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

14th
Jul 2017
Scott Leibrand
@scottleibrand
Jul 14 2017 00:22
@PieterGit how involved is it to get stretch? Is is as easy as flashing the right Raspbian image and running openaps-bootstrap.sh?
Robin Kirkman
@esialb
Jul 14 2017 00:23
@PieterGit updating to stretch is almost always going to be the better choice
@PieterGit @scottleibrand how do you feel about jubilinux moving to stretch? it's a pretty big upgrade.
maybe a parallel version of jubilinux for stretch
by the way, hi everybody, sorry for being gone so long
@scottleibrand i bought an explorer board for myself and will be able to (at some point) ensure that radio functionality still works with jubilinux as i do updates
two-way stuff between that and the board you provided
Scott Leibrand
@scottleibrand
Jul 14 2017 00:28
my main question about stretch for jubilinux would be: does that mean we can finally stop compiling bluez from source?
no rush, but if so, that might be worthwhile
Robin Kirkman
@esialb
Jul 14 2017 00:28
i suspect so
Scott Leibrand
@scottleibrand
Jul 14 2017 00:29
seems like it'd be worth testing then. you want to take first stab at an unstable image with stretch?
Robin Kirkman
@esialb
Jul 14 2017 00:29
yeah i'll see what i can do tonight
got time to kill
Dana Lewis
@danamlewis
Jul 14 2017 00:30
are we considering the pi3 stuff a blocker for oref0 dev-master release? or just warn those on pi3 about the workarounds
Scott Leibrand
@scottleibrand
Jul 14 2017 00:30
@esialb ok, cool. we're in the middle of trying to do the oref0 0.5.0 release, so it'll be a few days before we want to play with anything new
Robin Kirkman
@esialb
Jul 14 2017 00:30
understood, and i suspect it would be a few days before i'd have anything ready for other people
i can do the dist-upgrade tonight but there's probably more to it than that
Scott Leibrand
@scottleibrand
Jul 14 2017 00:31
I don't know that bluetooth functions have ever been tested on raspbian, since the Pi 2 (what we have) doesn't even have bluetooth. so unless this is a regression (worked with master but doesn't work with dev) it's probably best to address it in a patch release and/or with workaround notes
Dana Lewis
@danamlewis
Jul 14 2017 00:32
@PieterGit is this a regression, or just was never tested previously?
Scott Leibrand
@scottleibrand
Jul 14 2017 00:33
@betluis so it looks like your problem is happening every time, with the same error, so you will need to drill down into the aliases and reports to figure out what exactly is happening. you said you ran one such report successfully, so it might be best to start at the top, with the command/alias being used by oref0-pump-loop, reproduce the problem that way, and then start drilling down until you isolate it.
Dirk Gastaldo
@dirkgastaldo
Jul 14 2017 01:25
@cameronrenwick exactly, thank you. Next year will mark 50 years with T1D. Admonition acknowledged and accepted. Yes, thank God for my wife.
Robin Kirkman
@esialb
Jul 14 2017 01:27
well, upgrading the installation image to stretch was relatively painless. i'm flashing it to an edison right now.
@scottleibrand @PieterGit the stretch bluez is 5.43-2, will that version work for OpenAPS?
Cameron Chunn
@applehat
Jul 14 2017 01:42
I think the current installer builds (in Dev) 5.44, but before that we were using 5.37 - so 5.43-2 is probably fine.
And not having to build it would simplify things.
Robin Kirkman
@esialb
Jul 14 2017 01:43
alright. this is requiring tweaks along the way; it was kind of silly to hope i could just dist-upgrade and be done. ;)
Matthias
@ceben80
Jul 14 2017 01:54
@scottleibrand just fyi, friend of mine is looping on a pi0 w/ bt connectitity :wink:
Robin Kirkman
@esialb
Jul 14 2017 02:34
alright
stretch is booting
:D
Scott Leibrand
@scottleibrand
Jul 14 2017 03:07
:+1:
aliceinnowonderlands
@aliceinnowonderlands
Jul 14 2017 03:08
hi everyone =) so i'm looping for 3 days already and have some more newbie questions. i'm trying hard not to mess now with bolus myself except meals and watch what happens. i feel like may be i did set up some settings wrong somewhere. it seems like temp basal from pump can't go too far from my original basal. for example last night i went to sleep about 11pm with bg 145. unfortunately bg was keep growing slowly to 250 at 1am, and keep being +- this number all night until i awoke about 6am at morning and fixed it with bolus. i'm looking into nightscout now and can see that it was up and running and keeping my temp basal about 1u/h (it is what i used before openasp as basal), with rare changes to 1.05 or 1.175 or something like this. any suggestion what i set up wrong?
Dana Lewis
@danamlewis
Jul 14 2017 03:09
@aliceinnowonderlands what do you have maxIOB set to?
aliceinnowonderlands
@aliceinnowonderlands
Jul 14 2017 03:15
ops
pump logs telling me that its 0
max basal rate on pumps is 5
Scott Leibrand
@scottleibrand
Jul 14 2017 03:15
totally different things. one is an amount. the other is a rate.
Dana Lewis
@danamlewis
Jul 14 2017 03:16
@aliceinnowonderlands you’ll definitely want to change that if you expect OpenAPS to tackle highs, then. http://openaps.readthedocs.io/en/latest/docs/walkthrough/phase-3/beyond-low-glucose-suspend.html#going-beyond-low-glucose-suspend-mode
aliceinnowonderlands
@aliceinnowonderlands
Jul 14 2017 03:16
oh, yes
thanks a lot Dana and Scott
Dana Lewis
@danamlewis
Jul 14 2017 03:17
:+1:
Sebastien Lussier
@SebastienLussier
Jul 14 2017 03:18
hello! first, we've been looping for more than a week now and i want to thanks @danamlewis @scottleibrand and everyone involved that made this awesome project possible :clap:
Dana Lewis
@danamlewis
Jul 14 2017 03:18
@SebastienLussier :confetti_ball: :raised_hands:
Sebastien Lussier
@SebastienLussier
Jul 14 2017 03:19
i'll never get tired of seeing my son wake up with a perfect bg in the morning :)
Dana Lewis
@danamlewis
Jul 14 2017 03:19
<3 :)
Sebastien Lussier
@SebastienLussier
Jul 14 2017 03:19
we still have works to do for the meal & days, as the rig is not really portable right now, but it still helps
i have a question regarding AMA... i think something is wrong for us
aliceinnowonderlands
@aliceinnowonderlands
Jul 14 2017 03:20
@SebastienLussier it so awesome! congrats!
Sebastien Lussier
@SebastienLussier
Jul 14 2017 03:20
ama_off.jpg
Marcus Lindfors
@Mackwe
Jul 14 2017 03:20
Hi! Accidentally changed an alias I think.. Can someone on "master" please provide me with the output of openaps alias show ns-meal-carbs? What is that alias supposed to be?
Sebastien Lussier
@SebastienLussier
Jul 14 2017 03:20
the predictions are almost always off like this
what is wrong, the CI ?
the logs at that time :
Jul 13 10:41:10 diabetor pump-loop.log: {"carbs":50,"boluses":2.5,"mealCOB":43,"currentDeviation":12.1,"maxDeviation":8.86,"minDeviationSlope":0}
Jul 13 10:41:10 diabetor pump-loop.log: {"iob":2.588,"activity":0.0088,"bolussnooze":2.059,"basaliob":0.209,"netbasalinsulin":0.35,"hightempinsulin":0.45,"microBolusInsulin":0,"microBolusIOB":0,"time":"2017-07-13T14:40:24.000Z","lastBolusTime":1499954969000}
Jul 13 10:41:10 diabetor pump-loop.log: {"delta":8,"glucose":169,"short_avgdelta":7.22,"long_avgdelta":4.02}
Jul 13 10:41:10 diabetor pump-loop.log: {"duration":21,"rate":0.4,"temp":"absolute"}
Jul 13 10:41:10 diabetor pump-loop.log: Autosens adjusting basal from 0.35 to 0.4; target_bg from 126 to 118; sens from 165 to 144.7 (autosens ratio 1.14)
Jul 13 10:41:10 diabetor pump-loop.log: min_bg unchanged: 110; target_bg unchanged: 118; max_bg unchanged: 126
Jul 13 10:41:10 diabetor pump-loop.log: Carb Impact: 13.6 mg/dL per 5m; CI Duration: 3.4 hours; remaining 4h+ CI: 0 mg/dL per 5m
Jul 13 10:41:10 diabetor pump-loop.log: Accel. Carb Impact: 10 mg/dL per 5m; ACI Duration: 4.7 hours
Jul 13 10:41:10 diabetor pump-loop.log: UAM Impact: 10.4 mg/dL per 5m; UAM Duration: 3.6 hours
Jul 13 10:41:10 diabetor pump-loop.log: minPredBG: -38 minIOBPredBG: 39 minCOBPredBG: 61 avgPredBG: -38 COB: 43 carbs: 50
Jul 13 10:41:10 diabetor pump-loop.log: BG projected to remain above 110 for 120 minutes
Jul 13 10:41:10 diabetor pump-loop.log: bgUndershoot: 57 zeroTempDuration: 120 zeroTempEffect: 102 carbsReq: -50
thanks @danamlewis & @aliceinnowonderlands :)
Scott Leibrand
@scottleibrand
Jul 14 2017 03:25
@SebastienLussier what does autotune say?
Sebastien Lussier
@SebastienLussier
Jul 14 2017 03:26
@scottleibrand you mean the CSF line ? CSF | n/a | 7.442
I don't know how this is computed yet, but since it's summer and the rig is not portable yet, my son tends to go out of range after meals so the value might be off ?
Marcus Lindfors
@Mackwe
Jul 14 2017 03:29
@SebastienLussier I did! Thanks!
Scott Leibrand
@scottleibrand
Jul 14 2017 03:32
@SebastienLussier CSF is just ISF/CR. how does it recommend changing those?
Sebastien Lussier
@SebastienLussier
Jul 14 2017 03:34

Parameter | Current | Autotune

ISF | 165.600 | 164.966
CSF | n/a | 7.442
Carb Ratio | 20.000 | 22.167
not much changes...
Scott Leibrand
@scottleibrand
Jul 14 2017 03:41
how many days' worth of data was that run on?
Marcus Lindfors
@Mackwe
Jul 14 2017 03:43
This has started showing up lately: BG data is too old (it's probably this), or clock set incorrectly. The last BG data was read at Tue Jul 11 2017 18:21:32 GMT+0200 (CEST) but your system time currently is Fri Jul 14 2017 05:39:18 GMT+0200 (CEST). Any ideas? I think it might be because I re-installed with another CGM source (that I'm not currently using). If so; do I need to do something or just leave it be? It does work using the backup NS as CGM source.
Sebastien Lussier
@SebastienLussier
Jul 14 2017 03:44
@scottleibrand i'd say it's been running for around a week
Scott Leibrand
@scottleibrand
Jul 14 2017 03:45
@Mackwe what other cgm source did you add? xdripaps?
Marcus Lindfors
@Mackwe
Jul 14 2017 03:46
@scottleibrand I had xDrip, but my bluetooth (in the Edison) freaked out, so I am currently using NS as source and have 'g4' in the run-again.
Scott Leibrand
@scottleibrand
Jul 14 2017 03:46
the g4 cgm source shouldn't cause old data
check your ns-glucose.json and other *glucose.json files to see if it's actually downloading CGM data from NS
Marcus Lindfors
@Mackwe
Jul 14 2017 03:49
the monitor/glucose.json seems to have picked up my parakeet via NS.
@scottleibrand Everything works. It's just annoying with err msgs telling me something is wrong, when the next time it iterates, it works.
Firebat45
@Firebat45
Jul 14 2017 04:00
Hi all. Anyone give me some advice? Trying to set up Bluetooth on an Edison, I follow the steps to kill the process, wait, restart, then the hciconfig command fails every time.
Doesn't matter if I wait less than 10, 10 seconds, or more than 10
I get "Can't get device info: No such device"
Matt Cooper
@MrCooper42
Jul 14 2017 04:02
@scottleibrand I got it working! I used the link you sent me earlier and did it that way and its working great! Thanks a bunch!
although I can't ssh into it now (on a pc with the ip address) but not the most troubling issue
Dana Lewis
@danamlewis
Jul 14 2017 04:12
Just opened a PR to merge the dev docs (http://openaps.readthedocs.io/en/dev/) to master. the PR is here: openaps/docs#942 (and that’s what I’ve been linking people to, so if you’ve been enjoying the ‘new install script’ or some of the more nuanced troubleshooting docs, you’ve probably experienced part of dev already
Sebastien Lussier
@SebastienLussier
Jul 14 2017 04:16
@scottleibrand Should i just turn off AMA until i have more data ?
Dana Lewis
@danamlewis
Jul 14 2017 04:18
(something something turtles, prepping 0.5.0: https://github.com/openaps/oref0/pull/424#issuecomment-315264829)
Scott Leibrand
@scottleibrand
Jul 14 2017 04:18
AMA should work ok even if its predictions are sometimes off. you would need to dig into what decisions it's making that you think are incorrect to see whether/how those predictions are actually affecting dosing
Dana Lewis
@danamlewis
Jul 14 2017 04:27
happy birthday oref0 0.5.0 :confetti_ball: https://github.com/openaps/oref0/releases/tag/v0.5.0
huge thank you to everyone who contributed to this release - code review, code development, code testing, testing again, testing again again, etc. Lots of awesome helping hands. :raised_hands: :fire: :sparkles:
esialb @esialb claps
Cameron Chunn
@applehat
Jul 14 2017 04:31
Yay! I did one thing!
Matthias
@ceben80
Jul 14 2017 04:32
@danamlewis cool, great community! Together we're strong :tada: :star2:
Cameron Chunn
@applehat
Jul 14 2017 04:32
:shipit: great job everyone =)
Cameron Chunn
@applehat
Jul 14 2017 04:37
Excited to see more people able to use oref1 stuff :tada:
PieterGit
@PieterGit
Jul 14 2017 04:44
@esialb bluez 5.43 has been used with oref0 for a couple of weeks. quite some people have reported succes, but others had problems. I don't think those problems are related to bluez 5.43, but are general problems related to installing bluetooth.
@esialb i'm +1 on creating a stretch jubilinux image for oref 0.6.0 (dev)
Matt Cooper
@MrCooper42
Jul 14 2017 04:47
congrats on the release :D
Robin Kirkman
@esialb
Jul 14 2017 04:49
@PieterGit sounds good. i'm ironing out bugs with bluetooth right now actually.
Scott Leibrand
@scottleibrand
Jul 14 2017 04:49
heh, guess you figured it out, but: npm list oref0 will show your oref0 version
Matt Cooper
@MrCooper42
Jul 14 2017 04:49
haha yeah I figured it out thanks
Matt Cooper
@MrCooper42
Jul 14 2017 05:04
should it be an issue if that didnt work for me as a command anywhere? all i ever got was (empty) i just viewed the oref package.json to see i was on 0.4.3
Scott Leibrand
@scottleibrand
Jul 14 2017 05:05
mine shows:
# npm list oref0
oref0@0.5.0-dev /root/src/oref0
└── (empty)
you probably need to do npm list -g oref0 instead
Matt Cooper
@MrCooper42
Jul 14 2017 05:07
global tag did the trick!
speaking of tricks is this the place to ask about getting the loop to show on my nightscout site?
Scott Leibrand
@scottleibrand
Jul 14 2017 05:08
lots of docs on that to read and follow first, but then yes
Matt Cooper
@MrCooper42
Jul 14 2017 05:10
I will go through them one more, I set it up a while back and haven't updated it, I tried to see if there was an opportunity to do a pull request but it claimed i was up to date with master. (I already have all the right config vars) but I will give it a run and come back if I can't get it working
Robin Kirkman
@esialb
Jul 14 2017 05:15
boom
bluetooth seems to be working
time to upload a test image
Matt Cooper
@MrCooper42
Jul 14 2017 05:21
LOL I guess I just never refreshed and changed the settings to show it after i updated my vars... you guys rock!
Dana Lewis
@danamlewis
Jul 14 2017 05:21
:+1: it's nice when it's the easy fix!
Matt Cooper
@MrCooper42
Jul 14 2017 05:24
Totally! Sometimes it's just the little things.
Cylon75
@Cylon75
Jul 14 2017 06:08
Want to start with OpenAPS, what will be the preferred hardware route to run my rig on because the Intel Edison is being discontinued?
Scott Leibrand
@scottleibrand
Jul 14 2017 06:09
the Intel Edison
there are still more Intel Edisons out there than loopable pumps (and probably always will be). plus, once you have your rig(s) you won't need to buy another, so the discontinuance won't affect you.
by the time Edison availability becomes an issue, there'll be a new platform with improved functionality (probably a small screen, for one thing), but that will be months down the road.
Cylon75
@Cylon75
Jul 14 2017 06:11
ok, thanks, then i will order the edison for now
PieterGit
@PieterGit
Jul 14 2017 07:04
@esialb : please let me know when i can test a new image based on stretch (debian 9)
Marcus Lindfors
@Mackwe
Jul 14 2017 07:22
Woohoo! Congrats and thanks to everyone involved in 0.5.x release! Can't wait to have a go!
Robin Kirkman
@esialb
Jul 14 2017 07:45
@PieterGit hopefully soon. i've given up on uploading it with hotel wifi and am just going to use my 4g lte hotspot, which has data to burn
Robin Kirkman
@esialb
Jul 14 2017 07:52
@PieterGit it's being uploaded to http://www.jubilinux.org/dist/unstable/ , when the .zip.part is just a .zip you can download it
The total size is 615mb
Robin Kirkman
@esialb
Jul 14 2017 08:48
@PieterGit issue tracking for stretch jubilinux at https://github.com/jubilinux/jubilinux-stretch
stefanakerblom
@stefanakerblom
Jul 14 2017 09:10
Does anybody know howto display the battery of the rig on nightscout (grilled cheese)? I see it in the pumplog but found nothing in the docs.
Tim Street
@tim2000s
Jul 14 2017 09:47
Exciting- a new image!
Does that make it BLE compatible with Dexcom?
Robin Kirkman
@esialb
Jul 14 2017 09:48
@tim2000s not sure, but it should at least mean not compiling bluez
i will note that the 0.3.0-SNAPSHOT with stretch is probably the most experimental version yet and almost certainly has problems
one possible example is that i deleted the obsolete mraa from /usr/local . I could add a non-obsolete mraa to the image though if that's a problem.
Marcus Lindfors
@Mackwe
Jul 14 2017 10:59
Tried asking this before, but I'll try again; Do we have a Bluetooth/Jubilinux Wiz around? I can't get the Bluez/bt-pan client to get an IP. It had worked for days but just stopped for no obvious reasons. I've tried anything I could think of: Uninstall/reinstall of Bluez (two different versions), phone change (different phones, different telcos). The phone sees the Edison and says BT tethering is up, but the Edison never gets an IP. dhclient release/refresh, restarting the networking Daemon, complete powerdown etc. I just can't get an IP? Are there drivers on anything in Jubilinux to be uninstalled/reinstalled? Help, please..
PieterGit
@PieterGit
Jul 14 2017 11:03
@Mackwe did you check bluetoothctl (e.g. paired-devices)
Marcus Lindfors
@Mackwe
Jul 14 2017 11:04
@PieterGit Yes. Can't find anything wrong. Tried remove/re-pair too. And the bt-pan client connects to the phone. :-/
I seem to get a connection, but no IP.
PieterGit
@PieterGit
Jul 14 2017 11:05
did you try wifi tethering on the phone first, to see if the phone tethers fine?
Marcus Lindfors
@Mackwe
Jul 14 2017 11:06
@PieterGit Works fine. And the BT had worked too..
PieterGit
@PieterGit
Jul 14 2017 11:07
bluez on ancient kernel can do crazy stuff...
i think the best way forward will be to move dev to debian9 / stretch (@esialb) just released first alpha 0.3.0 version based on stretch and debian bluez 5.43
i fear most of the weird bluetooth problems are kernel related. i still would like to go to a 4.12.1+ kernel on the edison, but that has quite some work to go.
kernel is progressing, see https://communities.intel.com/thread/75472?start=105&tstart=0. That would probably solve quite some bluetooth and USB issues, but create new issues (e.g. not using MRAA anymore)
Marcus Lindfors
@Mackwe
Jul 14 2017 12:02
@PieterGit Thanks for the info. So while I wait for a newer kernel to be developed, I need to re-flash the Edison to make it work again, I guess.. :-(
PieterGit
@PieterGit
Jul 14 2017 12:23
maybe somebody else can help debug your bluetooth problem, here in EU we don't have bluetooth dexcom G4 :worried:
Marcus Lindfors
@Mackwe
Jul 14 2017 12:58
@PieterGit I am in Sweden ;-). No, this is not a Dexcom problem. I simply want the Edison to tether to my phone again. That's all.
Russell Gardner
@RussGG
Jul 14 2017 13:21
@Mackwe I am not a tech guy so I don’t understand alot of this. When I first set up BT to an iPhone it had issues. My phones name had an apostrophe in it, eliminating that fixed it.
Diabetic_Dude
@diabetic_dude_twitter
Jul 14 2017 13:21
I had that same problem too
made me so mad
that is also covered in the docs as of about 2 months ago
Marcus Lindfors
@Mackwe
Jul 14 2017 13:24
I see your frustration. What I can't wrap my head around is that everything worked fine for days and suddenly just didn't. I haven't changed anything afaik and it just stopped working. :-( I can see and connect to both phones, but can't get any data to pass the (BT) wireless link.
Diabetic_Dude
@diabetic_dude_twitter
Jul 14 2017 13:27
yeah I think that is likely not your problem
because it worked before. I also know that Iphones are not 100% with the BT set up (it also use to say that in the docs) I use to have loads of problems with it and had to switch to wifi
then after a few months it just started working again after a fresh set up so I switched back
if you had access to an android phone, I would try testing the set up with that just to see if it works
then you would know if it is your rig or your phone
what is your phone named? (I changed mine from "James's IPhone" to "in10city"
Diabetic_Dude
@diabetic_dude_twitter
Jul 14 2017 13:36
@Mackwe this is a quote from the support docs, the problem I would bet is your phone "Users have experienced various levels of success with the iPhone bluetooth tethering and when the rig switches between wifi and BT"
sux but I have experienced what your are going through and it just magically fixed one day after a fresh re set up.
for the record I do not recommend you start over to see if it fixes it, just understand that others have gone through the same thing and the issues are attributed to their iphone
Marcus Lindfors
@Mackwe
Jul 14 2017 13:42
@diabetic_dude_twitter Thank you for sharing. This is two Android phones. One 4.2.2 called "Albins S3" and one 6.0(?) called "Nanna". Both Samsung phones. And I have modded the oref0-online script since I don't want the Eddie to ever let go of the BT link. WiFi & BT is good, but the BT should always be up, since I sent my xDrip data directly to the rig w/o Nightscout.
But the problem remains even if I stop Cron and do everything manually.
cameronrenwick
@cameronrenwick
Jul 14 2017 13:48
I have a question about profile editor. At the bottom of the page there is a target BG range. I have had this set as it was on my pump from 4.4 to 10.4. I've always interpreted this as the range throughout a 24hr period I want to keep my blood sugar in. On my pump, under bolus setup, my BG target is set to 5.6 to 6.5. I've taken that value as: here's where I want my pump to try and keep me. Can someone help me understand if this is correct and/or should be adjusted? For the record, my BGs are doing really well but I just want to understand these 2 settings.
Diabetic_Dude
@diabetic_dude_twitter
Jul 14 2017 13:49
I personally set mine to the same to stop confusion on my part and I dont want to fight with the APS code when using the bolus wizard on my pump
cameronrenwick
@cameronrenwick
Jul 14 2017 13:50
so then you've got a fairly narrow range on your profile editor I'm thinking?
Diabetic_Dude
@diabetic_dude_twitter
Jul 14 2017 13:51
I do, all over
my doctor even commented on the tight range but then regressed to "if it is working for you then keep it up"
cameronrenwick
@cameronrenwick
Jul 14 2017 13:54
ok so thinking further, my understanding is that the pump rules so despite the fact that NS had 4.4 to 10.4, my pump, with its settings at 5.6-6.5 has always been setting the pace I believe. I guess this is really what I'm hoping to confirm
Diabetic_Dude
@diabetic_dude_twitter
Jul 14 2017 14:00
I do not believe so. but kind of.. I keep my setting mirrored so when I loose connection to NS I am still getting something similar to what my rig would do in normal circumstances. NS when connected always rules what happens when working.
sorry I will talk more in a second just got a call
the pump will not set the pace only will use those setting when doing the bolus wizard or when you MANUALLY try to correct a high
but in short that is why I keep mine the same, to avoid any confusion.
Diabetic_Dude
@diabetic_dude_twitter
Jul 14 2017 14:09
also @cameronrenwick if you are running autotune, that adds a whole new element
cameronrenwick
@cameronrenwick
Jul 14 2017 14:10
@diabetic_dude_twitter yes I am... and it's working very well too.. so all in, I'm doing very well
Diabetic_Dude
@diabetic_dude_twitter
Jul 14 2017 14:13
yup so is mine.... just saying that because I was watching the logs and noticing small changes that I did not enact.
an example is my ISF is set to 37 in NS but on my rig it is 46 (a huge difference actually)
cameronrenwick
@cameronrenwick
Jul 14 2017 14:15
ya I've done that too... that's where understanding the code and what everything means and does is so important in my opinion
Diabetic_Dude
@diabetic_dude_twitter
Jul 14 2017 14:16
yup agreed. I have been just watching my logs for a few days now with everything working just to help me understand how it is operating
I always look at it when I have problems, but then have problems find the issues because I don't know what good normal data looks like until recently
please excuse my grammar "but then I have problems finding the issues"
cameronrenwick
@cameronrenwick
Jul 14 2017 14:19
the totally ironic part about looping is, part of the genesis of looping is to reduce our involvement in T1D care... I think I spend 4x as much time now as when I was simply on a pump... all for good reason and better health, but if I wasn't a computery-techy-nerd (as my wife and kids remind me daily) I don't think I'd be nearly as interested...
Diabetic_Dude
@diabetic_dude_twitter
Jul 14 2017 14:20
LOL TRUE
you are also right about the docs, I read and read and read. and they change and change and change. I learn the code and it changes. then I start all over again but all in good faith and for good reason
my wife calls me inspector gadget because of all the diabetic "tools" I have now
(yes honey we do need amazon alexa it is on sale and it will help me with my diabetes)
cameronrenwick
@cameronrenwick
Jul 14 2017 14:22
the other day someone asked me if the "red light glowing in my pocket" (I had my rig in my dress-shirt pocket) was "my heart"...nope..pancreas actually....
I got this look from her like...ummm.. what?. I just said look, if you really wanna know we need an hour....
Diabetic_Dude
@diabetic_dude_twitter
Jul 14 2017 14:27
lol
Dana Lewis
@danamlewis
Jul 14 2017 14:30
@cameronrenwick OpenAPS runs on the settings from the pump. Not Nightscout.
So target, dia, isf, basals, etc...all from the pump
Your call on whether you want NS to be same for peace of mind, or different for different visual . But regardless. That doesn't impact OpenAPS operations
Diabetic_Dude
@diabetic_dude_twitter
Jul 14 2017 14:32
(sorry dana)
Dana Lewis
@danamlewis
Jul 14 2017 14:33
@diabetic_dude_twitter not sure why you're apologizing - no need!
Diabetic_Dude
@diabetic_dude_twitter
Jul 14 2017 14:34
yeah but giving out wrong info is never a good thing
Cameron Chunn
@applehat
Jul 14 2017 15:58
@Mackwe fwiw, I often times have "no data" on the edison even when bluetooth tethered. Rebooting both tends to fix the issue (I think android causes it in my case, tho)
ArthurusDent
@ArthurusDent
Jul 14 2017 15:58
I'm sure this is quite off-topic, but I would like to get a second opinion, if I could. I just started a new sensor four hours ago and it's not behaving well. Would you replace it? https://www.dropbox.com/s/gzopzi0oiln8wcx/Screenshot_20170714-175641.png?dl=0
Cameron Chunn
@applehat
Jul 14 2017 15:59
@ArthurusDent I would personally give it a little more time to see if it clears up. I've had sensors go wonky and revive. (Im also super cheap, and the idea of replacing a dexcom sensor before 20+ days bothers me, tho =P)
Also, probably not, but are you in a place you generally wouldn't be? My moms back yard has power lines near by and my dexcom jumps like crazy pack there from all the EMI.
ArthurusDent
@ArthurusDent
Jul 14 2017 16:00
@applehat ok, thanks
no, I'm at home so I don't expect any EMI distortion
the last sensor was super stable and worked 20 days
cameronrenwick
@cameronrenwick
Jul 14 2017 16:14
@ArthurusDent I'd be waiting a bit too but like @applehat I'm cheap (it's a common quality among Scottish ppl ; )) If it doesn't settle down though, I'd call your supplier. I've probably had 4 or 5 that have misbehaved like yours and my supplier in Canada (animas) is pretty good about providing a free replacement. Can't do it every month mind you but if you've done a normal insertion etc and it's giving out bad signal... complain!
and just to be clear..."Cameron" implies Scottish
ArthurusDent
@ArthurusDent
Jul 14 2017 16:14
@cameronrenwick Ok, thanks! :)
Scott Leibrand
@scottleibrand
Jul 14 2017 16:34
@cameronrenwick @diabetic_dude_twitter I believe the Nightscout profile editor BG range is what is used to calculate time in range in the NS reports. So typically people set that to something like 80 to 160 mg/dL or thereabouts. the BG target in the pump is used by OpenAPS to know what to correct to. Typically people set that to a really narrow range (often min == max) of 100-100 or 100-110 mg/dL or similar.
cameronrenwick
@cameronrenwick
Jul 14 2017 16:37
@scottleibrand perfect...I was just looking at that this very instant and was thinking that changing it to the narrow range WAS going to really have an impact on my report. Thanks so much for clarifying my thoughts. Is there a place I should be PRing this topic maybe?
Scott Leibrand
@scottleibrand
Jul 14 2017 16:38
yes, probably. I would take a look through the new dev docs (openaps.readthedocs.io/en/dev) to find the best place to mention it.
cameronrenwick
@cameronrenwick
Jul 14 2017 16:38
@scottleibrand will do
Scott Leibrand
@scottleibrand
Jul 14 2017 16:40
also, if everyone could read through those docs and comment on the open PR (openaps/docs#942) we'd like to get that merged ASAP if it looks better than the old/current docs
RedCyclist1
@RedCyclist1
Jul 14 2017 17:21
Is there any reason to update if we are on .5.0-dev?
Scott Leibrand
@scottleibrand
Jul 14 2017 17:22
depends on how recently you updated to dev
the most recent fixes went in on July 12
Luis Betancourt
@betluis
Jul 14 2017 17:24
@scottleibrand I have been trying to understand the loop issue going by doing manual instructions... I got this error when running "openaps report invoke settings/pumphistory-24h.json"
root@edison1:~/myopenaps# openaps report invoke settings/pumphistory-24h.json
pump://JSON/iter_pump_hours/settings/pumphistory-24h.json
settings/pumphistory-24h.json raised 192
Traceback (most recent call last):
File "/usr/local/bin/openaps-report", line 82, in <module>
app( )
File "/usr/local/lib/python2.7/dist-packages/openaps/cli/init.py", line 51, in call
self.run(self.args)
File "/usr/local/bin/openaps-report", line 75, in run
output = app(args, self)
File "/usr/local/lib/python2.7/dist-packages/openaps/cli/subcommand.py", line 52, in call
return self.method.main(args, app)
File "/usr/local/lib/python2.7/dist-packages/openaps/reports/invoke.py", line 40, in main
output = task.method(args, app)
File "/usr/local/lib/python2.7/dist-packages/openaps/uses/use.py", line 45, in call
output = self.main(args, app)
File "/usr/local/lib/python2.7/dist-packages/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 158, in find_records
records = decoder.decode( )
File "/usr/local/lib/python2.7/dist-packages/decocare/history.py", line 823, in decode
record = parse_record(self.stream, B, larger=larger, model=self.model)
File "/usr/local/lib/python2.7/dist-packages/decocare/history.py", line 755, in parse_record
record.parse( head + date + body )
File "/usr/local/lib/python2.7/dist-packages/decocare/records/base.py", line 66, in parse
return self.decode( )
File "/usr/local/lib/python2.7/dist-packages/decocare/history.py", line 571, in decode
temp = { 0: 'absolute', 1: 'percent' }[self.head[1]]
KeyError: 192
root@edison1:~/myopenaps#
Scott Leibrand
@scottleibrand
Jul 14 2017 17:24
which temp basal mode is your pump in?
Luis Betancourt
@betluis
Jul 14 2017 17:25
this is the brand new master
it's absolute (not %)
Scott Leibrand
@scottleibrand
Jul 14 2017 17:27
k. does openaps report invoke monitor/pumphistory.json fail the same way?
RedCyclist1
@RedCyclist1
Jul 14 2017 17:27
@scottleibrand Updated on Sunday so it sounds like I am good.
Luis Betancourt
@betluis
Jul 14 2017 17:27
no, this one does not fail
Scott Leibrand
@scottleibrand
Jul 14 2017 17:28
interesting. so the difference between those two commands is:
root@edison-eb4 ~/myopenaps # openaps report show settings/pumphistory-24h.json --cli
openaps use pump iter_pump_hours 27.0
root@edison-eb4 ~/myopenaps # openaps report show monitor/pumphistory.json --cli
openaps use pump iter_pump_hours 2.0
try running that with different numbers of hours and see where the cutoff is?
Luis Betancourt
@betluis
Jul 14 2017 17:29
root@edison1:~/myopenaps# openaps report invoke monitor/pumphistory.json
pump://JSON/iter_pump_hours/monitor/pumphistory.json
reporting monitor/pumphistory.json
root@edison1:~/myopenaps# echo $?
0
ok
Scott Leibrand
@scottleibrand
Jul 14 2017 17:30
@RedCyclist1 Sunday was July 9, so you're missing the unsuspend fix at the very least. if you have that feature enabled you'll want to update so it works.
RedCyclist1
@RedCyclist1
Jul 14 2017 17:30
@scottleibrand Ok, wow, I was off a few days
Luis Betancourt
@betluis
Jul 14 2017 17:32
root@edison1:~/myopenaps# openaps report show settings/pumphistory-24h.json --cli
openaps use pump iter_pump_hours 27.0
root@edison1:~/myopenaps# openaps report show monitor/pumphistory.json --cli
openaps use pump iter_pump_hours 2.0
Scott Leibrand
@scottleibrand
Jul 14 2017 17:34
those commands just display the commands you want to run
so run openaps use pump iter_pump_hours 27.0 and it should error, but with 2.0 it should succeed
Luis Betancourt
@betluis
Jul 14 2017 17:34
I see
Scott Leibrand
@scottleibrand
Jul 14 2017 17:35
and if that does what we expect, then try numbers in between 2 and 27 until you figure out where the cutoff is
Luis Betancourt
@betluis
Jul 14 2017 17:35
I see... yes
with 12 it works, with 13 it fails
Scott Leibrand
@scottleibrand
Jul 14 2017 17:38
ok. if you look in the pump's bolus history, can you see anything odd at 12-13 hours ago that might contribute to that?
Luis Betancourt
@betluis
Jul 14 2017 17:39
ahh good question... jaja I am at the other side of the world than my daughter righ now.. I need to call her
but what kind of oddness you would expect?
Scott Leibrand
@scottleibrand
Jul 14 2017 17:39
is there anything weird with timezones then?
I'm guessing there's a record in the pumphistory that decocare is failing to parse correctly
but no idea what that record might be: if it were a known one it would probably parse fine
Luis Betancourt
@betluis
Jul 14 2017 17:40
ok
Scott Leibrand
@scottleibrand
Jul 14 2017 17:41
this has been going on for more than 24h, right? if so, you might also want to try setting a bunch of manual temp basals (either via the pump buttons or via manual openaps commands) to see if you can push any problematic records off the more recent history pages and get longer iter_pump_hours queries to succeed
you might also try openaps use pump iter_pump <# of records> to see how far back the problematic one is
Luis Betancourt
@betluis
Jul 14 2017 17:42
interesting, I will try this one
will try to find out how many records to the problematic one
Scott Leibrand
@scottleibrand
Jul 14 2017 17:43
@cameronrenwick looking more closely at the NS profile editor and your PR, I think I may have misled you. Target BG range looks to be different from BG_TARGET_TOP etc. So perhaps Target BG range is just used in the Bolus Wizard Preview in NS...
Luis Betancourt
@betluis
Jul 14 2017 17:46
record 17th is the problematic one
Andy Sharrow
@dramageek
Jul 14 2017 17:47
@scottleibrand @cameronrenwick when you pull reports in Nightscout, there are entries for what to consider in-range.
Scott Leibrand
@scottleibrand
Jul 14 2017 17:50
@betluis then you might want to try adding a few more history records (by repeatedly setting an appropriate temp basal) and see if that number increments
Luis Betancourt
@betluis
Jul 14 2017 17:51
I see
the last ok record is from yesterday near midnight
so from that point to now there are 12 hrs (the exact number of hours the other command is ok)
would this mean we can wait 15 more hours for the 27 hours of the openaps use pump iter_pump_hours 27.0 command to work?
Scott Leibrand
@scottleibrand
Jul 14 2017 17:53
maybe, but hasn't it been longer than that already since the problem started?
Luis Betancourt
@betluis
Jul 14 2017 17:54
let me see...
in fact yes
Scott Leibrand
@scottleibrand
Jul 14 2017 17:54
If you want to manually add some records to pumphistory, you can send a bunch of "cancel temp basal" commands:
echo '{ "temp": "absolute", "rate": 0, "duration": 0 }' > canceltemp.json
killall -g oref0-pump-loop; openaps use pump set_temp_basal canceltemp.json
and then repeat the last command as many times as you like
do that from inside your myopenaps directory
Luis Betancourt
@betluis
Jul 14 2017 17:55
I can try this, but would this make an hours-related record to change?
by the other hand I wonder if we are getting new problematic records being generated... extending the problem
Scott Leibrand
@scottleibrand
Jul 14 2017 17:56
yeah, if the problem is truly a record from 12h ago, then it might be something that's getting recorded by the pump at midnight every night, in which case it won't age off
but the other possibility is that it's another record on the same "page" of data as the 12h-ago record
Luis Betancourt
@betluis
Jul 14 2017 17:56
what the h.. could that be??
Scott Leibrand
@scottleibrand
Jul 14 2017 17:56
in that case, adding more records would push that back
Luis Betancourt
@betluis
Jul 14 2017 17:57
ok, will work on it now
Scott Leibrand
@scottleibrand
Jul 14 2017 17:57
the pump does record a "daily totals" record every day at midnight
one other thing: are the pump and rig set to the same time?
Luis Betancourt
@betluis
Jul 14 2017 17:59
mmm I need to check, now that you mention it possibly we have a summer time one hour shift
Eric
@ecc1
Jul 14 2017 18:04
Anyone in the Denver area? I'm visiting for the week and a wire on my rig has detached. Any suggestions on makerspaces or other places where I could resolder it? Please tag me in any responses so I don't miss it. Thanks. Will build a backup before my next trip!
dan lightner
@danimaniac
Jul 14 2017 18:11
My current rig / config is giving me great results and I am completely satisfied with it's operation. I realize there is no need to update, however, I will never know if I like the newly released 0.5.1 BETTER unless I try it. I know how to update, but how do I roll back to my current release if I try 0.5.1 and am unhappy?
Diabetic_Dude
@diabetic_dude_twitter
Jul 14 2017 18:11
lol
I was in the same spot as you but did just end up updating both of my rigs
@ecc1 i dont live there but hope this helps you a bit http://coloradomakerhub.org/makerspace-map/
looks like there is a few in denver
Luis Betancourt
@betluis
Jul 14 2017 18:15
@scottleibrand I can now achieve 100+ records but no more than 14 hrs
what do you think
could it be possible to modify the code so that "pumphistory-24h" goes only for 14 hrs
and then get it back to 27 hrs the next day and try again?
Scott Leibrand
@scottleibrand
Jul 14 2017 18:18
yes, you could do that temporarily. autosens might be a bit off
Luis Betancourt
@betluis
Jul 14 2017 18:18
right, I understand that
ok, so I need to find where to do this
Scott Leibrand
@scottleibrand
Jul 14 2017 18:18
try searching your openaps.ini for 27
Luis Betancourt
@betluis
Jul 14 2017 18:18
ok
Luis Betancourt
@betluis
Jul 14 2017 18:27
ok, changed to 13hrs now, restarted cron and running the log...
let's se.......
jmcrawford
@jmcrawford
Jul 14 2017 18:32
We're running Dev branch circa last Saturday. Saw a situation this morning where IOB went from 1.32 to -0.17 on the next loop. This occurred a little more than 2 hours from last bolus (DIA is 4, so about half?). I haven't seen that before, but doesn't seem quite right. Possibly something that has been fixed in the last week?
Luis Betancourt
@betluis
Jul 14 2017 18:33
yep! finally..... I never thought seeing the openaps pill alive would mean so much!!!
@scottleibrand I will try to change the hours tomorrow back to 27, I wonder if this is something to address as an issue?
Scott Leibrand
@scottleibrand
Jul 14 2017 18:34
@jmcrawford yes, there was a probably-relevant fix for that recently: openaps/oref0#540
@betluis if you want to figure it out, you'll need to get familiar with decocare. I'm not, really
jmcrawford
@jmcrawford
Jul 14 2017 18:35
That references COB, not IOB. The IOB drop was 30 minutes ago, hasn't recovered
Scott Leibrand
@scottleibrand
Jul 14 2017 18:36
@betluis there are some tools for pulling the raw pumphistory data off the pump so we can figure out what's being decoded improperly. Ben was really good with that, but there are a few others who've tried their hand at it since he mostly moved on....
Luis Betancourt
@betluis
Jul 14 2017 18:36
what puzzles me is that this doesn't seem to be an issue at all, I only saw one instance when someone else complained about this
Scott Leibrand
@scottleibrand
Jul 14 2017 18:37
@jmcrawford not sure if it would also affect IOB or not, but to eliminate variables, I would upgrade and see if the problem persists.
Luis Betancourt
@betluis
Jul 14 2017 18:38
if you think it's worth doing it I will try...
Scott Leibrand
@scottleibrand
Jul 14 2017 18:38
it's your issue, so totally your call
Luis Betancourt
@betluis
Jul 14 2017 18:38
can you suggest a person to go to?
Scott Leibrand
@scottleibrand
Jul 14 2017 18:39
I would start with a decocare issue and see if anyone chimes in there
Luis Betancourt
@betluis
Jul 14 2017 18:56
Ok
Good
Marco
@CaptainBalou
Jul 14 2017 22:12
Tried to install latest oref0 dev 0.5.1. But while doing global-install it says „oref0@0.4.3“. Is that correct? I would assume to see @0.5.1?
wharper11
@wharper11
Jul 14 2017 22:14
looking for the directions for mongo lab clean up, anyone have that link?
Marco
@CaptainBalou
Jul 14 2017 22:27
Oh, I should probably take master branch to get 0.5.1...
wharper11
@wharper11
Jul 14 2017 22:32
BOOM! Thanks @danamlewis
reports are not rendering in NS, I think it needs a clean up.
Dana Lewis
@danamlewis
Jul 14 2017 22:33

Just merged the dev docs. Don't freak out, the docs now look different! http://openaps.readthedocs.io/en/latest/index.html

Please take a look, PRs allllways welcome and especially after a re-org like this :smile:

Old docs if needed: http://openaps.readthedocs.io/en/2017-07-13/

katie disimone
@Kdisimone
Jul 14 2017 22:33
@danamlewis wait…i will freak out
I wasn’t ready yet. LOL…change. AGH. ;)
Dana Lewis
@danamlewis
Jul 14 2017 22:34
:D
Big thanks to @Kdisimone for huge help on this doc re-org and lots of new content added
Greg Eden
@skaffen
Jul 14 2017 22:37
quick question - following the docs resulted in an ancient version of node being installed (0.10.29). this produces warnings during oref0 install. Am i fine to install 6.11.1?
Dana Lewis
@danamlewis
Jul 14 2017 22:39
@skaffen which docs when and doing what? :smile: (just merged docs so want to clarify what you're following )
francesc0-cgm
@francesc0-cgm
Jul 14 2017 22:40
@scottleibrand is cob limit still same as dia on last dev/master?
Tia
which i ran through initially about 2 weeks ago.
and then some stages today though nothing involving node particularly
if i were installing node on a server i would normally follow these instructions https://nodejs.org/en/download/package-manager/#debian-and-ubuntu-based-linux-distributions
wharper11
@wharper11
Jul 14 2017 22:44
trying to run reports on NS and it's locking up in the render phase. looking at MLab I have 112MB. anyone else having NS issues?
Greg Eden
@skaffen
Jul 14 2017 22:46
I like the new oref0 setup questions btw - the MaxIOB and nightscout token questions address two of my primary gripes/confusions
Dana Lewis
@danamlewis
Jul 14 2017 22:47
:+1: thx for feedback. please also take a look at the new docs and flag anything missing and/or submit PRs to help fix.
francesc0-cgm
@francesc0-cgm
Jul 14 2017 22:51
Because with my son i had to put carb absorption to 4 from 8 on preferences.json but he had 7 gr of cob and they suddenly went to 0 after dia
I believed now it was moved to 6 hrs to avoid zombies
Greg Eden
@skaffen
Jul 14 2017 22:52
ok will do. any reason not to install latest LTS node? there was a major security advisory issued yesterday...
PieterGit
@PieterGit
Jul 14 2017 22:54

i tested oref0 with v6.11.1 and I upgraded all the dependencies in package.json. 104 tests are succesfull, but there is one problem with determine-basal within the AMA code.

Problem with iobArray.  Optional feature Advanced Meal Assist disabled: TypeError: iobArray.forEach is not a function
    at determine_basal (/root/src/oref0/lib/determine-basal/determine-basal.js:344:18)
    at Context.<anonymous> (/root/src/oref0/tests/determine-basal.test.js:89:22)
    at callFn (/root/src/oref0/node_modules/mocha/lib/runnable.js:348:21)
    at Test.Runnable.run (/root/src/oref0/node_modules/mocha/lib/runnable.js:340:7)
    at Runner.runTest (/root/src/oref0/node_modules/mocha/lib/runner.js:443:10)
    at /root/src/oref0/node_modules/mocha/lib/runner.js:549:12
    at next (/root/src/oref0/node_modules/mocha/lib/runner.js:361:14)
    at /root/src/oref0/node_modules/mocha/lib/runner.js:371:7
    at next (/root/src/oref0/node_modules/mocha/lib/runner.js:295:14)
    at Immediate.<anonymous> (/root/src/oref0/node_modules/mocha/lib/runner.js:339:5)
    at runCallback (timers.js:672:20)
    at tryOnImmediate (timers.js:645:5)
    at processImmediate [as _immediateCallback] (timers.js:617:5)

Anybody know how ti rewrite line 344 iobArray.forEach(function(iobTick) { of determine-basal so that it is node v6.11 compatible?

Scott Leibrand
@scottleibrand
Jul 14 2017 22:58
@skaffen @PieterGit I've created a bootstrap and install script for nodejs-6: https://raw.githubusercontent.com/openaps/docs/nodejs-6/scripts/openaps-bootstrap.sh if you want to help test it
I've verified it installs ok, but haven't gotten as far as @PieterGit in testing it with all of determine-basal etc.
that said, I don't think there's much risk in continuing to run node 0.10 on non-development rigs, as oref0 doesn't interact with the outside world in a way that a DoS vulnerability should matter.
PieterGit
@PieterGit
Jul 14 2017 23:05
@scottleibrand : found your script already. I noticed i had problems with uploading data from the rig (e.g. upload-first). It was complaining about JSON files not valid, when I could not find any problems (and jq validated them as well). Those problems disappeared after upgrading the oref0 dependencies, see https://github.com/openaps/oref0/pull/528/commits/7aa73f33a42d93175ef998a4870dc90d8cbb14bd ) . Note: i'm testing this on a jubi 0.3.0 snapshot with stretch / debian9 on an explorer board. Note this is experimental, and should only be tried on a spare rig.
Scott Leibrand
@scottleibrand
Jul 14 2017 23:07
ok. when we get a chance we should also try the same thing on jubilinux 0.1.1 with the nodejs-6 stuff.
Greg Eden
@skaffen
Jul 14 2017 23:13
not too concerned about the DoS vuln, more curious about the stability offered by tracking an LTS branch, which is my instinct on a server. i understand it's prob not a big priority. alas my meagre sysadmin skills not really up to helping currently.
katie disimone
@Kdisimone
Jul 14 2017 23:19
@wharper11 Not having issues here. Have you also checked your “size on disk” in mlab? there’s TWO parts to mlab size that need to be checked.
PieterGit
@PieterGit
Jul 14 2017 23:20
@scottleibrand upgrading node to v6.11.1 and dependencies with jubi 0.1.1 is on my list for this weekend (I hope). But first need to get the Javascript issue with determine-basal out of the way ( https://gitter.im/nightscout/intend-to-bolus?at=59694bc2329651f46e8ab3f2 ) , because I suspect that otherwise node v6.11.1 (or the dependencies) break AMA.
Scott Leibrand
@scottleibrand
Jul 14 2017 23:20
@skaffen for "stability" the general approach of "if it ain't broke don't fix it" probably applies here. we'll get node 6 (or 8) tested and start using it eventually, though.
@PieterGit are you sure that error is caused by 6.11.1 itself and not by the way to installed it or other idiosyncrasies? that's why I was suggesting testing it with jubi 0.1.1 and everything else as similar as possible, to see if it's a node version thing or something else
I do have a jubi 0.1.1 rig I installed from the nodejs-6 branch, so I'll keep an eye out for the problem there
PieterGit
@PieterGit
Jul 14 2017 23:23
and before releasing it to the public, we must see what problems we encounter. I must say that @esialb work on upgrading to debian9 looks promising
Iain
@iainct
Jul 14 2017 23:23

I am experiencing a recurring issue. Using a explorer board intel edison. oref1 (hence on the dev branch) my will function for a short time and then keep hitting this fault

mmeowlink.exceptions.CommsException: No acknowledgement from pump on wakeup. Is it out of range or is the battery too low?

any ideas? A reboot tends to clear it, but its not ideal....

I am not sure what else to check at this point....
PieterGit
@PieterGit
Jul 14 2017 23:26
@iainct : there are some fixes for that in the 0.5.1 release. In case of an eror the pump loop should mmtune in about 25% of the cases. After the mmtune the pumploop usually runs fine. I think I also experience more of those exceptions than before, but am not very sure.
Iain
@iainct
Jul 14 2017 23:26
my rig is going through the mmtune sequence but continues to fail after that....
i normally notice when i have not been looping for an hour or so of failed retry's....
PieterGit
@PieterGit
Jul 14 2017 23:27
if you're watching the pump loop, sometimes /etc/init.d/cron stop and openaps mmtune fixes it sooner. Check if there are no hanging openaps/pump processes
Iain
@iainct
Jul 14 2017 23:28
i just killall openaps and managed to monitor-pump
Scott Leibrand
@scottleibrand
Jul 14 2017 23:28
I would upgrade to 0.5.1. depending on when you installed dev, some of the fixes @PieterGit is talking about might help
Iain
@iainct
Jul 14 2017 23:29
remind me how do i check my current version?
Greg Eden
@skaffen
Jul 14 2017 23:29
@scottleibrand yes. and also... "keep breaking until there's nothing left to fix"
Scott Leibrand
@scottleibrand
Jul 14 2017 23:29
@iainct if you were on dev, you'll be 0.5.0-dev. npm list -g oref0 though
we just released 0.5.0 to master yesterday and then immediately had to patch release to 0.5.1 due to a bad merge conflict resolution syntax error
Iain
@iainct
Jul 14 2017 23:31
yep i am on 0.5.0 i will upgrade now
Scott Leibrand
@scottleibrand
Jul 14 2017 23:31
but the code should be more or less identical to 0.5.0 as of yesterday. there were just lots of updates to that in the last few weeks, so hard to tell exactly which ones you have
Iain
@iainct
Jul 14 2017 23:32
thanks i will give this version ago and see what crops up
PieterGit
@PieterGit
Jul 14 2017 23:32
I'm out, way past normal bedtime here. Please fix all the remaining problems and add tons of new features while I'm asleep, and I'll join later this weekend.
Scott Leibrand
@scottleibrand
Jul 14 2017 23:35
lol sure
Marco
@CaptainBalou
Jul 14 2017 23:42
You are all working so hard here. Thanks guys. You keep all of us on stable 100. 👍🏻
Dana Lewis
@danamlewis
Jul 14 2017 23:44
:+1:
@PieterGit ;) thanks for all your work, too