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

18th
Jul 2017
LilDucky
@LilDucky
Jul 18 2017 00:00 UTC
oref0-setup --dir=/home/pi/lilducky --serial=666666 --cgm=MDT --ns-host=https://vvvv.herokuapp.com --api-secret=xxxx --max_iob=3 --enable=' meal ' --radio_locale='US'
bare bones
Dana Lewis
@danamlewis
Jul 18 2017 00:00 UTC
Ok good
LilDucky
@LilDucky
Jul 18 2017 00:01 UTC
Starting pump-loop at Tue 18 Jul 09:11:04 AEST 2017:
MDT CGM configured; not waiting
ImportError: No module named spi_serial
Listening for 30s silence before mmtuning: ......
Starting pump-loop at Tue 18 Jul 09:12:30 AEST 2017:
MDT CGM configured; not waiting
ImportError: No module named spi_serial
Listening for 30s silence before mmtuning: ................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................mmtune:
Starting pump-loop at Tue 18 Jul 09:44:05 AEST 2017:
MDT CGM configured; not waiting
ImportError: No module named spi_serial
Listening for 30s silence before mmtuning: ....................................................................................................................................................................................................................................................................................................................................................................................^C
Dana Lewis
@danamlewis
Jul 18 2017 00:01 UTC
Just making sure saying carelink doesn't still leave remnants of ttyport
live4sw
@live4sw
Jul 18 2017 00:09 UTC
Another quick question - should a 0.5.1 rig play nice with 0.4 rigs, provided that SMB/UAM are not turned on? Would assume yes but want to make sure.
Scott Leibrand
@scottleibrand
Jul 18 2017 00:26 UTC
@LilDucky looks like the logic in oref0-setup isn't quite right for carelink + MDT CGM. If you'd like, some investigation might be warranted to figure out what's wrong and how easy it'd be to fix.
@live4sw yes, without SMB they'll play nice together. disagreements about temp basals should be rare, and if they occur you'll just have some sawtoothing and end up with an end result in between the two.
LilDucky
@LilDucky
Jul 18 2017 00:28 UTC
thanks @danamlewis
Trust me to find any bugs.
Dirk Gastaldo
@dirkgastaldo
Jul 18 2017 00:28 UTC
Hello all. My rig wouldn't connect to the work WiFi. My phone connection had been flaky, too. I put an access point on my desk, and my phone connected immediately, and the signal is full. But my rig still won't connect. I finally turned my phone's hotspot off, and the rig still won't connect. It connected last week, and I haven't changed anything. There is no activity in network log after I turned off the hotspot. The last thing it shows is the 172.xxx IP address from my phone. No activity. Any ideas?
blob
LilDucky
@LilDucky
Jul 18 2017 00:28 UTC
is there a quick hack to kill it?
Dana Lewis
@danamlewis
Jul 18 2017 00:28 UTC
heh. trust anyone with that setup, because it’s not one we’re actively testing. @LilDucky
LilDucky
@LilDucky
Jul 18 2017 00:29 UTC
@danamlewis I am in IT and renowned for breaking things when testing.
live4sw
@live4sw
Jul 18 2017 00:30 UTC
In addition, is it a new feature in 0.5 to set a temp basal of >30mins? I did a doubletake when I just saw a 2hr zero temp set.
scottleibrand @scottleibrand finds the chaos monkey hat to give to @LilDucky
Scott Leibrand
@scottleibrand
Jul 18 2017 00:30 UTC
@live4sw yep
live4sw
@live4sw
Jul 18 2017 00:30 UTC
No objections, but was very surprised to see that after a year and a half with OpenAPS
LilDucky
@LilDucky
Jul 18 2017 00:30 UTC
@scottleibrand they used to call me Destructor in a previous job
Scott Leibrand
@scottleibrand
Jul 18 2017 00:31 UTC
it's most common with SMBs, but if you really start dropping and look like you'll need it, it'll set a long zero temp even w/o SMBs enabled
might be worth PR'ing in a note about that to the docs
live4sw
@live4sw
Jul 18 2017 00:31 UTC
It seems like there are a bunch of subtle improvements in 0.5 to (i) better compensate for lost connectivity and (ii) reduce the number of times something has to be enacted (which I assume is probably very good for pump battery life)
Dana Lewis
@danamlewis
Jul 18 2017 00:31 UTC
lots of improvements, as described in the release notes: https://github.com/openaps/oref0/releases
live4sw
@live4sw
Jul 18 2017 00:32 UTC
Ah, I see it - "allow non-SMB code paths to set temps up to 120 minutes when needed"
Thanks, good stuff
Scott Leibrand
@scottleibrand
Jul 18 2017 00:34 UTC
I don't think any of our changes help much with battery life.
I've mostly given up on optimizing for that.
Martin Haeberli
@mhaeberli
Jul 18 2017 02:11 UTC
so - can I install new -master- in a directory alongside old -dev- or -master- and switch back and forth basically by re-running the appropriate oref0-rerun and disabling and re-enabling cron ? or will I break something because of the different npm installs?
also - crazy question - somehow as I mentioned COB at the moment is always 0. but graph ad hoc looks to be more ‘stable’. so the question is, suppose I wanted to configure oref0 to run as if COB was always 0 - is there an easy way to do that? or rtfc (code vs doc) and tweak there...
yes, too many degrees of freedom
Martin Haeberli
@mhaeberli
Jul 18 2017 02:26 UTC
ok - or another phrasing - what is the best practice for running two different versions of oref0 at different times on the same Edison (switching back and forth as desired from the command line)
woodtrevor
@woodtrevor
Jul 18 2017 02:47 UTC
On a related note (batteries) I tried a different brand of lithiums that I stumbled across on Amazon. Same specs as energizer and cheaper. I got the same battery life = Win! Except...at end of life they did not die gracefully with a low battery warning and a working pump. Instead the pump silently went into a brown out reset cycle. I'm trying another one to see if it acts the same but that behaviour, even once, is pretty much a deal killer. :worried:
rodzilla55
@rodzilla55
Jul 18 2017 02:49 UTC
Just received the parts for my 2nd rig. Ordered same boards from same place (Hamshield). Both with jubilinux... but seems like very different versions. I'm relatively novice with linux. Can someone help me understand which of the following (Rig#1 or Rig #2) is the correct version? Or both? [pasting logs]

RIG #1
Debian GNU/Linux 8 rodsoaps ttyMFD2

rodsoaps login: root
Password:
Last login: Mon Jul 17 18:51:09 EDT 2017 from 10.0.0.236 on pts/0
Linux rodsoaps 3.10.17-poky-edison+ #6 SMP PREEMPT Wed Mar 23 21:47:59 EDT 2016 i686

The programs included with the Debian GNU/Linux system are free software;
the exact distribution terms for each program are described in the
individual files in /usr/share/doc/*/copyright.

Debian GNU/Linux comes with ABSOLUTELY NO WARRANTY, to the extent
permitted by applicable law.
root@rodsoaps:~# uname -r
3.10.17-poky-edison+
root@rodsoaps:~# lsb_release -a
No LSB modules are available.
Distributor ID: Debian
Description: Debian GNU/Linux 8.8 (jessie)
Release: 8.8
Codename: jessie
root@rodsoaps:~#

RIG #2
Debian GNU/Linux 8 jubilinux ttyMFD2

jubilinux login: root
Password:
Last login: Sat Jan 1 00:16:03 UTC 2000 on ttyMFD2
Linux jubilinux 3.10.98-poky-edison #5 SMP PREEMPT Tue Apr 4 22:42:44 EDT 2017 i686

The programs included with the Debian GNU/Linux system are free software;
the exact distribution terms for each program are described in the
individual files in /usr/share/doc/*/copyright.

Debian GNU/Linux comes with ABSOLUTELY NO WARRANTY, to the extent
permitted by applicable law.
root@jubilinux:~# uname -r
3.10.98-poky-edison
root@jubilinux:~# lsb_release -a
No LSB modules are available.
Distributor ID: Debian
Description: Debian GNU/Linux 8.3 (jessie)
Release: 8.3
Codename: jessie

parenthetic
@parenthetic
Jul 18 2017 02:51 UTC
I'm struggling to understand the logic in determine-basal. When bg > target_bg my basal rates are consistently getting set at a rate that is 0.025 U/hr below the profile.current_basal rate, but because of the rounding that is done, almost all the reporting in the pump-loop.log reports the profile.current_basal rate. So, for example, if my profile.current_basal= 0.6, the pump-loop.log will report: "setting current basal of 0.6 as temp. Setting neutral temp basal of 0.6U/hr", but the rate will actually be 0.575 (and actually, when a temp of 0.575 is sent to my pump (Paradigm 522 Ver 2.2A 1.1 OB OB), my pump enacts a temp basal rate of 0.58). Any help understanding how/why this is happening would be GREATLY appreciated. Thanks!
francesc0-cgm
@francesc0-cgm
Jul 18 2017 03:36 UTC
Set true on prefernces.json in set_neutral basal rate
It sets rounded basal to Make you sure it is working when you can watch only the pump
skip_neutral_temps:
Defaults to false, so that OpenAPS will set temps whenever it can, so it will be easier to see if the system is working, even when you are offline. This means OpenAPS will set a “neutral” temp (same as your default basal) if no adjustments are needed. If you are a light sleeper and the “on the hour” buzzing or beeping wakes you up (even in vibrate mode), you may want to turn this to “true” to skip this setting. However, we recommend it for most people who will be using this system on the go and out of constant connectivity.
katie disimone
@Kdisimone
Jul 18 2017 03:54 UTC
@rodzilla55 they both are fully functional versions of jubilinux. Your rig #1 has a slightly older version of jubilinux, but that will work just fine. That version is the March 23rd version. Your Rig #2 has the April 4th version. Totally fine too. http://www.jubilinux.org/dist/
and you can use uname -a to check the jubilinux version all in one swoop
rodzilla55
@rodzilla55
Jul 18 2017 03:59 UTC
@Kdisimone THANKS so much for the info... I notice Rig 1 is actually March 23 of 2016 vs Rig 2 April of 2017. So I was concerned about Rig 1. Thanks for the tip on uname -a.
katie disimone
@Kdisimone
Jul 18 2017 04:00 UTC
The only difference that I notice between the two releases is that the March 23rd release required a control-d in the middle of flashing. The real changes under the hood between the two systems have been seamless to the average user.
rodzilla55
@rodzilla55
Jul 18 2017 04:00 UTC
@Kdisimone 1 more quick Q... if I follow the cloning procedure at https://github.com/oskarpearson/mmeowlink/wiki/Backing-up-and-cloning-your-OpenAPS-Edison, will the 2 different versions of jubilinux cause me any issues?
katie disimone
@Kdisimone
Jul 18 2017 04:25 UTC
No I don't think that would cause a problem. HOWEVER...I don't think that's worth doing anymore. The old way of building an OpenAPS installation was quite cumbersome and took quite awhile. With the new setup scripts and runagain options...it would take less time to build your rig from scratch than it would to make a backup and then retrieve it. The backup instructions were very valuable in 2016...not so much in 2017.
rodzilla55
@rodzilla55
Jul 18 2017 04:26 UTC
Got it. Thanks.
Scott Leibrand
@scottleibrand
Jul 18 2017 05:15 UTC
@mhaeberli best practice is: don't do that (try to run two versions of the same software at the same time on the same computer). If you absolutely must run oref0 0.4.x and 0.5.x side by side, use a second rig. but far better IMO is to upgrade to 0.5.x, and adjust preferences.json to enable, disable, or adjust behavior you want to tune.
@parenthetic that is due to some weird floating point math and rounding in decocare. it was investigated at one point, and I think something was fixed, but we mostly don't care because it's within the margin of what oref0 cares about (0.1U/h), and the IOB gets calculated correctly.
cameronrenwick
@cameronrenwick
Jul 18 2017 11:43 UTC
@rodzilla55 if I can offer any sort of advice here, in my 2 rig world I got rig one running exactly like I wanted it (took maybe 2 weeks going systematically thru various changes step by step) then documented what it looked like in prefs etc THEN I duplicated rig 2 using the regular install (not clone method). Doing this again really helps to get you familiar with the rig and various methods of interfacing with it. Believe me, openAPS isn't a set it and forget it thing. You'll always be fine tuning, updating, changing... it's not your standard rodeo. Once you've got rig 2 mirrored your rig one, then 2 can become your dev rig and if there's something you want to experiment with a bit.. you're safe to do so having rig 1 to fall back on should something now behave as planned. I've taken this approach and it seems to work pretty well - I think others have done similar. As always YDMV
Martin Haeberli
@mhaeberli
Jul 18 2017 12:23 UTC
@scottleibrand thx
stefanakerblom
@stefanakerblom
Jul 18 2017 13:11 UTC
Will the xDrip+/AAPS apps/watchfaces work on Tizen (samsung gear 3 OS)
Colin Lennon
@colinlennon
Jul 18 2017 13:55 UTC
@tepidjuice Sorry for the late reply, I've been offline for a few days while we move apartment. Back online now. Hopefully the re-install fixes your issue but if not, let me know
parenthetic
@parenthetic
Jul 18 2017 13:59 UTC
@francesc0-cgm Thanks!
@scottleibrand Thanks. I was unaware of this 0.1U/h margin. To me, that's kind of a big deal. My basal rates range from 0.2U/h to 0.7U/h, so that's a 14-50 margin of error for me.
parenthetic
@parenthetic
Jul 18 2017 14:04 UTC
@scottleibrand "14-50%"
@scottleibrand More generally, though, could you point me to where the "reporting enact/suggested.json" and "reporting enact/enacted.json" are getting written out to the pump-loop.log file? I know nothing about decocare, but I'm guessing that's some python code that actually transmits the basal rate changes to the pump? Based on my debugging, it seems that it's in this step (between those two reporting statements) that the basal rate rounding issue takes place.
Scott Leibrand
@scottleibrand
Jul 18 2017 14:28 UTC
@parenthetic if it sets a temp that is 0.025U/h lower than desired, then over 30m you'll get 0.0125U less insulin than otherwise. If your ISF is 400 mg/dL/U, that would raise your BG by 5 mg/dL over 3-4 hours. That's noise.
And we'll before that, it will be recalculating a new temp basal to set, which will be based on actual insulin delivered at the slightly lower rate. So it will self correct even before it manages to get off by 0.01U.
The basals are executed by running openaps use pump set_temp_basal and passing it the appropriate json.
Brian Rabinovitz
@bdr1177
Jul 18 2017 15:26 UTC
Is there a ssh program for android that is like terminus? Tried reading the docs but didn't see anything (but I'm on my phone so might have missed it).
Scott Leibrand
@scottleibrand
Jul 18 2017 15:28 UTC
*termius ;-)
JuiceSSH is one
Brian Rabinovitz
@bdr1177
Jul 18 2017 15:29 UTC
Autocorrect 😅
Cameron Chunn
@applehat
Jul 18 2017 15:29 UTC
Termius is available on android as well
Brian Rabinovitz
@bdr1177
Jul 18 2017 15:29 UTC
Ok, I'll try that. Thanks!
Dana Lewis
@danamlewis
Jul 18 2017 15:30 UTC
might be worth PR’ing that page. I know it’s heavily Iphone explained, so someone who can update what applies to android/not woudl be great
Brian Rabinovitz
@bdr1177
Jul 18 2017 15:30 UTC
@danamlewis I was just going to suggest that.
I'm out of town this week, but if it still needs updated next week I'll pr it.
parenthetic
@parenthetic
Jul 18 2017 15:33 UTC
@scottleibrand Fair enough. The problem I've been seeing, though, is that when I'm above my target overnight and my basal rate is set at 0.025U/h lower for a full 8-10 hours overnight. Granted, even with my ISF of 50 mg/dL/U, that only translates into a BG reading that is 10-12mg/dL higher than it would have potentially been without the rounding error. I'm likely just getting hung up on the cognitive dissonance of having the rate being reported in the logs as 0.025U/h lower than my programmed rate when I'm above target, plus the same log reporting that the algorithm is setting my rate current rate.
Brian Rabinovitz
@bdr1177
Jul 18 2017 15:34 UTC
The next question is whether it is possible to ssh in from android when the rig is Bluetooth tethered. The logs report the network as null when the rig is tethered.
Scott Leibrand
@scottleibrand
Jul 18 2017 15:43 UTC
@parenthetic do you have your max_iob set to something non-zero?
@bdr1177 which logs? the wifi log (/var/log/openaps/network.log) should show your local IP, whether it came from wifi or Bluetooth (or occasionally both, as it transitions back to wifi)
Brian Rabinovitz
@bdr1177
Jul 18 2017 15:49 UTC
@scottleibrand papertrail network log. When it is Bluetooth tethered the wifi network name is blank. It does report a public IP address.
parenthetic
@parenthetic
Jul 18 2017 15:51 UTC
@scottleibrand No, I don't. It's been 4+ months since I started with this, but I set some goals for myself before bumping up my max_iob: 1) Understand the IOB calculations and 2) Understand the determine-basal logic. I'm getting close. :smile: (In my meager defense, I've had very little time to give to this over the past several months.)
Brian Rabinovitz
@bdr1177
Jul 18 2017 15:51 UTC
Whenever it is on a real wifi network, it reports the network name.
parenthetic
@parenthetic
Jul 18 2017 15:56 UTC
@scottleibrand I wonder if changing line 51 in basal-set-temp.js from rT.rate = suggestedRate; to rT.rate = profile.current_basal;would be appropriate if the conditional (suggestedRate === profile.current_basal) at line 37 doesn't actually hold in practice? Also, do you know whether the 0.025U/h thing always has a downward bias?
Scott Leibrand
@scottleibrand
Jul 18 2017 15:59 UTC
yes, the 0.025U/h thing is always downward, and only happens for specific basal rates where the floating point math happens to result in something like 0.599999999999999998. (other basal rates like 0.7000000000001 would just round properly to 0.7)
@parenthetic if you want, you could also start incrementally increasing your max IOB and watching how it performs. even a 0.5U max iob would allow it to easily counteract any rounding effects (and gradually help bring down any high BG). if you want you could also set a max basal just a little lower than your highest basal of the day until you're more comfortable with it.
Brian Rabinovitz
@bdr1177
Jul 18 2017 16:06 UTC
@scottleibrand ok I figured it out. The logs also report a local ip and that's what I needed. When I get back next week I'll pr instructions to the docs.
Terri Lyman
@tatigger
Jul 18 2017 16:27 UTC
Hi guys, I am updating one of my Edisons to the newest version of oref0, without adding the new oref1 stuff, and I see as it updates occasional
Checking oref0 installation
* master
Installing latest oref0 package
sudo: unable to resolve host Thing1
Do you guys know what is up with this? Do I need to do something differently to login to my Edison. I always login as root
Scott Leibrand
@scottleibrand
Jul 18 2017 16:43 UTC
sudo: unable to resolve host is no big deal. it happens when you didn't set the hostname properly in both /etc/hosts and /etc/hostname, or you haven't yet rebooted since doing so.
some of the scripts use sudo for backwards compatibility with Pi based setups, and that's just a warning sudo likes to spew when things aren't configured perfectly. it doesn't break anything though
Terri Lyman
@tatigger
Jul 18 2017 16:46 UTC
ok, thanks @scottleibrand
parenthetic
@parenthetic
Jul 18 2017 17:00 UTC
@scottleibrand Thanks, Scott. That's a good recommendation. I've gotten comfortable with a lot of how oref0 works, so I'm not all that timid about bumping up my max_iob -- it was more about setting some artificial goal for myself before taking the next steps. Any thoughts on the suggested change to basal-set-temp.js I mentioned?
Michael Spradling
@M1Sports20
Jul 18 2017 17:12 UTC
Hm, it seems my rig is not uploading data to nightscout. (It did until 8am today) It seems IOB is always 0 and temp basals are always 0. The graph doesn't update
still debugging it
It is still issuing temp basals to my pump, just doesn't seem be uploading them. Hope it still tracks them internally and uses that data for looping
Bolus wizard entrys are also not being uploaded
Scott Leibrand
@scottleibrand
Jul 18 2017 17:14 UTC
@parenthetic not sure if that would make a difference or not. you can try it: the rounding behavior should be deterministic, so fairly easy to reproduce
@M1Sports20 whenever you're troubleshooting an issue, always look at your pump-loop.log to see what it's really doing and what it really sees as far as IOB etc.
Shawn
@dyjaks
Jul 18 2017 17:19 UTC
So maybe a stupid question, and one I might solve if possible and I have time, but has anyone tried to get a dexcom to talk to the explorer directly? If xDrip+ can do it that means we understand the protocol, right?
Michael Spradling
@M1Sports20
Jul 18 2017 17:20 UTC
IOB in pump-log are also 0, thats not good. I know I have insulin
Shawn
@dyjaks
Jul 18 2017 17:20 UTC
Would work for a better offline loop I feel
Dana Lewis
@danamlewis
Jul 18 2017 17:20 UTC
Yup. There’s openxshareble that used to work (g4 receiver to rig), but no one’s gotten it to work in the 8+ months since we switched to oref0-setup. the other option is to put nate’s xdrip sniffer for g5 on linux so the rig can read from a g5 transmitter.
@thebookins has done some good work figuring out how to use a second Loop app for pulling G5 locally down to the rig, but that requires stuff (iphone, apple developer account, etc) vs. straight to the rig.
So it would be great if someone figured that out, with an eye toward whatever hardware we switch to next (potentially pi zeroW) also being compatible with that solution
Martin Haeberli
@mhaeberli
Jul 18 2017 17:22 UTC
@danamlewis I like the openxshareable and the xdrip sniffer plans … and/but, fyi, Ben West in San Diego at Hack NS mentioned something to me about a needed jubilinux dexcom 5 driver. we here may also look into @thebookins ‘ approach - in our ‘infinite spare time’ Ha!
Dana Lewis
@danamlewis
Jul 18 2017 17:23 UTC
::dreams of spare time:: ::)
Shawn
@dyjaks
Jul 18 2017 17:35 UTC
Been a while since I looked at java OR python but openxsharable is smaller so how hard can it be /s
(first file I open has magic guids)
Marco
@CaptainBalou
Jul 18 2017 19:14 UTC
@scottleibrand Hello Scott. Thanks for offering your help concerning decocare. I already saw that some have tested it and that you put it into dev. Because I am on dev I thought I get that automatically. I re-ran setup but don’t see any change. @ceben80 told me I have to get the dev of decocare to see that. But even after his commands I seem to be to stupid to get it. I think on some day I’ll get an update like through a magic hand and then the pills are on the correct position. Until this day I’ll keep ignoring the position. :-D
Michael Spradling
@M1Sports20
Jul 18 2017 19:18 UTC
@scottleibrand Sorry, just seen your message now, yes I looked at pump-log. I figured I would look there and get to the root cause. For some reason my IRC to gitter bridge is failing and I am not getting your replys.
Anyways pump-log ns-upload look free of errors. Although, I don't really know what is normal yet.
But IOBs are 0 even if I just bolus
Dana Lewis
@danamlewis
Jul 18 2017 19:19 UTC
@M1Sports20 why don’t you post some of your logs if you don’t know what normal looks like?
Michael Spradling
@M1Sports20
Jul 18 2017 19:20 UTC

determine-basal://text/shell/enact/suggested.json
reporting enact/suggested.json
{"temp":"absolute","bg":107,"tick":-2,"eventualBG":85,"snoozeBG":92,"insulinReq":0,"reservoir":null,"deliverAt":"2017-07-18T19:15:17.185Z","minPredBG":999,"predBGs":{"IOB":[107,105,102,100,98,97,96,94,94,93,92,92,92]},"COB":0,"IOB":0,"reason":"COB: 0, Dev: -22, BGI: 0, ISF: 53, Target: 99, minPredBG 92, IOBpredBG 92; Eventual BG 85 < 86, bolus snooze: eventual BG range 85-92; setting current basal of 1 as temp. ","duration":30,"rate":1}
enact/enacted.json: {"insulinReq":0,"bg":107,"reservoir":null,"temp":"absolute","snoozeBG":92,"rate":1,"predBGs":{"IOB":[107,105,102,100,98,97,96,94,94,93,92,92,92]},"minPredBG":999,"IOB":0,"reason":"COB: 0, Dev: -22, BGI: 0, ISF: 53, Target: 99, minPredBG 92, IOBpredBG 92; Eventual BG 85 < 86, bolus snooze: eventual BG range 85-92; setting current basal of 1 as temp. ","COB":0,"eventualBG":85,"recieved":true,"duration":30,"tick":-2,"timestamp":"2017-07-18T15:15:22.035794","deliverAt":"2017-07-18T19:15:17.185Z"}
glucose.json newer than pumphistory. : Refreshed pumphistory
{"carbs":0,"boluses":0,"mealCOB":0,"currentDeviation":-3.67,"maxDeviation":0,"minDeviationSlope":0}
{"iob":0,"activity":0,"bolussnooze":0,"basaliob":0,"netbasalinsulin":0,"hightempinsulin":0,"microBolusInsulin":0,"microBolusIOB":0,"time":"2017-07-18T19:14:25.000Z","lastBolusTime":0}
{"delta":-2,"glucose":107,"short_avgdelta":-2.72,"long_avgdelta":-3.63}
{"duration":30,"rate":1,"temp":"absolute"}
Autosens adjusting basal from 1.3 to 1; target_bg from 90 to 99; sens from 41 to 53.2 (autosens ratio 0.77)
Carb Impact: -2.7 mg/dL per 5m; CI Duration: 0 hours; remaining 4h+ CI: 0 mg/dL per 5m
Accel. Carb Impact: 10 mg/dL per 5m; ACI Duration: 0 hours
UAM Impact: -3.6 mg/dL per 5m; UAM Duration: 0 hours
minPredBG: 92 minIOBPredBG: 92 avgPredBG: 92 COB: 0 carbs: 0
BG projected to remain above 86 for 240 minutes
bgUndershoot: -8 zeroTempDuration: 240 zeroTempEffect: 277 carbsReq: -64
determine-basal://text/shell/enact/suggested.json
reporting enact/suggested.json
{"temp":"absolute","bg":107,"tick":-2,"eventualBG":85,"snoozeBG":92,"insulinReq":0,"reservoir":null,"deliverAt":"2017-07-18T19:15:49.285Z","minPredBG":999,"predBGs":{"IOB":[107,105,102,100,98,97,96,94,94,93,92,92,92]},"COB":0,"IOB":0,"reason":"COB: 0, Dev: -22, BGI: 0, ISF: 53, Target: 99, minPredBG 92, IOBpredBG 92; Eventual BG 85 < 86, bolus snooze: eventual BG range 85-92, temp 1 ~ req 1U/hr. "}
enact/enacted.json: {"insulinReq":0,"bg":107,"reservoir":null,"temp":"absolute","snoozeBG":92,"rate":1,"predBGs":{"IOB":[107,105,102,100,98,97,96,94,94,93,92,92,92]},"minPredBG":999,"IOB":0,"reason":"COB: 0, Dev: -22, BGI: 0, ISF: 53, Target: 99, minPredBG 92, IOBpredBG 92; Eventual BG 85 < 86, bolus snooze: eventual BG range 85-92; setting current basal of 1 as temp. ","COB":0,"eventualBG":85,"recieved":true,"duration":30,"tick":-2,"timestamp":"2017-07-18T15:15:22.035794","deliverAt":"2017-07-18T19:15:17.185Z"}
Settings less than 10m old
Edison on battery: 58%. Pumphistory-24 < 90m old
Completed pump-loop at Tue Jul 18 15:15:50 EDT 2017

Starting pump-loop at Tue Jul 18 15:16:02 EDT 2017:
Waiting up to 4 minutes for new BG: ........................Radio ok. Listening: .%

I just took a small bolus, I guess I expected IOB to be positive
Looks to be also throughing off autosense calculations. Not knowing the source yet, I am assuming this is because it is not recording IOB
Katie Aldridge
@kcrcgm
Jul 18 2017 19:48 UTC
image.png
@scottleibrand Hi! I ran the tail -f ~/myopenaps/monitor/glucose.json and wasn't seeing it get truncated. So I did tail -f ~/myopenaps/xdrip/glucose.json and it IS getting truncated. Is that good enough evidence before I procede with the other xdrip debugging instructions?
Scott Leibrand
@scottleibrand
Jul 18 2017 20:05 UTC
@kcrcgm sure, you can proceed, especially since you have another problem that you'll be able to see whether this fixes.
Katie Aldridge
@kcrcgm
Jul 18 2017 20:05 UTC
Ok, great. Thanks!
@scottleibrand Dumb question, do I run each of the next lines one at a time or copy and paste the whole block?
Tim Street
@tim2000s
Jul 18 2017 20:09 UTC
@tepidjuice - Are you using a recent version of the xdripaps with the download from night scout code in it? Have you tried disabling this?
PieterGit
@PieterGit
Jul 18 2017 20:42 UTC
@LilDucky : i think you hit openaps/oref0#552
Martin Haeberli
@mhaeberli
Jul 18 2017 20:56 UTC
@danamlewis belated :+1:
LilDucky
@LilDucky
Jul 18 2017 20:58 UTC
@PieterGit thnaks. Just read it and seems like that is the conversation I had with @danamlewis yesterday. I will go and look at the rig shortly and see if I can work out what is mentioned in it.
I am no *nix guru
PieterGit
@PieterGit
Jul 18 2017 20:59 UTC
you don't have to be a *nix or python guru to contribute
LilDucky
@LilDucky
Jul 18 2017 21:01 UTC
Yep, but it helps loads. I always seem to find bugs.... work hates me for this.
Am I correct to assume that I can hack the aliases.json file and remove the offending line as a temp,fix?
PieterGit
@PieterGit
Jul 18 2017 21:05 UTC
i wrote an easier workaround in the issue
we don't hate you for finding bugs
and reporting them, and help fix them
Dana Lewis
@danamlewis
Jul 18 2017 21:05 UTC
:+1: :clap:
LilDucky
@LilDucky
Jul 18 2017 21:08 UTC
Thanks. Will haul my butt out of bed shortly and have a tinker. It is 7 am here
francesc0-cgm
@francesc0-cgm
Jul 18 2017 21:08 UTC
Using online monitoring via urchin and selecting openaps as data you can have last temping time, iob and current temp. Anyone have a sample to add it as custom text to take all the info from openaps pill on my own ns site?
PieterGit
@PieterGit
Jul 18 2017 21:10 UTC
it's 11pm here, so I'm about to quit. But others can help you make the change. Please tag me, or update the issue. I changed the issue, because it's not related to your MDT CGM, but it's related to a non explorer board (which I thought it was). Am I right that the rig-pump communication works fine (at least: as far as possible with those broken Carelink USB sticks)
@Kdisimone : thanks for finding the link to openaps/oref0#522 I only used the GitHub blame function :smile: I knew it was for a good cause
katie disimone
@Kdisimone
Jul 18 2017 21:13 UTC
@PieterGit no worries
@LilDucky yes, you can edit your alias.
vi ~/src/oref0/lib/oref0-setup/alias.json
PieterGit
@PieterGit
Jul 18 2017 21:15 UTC
@LilDucky : I'm out. Others will help you after you taken a coffee or tea first.
LilDucky
@LilDucky
Jul 18 2017 21:15 UTC
@PieterGit yes, pump Comms are fine. I will update #522 with results soon.
katie disimone
@Kdisimone
Jul 18 2017 21:16 UTC
#522 is technically closed. May need to open new?
LilDucky
@LilDucky
Jul 18 2017 21:17 UTC
@Kdisimone thanks, I remember using vi from my job 25+ years ago. I will report findings here.
katie disimone
@Kdisimone
Jul 18 2017 21:17 UTC
nano would also work if you prefer that to vi
LilDucky
@LilDucky
Jul 18 2017 21:20 UTC
That is my destruction tool of choice
PieterGit
@PieterGit
Jul 18 2017 21:28 UTC
BTW i still experience radio errors with 0.5.0, most of them are solved after a mmtune run. They can cause rig-pump communication to not work for about an hour at most. On the other hand I don't experience lot's of reboots of the rig anymore due to pump errors. Do others also encounter rig-pump communication errors? (754WW and Explorerboard setup)
Dana Lewis
@danamlewis
Jul 18 2017 21:31 UTC
0.5.0 or are you on 0.5.1?
Paul Dickens
@thebookins
Jul 18 2017 21:38 UTC
@dyjaks FYI my attempt at a linux version of xDripG5 is at https://github.com/thebookins/xdrip-js/. It's not working yet, mainly because I can't quite get the G5 to pair, but I have had it working on MacOS. Most of the logic for reading from the transmitter is in there. I suspect that we might need need to upgrade the kernel on the Edison to get pairing to work. I agree with you, having the Edison read from the G5 directly should definately improve the offline experience. Plus the ability to loop during sensor restarts etc.
PieterGit
@PieterGit
Jul 18 2017 21:38 UTC
0.5.1 (well dev branch to be exact)
LilDucky
@LilDucky
Jul 18 2017 22:14 UTC
@PieterGit @Kdisimone I am editing my openaps.ini file and the mmtune alias only contains
mmtune = report invoke monitor/temp_basal.json
katie disimone
@Kdisimone
Jul 18 2017 22:15 UTC
well, that’s not the file I referenced...
LilDucky
@LilDucky
Jul 18 2017 22:15 UTC
ok, @PieterGit had that in #522
katie disimone
@Kdisimone
Jul 18 2017 22:18 UTC

see my link that I posted before? That gave the file to edit. Or, if you don’t like that link…this explicitly shows the edit (in reverse form) that Pieter is saying caused the problem

https://github.com/openaps/oref0/pull/498/files

PieterGit
@PieterGit
Jul 18 2017 22:21 UTC
@LilDucky still awake, can you grep reset openaps.ini to see what if the reset_spi_serial.py script pops up?
LilDucky
@LilDucky
Jul 18 2017 22:22 UTC
nada
PieterGit
@PieterGit
Jul 18 2017 22:24 UTC
Puzzled... What oref0 version are you using? Can you grep for grep oref0_init_pump_comms openaps.ini?
LilDucky
@LilDucky
Jul 18 2017 22:24 UTC
0.5.1
nada again
just built it yesterday
edited the aliases fiel and rebooting now
still mmtuning...
and still has import error too
PieterGit
@PieterGit
Jul 18 2017 22:29 UTC
found it... see coment in openaps/oref0#552
(let's call it 552 instead of 522 :smile: )
LilDucky
@LilDucky
Jul 18 2017 22:31 UTC
so, do i restore the alias?
PieterGit
@PieterGit
Jul 18 2017 22:34 UTC
doesn't matter, because the alias.json is apparently not used for MDT CGM
i forgot that
@LilDucky : do you have a WW-pump?
LilDucky
@LilDucky
Jul 18 2017 22:36 UTC
restored aliases.json and commented line out. killed all processes.
should I reboot?
yep, WW pump on 916.5
cleaned out the log and waiting for it
PieterGit
@PieterGit
Jul 18 2017 22:40 UTC
Ok, 916 is fine. That means you also can/must comment out oref0_init_pump_comms.py in the pump loop script. Because init WW pump at 816 is not needed for you and the oref0_init_pump_comms.py will also show the false error.
LilDucky
@LilDucky
Jul 18 2017 22:43 UTC
still doing same in log.
editing loop now - should not all the mmtune be murdered in this case?
PieterGit
@PieterGit
Jul 18 2017 22:44 UTC
what file are you editting?
LilDucky
@LilDucky
Jul 18 2017 22:44 UTC
because carelink does not support it?
~src/oref0/bin/oref0-pump-loop.sh
PieterGit
@PieterGit
Jul 18 2017 22:45 UTC
to install it use npm run global-install. otherwise your pump loop won't use it.
it will be fixed with next cron job of the pump loop
LilDucky
@LilDucky
Jul 18 2017 22:47 UTC
getting this
PieterGit
@PieterGit
Jul 18 2017 22:47 UTC
i'm now really going to bed. I think I solved your issue, but I need to check with @scottleibrand about the proper fix(es) for issue openaps/oref0#552 Thanks for testing.
LilDucky
@LilDucky
Jul 18 2017 22:47 UTC

pi@raspberrypi:~/lilducky $ sudo npm run global-install
npm ERR! Error: ENOENT, open '/home/pi/lilducky/package.json'
npm ERR! If you need help, you may report this entire log,
npm ERR! including the npm and node versions, at:
npm ERR! http://github.com/npm/npm/issues

npm ERR! System Linux 4.9.35+
npm ERR! command "/usr/bin/nodejs" "/usr/bin/npm" "run" "global-install"
npm ERR! cwd /home/pi/lilducky
npm ERR! node -v v0.10.29
npm ERR! npm -v 1.4.21
npm ERR! path /home/pi/lilducky/package.json
npm ERR! code ENOENT
npm ERR! errno 34
npm ERR!
npm ERR! Additional logging details can be found in:
npm ERR! /home/pi/lilducky/npm-debug.log
npm ERR! not ok code 0

PieterGit
@PieterGit
Jul 18 2017 22:47 UTC
is lilducky your openaps dir?
LilDucky
@LilDucky
Jul 18 2017 22:47 UTC
yessir
PieterGit
@PieterGit
Jul 18 2017 22:47 UTC
try cd ~/src/oref0 ; npm run global-install
that would work better
LilDucky
@LilDucky
Jul 18 2017 22:49 UTC
so, that has restored to original?
goodnight
PieterGit
@PieterGit
Jul 18 2017 22:50 UTC
it will install the /usr/local/bin/oref0-pump-loop. You could have also editted that file.
LilDucky
@LilDucky
Jul 18 2017 22:50 UTC
thanks
tepidjuice
@tepidjuice
Jul 18 2017 22:55 UTC
@colinlennon I've reflashed and reinstalled openaps and it's not tethering to my phone.
Dana Lewis
@danamlewis
Jul 18 2017 23:01 UTC
@audiefile did you have any luck looking at battery stuff related to pills not showing up?
If you did or didn't - will you start an issue sharing what you know? I'm seeing enough people with this to think it's 0.5.1 related, but it's not universal and I'm not seeing it so having a hard time figuring it where we should drill in. @audiefile
Ditto for anyone else who's seeing unknown OpenAPS pill and confirmed it's not mlab size related - please share as much as you can about your setup and what devices etc you have. @stephenc95 etc ^
Katie Aldridge
@kcrcgm
Jul 18 2017 23:06 UTC
@scottleibrand WOOHOO You fixed it! My Autosens is working! Thanks so much! (I figured out to enter each command one at a time... ;)) Later tonight I will check xdrip/glucose.json and monitor/glucose.json. Don't have time to watch them right now, but I suspect that all is working fine now.
Debugging xdrip 3.JPG
Dana Lewis
@danamlewis
Jul 18 2017 23:07 UTC
@kcrcgm nicely done!
Stephen Chandler
@stephenc95
Jul 18 2017 23:09 UTC
@danamlewis I am running G5 with xdrip+ and an edison with 0.5.1
Doran Barton
@fozzmoo
Jul 18 2017 23:09 UTC
@danamlewis my mLab instance is at 496MB on disk. :)
Dana Lewis
@danamlewis
Jul 18 2017 23:09 UTC
@stephenc95 what version of NS are you on?
Stephen Chandler
@stephenc95
Jul 18 2017 23:10 UTC
@danamlewis Version 0.9.2
Dana Lewis
@danamlewis
Jul 18 2017 23:11 UTC
@fozzmoo compacting should help, but let me know if it doesn't, and what you're gear and NS details are in that case
Doran Barton
@fozzmoo
Jul 18 2017 23:12 UTC
BTW, I've been working with some people at Red Hat on their Openshift.io project. They're interested in featuring a one-click Nightscout container bundle for free. Essentially that would be a one or two click install of MongoDB, a web server, and Nightscout from github. One of their consultants attended my Hacking Diabetes presentation and got thinking how Openshift.io could help our community.
@danamlewis I don't know how long compacting takes. When I run the command, mLab says "The repairDatabase command was issued, but the request timed out."
@danamlewis I'm running NS 0.9.2 on my own linode VM with mLab, of course. My OpenAPS rig is running on an Intel Edison+Explorer board. I have a Dexcom G4+Share and an LG G4 Android phone running xdrip+. Whew.
Dana Lewis
@danamlewis
Jul 18 2017 23:31 UTC
@fozzmoo should introduce them to/talk with @jwedding and the NS Foundation folks
tepidjuice
@tepidjuice
Jul 18 2017 23:32 UTC
@colinlennon I got the tether working. I still have the same rest API upload issue.
Dirk Gastaldo
@dirkgastaldo
Jul 18 2017 23:50 UTC
@Kdisimone Katie, are you on?
Doran Barton
@fozzmoo
Jul 18 2017 23:57 UTC
@danamlewis Wahoo! \o/ I dumped the devicestatus collection to my local box and deleted the collection. Shazam! OpenAPS and Pump pills magically reappeared!
Dana Lewis
@danamlewis
Jul 18 2017 23:58 UTC
So some people are solved by compacting. A few are not.
Not sure though if it's coincidental that 0.5.1 sends ppl to edge of size limit or if there's something going on