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

6th
Jan 2017
Scott Leibrand
@scottleibrand
Jan 06 2017 00:04
@bdr1177 have you ordered an Explorer board and/or TI stick yet?
Brian Rabinovitz
@bdr1177
Jan 06 2017 00:04
Yes, just ordered yesterday.
Scott Leibrand
@scottleibrand
Jan 06 2017 00:04
Back when Carelink was the only option, we had a method for automatically detecting that scenario and running oref0-reset-usb
It's not in oref0-setup yet because everyone doing dev has moved on to mmeowlink.
But if you want to read through my old openaps-sh repo (dev branch maybe) it's still there.
And then maybe you could get it added to oref0-setup for Carelink users.
Aaron Michelson
@Aaybob
Jan 06 2017 00:06
@scottleibrand was that something we did in cron?
@Aaybob a few people have done something like that, either in cron or in their pump-loop. No one has automated the setup yet though.
Aaron Michelson
@Aaybob
Jan 06 2017 00:10
@scottleibrand @bdr1177 yup put a call to that code alias in cron, we were doing every five minutes then.
Scott Leibrand
@scottleibrand
Jan 06 2017 00:11
If one of you could get that added to oref0-setup and PR in the change to dev, that would be great.
Aaron Michelson
@Aaybob
Jan 06 2017 00:11
...... preflight && loop .......
Brian Rabinovitz
@bdr1177
Jan 06 2017 00:11
Ok thanks. I'm sure I'll be ditching the carelink as soon as I get the Edison set up with explorer board. The documentation still says in some places that the carelink is default. It might be good to update this since it doesn't seem like a recommended method anymore.
Scott Leibrand
@scottleibrand
Jan 06 2017 00:11
Yes, please go ahead and PR fixes for those too.
Brian Rabinovitz
@bdr1177
Jan 06 2017 00:11
Will do.
Scott Leibrand
@scottleibrand
Jan 06 2017 00:12
Thx
Aaron Michelson
@Aaybob
Jan 06 2017 00:13
I'll have some updates for BLE tether instructions but want it to actually work first
Aaron Michelson
@Aaybob
Jan 06 2017 00:27
In the dev branch oref0-setup interactive
This error each try:
Checking mraa installation
usage: openaps-device [-h] {add,remove,show} ...
openaps-device: error: unrecognized arguments: US
Can't add pump

Are you using a USA/North American pump? If so, just hit enter. Otherwise enter WW:
Ok, -n US it is

I'll try space,enter

Scott Leibrand
@scottleibrand
Jan 06 2017 00:31
I updated the dev branch this week to install the latest openaps dev to fix that. Did you pull the latest oref0 dev?
Aaron Michelson
@Aaybob
Jan 06 2017 00:32
maybe not
root@Fun:~# $ mkdir -p ~/src; cd ~/src && git clone -b dev git://github.com/openaps/oref0.git || (cd oref0 && git checkout dev && git pull)
-bash: $: command not found
fatal: destination path 'oref0' already exists and is not an empty directory.
M bin/bt-pan
M bin/oref0-dexusb-cgm-loop.py
M bin/oref0-find-ti.sh
M bin/oref0-subg-ww-radio-parameters-timeout.sh
M bin/oref0-subg-ww-radio-parameters.sh
Already on 'dev'
Your branch is up-to-date with 'origin/dev'.
Already up-to-date.
root@Fun:~/src#
Tim Street
@tim2000s
Jan 06 2017 00:36
@scottleibrand I was having issues with that this morning. I flipped over to master as when I tried to use the WW pump I got the same error, so it's doesn't seem to be region linked.
I had pulled the latest dev, as far as I am aware.
Scott Leibrand
@scottleibrand
Jan 06 2017 00:38
Maybe I only fixed in the autotune branch.
Want to make that same fix and PR it to dev?
Or we could just merge autotune to dev I suppose. ;-)
Aaron Michelson
@Aaybob
Jan 06 2017 00:42
Sure, but it'd be a first for me. I would just change docs to do what I did which was space space space enter
that worked
Scott Leibrand
@scottleibrand
Jan 06 2017 00:42
openaps/oref0#313
If anyone wants to review and :+1: that (mainly for any effect on anyone not using autotune) we can get it merged.
Aaron Michelson
@Aaybob
Jan 06 2017 00:44
will do, right now , just finished my space run successful, script will delete pump and reinstall so should be an okay test
Scott Leibrand
@scottleibrand
Jan 06 2017 00:44
Was mostly just waiting to see if @PieterGit got a chance to PR his suggested changes.
Aaron Michelson
@Aaybob
Jan 06 2017 01:25
BLE tethering didn't work.
How long to charge a new 2000mAh 3.7V LiPo from explorer?
Dana Lewis
@danamlewis
Jan 06 2017 01:28
@Aaybob fwiw, I still haven’t gotten the tethering to work with my iphone yet, going to try again next week when I have control over my home router
Dana Lewis
@danamlewis
Jan 06 2017 01:35
finally got around to adding a section explaining preferences.json - please PR any additional stuff missed to that new section :) http://openaps.readthedocs.io/en/latest/docs/walkthrough/phase-3/beyond-low-glucose-suspend.html#understanding-your-preferences-json
Aaron Michelson
@Aaybob
Jan 06 2017 01:37
@danamlewis the iphone instructions on Hotspot connect using bluetoothe 1.) Pair Iphone with your computer ;-<{
Dana Lewis
@danamlewis
Jan 06 2017 01:38
ah, sorry, I was referring to bluetooth tethering edison/explorer to my iphone
Aaron Michelson
@Aaybob
Jan 06 2017 01:40
@danamlewis isn't bluetooth tethering the same as bluetooth hotspot?
Aaron Michelson
@Aaybob
Jan 06 2017 01:46
@danamlewis settins>general>about> 13 lines down is Bluetooth address, one number away from wi-fi address settings>Personal Hotspot has instructions to connect with wifi, bluetooth, and usb but the bluetooth instructions don't make sense to me (pair iphone)(tap code on computer)(3.connect iphone from computer is what "Configure Bluetooth....." says to do with the pair AA:BB:CC... command
ejoe132
@ejoe132
Jan 06 2017 01:46
@Aaybob I do not unfortunately
Aaron Michelson
@Aaybob
Jan 06 2017 01:50
@ejoe132 @danamlewis Sprint tells me that BLE tethering is part of the Hotspot feature. you just sellect bluetooth or wifi.
was that your understanding
@ejoe132 when you say the company wants $30/month for tethering do they use the word "hotspot"?
Aaron Michelson
@Aaybob
Jan 06 2017 02:02
pair command returns "address unavailable"
expected result is "Request confirmation [agent] Confirm passkey 123456 (yes/no): "
[bluetooth]# power on
[CHG] Controller 98:4F:EE:04:5A:D0 Class: 0x400110
Changing power on succeeded
[CHG] Controller 98:4F:EE:04:5A:D0 Powered: yes
[bluetooth]# discoverable on
Changing discoverable on succeeded
[CHG] Controller 98:4F:EE:04:5A:D0 Discoverable: yes
[bluetooth]# agent on
Agent registered
[bluetooth]# default-agent
Default agent request successful
[bluetooth]# pair 1C:5C:F2:72:B7:5F
Device 1C:5C:F2:72:B7:5F not available
[bluetooth]#
ejoe132
@ejoe132
Jan 06 2017 02:07
@Aaybob when i spoke to them i used the term wifi tethering, didnt know about bluetooth tethering? Im guessing it just moves data over bluetooth band waves vs wifi range
Aaron Michelson
@Aaybob
Jan 06 2017 02:11
yup, sounds like hotspot and tethering are the same thing
How to make iPhone available to BLE tethering?
I make my edison discoverable for three minutes but what do I do on the iPhone?
[bluetooth]# discoverable on
Changing discoverable on succeeded
Aaron Michelson
@Aaybob
Jan 06 2017 02:16
edison does not appear in settings>bluetooth>otherdevices
why is it instructed to make the edison discoverable when it is edison that is connecting to iPhone, iPhone is allways discoverable
Aaron Michelson
@Aaybob
Jan 06 2017 02:31
root@Fun:~/ble# sudo hciconfig hci0 name $HOSTNAME
Can't change local name on hci0: Network is down (100)
root@Fun:~/ble#
ejoe132
@ejoe132
Jan 06 2017 02:32

Im at Phase 2 of orefo, put the link into PUTTY, first time it did a long list ran for two minutes then stuck at 93%, waited 15 min and it didnt do anything, figured wifi froze, re ran it and it gave me this, says 100% not showing what walkthrough says it should

pi@raspberrypi:~ $ curl -s https://raw.githubusercontent.com/openaps/docs/master/scripts/quick-packages.sh | bash -
Hit http://mirrordirector.raspbian.org jessie InRelease
Hit http://mirrordirector.raspbian.org jessie/main armhf Packages
Hit http://mirrordirector.raspbian.org jessie/contrib armhf Packages
Hit http://mirrordirector.raspbian.org jessie/non-free armhf Packages
Hit http://mirrordirector.raspbian.org jessie/rpi armhf Packages
Ign http://mirrordirector.raspbian.org jessie/contrib Translation-en_GB
Ign http://mirrordirector.raspbian.org jessie/contrib Translation-en
Ign http://mirrordirector.raspbian.org jessie/main Translation-en_GB
Ign http://mirrordirector.raspbian.org jessie/main Translation-en
Ign http://mirrordirector.raspbian.org jessie/non-free Translation-en_GB
Ign http://mirrordirector.raspbian.org jessie/non-free Translation-en
Ign http://mirrordirector.raspbian.org jessie/rpi Translation-en_GB
Ign http://mirrordirector.raspbian.org jessie/rpi Translation-en
100% [Connecting to archive.raspberrypi.org (2a00:1098:0:82:1000:13:0:7)]

Tim Mecklem
@tmecklem
Jan 06 2017 02:32
@scottleibrand you're too fast for me on PRs :smile:
ejoe132
@ejoe132
Jan 06 2017 02:35
just came back with a bunch of stuff, Ill try to copy it here, thinking I missed something
Bringing up different "change logs" right now on "proccessing change logs for triggers man-"
jmcrawford
@jmcrawford
Jan 06 2017 02:46
@Aaybob I'm at the same spot, trying to get BT tethering working. Ran into the same error last night. Need to do "hciconfig hci0 up", then you can do the name command
Aaron Michelson
@Aaybob
Jan 06 2017 02:47
I'm having some luck here @JELCRAWFORD https://wiki.debian.org/BluetoothUser
jmcrawford
@jmcrawford
Jan 06 2017 02:48
In my mind, there is some stuff missing from the docs. Need a way to start BT every time the edison boots. There is a section in the Intel Edison docs, but I haven't had success yet: http://download.intel.com/support/edison/sb/edisonbluetooth_331704007.pdf
Scott Leibrand
@scottleibrand
Jan 06 2017 02:49
@jmcrawford are you working off of oref0 dev? lots of work there on getting bluez to start reliably on boot
ejoe132
@ejoe132
Jan 06 2017 02:49
Got my problem resolved.
jmcrawford
@jmcrawford
Jan 06 2017 02:51
yes, 0.2.0-dev
Aaron Michelson
@Aaybob
Jan 06 2017 02:52
I connect with bluetooth but it dissconnects quickly
jmcrawford
@jmcrawford
Jan 06 2017 02:53
I can make a connection, but no network device is set up. the "bt-pan client ..." command doesn't work as in the document. That's where I'm at.
Scott Leibrand
@scottleibrand
Jan 06 2017 02:53
@jmcrawford 0.2.0-dev is an openaps version number, not oref0
Aaron Michelson
@Aaybob
Jan 06 2017 03:01
@jmcrawford would you see if you can duplicate this successful result. I am paired with BLE now. I went into bloothctl,
trust iphoneaddress
connect iphoneaddress
jmcrawford
@jmcrawford
Jan 06 2017 03:02
chasing the kids towards bed, will be back in a while
Aaron Michelson
@Aaybob
Jan 06 2017 03:03
in my iphone settings>bluetooth I have listed edison . I connect any time i tap the row. it stays connected about five seconds.
bluetoothctl, commands: power on, scan on, pair <phoneaddress>, connect <phadd>, trust <phadd>.
jmcrawford
@jmcrawford
Jan 06 2017 03:08
@scottleibrand oref0@0.3.4-dev
Scott Leibrand
@scottleibrand
Jan 06 2017 03:09
@jmcrawford ok good. not sure if all the attempts have been successful yet, but take a look at what oref0-setup does to run the bluetoothd stuff
this stuff is all still a work in progress, as you probably noticed
Aaron Michelson
@Aaybob
Jan 06 2017 03:22
the only thing I added was connect <phone address> during bluetoothctl.
and I never saw edison ask to confirm pair request
I have a new problem...
nightscout is not getting pump information
I am using a pump I haven't used for a while
Aaron Michelson
@Aaybob
Jan 06 2017 03:28
edison is setting temp basals and nightscout is getting dexcom
Dana Lewis
@danamlewis
Jan 06 2017 03:28
did you change the serial number?
Aaron Michelson
@Aaybob
Jan 06 2017 03:28
edison serial number is correct. Is there a nightscout serial number input?
Dana Lewis
@danamlewis
Jan 06 2017 03:29
pump serial.. ah, you said it is successfully setting temps? good
Aaron Michelson
@Aaybob
Jan 06 2017 03:34
would the only possiblilty be my input of nightscout login during oref0-setup?
I had two rigs transmitting to nightscout one RPi one edison. RPi was my working system. I'm switching to edison but in the last four hours nightscout stopped getting loop or pump data
when I plugged a battery into edison for the first time??
I want to go out with the dog and test my ble tether
Nightscout stopped getting loop data while RPi was still in use.
Aaron Michelson
@Aaybob
Jan 06 2017 03:42
maybe im going over data usage on azure but im still getting cgm data on nightscout
also when i got success with ble tethering i had done "forget" all other bluetooth connections
Aaron Michelson
@Aaybob
Jan 06 2017 03:50
azure went to yellow zone in memory usage, reset in 10 more minutes so will see
Aaron Michelson
@Aaybob
Jan 06 2017 04:24
nightscout is back up with RPi :smile: I think I used wrong secret name with edison install (a capital letter) but is it okay if two rigs send nightscout data?
BLE tether tested good
Scott Leibrand
@scottleibrand
Jan 06 2017 04:35
Two rigs is fine. Two pumps pushing to one NS not so good.
jmcrawford
@jmcrawford
Jan 06 2017 04:38
@Aaybob If you reboot your edison, what steps do you need to do to get BT tethered again?
Mitchell Slep
@moomoobloo
Jan 06 2017 05:21

Anyone have suggestions for how to debug a bad COB calculation from AMA?

According to git, here's what happened to monitor/meal.json after bolusing 0.7 U for 18g of carbs using the pump's bolus wizard:

- {"carbs":37,"boluses":1.2,"mealCOB":0}
+{"carbs":55,"boluses":1.9,"mealCOB":53}

carbs and boluses were updated correctly but mealCOB is totally wrong, which lead to a temp basal that was way too high.

I synced back to the bad commit, but when I run openaps report invoke monitor/meal.json off of that state, I get something that looks more reasonable (although maybe still wrong- why did mealCOB go up by more than 18?):

- {"carbs":37,"boluses":1.2,"mealCOB":0}
+{"carbs":55,"boluses":1.9,"mealCOB":21}
Scott Leibrand
@scottleibrand
Jan 06 2017 05:29
The carb stacking code is complicated, unfortunately. Were there any gaps in BG data or anything else that was "odd"?
You might eventually need to start adding unit tests to debug this thoroughly.
Mitchell Slep
@moomoobloo
Jan 06 2017 05:35
Screen Shot 2017-01-05 at 9.31.56 PM.png
No gaps, but immediately before the BG jumped form 67 to 116 in one reading, almost like a calibration (it's a new sensor)
Aaron Michelson
@Aaybob
Jan 06 2017 05:41
how close to edison do you keep the pump?
Mitchell Slep
@moomoobloo
Jan 06 2017 05:42
Aside from the underlying bug, any idea why it would not be reproducible after syncing back in git? Only thing I can think of is that carbhistory.json is not an openaps report and therefore not tracked in git, but I'm using a pretty recent version and if that were the issue I think carbs wouldn't match up when I try to repro.
Unless there is something non-deterministic about the code.
Aaron Michelson
@Aaybob
Jan 06 2017 05:57
loop with bluetooth tethering errors out with
mmeowlink.vendors.serial_interface.AlreadyInUseException: /dev/spidev5.1 already in use
.mmtune:mmtune: Starting pump-loop at Thu Jan 5 21:52:03 PST 2017:
Radio ok, Listening: .Usage: grep [OPTION]... PATTERN [FILE]...
Try 'grep --help' for more information.
Listening: ...........................................................No pump comms detected from other rigs
Old pumphistory: .
Error, retrying
Listening: .......
Scott Leibrand
@scottleibrand
Jan 06 2017 06:02
@moomoobloo perhaps you didn't manage to get the "now" pointer sync'd back? Not sure.
Mitchell Slep
@moomoobloo
Jan 06 2017 06:04
What do you mean by "now" pointer? I did git checkout <commit>, anything else I need to do?
I also copied the entire openaps directory to a new location and reproduced from there, so the active loop would not interfere.
Aaron Michelson
@Aaybob
Jan 06 2017 06:09
returning to master git wants me to commit or stash changes. Is there a simple workaround for mkdir -p ~/src; cd ~/src && git clone git://github.com/openaps/oref0.git || (cd oref0 && git checkout master && git pull)
I think I'll go another round with ble tethering. I had two competing crons messing things us.
Aaron Michelson
@Aaybob
Jan 06 2017 06:33

what happened?

Continue? y/[N] y
Checking /root/myopenaps: Initialized empty Git repository in /root/myopenaps/.git/
[master (root-commit) 524ff93] initializing openaps 0.1.5
1 file changed, 0 insertions(+), 0 deletions(-)
create mode 100644 openaps.ini
/root/myopenaps initialized
/root/src/oref0/ already exists; pulling latest
fatal: unable to connect to github.com:
github.com: Name or service not known

Couldn't pull latest oref0

Scott Leibrand
@scottleibrand
Jan 06 2017 06:46
You have no Internet.
@moomoobloo what determines whether it's calculating COB as of the commit time, vs. as of the rig time?
I haven't tested retrospective runs much.
Aaron Michelson
@Aaybob
Jan 06 2017 06:49
home network has internet, RPi has internet, internet cant get to edison. interesting
maybe its time to reboot home computer
or call it a night
Thomas Emge
@ThomasEmge
Jan 06 2017 06:58
I rebuild my RPI2 w TI with the setup script and during the night I am seeing that the rig is adjusting the temps downwards where I would expect it to go the opposite way. Here is the content of the suggested.json {"temp":"absolute","bg":221,"tick":"+1","eventualBG":227,"snoozeBG":227,"predBGs":{"IOB":[221,222,223,224,224,225,225,226,226,226,226,227,227]},"IOB":0,"reason":"COB: undefined, Dev: 6, BGI: 0, ISF: 40, Target: 100; Eventual BG 227 >= 100, max_iob 0, 26m@0.000 = 0.000 > 2 * req 0. Setting temp basal of 0.8U/hr, but 26m left and 0.8 ~ req 0.8U/hr: no action required"}. What am I missing in my setup?
JELCRAWFORD
@JELCRAWFORD
Jan 06 2017 07:16
More questions here about COB - supper here was finished around 7:30pm, COB had zeroed out (per APS calculation) around 11pm. Now, at 12:15am, we have a reading of 56 COB - son has been sleeping since about 8:45pm. Unless dreaming about hamburgers can result in actual COB, I'm a bit confused. We've also had a steady rise into the middle 14's-15's (250-270's) in the past few nights - this is not new, but previously OpenAPS had held it to lower numbers - mid 9's - mid 10's (up to 180) pretty nicely. Feeling we are missing something with the COB and not sure how it relates to the relentless spike which was previously held nicely by OpenAPS, but not in the past few nights. Any insights/thoughts? Is the experienced change in COB to be expected or something amiss?
JELCRAWFORD
@JELCRAWFORD
Jan 06 2017 07:25
now at 69 COB. Sound asleep.
jmcrawford
@jmcrawford
Jan 06 2017 07:35
Is there a trick to mounting a flash drive via OTG on Explorer? I see messages logged that it detects something connected, but no device is created.
Scott Leibrand
@scottleibrand
Jan 06 2017 07:57
@ThomasEmge you pasted it. max_iob 0
@JELCRAWFORD sounds like a bug, but not sure what it'd be, and shouldn't really affect anything unless BG starts to rise (and then it'll actually cause it to high-temp more, not less)
jmcrawford
@jmcrawford
Jan 06 2017 08:06
blob
@scottleibrand @JELCRAWFORD really messes with AMA projections
Scott Leibrand
@scottleibrand
Jan 06 2017 08:07
The top one doesn't matter for anything. It's just that middle one it actually uses.
JELCRAWFORD
@JELCRAWFORD
Jan 06 2017 08:08
@scottleibrand - well, BG has climbed up high - but our max safe basal rate prevented what the pancreas was asking for, (though suggested rates didn't seem unreasonable, and certainly no less reasonable than my tendency to rage bolus in response to similar frustrating spikes that are painfully slow to respond to correction) so its a good test of safety measures, if nothing else. Truthfully, might have appreciated a sharp drop so bedtime for the adults was sooner though...... thx for thoughts. I'm not sure how to debug, but will look through data and see if I can provide anything useful. See graph above from @jmcrawford w/pill indicting current COB. Must be a wicked good food dream :)
Scott Leibrand
@scottleibrand
Jan 06 2017 08:08
And looks like that COB hasn't prevented it from low temping for the drop.
Although looks like you have basically no basal overnight?
JELCRAWFORD
@JELCRAWFORD
Jan 06 2017 08:09
nope, but insulin needs after our spike are generally very low
Thomas Emge
@ThomasEmge
Jan 06 2017 08:09
@scottleibrand hmm, so where is this max_iob coming from? I am not seeing it in the suggested.json on my Edison configuration.
Scott Leibrand
@scottleibrand
Jan 06 2017 08:10
Take another look at the docs
There is a whole section on that.
JELCRAWFORD
@JELCRAWFORD
Jan 06 2017 08:10
so not much to low temp - makes correction tricky, but OpeanAPS had smoothed things out significantly prior to past few nights again... growth spurt maybe, gremlins, food dreams - who knows. Pancreas is still better than us at trying to control. NS alarm lets us know if we need to carb correct.
Scott Leibrand
@scottleibrand
Jan 06 2017 08:11
I wonder what autotune would think of that.
JELCRAWFORD
@JELCRAWFORD
Jan 06 2017 08:14
haven't turned on at this point... will look at details. It's in the dev file, right? expect that to be on edison once we get thigns sorted there... @jmcrawford still working hard on BLE, but may start with wifi - harder with shifts to and from school, recess etc. BLE setup would be ideal. One step at a time. Really appreciate the thoughts though - many thx.
Thomas Emge
@ThomasEmge
Jan 06 2017 08:16
@scottleibrand I might have been starring at my preferences file for too long. Thanks for pointing it out.
Scott Leibrand
@scottleibrand
Jan 06 2017 08:17
:+1:
PieterGit
@PieterGit
Jan 06 2017 08:38
@scottleibrand : sorry: had no time to PR stuff. first trying to get dexusb on edison working and having other priorities the coming days.
Tim Street
@tim2000s
Jan 06 2017 11:21
Has anyone with a WW pump got the Dev set-up branch running, and using the bluetooth tethering? Obviously there were a few issues with the pump not being found previously.
Tim Street
@tim2000s
Jan 06 2017 12:18
Interesting! My first Edison/Explorer flashed with no issues whatsoever. The second one has gone completely up the swanny.
Chris Oattes
@cjo20
Jan 06 2017 12:53
Is "up the swanny" a technical term? :P What's it doing wrong?
Aaron Michelson
@Aaybob
Jan 06 2017 12:56

The latest build from git/openaps does not have "preflight" support for MiniMed Carelink. Would someone help me PR to add:

openaps alias add preflight '! bash -c "rm -f monitor/clock.json && echo -n \"PREFLIGHT \" && openaps report invoke monitor/clock.json 2>/dev/null >/dev/null && grep -q T monitor/clock.json && echo OK || ( ( mm-stick warmup 2>&1 || sudo oref0-reset-usb ) | egrep -v \"^ \"; echo FAIL; openaps get-bg; sleep 120; exit 1 )"' || die "Can't add preflight"

and then put it in cron or a loop for the rigs using CareLink. I have only PR'd to docs.

Aaron Michelson
@Aaybob
Jan 06 2017 13:06

I could PR a few steps to docs for CareLink users. with a cron edit

          • cd /home/pi/myopenaps && ( ps aux | grep -v grep | grep -q 'openaps pump-loop' || openaps preflight && openaps pump-loop ) 2>&1 | tee -a /var/log/openaps/pump-loop.log

tested pump-loop.log result is:
Completed pump-loop at Fri 6 Jan 05:04:00 PST 2017

PREFLIGHT OK
Starting pump-loop at Fri 6 Jan 05:04:10 PST 2017:

Aaron Michelson
@Aaybob
Jan 06 2017 13:12
I have only experience PRing to docs without a full grasp of the process therefore I am asking someone to help me PR to software. Do you think a docs PR is preferable?
Brian Rabinovitz
@bdr1177
Jan 06 2017 13:52
@Aaybob I don't have the experience to do the software PR either. For the docs cron edit, where would the preflight code be inserted?
Aaron Michelson
@Aaybob
Jan 06 2017 13:53
look to the right in this gitter.im window. do you see Aaybob edited #523?
go into that link and sellect the tab that shows three files edited, the most recent file added there has my updates to the documentation.
Brian Rabinovitz
@bdr1177
Jan 06 2017 13:55
@Aaybob Yes, I see it now. I was on my phone before so I didn't have full gitter functionality.
Aaron Michelson
@Aaybob
Jan 06 2017 14:15
congratulations on first PR @bdr1177
Brian Rabinovitz
@bdr1177
Jan 06 2017 14:16
@Aaybob Thanks!
Tim Street
@tim2000s
Jan 06 2017 14:23

Is "up the swanny" a technical term? :P What's it doing wrong?

It went into a loop then appeared to miss part of the dry set up - I need to run through the troubleshooting. The roots never properly loaded and I can't get console access to work now.

*dry = dfu
Chris Oattes
@cjo20
Jan 06 2017 14:26
Have you tried bringing up the serial connction before turning the board on? You should be able to do that, then stop the autoboot. That'll give you a console where you can enter run do_flash (I think) to try reflashing it
Tim Street
@tim2000s
Jan 06 2017 14:39
I get to
[  OK  ] Started File System Check on /dev/disk/by-partlabel/home.
         Mounting /home...
[  OK  ] Mounted /home.
         Starting Rescue Shell...
[  OK  ] Started Rescue Shell.
[  OK  ] Reached target Rescue Mode.
I've not been ale to stop autoboot
Chris Oattes
@cjo20
Jan 06 2017 14:42
at that point you should be able to type "reboot" and it'll restart the board, and ask you to press a key to stop autoboot
Does anyone know why we send 500 "PowerControl" messages to the pump over ~8 seconds to get it to wake?
Tim Street
@tim2000s
Jan 06 2017 15:04
Got it this time - wasn't hitting the enter button early enough
Tim Street
@tim2000s
Jan 06 2017 15:14
Now there's interesting:
jubilinux tim$ sudo ./flashall.sh
Error: Running Homebrew as root is extremely dangerous and no longer supported.
As Homebrew does not drop privileges on installation you would be giving all
build scripts full access to your system.
Tim Street
@tim2000s
Jan 06 2017 15:23
When it comes to installing the rootfs, i get this in the dfu log:
dfu-util: Device has DFU interface, but has no DFU functional descriptor
Deducing device DFU version from functional descriptor length
dfu-util: No DFU capable USB device available
dfu-util 0.9
Chris Oattes
@cjo20
Jan 06 2017 15:31
did you run do_flash?
on the edison
Aaron Michelson
@Aaybob
Jan 06 2017 15:42

I'm adding a lot to the documents but I don't see the changes happening. I experiment, edit the docs and the docs are not edited. I think that git is too complicated for text editing. So if someone wants to add this to the documentation it would be appreciated: the section is BLE Tethering (experimental) , the command is "npm run global install", the fix is "Running this code will install all of the dependencies for you automatically:

curl -s https://raw.githubusercontent.com/openaps/docs/master/scripts/quick-packages.sh | bash -
"
to install npm before trying to run npm.
There are other updates lost somewhere in a git repository for a git expert to do something with.

Aaron Michelson
@Aaybob
Jan 06 2017 15:51
Or not because I've been deleting my repository to present a new edit.
Chris Oattes
@cjo20
Jan 06 2017 15:52
Aaybob: what do you mean by "I don't see the changes happening"?
Dana Lewis
@danamlewis
Jan 06 2017 15:52
@Aaybob are you doing command line doc edits, or using the web gui? I use the web gui myself
Aaron Michelson
@Aaybob
Jan 06 2017 15:53
i can't learn git,
it's been a year its not happening
I suggest we take documentation off git.
Chris Oattes
@cjo20
Jan 06 2017 15:56
@Aaybob your changes are showing up in your pull request
Where aren't you seeing your changes appear?
Aaron Michelson
@Aaybob
Jan 06 2017 15:57
emailing text files would work
i'm not ready to try to use git, its making me cry
Chris Oattes
@cjo20
Jan 06 2017 16:00
@Aaybob I see your changes, I don't understand where you aren't seeing the changes happening. If you tell me what you were expecting to change, I might be able to help
Tim Street
@tim2000s
Jan 06 2017 16:04
@cjo20 Yes - ran it according to the instructions - re-ran the flashall.sh alongside doing a run do_flash on the edison
Dana Lewis
@danamlewis
Jan 06 2017 16:04
@Aaybob I think what is likely happening is since you have an open pull request, the additional changes you are making to files in the docs are getting added to the open pull request automatically. Take a look at openaps/docs#523 - you can scroll down and see some of the things you’ve been working on this morning
JELCRAWFORD
@JELCRAWFORD
Jan 06 2017 16:24
@danamlewis - headed to the Endo on Monday. Emailed in advance with OpenAPS docs link and a heads up that we've closed the loop at home. She is very open - but I think no experience at this point. Also, closed loop, even commercial, will be longer coming as we are in Canada, so not many docs/clinics going down this path as yet. At clinic there is very limited experience with any tech besides the pump. Aware of NS, but no experience there either. Even for CGM - very limited knowledge other than they can confirm it exists and give you contact numbers for purchase - so very big knowledge gap overall. Can you suggest, other than formal docs for both NS and Openaps any other specific resources that would be good to pass along? Do you have a current estimate of # PWD's on closed loops using OpenAPS. I did have coffee, with my son along, with one of the head hospital admins yesterday as we are friendly - she was very interested/curious about our devices and how they work and how they provide benefits also - when there is a personal connection I think it makes generating awareness/interest easier. Good resources go a long way to encourage interest and awareness once piqued. Anything specific you (or others) can recommend? Thx in advance for any pointers to help us support the broad cause with folks here.
Dana Lewis
@danamlewis
Jan 06 2017 16:25
in addition to the docs, I would provide a direct link to the reference design (https://openaps.org/reference-design/) for a high level explanation of how the system is designed. (docs are likely to overwhelm, although good for them to have access for any specific Qs).
the early self-reported outcomes study may or may not be useful, too - here’s the blog post version of the data (https://openaps.org/2016/06/11/real-world-use-of-open-source-artificial-pancreas-systems-poster-presented-at-american-diabetes-association-scientific-sessions/), there’s also a LTE in JDST you could print instead if you think the endo would care more about seeing it in a publication
Chris Oattes
@cjo20
Jan 06 2017 16:27
https://diyps.org/ has a lot of great information, presented at a much better level than the docs
Tim Street
@tim2000s
Jan 06 2017 16:29
you can also have a look at this: http://www.diabettech.com/artificial-pancreas/loop-and-openaps-down-and-dirty/ which gives a higher level overview of the algorithms involved.
Dana Lewis
@danamlewis
Jan 06 2017 16:30
(thanks @cjo20 :)) - fwiw, i’ve tried to create a list of the posts I see people discussing most often: https://diyps.org/top-posts/
Tim Street
@tim2000s
Jan 06 2017 16:30
@cjo20 Been trying to flash plain ubilinux as well - it hangs at the same point...
off to mmeowlink
Chris Oattes
@cjo20
Jan 06 2017 16:31
do you have a windows pc handy? I think I've had that problem on a linux box before
Dana Lewis
@danamlewis
Jan 06 2017 16:31
@JELCRAWFORD and re: number of people, out of the entire community of DIY loopers that have shared that they are looping using the variety of tools out there (so this is purely an estimate), looks like there is (n=1)*202+ DIY closed loopers worldwide.
Tim Street
@tim2000s
Jan 06 2017 16:33
@cjo20 Let me give it a go.
JELCRAWFORD
@JELCRAWFORD
Jan 06 2017 16:33
@danamlewis - Already provided ref. design. It is a very good and straight forward explanation - it is well done. Will print both your references for Endo and also for hospital admin - yes, I think in that setting, publications go further. Already printed the the abstract of your above self-report outcomes - also good. Thx also @tim2000s and @cjo20 Will review all that you've all provided and select as well as provide digitally. It's encouraging to me when those who may not have daily experiences with D show a genuine interest in what we are doing (as a community), particularly when some of them are decision makers at a higher level :) Thx for numbers. I expect Endo may ask. That estimate, from your comments, includes Loop also then in add. to Openaps I assume.
Dana Lewis
@danamlewis
Jan 06 2017 16:34
yes, of those who have told us they are looping from individual implementations; OpenAPS; Loop; AndroidAPS; etc.
:+1:
you’re probably way good to go for now, but if needing additonal materials, i’ve been trying to keep an updated list of mainstream news or publications here: https://openaps.org/in-the-news/ (and for anyone else who needs it)
JELCRAWFORD
@JELCRAWFORD
Jan 06 2017 16:42
Will bookmark all of those for going forward. If I compile anything I can send along or post for others also. At only 8mos in for us, after hooking up a used insulin pump in our living room(prev. long waits here, though we got traction on fixing that) and now implementing Opeanaps, it is a good think we drew the endo we did - she's very supportive and collaborative with us and I think will be a good starting point in our setting for pushing greater awareness (at least) of technology and the potential impact for those that choose it. So 2006 wasn't ALL lucky for us, but mostly :)
Dana Lewis
@danamlewis
Jan 06 2017 16:42
:) :clap:
JELCRAWFORD
@JELCRAWFORD
Jan 06 2017 16:47
@danamlewis documentary from Dec 1, 2016 - comes up needing a password... had been meaning to ask the community about this as we wanted to view. Am I missing some other way to view it that doesn't require a password?
Dana Lewis
@danamlewis
Jan 06 2017 16:48
I think Ali downloaded a copy , let me see if I can re-upload that later today. originally wasn’t locked. (rar)
JELCRAWFORD
@JELCRAWFORD
Jan 06 2017 16:49
thx!
Tim Street
@tim2000s
Jan 06 2017 17:45
Hmmm. I'm not quite portable yet. While I can see ed209 has connected to my iPhone and tethering is allowed, it's not getting data from Nightscout and not posting.
Meanwhile, itnworks when connected to my home wifi.
Rather odd.
Chris Oattes
@cjo20
Jan 06 2017 17:59
Really chrome? You can't put that image the right way round -_-
Chris Oattes
@cjo20
Jan 06 2017 18:20
@MilosKozak About detecting mistakes; if something isn't explicitly wrong or definitely a safety issue, it should probably be a warning that it might not be what was intended, rather than a hard error
Milos Kozak
@MilosKozak
Jan 06 2017 18:25
well it depends. 0.025 whole day is not supported by pump and it's dangerous to use because of possible jam in tubing, inaccuracy etc. i prefer in this case not to do any automated operation and make user to fix it first. at least i use this approach in AAPS
Chris Oattes
@cjo20
Jan 06 2017 18:26
It is supported by some pumps. You can set some pumps to deliver 0u/hr basal all day if you want.
Sgoldberg89
@Sgoldberg89
Jan 06 2017 18:28
So my second loop died today and now I'm loopless. :( wondering if anyone has had this problem, my loop just keeps repeating this: Starting pump-loop at Fri 6 Jan 10:26:41 PST 2017:
Radio ok, Listening: .No comms detected
Old pumphistory:
Error, retrying
Joe
@t1djoe
Jan 06 2017 18:29
+1 @Sgoldberg89 Working through this same issue right now.
Sgoldberg89
@Sgoldberg89
Jan 06 2017 18:30
@t1djoe any ideas?
Scott Leibrand
@scottleibrand
Jan 06 2017 18:30
@cjo20 I think this one is defined as a safety issue. our algorithm doesn't work safely if it can't low-temp, and doesn't know how high it can safely high-temp
Chris Oattes
@cjo20
Jan 06 2017 18:30
@scottleibrand is being able to low temp by 0.1u/hr enough then?
Scott Leibrand
@scottleibrand
Jan 06 2017 18:34
possibly. the way I think about it: if someone comes in with a pump profile that sorta makes sense, we'll do our best. but if they're coming in with something that is incompatible with closed looping (like zero basal for part of the day) we would need to actually examine that use case before we decided what to do. in 150+ users, including infants with tiny basals, we haven't needed to support anything less than 0.1U/hr. that check was added as a safety feature that makes sure that the data coming from decocare is sane. relaxing the safety checks for some future hypothetical situation we can't even define would make things less safe for everyone else, for no gain.
if this were prompted by someone with an actual need to set < 0.1U/hr basals, I would be more receptive.
Joe
@t1djoe
Jan 06 2017 18:36
@Sgoldberg89 Not so far. Working back through some changes I made earlier to see if anything there broke it.
Scott Leibrand
@scottleibrand
Jan 06 2017 18:37
@Sgoldberg89 @t1djoe I would killall -g openaps; openaps mmtune, killall -g openaps; openaps use pump model etc. to see if you get any more details on why it can't communicate
JELCRAWFORD
@JELCRAWFORD
Jan 06 2017 18:40
@scottleibrand @MilosKozak - we use .025 daily at times, as well as use a zero basal sometimes (not on vacation as he isn't active enough and insulin needs are greater) as our son has very low insulin needs sometimes. Have not had any difficulties with the pump. Know another family that also has used zero and very low basal - no problems. That said, we, at times, don't have much cushion for lows, but that is not different from pump alone, so we don't view it as an APS problem, just a diabetes problem.
Chris Oattes
@cjo20
Jan 06 2017 18:40
The part that makes it feel like it's a bit pointless trying to restrict it is because it's saying that 0.1u/hr all day is perfectly fine, but 0.075u/hr all day is unsafe. The difference between the two is so miniscule. If someone has an isf of 180, which is insanely sensitive, oref0 will be able to drop them by 18/hr if they're on 0.1u/hr, or 13.5/hr if they're on 0.075.
Scott Leibrand
@scottleibrand
Jan 06 2017 18:41
@JELCRAWFORD have you ever had the openaps complain your basal is too low?
JELCRAWFORD
@JELCRAWFORD
Jan 06 2017 18:42
openaps will not work with zero basal, we had to increase to .025 to make it work.
Scott Leibrand
@scottleibrand
Jan 06 2017 18:42
and in that situation, it will only high-temp as high as 0.1U/hr: is that sufficient?
(without changing the 4x max current basal multiplier)
JELCRAWFORD
@JELCRAWFORD
Jan 06 2017 18:46
@scottleibrand - no it is not sufficient in some cases - we have an additional setting for nighttime which increases or max current basal multiplier because of our aforementioned massive spike in the nighttime hours. We feel ok doing this because we always watch carefullly at this time (yes, limited sleep, like many others) and we revert to more conservative numbers in the daytime hours (12:30pm and onward till 7pm again) where he is not necessarily watched like a hawk, so to speak.
Scott Leibrand
@scottleibrand
Jan 06 2017 18:50
k. as you get experience with that, think about ways we could modify the system to make it easier for you to use (not have to change as many things manually) and/or anything we can do to make it safer for widely varying insulin needs like that. sounds like you won't be affected by the max daily basal minimum, as it's higher during the day
JELCRAWFORD
@JELCRAWFORD
Jan 06 2017 18:53
@scottleibrand actually, husband has written something to automate the change to our settings back and forth each night. Maybe could put that forward as a change or optional? The weird nighttime numbers (endo agrees the insulin needs are atypical - from very high needs and very low needs each day).
Sgoldberg89
@Sgoldberg89
Jan 06 2017 18:54
@scottleibrand trying to understand your tip, can you clarify how to use those commands? When I copy and paste "killall -g openaps; openaps mmtune" I get: [Errno 2] No such file or directory
openaps: 'mmtune' is not an openaps command.
Scott Leibrand
@scottleibrand
Jan 06 2017 18:55
@JELCRAWFORD @jmcrawford why don't you write up a github issue describing what you're doing and why, and we can brainstorm ways to bake that in properly with a schedule or something, so you don't have to have something change it all the time...
@Sgoldberg89 are you in your myopenaps directory?
Sgoldberg89
@Sgoldberg89
Jan 06 2017 18:56
@scottleibrand ah...
looks like the second of those "killall -g openaps; openaps use pump model" spat out a bunch of info, not sure if any of it is helpful though
Scott Leibrand
@scottleibrand
Jan 06 2017 19:04
I know it's not very helpful if I don't know what it says... ;-)
Aaron Michelson
@Aaybob
Jan 06 2017 19:05
@JELCRAWFORD At one point endo told me I would have to eat because we couldn't set basal low enough. However the 1/2 year I used afrezza I was able to analyze the bolus affect on the basal. I increased basal up to near my night rate and would bolus one time for two meals when I didn't use afrezza. The "tail" of my bolus adds so much to what I took to be basal effect. Afrezza is an insulin bolus directly into the blood stream in effect but caused respiratory problems.
Sgoldberg89
@Sgoldberg89
Jan 06 2017 19:05
"723"Traceback (most recent call last):
File "/usr/local/bin/openaps-use", line 4, in <module>
import('pkg_resources').run_script('openaps==0.1.5', 'openaps-use')
File "/usr/lib/python2.7/dist-packages/pkg_resources.py", line 534, in run_script
self.require(requires)[0].run_script(script_name, ns)
File "/usr/lib/python2.7/dist-packages/pkg_resources.py", line 1438, in run_script
execfile(script_filename, namespace, namespace)
File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/EGG-INFO/scripts/openaps-use", line 63, in <module>
app( )
File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/openaps/cli/init.py", line 52, in call
self.epilog( )
File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/EGG-INFO/scripts/openaps-use", line 50, in epilog
super(UseToolApp, self).epilog( )
File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/openaps/cli/init.py", line 75, in epilog
self.create_git_commit( )
File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/openaps/cli/init.py", line 82, in create_git_commit
self.git_repo( )
File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/openaps/cli/init.py", line 78, in git_repo
self.repo = getattr(self, 'repo', Repo(os.getcwd( ), odbt=GitCmdObjectDB))
File "/usr/local/lib/python2.7/dist-packages/git/repo/base.py", line 156, in init
raise InvalidGitRepositoryError(epath)
git.exc.InvalidGitRepositoryError: /home/pi/openaps
Scott Leibrand
@scottleibrand
Jan 06 2017 19:06
InvalidGitRepositoryError would be your issue. need to figure out why oref0-reset-git isn't fixing that
want to go ahead and run that manually?
Sgoldberg89
@Sgoldberg89
Jan 06 2017 19:10
Sure
Joe
@t1djoe
Jan 06 2017 19:10
Very peculiar. I ran the commands @scottleibrand recommended, with no errors. mmtune completes, I'm able to read model, reservoir, status, etc. with no issues.
Sgoldberg89
@Sgoldberg89
Jan 06 2017 19:10
Is there an easy way to do that?
Aaron Michelson
@Aaybob
Jan 06 2017 19:14
@Sgoldberg89 you could try : git config --global user.email on the command line to see if it has your email correct
then : git config --global user.name that should be the data you opened up an account on git with
Sgoldberg89
@Sgoldberg89
Jan 06 2017 19:14
That's all correct
*The readout is correct
PieterGit
@PieterGit
Jan 06 2017 19:15
@t1djoe all the commands that will change the git repository will fail.
@scottleibrand i also have a broken git repo that i need to find out why fix-git doesn't fix it. i think openaps should not fail if it can't commit to git and just proceed
Aaron Michelson
@Aaybob
Jan 06 2017 19:16
I've no git expertice. Try going to your account and see what you can find out.
Joe
@t1djoe
Jan 06 2017 19:18
@PieterGit Not sure a git issue is causing my problem.
PieterGit
@PieterGit
Jan 06 2017 19:22
@t1djoe maybe there was another openaps process locking a resource. are there any problems your currently facing? i tried scrolling back, but couldn't find a description of your problem (I found various remarks on git)
Aaron Michelson
@Aaybob
Jan 06 2017 19:24
@PieterGit checked crontab -l ? should see just one directory path in the scheduled jobs and about ten or so lines
Scott Leibrand
@scottleibrand
Jan 06 2017 19:25
@Sgoldberg89 you should just be able to run oref0-reset-git from your myopenaps directory
Aaron Michelson
@Aaybob
Jan 06 2017 19:25
@Sgoldberg89 if you've been repeatedly installing with new names check crontab -l
Scott Leibrand
@scottleibrand
Jan 06 2017 19:25
@t1djoe you probably need to start looking at the aliases (openaps alias show pump-loop or whatever) and start running their components manually until you can reproduce the problem
Joe
@t1djoe
Jan 06 2017 19:25
@PieterGit I'm seeing output like this in my pump-loop.log file: Radio ok, Listening: .No comms detected
Old pumphistory:
Error, retrying
@scottleibrand That's my current path. :D
PieterGit
@PieterGit
Jan 06 2017 19:29
@t1djoe what does openaps report invoke monitor/status.json return?
some aliases output stuff to /dev/null. You can see in openaps.ini what openaps is calling to debug what is working (and what is not)
Aaron Michelson
@Aaybob
Jan 06 2017 19:30
@t1djoe @PieterGit you should stop cron
Joe
@t1djoe
Jan 06 2017 19:32
@Aaybob Already stopped cron. :D
@PieterGit pump://JSON/status/monitor/status.json
reporting monitor/status.json
jmcrawford
@jmcrawford
Jan 06 2017 19:37
@scottleibrand @JELCRAWFORD just two copies of preferences.json, with a couple of cron entries to copy them back and forth. Don't like leaving current_basal_safety_multiplier that high all the time, defeats its purpose. But we need it high in the evenings to deal with the spike.
Scott Leibrand
@scottleibrand
Jan 06 2017 19:41
yeah. does the current_basal_safety_multiplier end up being the limiting factor at any point during the day? or are you always limited by the pump max basal rate and/or the max_daily_basal_safety_multiplier?
JELCRAWFORD
@JELCRAWFORD
Jan 06 2017 19:55
@Aaybob Dont' think we have that insulin option in Canada - we use Apidra. Hate 'feeding' insulin - prefer to feed hunger and try to maintain as close to 'normal' eating behaviours as possible. Our kiddo has wildly variable basal rates from zero (not when pancreas runs as it won't work then) to .8 - sometimes higher for our nighttime spike. OpenAPS has given us addt'l info as we watch it try to correct/manage throughout the day. As far as zero basal and concerns about that - our endo had no concerns, other families said they'd been told not to do that. Seems there are many 'rules' that you have to use experience and judgment to decide if they work in your individual situation.
Aaron Michelson
@Aaybob
Jan 06 2017 20:16
@JELCRAWFORD I wish I had an easier suggestion. Aren't we are finding that the rules did not work? You always take a bolus so adjust basal up, arrange eating schedule to match the bolus. Bolus breakfast eat carbs and protein as best to match, three hours later or when BG starts to drop make that the eat time. Seven hours later insulin is becoming basal dominant so you can restart that process. Sorry in advance.
JELCRAWFORD
@JELCRAWFORD
Jan 06 2017 20:19
@scottleibrand I think maxsafebasal is our usual limitation now that Mike changed current_basal_safety_multiplier. Needs in the day for bolus rates are generally much lower than the needs at 7:30pm - sometime around midnight.... on a bad night till 1am. Mostly APS manages it well, but for the spike which has always invoked a fair amount of swearing since the beginning (again, aps still usually does a better job than we do). Last night, for example, b/c of COB issue, I used Temp Target and exercising soon to get a zero basal as he came down quickly. Didn't completely avert the low - still dropped below a 4.4 and I treated, but mitigated what would have otherwise been a more moderate low if APS estimate had been correct - no real fears about that thou, as NS gives solid warning along with pushover warnings and the dex receiver itself. Overall the situation, independent of APS, needs careful monitoring and now less 'action' than before to not be sky high. Response, with our without APS sometimes results in lows. One very important factor for us which improved the situation was adjustment to DIA - longer than we were previously using - this made such a difference I have wondered if the documentation needs an addition to emphasize this as an important 'tweak' for best results... need to re-read and reflect about if it needs more emphasis or if we just missed the emphasis because of all the other things we were considering.
@aaybob - no need for sorry - we've quickly come to learn that the 'rules' are flexible and what seems to work today may not work tomorrow but may work next week. Black and white is easier, but almost nonexistent for all of us! The only black and white that I see is that if you are trying to maintain optimal bg levels (individual of course), it is a lot of work. APS makes it easier - YAY! - but doesn't remove that fact, which is plainly demonstrated by the fact that this board is humming most hours of the day, slowing down from 1am to 6am only.....
Scott Leibrand
@scottleibrand
Jan 06 2017 20:42
:+1:
Joe
@t1djoe
Jan 06 2017 20:43
@scottleibrand Root issue appears to be, after executing openaps use pump iter_pump_hours 27 I get the following error:
Traceback (most recent call last): File "/usr/local/bin/openaps-use", line 6, in <module> exec(compile(open(__file__).read(), __file__, 'exec')) File "/root/src/openaps/bin/openaps-use", line 63, in <module> app( ) File "/root/src/openaps/openaps/cli/__init__.py", line 51, in __call__ self.run(self.args) File "/root/src/openaps/bin/openaps-use", line 57, in run output = app(args, self) File "/root/src/openaps/openaps/uses/__init__.py", line 92, in __call__ return self.method.selected(args)(args, app) File "/root/src/openaps/openaps/uses/__init__.py", line 31, in __call__ return self.method(args, app) File "/root/src/openaps/openaps/uses/use.py", line 45, in __call__ output = self.main(args, app) File "/root/src/openaps/openaps/vendors/medtronic.py", line 640, in main for rec in self.range( ): File "/usr/local/lib/python2.7/dist-packages/decocare/models/__init__.py", line 67, in __call__ for record in page: File "/usr/local/lib/python2.7/dist-packages/decocare/models/__init__.py", line 46, in download_page for record in self.find_records(page): File "/usr/local/lib/python2.7/dist-packages/decocare/models/__init__.py", line 157, in find_records decoder = history.HistoryPage(response.data, self.inst) File "/usr/local/lib/python2.7/dist-packages/decocare/history.py", line 777, in __init__ if lib.BangInt(crc) != computed: File "/usr/local/lib/python2.7/dist-packages/decocare/lib.py", line 235, in BangInt ( x, y ) = ints ValueError: need more than 0 values to unpack
If it impacts things, I'm currently running Dev on this rig.
Scott Leibrand
@scottleibrand
Jan 06 2017 20:56
try doing something shorter, like iter_pump_hours 1, or even iter_pump 5
you might have an unknown pump history record that decocare is barfing on
Aaron Michelson
@Aaybob
Jan 06 2017 20:58
using Cygwin how does one log into edison?
Joe
@t1djoe
Jan 06 2017 21:05
Even going down to iter_pump 5 gives me the same error.
Scott Leibrand
@scottleibrand
Jan 06 2017 21:07
Is there anything abnormal about this pump's recent history that you can think of?
and do other non-history pump commands work normally?
Joe
@t1djoe
Jan 06 2017 21:16
Nothing abnormal, all non-history commands I've tried work properly. Loop was running fine all morning, I was working on the Edison at lunch trying to get BT tethering up. When I restarted my cron job, it began throwing this message. Unfortunately for me, I didn't keep detailed enough notes of what I was doing to determine if I hosed something up, but at this point, I'm assuming yes, as everything else appears to checkout.
Scott Leibrand
@scottleibrand
Jan 06 2017 21:21
if you can get the current temp basal, clock, model, and all that stuff, and only history fails, I'm not sure it could be anything you did
one thing I've tried in the past is injecting a bunch of history events into the pump by setting the same temp basal over and over dozens of times. sometimes that pushes any bad history off the first page so the queries work again...
but before you do that: do you have another openaps rig you could try?
Aaron Michelson
@Aaybob
Jan 06 2017 21:24
@t1djoe I reflashed edison after getting ble tether running. Not that I needed to because right after I started flashall.bat I thought my multiple cron schedules were walking on top of each other.
Joe
@t1djoe
Jan 06 2017 21:25
@scottleibrand Currently, no. My explorer board should be here Monday, but I was planning to setup a Pi rig this weekend for backup in the house. I'll try the temp basal trick to see if it clears up this rig. Worst case, I reflash it tonight and be back up.
Scott Leibrand
@scottleibrand
Jan 06 2017 21:28
k
Aaron Michelson
@Aaybob
Jan 06 2017 21:28
You could use a new microSD card for the redo and then troubleshoot this one later.
PieterGit
@PieterGit
Jan 06 2017 21:49
I'm still not having any luck with a newer kernel on an edison, see https://communities.intel.com/message/444831#444831 for my current problems. did others here manage to upgrade their edison's u-boot and kernel somehow? the default kernel does not seem to be stable enough to use with a usb connection (which is required for dexcom that does not support bluetooth)
Aaron Michelson
@Aaybob
Jan 06 2017 21:50
This is an updated CareLink solution @bdr1177 the cron file should be changed to this, with all the lines having the /4 added to . CareLink needs time to recover if it fails so loop runs every four minutes. In addition to th /4 an additional line is needed, the 2-59 line.
/4 cd /home/pi/myopenaps && ( ps aux | grep -v grep | grep -q 'openaps pump-loop' || openaps pump-loop ) 2>&1 | tee -a /var/log/openaps/pump-loop.log
2-59/4
* cd /home/pi/myopenaps && openaps preflight | tee -a /var/log/openaps/pump-loop.log
Rob Kresha
@rkresha
Jan 06 2017 22:15
Here is the Edison/Explorer case that I've been working on http://goo.gl/NiRQBb Have fun!
Aaron Michelson
@Aaybob
Jan 06 2017 22:17
what do you use to 3d print it....how much?
Rob Kresha
@rkresha
Jan 06 2017 22:18
I've got a 3d printer here at the house. It's a monoprice i3 v2.. how much for the printer?
Aaron Michelson
@Aaybob
Jan 06 2017 22:18
@moimemecloud did you consider heat dissipation?
I think we've got a winner!
Rob Kresha
@rkresha
Jan 06 2017 22:20
I assume that question is to me... look at the model on thingiverse, it's has ventilation holes :)
Aaron Michelson
@Aaybob
Jan 06 2017 22:21
@rkresha :fire:
jaylagorio
@jaylagorio
Jan 06 2017 22:42
No one else is using an Altoids tin, huh?
Rob Kresha
@rkresha
Jan 06 2017 22:42
Something about metal and circuit boards. :worried:
Scott Leibrand
@scottleibrand
Jan 06 2017 22:42
Makes a good Faraday cage, I hear... :)
jaylagorio
@jaylagorio
Jan 06 2017 22:43
@scottleibrand It does if you don't use one of those handy, tiny 2.4 GHz antennas on the Explorer!
@rkresha I lined the inside with parchment paper. The lipo battery is low profile enough that sticking it to the lid keeps it off the board as well.
Scott Leibrand
@scottleibrand
Jan 06 2017 22:45
are you actually doing that? what kind of antennas?
jaylagorio
@jaylagorio
Jan 06 2017 22:46
I'm on mobile, but it wad mentioned here last week. I got it from Adafruit, when I get home ill paste the link again. I'm still impressed it works well on the 900MHz freqs.
*was
Also it's my in-progress rig so I don't have it on me for a pic
Scott Leibrand
@scottleibrand
Jan 06 2017 22:48
cool. when you get a chance I'd like to see the details
we've only ever used on-board antennas and fairly minimalist cases so far
jaylagorio
@jaylagorio
Jan 06 2017 22:54
Definitely, I can do that later tonight.
Aaron Michelson
@Aaybob
Jan 06 2017 23:27

I have a screen capture video of the whole thing.
I used this set of instructions at https://wiki.debian.org/BluetoothUser
Using bluetoothctl
If bluetooth-agent is not available, try bluetoothctl:
Start the bluetoothctl interactive command. Enter "help" to get a list of available commands.
Turn the power to the controller on by entering "power on". It is off by default.
Enter "devices" to get the MAC Address of the device with which to pair.
Enter device discovery mode with "scan on" command if device is not yet on the list.
Turn the agent on with "agent on".
Enter "pair MAC Address" to do the pairing (tab completion works).
If using a device without a PIN, one may need to manually trust the device before it can reconnect successfully. Enter "trust MAC Address" to do so.
Finally, use "connect MAC address" to establish a connection.
Repeating oref0-setup stops with this error before finishing:

Checking for BT Mac or Shareble
{
"model": "G5"
}added cgm://JSON/oref0_glucose/raw-cgm/raw-entries.json
Removing any existing pump device:
removed medtronic://pump
Checking spi_serial installation
Checking mraa installation
Installing swig etc.
Reading package lists... Done
Building dependency tree
Reading state information... Done
E: Unable to locate package swig
Could not install swig etc.

Any suggestions are welcome to complete oref0-setup after installing dev and turning on BLE tethering. Dev was the first oref0-setup on this edison explorer.

jaylagorio
@jaylagorio
Jan 06 2017 23:30
@scottleibrand That's the one, from @rkresha
Sgoldberg89
@Sgoldberg89
Jan 06 2017 23:35
@t1djoe Did you ever get yours working again?
Mine also crapped out on its own, had been working fine for days and I wasn't messing with it.
Aaron Michelson
@Aaybob
Jan 06 2017 23:47
@rkresha @jaylagorio :sparkles: :smile: