, or is it fine this way?