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

4th
Oct 2017
garykidd
@garykidd
Oct 04 2017 00:01
travel kit. interlocking AA battery storage trays repurposed.....
20171003_195808_resized.jpg
jaylagorio
@jaylagorio
Oct 04 2017 00:59
@garykidd Are each of those top two cases independent rigs?
David
@davidkeddydb
Oct 04 2017 01:00
I am trying to put together another rig with a spark fun baseboard, Edison and TI stick. I have 2 other rigs with this set up. Usually very easy set up. This time I flashed the TI stick, plugged the TI stick into the base block and have run ls -al /dev/tty* but I am getting nothing different when I plug the TI stick in relative to the port addresses. Any clue what dumb thing I did? The TI stick is lightening up when plugged into the base block
jaylagorio
@jaylagorio
Oct 04 2017 01:03
@scottleibrand @danamlewis Adding 15 minutes to the default maxSMB preference took 20 points off the peak - and I had a random 20 minutes of not looping so I suspect it might have been better had that not happened.
image.png
garykidd
@garykidd
Oct 04 2017 01:13
@jaylagorio yup separate rigs each with 2500 mah batteries
@scottleibrand that sounds pretty achievable/elegant. I kinda know grep but will just need to probably google a bit on the arguments maybe
jaylagorio
@jaylagorio
Oct 04 2017 01:45
@scottleibrand @danamlewis I spoke too soon, more like 10 points off the peak but I stand by the part where the 20 minutes of failed looping contributed to that. More testing tomorrow.
@garykidd Did you shave the molding that holds the batteries in place down or something? That's pretty slick.
David
@davidkeddydb
Oct 04 2017 01:53
Screenshot 2017-10-03 21.50.51.png
@jaylagorio I just tried 120 mins, but at least on this meal,wasn't enough. I did enter the carb entry about 20 mins prior to eating, but I am going to try 135 mins next time
jaylagorio
@jaylagorio
Oct 04 2017 02:15
I wonder if this is a YDMV thing? I don't know, maybe this is one setting that will need to be tweaked individually.
Thacher
@thacherT1D
Oct 04 2017 02:28
Hey @scottleibrand or @danamlewis what's our estimated current # of openaps users? At a session at Health 2.0 and you're being cited :) -- she said 100 people... I know it's more than that? at MedX dana said ~400?
Scott Leibrand
@scottleibrand
Oct 04 2017 02:28
@davidkeddydb before raising maxSMBBasalMinutes any further, check whether it was actually hitting maxBolus at all. it might've just been limited by insulinReq
Dana Lewis
@danamlewis
Oct 04 2017 02:29
OpenAPS.org/outcomes always has the updated count. (400+) @thacherT1D
Thacher
@thacherT1D
Oct 04 2017 02:29
Thanks :)
Scott Leibrand
@scottleibrand
Oct 04 2017 02:29
that's "individuals around the world with various types of DIY closed loop implementations (that we know of)", including Loop and AndroidAPS users who filled out the form
Dana Lewis
@danamlewis
Oct 04 2017 02:30
That's all diy loopers, and an underestimate based on who self reports
@scottleibrand better bring me a diet soda when you pick me up ;)
garykidd
@garykidd
Oct 04 2017 02:41
rig #3 lost connectivity to wifi....may be a sledge hammer to thumb tack but I just reflashed and re-ran script. ...it'snot troubleshooting but it's highly probable to solve the problem.
Scott Leibrand
@scottleibrand
Oct 04 2017 02:42
the "nuke it from orbit" troubleshooting method? :)
garykidd
@garykidd
Oct 04 2017 02:44
yup ...something like that....not that I'm that handy but it kind of reminds me of how technicians troubleshoot pc's by swapping various boards...nobody knows what an oscilliscope is for these days...but who can blame them...doesnt ususally pay
Scott Leibrand
@scottleibrand
Oct 04 2017 02:48
yeah, when you automate manufacturing, replacement becomes way cheaper than repair
Sebastien Lussier
@SebastienLussier
Oct 04 2017 02:50
Hello! My "size on disk" on mlab is 496MB... I tried compacting (repairDatabase), got an "ok" response, but size on disk didn't change...
What are my options ? Do i need to delete data ? Size is ~260MB
garykidd
@garykidd
Oct 04 2017 03:04
@SebastienLussier there's a section for that in troubleshooting...give it a shot and I'm sure folks canhelp you if you cant take it further...
Scott Leibrand
@scottleibrand
Oct 04 2017 03:11
be sure to upload your data to the OpenAPS Data Commons on Open Humans before deleting anything if you decide that's necessary
br
@brmccollum
Oct 04 2017 04:07

Hi All,
I have been looping successfully for four days, however my battery got low on my pump and needed replacement. After replacing it, my rig is no longer looping consistently.

I keep seeing this in log, and I do not remember seeing it prior to replacing the battery: "need more than ) values to unpack"

```mmtune: "916.588", 5, -44
Starting pump-loop at Tue Oct 3 20:57:29 PDT 2017:
Waiting up to 4 minutes for new BG: glucose.json newer than pump_loop_completed
Radio ok. Listening: .No pump comms detected from other rigs
Old pumphistory: RefreshValueError: need more than 0 values to unpack

Error, retrying
pump_loop_completed more than 15m old; waiting for 30s silence before mmtuning
Radio ok. Listening: .No pump comms detected from other rigs
Listening for 30s silence before mmtuning: .No pump comms detected from other rigs
mmtune: "916.492", 4, -62 waiting for 4 second silence before continuing
Radio ok. Listening: .No pump comms detected from other rigs```

(Any direction would be appreciated.)

katie disimone
@Kdisimone
Oct 04 2017 04:14
That error @br usually clears out by itself eventually. The rig has good communication strength with pump, just needs to have a couple chances sometimes to get a good full pump history read.
br
@brmccollum
Oct 04 2017 04:20
Okay, thanks @Kdisimone
Craig Brenner
@cbrenner
Oct 04 2017 04:30
Is there a simple way to clone your app on Heroku under a single account so that I can create a separate NS sandbox/test site independent from our NS production site . I would rather not have to retype all of the settings, variables, etc. I didn't see any obvious way of doing this on Heroku.
Andrew Warrington
@andrew-warrington
Oct 04 2017 11:32
Hi all, I have a question about use of the Medtronic sensor. My rig is pulling glucose data straight from the pump/sensor instead of from Nightscout. I am wondering, in this configuration, does OpenAPS still consult Nightscout on each loop for info entered in the Care Portal, e.g. temporary targets or carb entries? Apologies I could not find this via search, and apologies I might be able to figure this out for myself by observation but was hoping for a quick "yes or no" answer. Thank you in advance!
Chris Oattes
@cjo20
Oct 04 2017 11:36
Not sure about carb entries, but it should try and pull temp targets.
Andrew Warrington
@andrew-warrington
Oct 04 2017 11:57
Thanks @cjo20 ! Good to know it tries to pull something. I can observe starting tonight and see what it pulls.
jaylagorio
@jaylagorio
Oct 04 2017 12:01
The latest dev build of 0.6.0 definitely pulls carbs from NS, that's how you announce meals without bolusing
Andrew Warrington
@andrew-warrington
Oct 04 2017 12:16
nightscout.jpg
I have my answer.
I infer from this that OpenAPS is not recognizing these carbs...
jaylagorio
@jaylagorio
Oct 04 2017 12:18
What version are you running? I think it's been picking up carbs for a while now.
Maybe the loop just hasn't hit NS yet? Or hasn't deposited the results of the loop into NS yet?
Andrew Warrington
@andrew-warrington
Oct 04 2017 12:19
It has made a number of changes to NS in the 2 hours since that carb entry, including updating the basal rates etc.
openaps version is 0.2.0-dev. not sure which command to run to get oref0 version.
I installed it, I thought from master, on Sunday.
jaylagorio
@jaylagorio
Oct 04 2017 12:22
You should check the running log on the rig and see what it says in the stream of entries that come through.
l at the command prompt after you log in
(The letter L, for log)
Andrew Warrington
@andrew-warrington
Oct 04 2017 12:24
ok I just entered 5 grams of carb in the care portal, and now 1) shows up in the pump-loop.log and 2) shows up in the COB pill. So I change my conclusion, YES it is picking those carbs up.
:P
nightscout2.jpg
jaylagorio
@jaylagorio
Oct 04 2017 12:24
Great!
Andrew Warrington
@andrew-warrington
Oct 04 2017 12:25
Thanks for the help! Really glad about this.
jaylagorio
@jaylagorio
Oct 04 2017 12:26
You got it! Always hit up someone here for help if you need it. This channel's a great resource.
cameronrenwick
@cameronrenwick
Oct 04 2017 12:29
need some help.. usual things don't get me looping forgot to tether when I left for work and now at work can't loop. Here's the log
Starting supermicrobolus pump-loop at Wed Oct 4 08:25:02 EDT 2017 with 12 second wait_for_silence:
Waiting up to 4 minutes for new BG: ls: cannot access /tmp/pump_loop_completed: No such file or directory
Radio ok. Listening: .No pump comms detected from other rigs
Preflight OK. Old pumphistory-24h, waiting for 12 seconds of silence: Radio ok. Listening: .No pump comms detected from other rigs
Old pumphistory-24h refreshValueError: need more than 0 values to unpack
SMB pump-loop failed. pump_loop_completed more than 15m old; waiting for 40s silence before mmtuning
Radio ok. Listening: .No pump comms detected from other rigs
Listening for 40s silence before mmtuning: .No pump comms detected from other rigs
mmtune: "916.684", 5, -37 Unsuccessful supermicrobolus pump-loop at Wed Oct 4 08:28:17 EDT 2017
any ideas? is there a "first-upload" command when using smb?
jaylagorio
@jaylagorio
Oct 04 2017 12:30
Does it do that over and over again?
cameronrenwick
@cameronrenwick
Oct 04 2017 12:30
yes
jaylagorio
@jaylagorio
Oct 04 2017 12:30
And you've tried rebooting?
cameronrenwick
@cameronrenwick
Oct 04 2017 12:30
yup
jaylagorio
@jaylagorio
Oct 04 2017 12:31
What CGM are you using?
cameronrenwick
@cameronrenwick
Oct 04 2017 12:31
g4
jaylagorio
@jaylagorio
Oct 04 2017 12:31
I don't have my rig handy, so bear with me here
cameronrenwick
@cameronrenwick
Oct 04 2017 12:31
with xdrip+... NS is getting BGs as per usual..
jaylagorio
@jaylagorio
Oct 04 2017 12:31
What's your openaps directory?
Ah, nevermind.
That was the first thing I was going to check
cameronrenwick
@cameronrenwick
Oct 04 2017 12:32
ya it's really strange... all things "seem" good
jaylagorio
@jaylagorio
Oct 04 2017 12:32
Old pumphistory-24h refreshValueError: need more than 0 values to unpack
So that's definitely the problem
The pump isn't responding right
Preflight OK. I think that means it's responding at all, which is good
cameronrenwick
@cameronrenwick
Oct 04 2017 12:32
ya.. never had that before
jaylagorio
@jaylagorio
Oct 04 2017 12:32
Have you tried pulling and replacing the pump battery?
cameronrenwick
@cameronrenwick
Oct 04 2017 12:33
ya.. gonna do a cold start on both pump and rig by pulling batt...
back soon
jaylagorio
@jaylagorio
Oct 04 2017 12:33
Good luck!
Katie Aldridge
@kcrcgm
Oct 04 2017 12:47
Quick autotune question. Could lows from exercise cause autotune to incorrectly raise my isf? I worked in the garden the last few afternoons and got low. Autotune has my isf has gone up each day. Now it's at at 78, which seems really high to me... (i adjusted my physical pump settings yesterday to match autotune also)
Chris Oattes
@cjo20
Oct 04 2017 13:11
Anyone had problems with their edison not being able to reconnect to wifi networks? Every time I move out of range of a network, I need to reboot to get it to reconnect
Steve Bernacki
@sbernacki
Oct 04 2017 13:43
@cjo20 I've had that happen with my Edison as well. Not 100% consistently, but there's times where moving from hotspot back to Wifi causes the device to kernel panic, requiring a hard reset.
Chris Oattes
@cjo20
Oct 04 2017 13:56
sbernacki: it doesn't kernel panic, it just completely refuses to get a new ip address. Everything else still runs
Meghan Rutledge
@mar5477
Oct 04 2017 14:12
Since 0.6.0 has no-git and I'm running it, should I still be receiving corrupt objects or is something not set up correctly?
jaylagorio
@jaylagorio
Oct 04 2017 14:13
Was it a clean install or a rerun of the setup script?
Meghan Rutledge
@mar5477
Oct 04 2017 14:15
I've tried both. Most recently I followed the instructions under the pull request here (openaps/oref0#650)
jaylagorio
@jaylagorio
Oct 04 2017 14:15
Hm, I'd page @scottleibrand with that one...
Meghan Rutledge
@mar5477
Oct 04 2017 14:16
log shows "Preflight git.exc.InvalidGitRepositoryError: /root/myopenaps
fail."
gregorschlump
@gregorschlump
Oct 04 2017 14:21
How was boarding a flight (better to say passing the security at an airport) with the OpenAPS Edison (or whatever computer you are using) for you?
jaylagorio
@jaylagorio
Oct 04 2017 14:21
Problem free, every time. Not a single look once. @danamlewis frequently tweets the same, with pictures.
gregorschlump
@gregorschlump
Oct 04 2017 14:29
Ahh, thanks, @jaylagorio
Glenn Primack
@primags
Oct 04 2017 14:42
Can anyone help me find where the wifi supplicant logs are? I need to review the logs but can't see any detailed way to do that (networklogs doesn't give enough info). Thanks.
Andrew Warrington
@andrew-warrington
Oct 04 2017 14:46
@gregorschlump I walked around with an old-school soldered together version (pre Explorer Board) for 18+ months. To my friends and family it looks like a scary device. Despite that I passed through security in Boston, Newark, London (multiple times) and Basel (multiple times) -- probably others as well which I don't remember. Not a hesitation anywhere. Before my first trip I went to Basel airport and showed it to security and told them I had a flight in a week and asked them if it would be a problem. They ran it through the x-ray and said "this looks nothing like a problem".
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 15:04
@gregorschlump I went on a flight a few weeks ago and threw all my items in the tray and let them go through xray, no questions asked. on the way back I kept my pump on and got pulled aside and had everything searched, no questions asked took an extra minute or two. (they tend to get a bit scared when they ID an item attached to your hip that you did not make them aware of)
David
@davidkeddydb
Oct 04 2017 15:07
@scottleibrand Thanks. I will look at that in the logs and see what it tells me on the insulinReq
Meghan Rutledge
@mar5477
Oct 04 2017 15:08
I did a reinstall, but myopenaps is under bin now, that shouldn't be there..? Did I run the setup script from the wrong prompt?
But I'm looping, which is making me wonder if everything is correctly pointing to it
Eric
@ecc1
Oct 04 2017 15:10
having it under ~/bin is nonstandard, but shouldn't break anything. you must have changed to the bin directory somehow before running the setup
(I could be totally wrong -- maybe there are scripts that haven't run yet that will expect it under your home directory ?)
jaylagorio
@jaylagorio
Oct 04 2017 15:32
Yeah, I would recommend rerunning setup but in your home directory. That way if there's weirdness later you don't have something extra that you have to rule out
Meghan Rutledge
@mar5477
Oct 04 2017 15:36
Yeah, I went ahead and did that. There's too many things that reference myopenaps to leave it in bin
gregorschlump
@gregorschlump
Oct 04 2017 15:50
@diabetic_dude_twitter @andrew-warrington I never had problems with my pump in security areas. But I was wondering if they get a little more suspicious if you carry a small computer with you... Thanks for your answers
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 15:56
I just tell them it is a remote control for my pump. I actually have more issues with my pump than I do with my edisons and I carry two of them along with extra batteries and like minded items in my backpack
cables, meters, food, meds, needles.
Allison
@allisong4
Oct 04 2017 16:05
Hey there! Just started looping last night, but began wondering how the BG target is determined at any given time. The loop seems to be less aggressive than I was for myself prior to looping... I would prefer to be under 120 most of the time (except after meals with too many carbs and/or meals for which I haven't prebolused for). My actual target for fasting/between meals is really closer to 80-90. Do I need to set a constant temp target to get it to adjust down?
Glenn Primack
@primags
Oct 04 2017 16:07
@allisong4 I'm pretty new too. I adjusted some of the safety multipliers in the preferences file, and that seems to have made it more aggressive. Read the part in the manual about how edison will show if it is restricted by the safety limits and you can see how much it would've really temp basal'd for if it had no safety mechanisms.
@allisong4 Be careful though until you get used to it.
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 16:09
Glenn Primack
@primags
Oct 04 2017 16:10
@diabetic_dude_twitter No, not yet....Its on my list :)
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 16:10
if you set up the eating soon button it will set a high and low target of 80 when you are about to eat. works well for me
you also want to keep the target closer to 110-120 to give a bit of a cushion for you and allow openaps to try to suspend to correct your lows
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 16:16
@allisong4 I also would recommend watching the logs for like an hour or two then it will make more sense as to what is happening and why
I watch mine a lot
enact/smb-suggested.json: {"insulinReq":-0.26,"bg":89,"reservoir":"251.9","temp":"absolute","snoozeBG":175,"predBGs":{"COB":[89,88,88,88,88,89,90,92,93,95,97,99,101,103,105,107,109,111,112,114,116,117,118,120,121,122,123,125,126,127,127,128,129,130,131,132,132,133,134,134,135,135,136,136,137,137,138],"aCOB":[89,95,98,98,98,99,100,101,103,104,106,108,110,112,114,115,117,119,120,122,123,124,125,126,127,128,129,130,131,131,132,132,133,133,134,134,134,135,135,135,135,135,136],"UAM":[89,88,88,88,88,89,90,91,93,94,96,98,100,102,104,105,107,109,110,112,113,114,115,116,117,118,119,120,121,121,122,122,123,123,124,124,124,125,125,125,125,125,126],"IOB":[89,88,88,88,88,89,90,91,93,94,96,98,100,102,104,105,107,109,110,112,113,114,115,116,117,118,119,120,121,121,122,122,123,123,124,124,124,125,125,125,125,125,126]},"IOB":-0.698,"reason":"COB: 2, Dev: 6, BGI: 1.65, ISF: 73, Target: 124, minPredBG 105, minGuardBG 88, IOBpredBG 126, COBpredBG 138, UAMpredBG 126;  insulinReq -0.26. ; SMB bolus snooze: setting current basal of 0.975 as temp.  28m left and 0.975 ~ req 0.975U/hr: no temp required","COB":2,"eventualBG":138,"tick":-1,"deliverAt":"2017-10-04T16:14:54.135Z"}
No smb_enact needed. Temp refreshed: monitor/temp_basal.json: {"duration":28,"rate":0.975,"temp":"absolute"}
No bolus needed (yet). Settings less than 15 minutes old. Edison battery at 69% is charged (>= 98%) or likely charging (60-70%). Pumphistory-24 < 15m old
Completed supermicrobolus pump-loop at Wed Oct 4 11:14:57 CDT 2017:
Meghan Rutledge
@mar5477
Oct 04 2017 16:17
@primags @diabetic_dude_twitter @allisong4 something I just read in the docs was that SMB won't be enabled with temp targets if the temp target is above 100. Wondering why that would be since it removes the room for buffer.
Allison
@allisong4
Oct 04 2017 16:18
Thanks for the quick replies! I did notice that I was hitting the cap of my max_iob so that, along with the multipliers, may need adjusting. I have set up IFTTT buttons, love that! Is there a way I can explicitly set my target to 110-120? I can't figure out if it's some of the config vars in Heroku or if it's the range in the Nightscout profile. Watching my logs when I can, hopefully I can set some time aside when my sugars are trending a bit higher and see how it intends to nudge them back down.
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 16:18
exercise @mar5477
would be my first assumption
@allisong4 set that target on your pump
@allisong4 you want a bigger range to start then slowly start to close the gap
Meghan Rutledge
@mar5477
Oct 04 2017 16:21
@diabetic_dude_twitter do you think that the temp target shouldn't qualify the enableSMB then?
would it be better just to have an "enableSMB" for the next X minutes/hours and then a separate temp target (through IFTTT or otherwise)?
Glenn Primack
@primags
Oct 04 2017 16:29
I ended up getting my wireless config at work to work, thanks! I'd still like to know where the debug logs are though....
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 16:30
@mar5477 I am not sure on that one at all, I think that you are referring to an example listed here : http://openaps.readthedocs.io/en/latest/docs/Customize-Iterate/oref1.html?highlight=temp%20SMB
"For example, if your target is usually 100mg/dL, a temp target of 99 (or 80, the typical eating soon target) will enable SMB.
Conversely, a higher temp target (101 if your target is 100) will disable SMB. "
all that is really saying is that SMB will not work when it is not needed.
@primags if i knew I would tell you
Meghan Rutledge
@mar5477
Oct 04 2017 16:37
@diabetic_dude_twitter Right! But what if you were in the 300s and wanted to SMB to aggressively bring it down, but wanted to do so conservatively with a target of 150? I understand it wouldn't work with this, but why not? Just because a higher target implies we wouldn't want to quickly lower our BS to that?
Scott Leibrand
@scottleibrand
Oct 04 2017 16:39
@kcrcgm what version of oref0 are you using for autotune? 0.5.4 and prior have a high-ISF bias that is fixed in 0.6.0-dev.
@mar5477 you could set your regular target on your pump higher in that situation.
If you set a high temp target, we assume that means that there will be something (activity etc.) happening that might cause BG to drop faster than predicted, so the safest thing to do is to only high-temp, not SMB, so they can be canceled if BG starts to drop faster than expected.
Allison
@allisong4
Oct 04 2017 16:44
@diabetic_dude_twitter thanks! I did have my target set (albeit a low one) on the pump so I raised it a bit and we'll see how it goes. Will probably also increase max_iob and potentially the multipliers. Does SMB correct rising BGs more quickly than regular oref0? Just holding off on enabling SMBs per instructions to try out oref0 for 2 weeks first.
Scott Leibrand
@scottleibrand
Oct 04 2017 16:45
Yes, SMBs are more aggressive, because they can deliver the whole insulinReq in about 15m instead of taking 30m for the high temp to run to completion.
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 16:45
I would change things slowly so you know what to roll back if things don't work well for you
Scott Leibrand
@scottleibrand
Oct 04 2017 16:45
Agreed.
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 16:46
I normally change one thing, wait 3 days look at logs and then go forward from there
Allison
@allisong4
Oct 04 2017 16:46
That's what I thought on SMBs, thanks. Then I'll only increase max_iob, and wait on the multipliers. I'm not known for my patience, but I do appreciate the scientific process :-)
Brian Rabinovitz
@bdr1177
Oct 04 2017 16:48
Can someone explain whey there was no bolus snooze here preventing the low temp? It looks like the zero temp started almost immediately after the meal bolus. I understand that OpenAPS was predicting a low, but I would have thought that bolus snooze would have prevented the zero temp. (I believe this rig is running 0.5.3)
Capture.JPG
Scott Leibrand
@scottleibrand
Oct 04 2017 16:49
SnoozeBG was only 76. Is this running SMB?
SMB bolus snooze only kicks in if the eventualBG is above target.
Velibor Maric
@vebaba
Oct 04 2017 16:50

Hi all. I intend to bolus :) Yet, my Nightrider is still in customs and I try to set up edison to work. I'm following instructions to install OAPS https://openaps.readthedocs.io/en/latest/docs/Build%20Your%20Rig/OpenAPS-install.html# and I run into a problem after initializing installation oref0 script


Some packages could not be installed. This may mean that you have
requested an impossible situation or if you are using the unstable
distribution that some required packages have not yet been created
or been moved out of Incoming.
The following information may help to resolve the situation:

The following packages have unmet dependencies:
nodejs : Conflicts: nodejs-legacy
E: Unable to correct problems, you have held broken packages.


I searched and found one thread in this discussion from 2016 with the same issue but I couldn't manage to find solution in that discussion.
I'm not Linux literate and would appreciate a bit more info in case you try to explain me what to do.
Any help appreciated, thanks :)

Scott Leibrand
@scottleibrand
Oct 04 2017 16:51
Also, bolus snooze is probably going away. The 0.6.0-dev version has autotune fixes that should fix the CR issues that contribute to many of the mispredictions.
@vebaba it's probably easiest to reflash the Edison and follow the install instructions rather than trying to manually fix your nodejs install.
Velibor Maric
@vebaba
Oct 04 2017 16:54
Reflash and start over with oref0 instalation? Did I understand correctly?
garykidd
@garykidd
Oct 04 2017 16:55
I keep thinking MSB v s SMB....wondering if it might of crossed your mind to call them MSBs as oposed to SMB? I know it's a stupid comment.
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 16:55
I bet it fixes your issue a low faster @vebaba
oops low = lot
Brian Rabinovitz
@bdr1177
Oct 04 2017 16:55
@scottleibrand No, we're not running SMB because of too low basals. I understand that bolus snooze may go away. I'm not sure I understand the ins and outs of it, but I'm concerned about situations like this without bolus snooze. OpenAPS temped at zero right after the meal bolus. BG ended up going high shortly after this, even though we used ES and other settings are autotuned.
Velibor Maric
@vebaba
Oct 04 2017 16:57
OK, thanks :)
Dana Lewis
@danamlewis
Oct 04 2017 16:57
@andrew-warrington cob pill is known to be laggy in NS. We recommend not using it because it causes confusion. OpenAPS pill and or logs are source of truth.
Scott Leibrand
@scottleibrand
Oct 04 2017 16:58
@bdr1177 your problem is probably at least partly that 0.5.3 autotunes CR too high. 0.6.0 switches over to autotuning CR directly (not based on ISF) and fixes the high bias.
Brian Rabinovitz
@bdr1177
Oct 04 2017 17:00
@scottleibrand I have a test rig with the basal-vs-bolus 0.6.0 dev installed. Is the main 0.6.0 dev branch stable enough to recommend installing on our main rig?
David
@davidkeddydb
Oct 04 2017 17:11
@scottleibrand (You can hit me later for being lazy and not looking at the code) Does 0.6.0 dev branch autotune work with just using a single CR?
gregorschlump
@gregorschlump
Oct 04 2017 17:13
@danamlewis Awesome. There are posts about everything somewhere ;)
Dana Lewis
@danamlewis
Oct 04 2017 17:17
@davidkeddydb all Autotune versions are still one CR
@gregorschlump ha ;p
Scott Leibrand
@scottleibrand
Oct 04 2017 17:39
@bdr1177 if you haven't seen any issues on your dev rig and are comfortable updating code once a week or so, no reason you can't switch your main rig to dev as well (or switch over to using your dev rig as your main one)
Glenn Primack
@primags
Oct 04 2017 17:43
@diabetic_dude_twitter Just finished the IFTTT setup....awesome stuff! Thanks for the reminder and it got me finally working on it :)
Craig Brenner
@cbrenner
Oct 04 2017 18:36
I set up two basal profiles on the 722 (Standard and Pattern A). Pattern A is 20% higher than standard and is what we are using now because my son is sick. So, even though Patten A is currently active on the pump, it seems that the OpenAPS is still using the 'Standard' basal profile for a baseline. Oddly, when i look at the basal profile json file in settings, the pattern A profile is what is recorded.
Meghan Rutledge
@mar5477
Oct 04 2017 18:37
@cbrenner you probably need to change the basal profiles in your Nightscout profile?
Craig Brenner
@cbrenner
Oct 04 2017 18:38
@mar5477 NS is changed for reference but that doesn't impact OpenAPS so it shouldn't matter
Dana Lewis
@danamlewis
Oct 04 2017 19:08
@cbrenner are you using autotune?
Craig Brenner
@cbrenner
Oct 04 2017 19:09
@danamlewis Yes
Dana Lewis
@danamlewis
Oct 04 2017 19:20
so, autotune tuned off the basal running at midnight-ish. you may want to temporarily disable autotune, or re-run manually so it tunes off the current profile, if you change basal patterns like this mid-day. because openaps knows, but it's working from the autotune profile. does that help clarify?
(Trying to figure out a better way to visualize that flow, beyond words. Suggestions welcome.)
Craig Brenner
@cbrenner
Oct 04 2017 19:23
@danamlewis I changed the basal profile two days ago so Autotune has run twice at midnight. Looking at the Autotune recommendations file, it shows the pattern A profile as 'current' but still pulling the 'standard' profile
Dana Lewis
@danamlewis
Oct 04 2017 19:24
shows pattern A where, and is "still pulling standard" where?
Craig Brenner
@cbrenner
Oct 04 2017 19:26
@danamlewis shows pattern A as current in the latest 'autotune_recommendations.log'. OpenAPS pill shows it is using the current rate (which are the rates in the standard profile not pattern A)
Dana Lewis
@danamlewis
Oct 04 2017 19:27
is that current rate matched to what autotune's recommendation? Can you maybe paste your autotune log, and your pill?
Tore Bjørndalen
@BjoTore_twitter
Oct 04 2017 19:28
Where can I find a current description on how to interpret the content of the OpenAps pill?
Dana Lewis
@danamlewis
Oct 04 2017 19:29
@BjoTore_twitter this is the long version of determine-basal, but it's missing a few things from recent versions: http://openaps.readthedocs.io/en/latest/docs/While%20You%20Wait%20For%20Gear/Understand-determine-basal.html (so, please feel free to paste a pill and ask questions, and would love someone's help getting that page update, perhaps with pictures showing logs in log form as well as pill form to better help understand these)
Glenn Primack
@primags
Oct 04 2017 19:30
@danamlewis Just wanted to mention the documentation for this project is nothing short of amazing! I know its quite a chore to keep up with...
Dana Lewis
@danamlewis
Oct 04 2017 19:33
@primags heh, thanks :) it always needs more work, and it's never perfect, so appreciate hearing when it's helpful (as well as when it's not)! :)
David
@davidkeddydb
Oct 04 2017 19:35
@danamlewis @scottleibrand I don't know whether to open an issue or maybe one is already open, but I installed 0.6.- dev no git following the instructions, it was working fine until today and now I am getting a loose object git fail which I thought was impossible now that I am no git?
Sparkfun base block with Edison and TI stick
Tore Bjørndalen
@BjoTore_twitter
Oct 04 2017 19:39
image.png
dan lightner
@danimaniac
Oct 04 2017 19:40
River Bluff High School in Lexington, SC is evacuated due to a beeping OmniPod. I always figured my OpenAPS rig would be problematic at the airport...poor kid. Sorry, too good not to share.
Dana Lewis
@danamlewis
Oct 04 2017 19:40
@danimaniac i saw that. (and for anyone who's worried about airports, don't. normal electronics in bag are fine. it's just like a cell phone or other mini computer.
Tore Bjørndalen
@BjoTore_twitter
Oct 04 2017 19:40
@danamlewis what's the MinGuardBG?
Craig Brenner
@cbrenner
Oct 04 2017 19:41
@danamlewis Yes, current rate is matched to autotune. See pictures.
OpenAPS Pill.jpg
Autotune.jpg
Dana Lewis
@danamlewis
Oct 04 2017 19:43
@BjoTore_twitter heh, exhibit A of how that page is about to be even more behind. It's something we added in dev as a safety threshold, to make sure that SMB's aren't delivered when you're low or predictied to be low in the short term
@cbrenner ok: so it's setting the neutral temp based on Autotune, which is as expected. (autotune only makes small changes each day, so if your pattern is very different, it'll likely take a few days to catch up). If you want the current pattern to be baseline/what it temps from, I'd disable autotune temporarily from running (remove from enable line and bash runagain).
Also, are you running autosens?
Craig Brenner
@cbrenner
Oct 04 2017 19:44
@danamlewis yes, running autosens as well
Dana Lewis
@danamlewis
Oct 04 2017 19:45
k good. that will help with resistance from the sickness, too, but if you're that confident in switching patterns, I'd do the above and run without autotune for the few days while sick.
Tore Bjørndalen
@BjoTore_twitter
Oct 04 2017 19:46
My target is 5.3 and what indicates a risk of going low?
Dana Lewis
@danamlewis
Oct 04 2017 19:46
(and if you disable autotune and do bash runagain, remember to re-set your preferences afterward)
Tore Bjørndalen
@BjoTore_twitter
Oct 04 2017 19:48
The current IOB?
Dana Lewis
@danamlewis
Oct 04 2017 19:48
@BjoTore_twitter it's showing eventualBG as above target; so it's setting a low temp and miccrobolusing to get you to the insulinReq needed
it's not saying you're going to go low; just showing those variables and because they're above target, those aren't cutting off insulin; it's starting to SMB you
Craig Brenner
@cbrenner
Oct 04 2017 19:48
@danamlewis Ok, so I will disable autotune in the prefs. When you say bash runagain, do you mean reboot the system or what am i running again as autotune will be disabled?
Tore Bjørndalen
@BjoTore_twitter
Oct 04 2017 19:50
image.png
Ok, and indeed those SMB's have finally arrived
Craig Brenner
@cbrenner
Oct 04 2017 19:51
oh wait, sorry, you mean to re run my setup script...got it
editing the runagain file to remove autotune will be faster than doing oref0-setup.sh via the interactive script
Craig Brenner
@cbrenner
Oct 04 2017 19:51
@danamlewis oh wait, sorry, you mean to re run my setup script...got it
Dana Lewis
@danamlewis
Oct 04 2017 19:52
@BjoTore_twitter yea, if you look at the last line from the first pill you pasted, it said "Microbolusing 0.2u"
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 20:01
can someone please do an ls in myopenaps so I can see what needs to be in there? I was moving my autotune logs and accidently put them in the root of myopenaps
scroll down on this page for it
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 20:02
thnx
@danamlewis what about profile.json and profile.pump.json ?
Dana Lewis
@danamlewis
Oct 04 2017 20:06
if it's a setting, it's under settings
Screen Shot 2017-10-04 at 1.06.48 PM.png
(keep scrolling to see all the ls examples on that page if you haven't
if you moved a folder or a file, it should be fairly easily to move back the same way you moved it
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 20:10
yup but I did this mv autotune/* -v ~/myopenaps/ like a true champ
so if there was a settings folder in autotune.... well ...
I need to re make it
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 20:16
wow I did something just freaking awesome :worried: looking at my other rig in awe
Dana Lewis
@danamlewis
Oct 04 2017 20:17
(don't fret, you can always copy over your autotune file from the RigOk to RigNotOk)
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 20:18
I am not worried but thanks I was not paying attention to the settings folder at all but for some reason it seems to be gone on one of my rigs
who needs a settings folder anyway??? Just wing it right? profile.json??? profiling is just wrong
Andrew Warrington
@andrew-warrington
Oct 04 2017 20:25
Question -- do any of the orefx implementations include gain scheduling? i.e. as the glucose level gets higher, the ISF changes?
Dana Lewis
@danamlewis
Oct 04 2017 20:25
no
we observe/handle resistance (perceived or whatever factors causing) or sensitivity changes w/ autosens
Andrew Warrington
@andrew-warrington
Oct 04 2017 20:26
got it.
thank you!
Dana Lewis
@danamlewis
Oct 04 2017 20:27
:+1:
jaylagorio
@jaylagorio
Oct 04 2017 20:51
@danamlewis When an SMB happens can the SGV at the time be entered into that record? Because that would be good to know for when I can't hover over the entry after all the boluses are on the display. Should I open an Issue? I'm not confident I could make the mod myself.
Scott Leibrand
@scottleibrand
Oct 04 2017 20:54
@jaylagorio SMBs are read back from the pumphistory and uploaded to NS. I don't know if there's a way to provide the current BG to the pump when issuing a bolus. If not, then doing so after the fact when reading pumphistory and uploading it to NS would be... complicated.
probably better to fix the mouse-over tooltips on NS to have a smaller "target area" around each bolus
jaylagorio
@jaylagorio
Oct 04 2017 20:57
Oh right, you mentioned that. I did put that issue in for Nightscout and was hoping to get a quicker win.
Well, someone else apparently put that issue in, I commented on it hoping for some visibility.
Scott Leibrand
@scottleibrand
Oct 04 2017 20:59
I wonder if that's the kind of tweak one of us could do ourselves to the NS c-r-m code
Scott Leibrand
@scottleibrand
Oct 04 2017 21:14
Samuchco
@Samuchco
Oct 04 2017 21:20
I'm trying to set IFTTT. I'm running in a few issues:
1- The instructions say to search "Maker Webhooks". There is no such thing
There is Webhooks
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 21:22
that is what you use
the docs need to be updated to reflect the new name on ifttt
Samuchco
@Samuchco
Oct 04 2017 21:23
Then there must be an issue with the heroku settings, since the Key is MAKER_KEY. not WEBHOOKS_KEY
and in enable, it should be Webhooks......am I right?
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 21:24
hold on a sec
are you doing this for ... your phone ?
using the button ?
or alexa?
what happens when you test it on the IFTTT site ?
you should be able to click on my applets, choose the one you made, click the gear in the upper right hand corner then click view activity log
there you should get an idea of what is wrong.
Samuchco
@Samuchco
Oct 04 2017 21:28
I'm trying to set it up for phone and website
Even though the button is active? I don't see where tp press (on the website)
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 21:29
gimme one sec
if anyone could tell me how to remove a carb announcement from NS that would be really cool
Dana Lewis
@danamlewis
Oct 04 2017 21:32
@diabetic_dude_twitter pencil button in NS to edit and delete treatments
Samuchco
@Samuchco
Oct 04 2017 21:33
the activity tells me "Applet failed" When I open "show Details" I get Google coordinates and at the bottom " ShareURL { } "
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 21:33
@danamlewis thank you
Samuchco
@Samuchco
Oct 04 2017 21:34
@diabetic_dude_twitter another way to delete a treatment is via Report --> Treatments and x out the treatment you want to delete
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 21:34
did you click on .... add ingredient at the bottom of your recipe? You should not
Samuchco
@Samuchco
Oct 04 2017 21:35
no I didn't
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 21:36
what button are you trying to make?
@Samuchco normally the hashed secret is wrong or.... the { at the beginning is missing or the } at the end
{"enteredBy": "IFTTT-button", "reason": "low treatment", "carbs": 10, "secret": "1000100101001101001001010101010"}
Samuchco
@Samuchco
Oct 04 2017 21:43
all settings on my IFTTT are correct and rechecked. hashed srecret copied and pasted. All { } are in place. I wonder if the heroku settings are wrong since they are named MAKER, but MAKER is no longer the correct name
@diabetic_dude_twitter Is your IFTTT working properly?
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 21:43
give it 5 minutes
I am also going to check my heroku site
yes that is why i needed to remove the treatment, I announced a bug meal to test and god love it... it worked but I have no food
excuse me a big meal
Samuchco
@Samuchco
Oct 04 2017 21:45
where you successful in removing the test via the "pencil" icon?
were*
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 21:46
yes thank god
@Samuchco this is what my enable looks like "bridge loop pump iob basal careportal iage sage rawbg openaps maker alexa pushover"
maker works for me
Samuchco
@Samuchco
Oct 04 2017 21:47
I don't have alexa or pushover
are you using your phone or website to send the temp targets
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 21:48
I have an alexa at home that I set up so when I wake up I can tell her how to manage my settings
Alexa and my phone
Samuchco
@Samuchco
Oct 04 2017 21:48
have you tried from the website?
iPhone or Android
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 21:50
tested two from the web
did not get notifications for either and ... NS has not shown the entry yet
(I normally test from my phone)
trying from my watch now
showed up
I use an iphone
so works from my phone or watch but not the website
Samuchco
@Samuchco
Oct 04 2017 21:56
Thanks!!! from my phone, Android, I tried and got "Applet Failed" and then "Action Skipped"
Diabetic_Dude
@diabetic_dude_twitter
Oct 04 2017 21:57
what button are you making ?
a custom one ? or one there is a template for?
Samuchco
@Samuchco
Oct 04 2017 21:59
the one on the site
I am going to recheck the settings
Samuchco
@Samuchco
Oct 04 2017 22:12
@diabetic_dude_twitter got it to work from the phone. I deleted the Eating Soon applet and re-did it. Thanks!!!!
Paul Dickens
@thebookins
Oct 04 2017 23:39
Hoping someone can help with offline looping and xDripAPS. We are using an iPhone and custom Loop app to send glucose values to the xDripAPS microservice. Yesterday we went caving and noticed that looping stopped while underground. It seems that Bluetooth tethering is lost when the phone doesn't have internet. The xDripAPS local endpoint (172.20.10.x:5000/api/v1/entries) wasn't available. Is there any way to make the tethering persistent?
Dana Lewis
@danamlewis
Oct 04 2017 23:39
I think you found what everyone else has found - that route doesn't work when you don't have any connectivity :(
Paul Dickens
@thebookins
Oct 04 2017 23:42
Mmmm. But that's the whole point, isn't it? We need that route when we don't have connectivity! Might have to double down on the direct Edison G5 connection.
Or make the Edison a BLE peripheral with a protocol we define.
Dana Lewis
@danamlewis
Oct 04 2017 23:45
+1 for doubling down on G5 ;) @scottleibrand and I are also working on a rig that will spin up a hotspot that the phone can connect to, so you can read from it and know what's going on (and ideally send temp targets to it, but that's phase 2)
Paul Dickens
@thebookins
Oct 04 2017 23:48
Great idea! Is that project public yet? I'll need a way to send calibrations and sensor start/stop events to the rig, was hoping to use the Bluetooth tethered address, but will need another way after yesterday's experience.
Scott Leibrand
@scottleibrand
Oct 04 2017 23:54
there's an oref0 branch with the work we've done so far, but no documentation so far
we have the oref0-online parts working, but are just getting started on ways to actually use the local-only hotspot connection (beyond ssh'ing into the rig)
Paul Dickens
@thebookins
Oct 04 2017 23:56
Cool, I'll check it out.