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

14th
May 2017
rsilvers129
@rsilvers129
May 14 2017 00:09
Oh. For just a short time. Isn't that just how OpenAps brings down a high? I don't think that means he needs a 5 u/hr rate for any specific basal setting.
Scott Leibrand
@scottleibrand
May 14 2017 01:17
@rsilvers129 her point was they had to override the oref0 maxSafeBasal setting for that time period to allow OpenAPS to high temp enough to handle that rise (when it did happen).
rsilvers129
@rsilvers129
May 14 2017 01:17
Yeah. That makes sense.
tepidjuice
@tepidjuice
May 14 2017 02:01
even with "remainingCarbFraction": 1 openAPS is still 0 temping me after bolusing causing delayed highs.
Scott Leibrand
@scottleibrand
May 14 2017 02:12
Can you paste a NS screenshot from when it was zero temping and you think it shouldn't have been?
Bob Duerr
@Duer0049
May 14 2017 02:12

I am lost, here is where I am, I can't get the openaps rig to communicate with the pump, I checked that the pump is the correct type with the correct firmware, but it continues to say no pump comms detected.

Here is the code:

Starting pump-loop at Sat May 13 21:02:03 CDT 2017:
Radio ok, Listening: .No pump comms detected from other rigs
Profile less than 60m old
temp_basal.json more than 5m old
Temp refreshError, retrying
Starting pump-loop at Sat May 13 21:02:52 CDT 2017:
Radio ok, Listening: .No pump comms detected from other rigs
Profile less than 60m old
temp_basal.json more than 5m old
Temp refreshed
Could not parse input data: { [Error: Cannot find module '/root/myopenaps/monitor/glucose.json'] code: 'MODULE_NOT_FOUND' }
determine-basal://text/shell/enact/suggested.json
reporting enact/suggested.json
cat: enact/enacted.json: No such file or directory
pumphistory more than 5m old: Refreshed pumphistory
Could not parse input data: { [Error: Cannot find module '/root/myopenaps/monitor/glucose.json'] code: 'MODULE_NOT_FOUND' }
determine-basal://text/shell/enact/suggested.json
reporting enact/suggested.json
cat: enact/enacted.json: No such file or directory
Settings less than 10m old
Pumphistory-24 less than 20m old
Completed pump-loop at Sat May 13 21:04:26 CDT 2017

tepidjuice
@tepidjuice
May 14 2017 02:16
@scottleibrand I'm unable to take screenshots sorry.
Helene Brashear
@sillystring13
May 14 2017 02:18
@Duer0049 I'm a newbie myself, but it looks like you've got an error on loading a module - if you look in the directory /root/myopenaps/monitor/ can you see glucose.json? If not, I think that not all the files got installed correctly
Dana Lewis
@danamlewis
May 14 2017 02:33
Yup @sillystring13 , @Duer0049 what kind of CGM are you using to get CGM data, and how are you expecting to get that data to the rig?
katie disimone
@Kdisimone
May 14 2017 02:33
@tepidjuice if you're on Windows, use Snipping Tool. If you're in mac, use an app called Skitch that you can download for free.
Dana Lewis
@danamlewis
May 14 2017 02:33
(Also hi @sillystring13 :smile: )
Bob Duerr
@Duer0049
May 14 2017 02:35
@danamlewis I am using a Dexcom G4, with the Dexcom with Share receiver.
Dana Lewis
@danamlewis
May 14 2017 02:37
Ok. Is it plugged in or are you expecting to pull bgs from Nightscout?
Bob Duerr
@Duer0049
May 14 2017 02:38
I would like to pull bgs from Nightscout.
Dana Lewis
@danamlewis
May 14 2017 02:39
K. Check your NS URL and API are right, then, (assuming your NS successfully has your bgs already)
Bob Duerr
@Duer0049
May 14 2017 02:40
ok, thanks.
Scott Leibrand
@scottleibrand
May 14 2017 02:40
@Duer0049 and re: the original message, it says "from other rigs": that message is a good thing. It means you don't have any other rigs interfering, so it's safe to continue talking to the pump.
Helene Brashear
@sillystring13
May 14 2017 02:42
@danamlewis We're doing it! We've been looping for ~ 2 hours!
tepidjuice
@tepidjuice
May 14 2017 02:44
@Kdisimone on gentoo.
output.pdf
@scottleibrand that was harder than it needed to be but I managed to get a screen shot
output.pdf
Scott Leibrand
@scottleibrand
May 14 2017 02:47
Any chance you could scrub back to the time of the zero temping and catch the purple prediction lines from then?
tepidjuice
@tepidjuice
May 14 2017 02:49
output.pdf
Scott Leibrand
@scottleibrand
May 14 2017 02:49
It looks like it started SMBing just before you did your second bolus...
tepidjuice
@tepidjuice
May 14 2017 02:49
output.pdf
Overall I think it's taking insulin away when it's needed and then trying to catch up after it sees the bgl start to rise.
Scott Leibrand
@scottleibrand
May 14 2017 02:52
Yeah, it's doing that for safety.
My current thinking is to bring back bolus snooze only when the eventual COBpredBG is above target.
tepidjuice
@tepidjuice
May 14 2017 02:52
It's causing me a lot more post-prandial hyperglycaemia than I'm used to.
Helene Brashear
@sillystring13
May 14 2017 02:55
@tepidjuice don't know if it helps, but gimp will get you an easy screenshot in jpg format on linux boxes
tepidjuice
@tepidjuice
May 14 2017 02:56
@sillystring13 thanks. I ended up doing it from the command line.
Helene Brashear
@sillystring13
May 14 2017 02:57
Now that we are looping - should I do any BG corrections? little dude is running a little high
tepidjuice
@tepidjuice
May 14 2017 02:58
@sillystring13 I still correct all the time. Not sure what others do. I rely on openAPS to keep me in range not to get me there.
@scottleibrand I'm happy to test out the bolus snooze when COBpredBG and report back.
Helene Brashear
@sillystring13
May 14 2017 03:01
@tepidjuice thanks! the docs are bit fuzzy on that part, so i wasn't sure where i should be active and where i should be patient
Scott Leibrand
@scottleibrand
May 14 2017 03:01
The reason I don't want to bring back full bolus snooze is because too many people do manual corrections when SMB or high temps have already taken care of the high, so bolus snooze would allow the overcorrection to cause a low.
Am traveling, so haven't gotten to coding it yet, but I think I know what I want it to do.
tepidjuice
@tepidjuice
May 14 2017 03:03
Could overcorrecting be solved by having a correction option through NS or openAPS?
katie disimone
@Kdisimone
May 14 2017 03:04
@scottleibrand I've been using it with bolus booze enabled and it's been much improved to prevent the later rebounds.
tepidjuice
@tepidjuice
May 14 2017 03:04
@sillystring13 that's just me. You might want to see what others do as well.
Dana Lewis
@danamlewis
May 14 2017 03:04
@sillystring13 have you change your maxiob above 0 yet?
Scott Leibrand
@scottleibrand
May 14 2017 03:04
@Kdisimone how did you enable that?
Dana Lewis
@danamlewis
May 14 2017 03:04
Also :confetti_ball: for looping :smile:
OpenAPS will correct you to target @tepidjuice . But how much varies on how much maxiob you decide to allow it
If you do too much manual correction OpenAPS can try to low temp but you may still end up w a low
Helene Brashear
@sillystring13
May 14 2017 03:06
@danamlewis i was trying to figure that out - docs say you might want to change it, but i'm not sure how
Dana Lewis
@danamlewis
May 14 2017 03:06
I recommend setting a temporary lower target for an hour if you want to give it a nudge, vs manually bolusing
@sillystring13 nano preferences.json to change it
If it's zero, it'll just low temp and high temp until netIoB is back to zero. If a bit high, you'll need maxiob above zero for it to high temp down to target
Scott Leibrand
@scottleibrand
May 14 2017 03:07
@Kdisimone your blog post was actually what gave me the idea to only bolus snooze when there's enough COB to get eventual COBpredBG above target, to help prevent lows after correction boluses late in the absorption.
katie disimone
@Kdisimone
May 14 2017 03:07
@scottleibrand and @danamlewis used the new aliases today too!' So nice. Thanks!
tepidjuice
@tepidjuice
May 14 2017 03:08
@danamlewis I know but I'm too impatient so I prefer the manual bolus method.
Scott Leibrand
@scottleibrand
May 14 2017 03:08
Do you ever go low from corrections?
Dana Lewis
@danamlewis
May 14 2017 03:08
Fair enough. I just wouldn't recommend that to others, especially new loopers.
tepidjuice
@tepidjuice
May 14 2017 03:09
@scottleibrand No. Although I am usually quite mindful of my iob and future exercise/food.
Helene Brashear
@sillystring13
May 14 2017 03:10
@danamlewis thanks! can i change the json file without restarting things?
Dana Lewis
@danamlewis
May 14 2017 03:10
@sillystring13 yup!
Helene Brashear
@sillystring13
May 14 2017 03:11
@danamlewis groovy!
Scott Leibrand
@scottleibrand
May 14 2017 03:13
@tepidjuice cool. I've noticed it's super easy to bolus more than necessary when on the go or otherwise distracted, so we've been attempting to eliminate the need for correction boluses in favor of low temp targets.
Dana Lewis
@danamlewis
May 14 2017 03:13
Happy first night of looping! :smile:
Helene Brashear
@sillystring13
May 14 2017 03:15
Thanks! I just changed the file and watching all the math propagate was very cool.
tepidjuice
@tepidjuice
May 14 2017 03:20
If I raised max_iob, max_daily_safety_multiplier and current_basal_safety_multiplier it'd eliminate the need for correction boluses but I like having a minimally invasive loop.
JELCRAWFORD
@JELCRAWFORD
May 14 2017 03:45
@sillystring13 - we often use the temporary target to bump up or down. We have, inadvertently, caused DS to go low from correcting without considering IOB. He's also caused this trouble himself at school when using the bolus wizard and correcting bg with a meal that had already been corrected by OpenAPs, but hadn't fully taken effect yet. I think @kdsimone, for this reason, has her daughter enter only carbs into the pump bolus wizard - thenless chace of mistake (pls correct if I misunderstood @kdsimone). Most often, we only correct if we are pretty sure we made a carb counting error and we catch it before Openaps does a lot to fix it. Probably other solutions that work too though. Just our method!
@tepidjuice - husband has created a cron job for day and one for evening. Day is more conservative as DS is not in our care. Night, up till when we usually go to bed, more aggressive when we're watching.
JELCRAWFORD
@JELCRAWFORD
May 14 2017 03:50
The power to correct bg by two rather when DS is away from us via fttt is pretty satisfying :)
BG by temp target
Helene Brashear
@sillystring13
May 14 2017 03:57
Thanks everybody - this discussion is really informative :)
katie disimone
@Kdisimone
May 14 2017 04:54
@scottleibrand is there a part of oref0 that decays carbs without seeing a BG change? Like if daughter enters carbs and boluses while BGs are not coming into NS (sensor restart)....it seems like all those (or a lot of them) are sitting undecayed in OpenAPS now that she's come back online.
She's been offline for nearly 8 hours...so more than the usual 2 hr restart.
Dana Lewis
@danamlewis
May 14 2017 04:55
Yes there's a minimum carb absorption
katie disimone
@Kdisimone
May 14 2017 04:56
πŸ‘πŸ»
Scott Leibrand
@scottleibrand
May 14 2017 04:57
But by default it's not very high. You might want to raise it. min_5min_carbimpact
If you're on dev it only uses that for decaying old carbs, so set it as high as you like. We've bumped it up to 6 so far.
You could just keep incrementing it by 1 and then waiting for the next COB calculation to see if it's lowered it enough yet.
Or you could do the conversion from g/hr
Scott Leibrand
@scottleibrand
May 14 2017 05:03
@tepidjuice sounds like maybe you don't really want oref1 then?
Kevin Marshall
@ruess
May 14 2017 06:27
Anyone know of any good tools or methods for making a backup of the edison rig?
Just in case it’s lost and needs to be rebuilt?
might save a bit of time
elodaille01
@elodaille01
May 14 2017 08:12
Hi everyone ! What is the benefit of IFTTT vs Care portal to set a temporary target ? thank you ;)
Lorenzo Sandini
@lsandini
May 14 2017 10:03
After a long night at work, and half-asleep, I flashed a new edison and started configuring it. All went well, I get a prompt when connected with UART and can login as root or edison normally. However when I try to login using wifi, the passwords are refused. I get a prompt though, so I know it' s connected to my network...
Lorenzo Sandini
@lsandini
May 14 2017 10:22
fixed, I had an IP address conflict in my home router, duh
dan lightner
@danimaniac
May 14 2017 11:56
FYI - I can confirm the Samsung Galaxy J3 Emerge on Virgin Mobile bt tethers consistently well. Now I just have to configure my rig to switch properly between bt and Wi-Fi and for OpenAPS/xdrip+ uploads to Ns...Search icon, here I come.
francesc0-cgm
@francesc0-cgm
May 14 2017 12:14
Any news on g5 to openaps?
dan lightner
@danimaniac
May 14 2017 12:23
...sorry, I meant xDripAPS/xDrip+ uploads to NS.
rsilvers129
@rsilvers129
May 14 2017 12:38
@ruess I would probably back up with tar, and then move the archive off the Edison with scp. Or maybe set up an rsync to periodically copy to another system.
katie disimone
@Kdisimone
May 14 2017 13:25
@ruess I just made a spare rig by following the same procedures as making the original. The only difference is to give it a new edisonhost name, different than your main rig. If you make the preferences the same as your other rig, it will be seamless if you need to swap. However, if you then also buy another explorer board, you can just keep it plugged into a wall socket, without a lipo, in your house to provide backup coverage or extended range.
Lorenzo Sandini
@lsandini
May 14 2017 13:29
still struggling to get SMB activated. New rig, fresh master install, re-run the setup with only microbolus enabled. No mention of SMB in preferences.json though... and the logs don't mention "starting supermicrobolus pump-loop" as it said in the first attempt in dev. I must be dumb, but I read #424...
I understand I don't have to switch to dev, this functions have been merged from dev to master already, right ?
tepidjuice
@tepidjuice
May 14 2017 13:37
@lsandini I think SMB is only dev at the moment.
Lorenzo Sandini
@lsandini
May 14 2017 13:37
aah ok
Tim Mecklem
@tmecklem
May 14 2017 14:18
Recently had the Edison connector break its solder joints on the explorer board. Bridging the joints and wicking the excess off fixed it. If you experience the same thing (it's a weak point on the board) there's hope.
Dana Lewis
@danamlewis
May 14 2017 15:19
@francesc0-cgm not sure what you mean by news; G5 has always worked with OpenAPS
francesc0-cgm
@francesc0-cgm
May 14 2017 15:20
Yes i know. I heard someone is working on a direct connection between g5 and openaps without a phone
Kate Farnsworth
@ELUTE
May 14 2017 15:33
@Duer0049 did you get your issue sorted? Which pump are you using?
francesc0-cgm
@francesc0-cgm
May 14 2017 15:40
Thank you @Dana is the project on github to take a look to progress?
Scott Leibrand
@scottleibrand
May 14 2017 15:50
@ruess I wouldn't bother backing everything up. Just make sure your oref0-runagain.sh is up to date, and then copy and paste that somewhere. If you've modified your preferences.json manually and don't want to update oref0-runagain with the changes, you can save a copy of that too.
Lorenzo Sandini
@lsandini
May 14 2017 16:25
dev doesn't work for me at all. Basic edison/explorer rig with 722WW pulling CGM from NS. Works fine on master. When I switch to dev and re-run the install script, I get nothing but trouble. Even with no flags enabled. Comms with the pump fail, seems to be stuck without being able to complete a single loop. :(
Dana Lewis
@danamlewis
May 14 2017 16:28
Did you do nom run global install?
Lorenzo Sandini
@lsandini
May 14 2017 16:30
yes I did, install runs fine, warns about some outdated versions of various packages though
Tim Street
@tim2000s
May 14 2017 17:26
@lsandini - we know it works with WW pumps as a few of us are using it. Can you post your logs showing where the issue is?
Lorenzo Sandini
@lsandini
May 14 2017 17:28
@tim2000s I think my first SMB loop ever has just completed. It was stuck just after mmtune, waiting for 30s silence...
I just let it sit there for almost 30 min, and then it started working.
wharper11
@wharper11
May 14 2017 18:07
hi. switching over from azure to heroku and having problems. is there a gitter site specific for these types of questions? or ask here?
Peter Miller
@ochenmiller
May 14 2017 18:08
@LadyViktoria , thanks for the advice on BT connected glucometers! Am successful with gatttool reading from the Bayer Contour One!
Tim Street
@tim2000s
May 14 2017 18:10
:+1:
Peter Miller
@ochenmiller
May 14 2017 18:12
$ gatttool -I [ ][LE]> connect CC:78:AB:24:77:A7 Attempting to connect to CC:78:AB:24:77:A7 Connection successful [CC:78:AB:24:77:A7][LE]> #set security [CC:78:AB:24:77:A7][LE]> sec-level medium [CC:78:AB:24:77:A7][LE]> #set notify on 2a18 GLUCOSE_MEASUREMENT_SERVICE_UUID [CC:78:AB:24:77:A7][LE]> char-write-req 0x000c 0100 [CC:78:AB:24:77:A7][LE]> #set notify on 2a34 [CC:78:AB:24:77:A7][LE]> char-write-req 0x000f 0100 [CC:78:AB:24:77:A7][LE]> #set indicate on 2a52 RECORD_ACCESS_CONTROL_UUID not necessary [CC:78:AB:24:77:A7][LE]> char-write-req 0x0014 0200 [CC:78:AB:24:77:A7][LE]> #call for all records [CC:78:AB:24:77:A7][LE]> char-write-cmd 0x0013 0101 Characteristic value was written successfully Characteristic value was written successfully Characteristic value was written successfully Notification handle = 0x000b value: 13 00 00 e1 07 04 15 16 30 1f 10 ff 61 b0 f1 Notification handle = 0x000e value: 02 00 00 01 Notification handle = 0x000b value: 03 01 00 e1 07 04 15 16 32 12 10 ff 5e b0 f1 Notification handle = 0x000b value: 03 02 00 e1 07 04 15 17 1a 18 10 ff 59 b0 f1 Notification handle = 0x000b value: 03 03 00 e1 07 04 15 17 35 2b 10 ff 4d b0 f1 Notification handle = 0x000b value: 03 04 00 e1 07 04 15 17 38 33 10 ff 54 b0 f1 Notification handle = 0x000b value: 03 05 00 e1 07 04 16 00 23 21 10 ff 77 b0 fa Notification handle = 0x000b value: 03 06 00 e1 07 04 16 01 39 00 10 ff ec b0 f1 Notification handle = 0x000b value: 03 07 00 e1 07 04 16 01 3a 13 10 ff cc b0 f1 Notification handle = 0x000b value: 03 08 00 e1 07 04 16 03 2a 20 10 ff 95 b0 f1 Notification handle = 0x000b value: 03 09 00 e1 07 04 17 03 07 34 10 ff d7 b0 f1 Notification handle = 0x000b value: 03 0a 00 e1 07 04 1b 03 1c 0d 10 ff ba b0 f1 Notification handle = 0x000b value: 03 0b 00 e1 07 05 0e 02 19 08 10 ff 95 b0 f1 Notification handle = 0x000b value: 03 0c 00 e1 07 05 0e 02 37 0d 10 ff 99 b0 f1 Notification handle = 0x000b value: 03 0d 00 e1 07 05 0e 10 21 15 10 ff 81 b0 f1 Notification handle = 0x000b value: 03 0e 00 e1 07 05 0e 10 26 14 10 ff 81 b0 f1 Notification handle = 0x000b value: 03 0f 00 e1 07 05 0e 10 3b 18 10 ff 8f b0 f1 Indication handle = 0x0013 value: 06 00 01 01 [CC:78:AB:24:77:A7][LE]> exit
third set of values from the end are glucose readings.
katie disimone
@Kdisimone
May 14 2017 18:13
@wharper11 ask here or message me if it's just an extension of our previous convo too.
Now to parse the records
All this is to complete the fully self contained offline system of glucometer, dexcom g4, wixel, explorer, edison, pebble
oh, and medtronic 722 of course
wharper11
@wharper11
May 14 2017 18:28
@Kdisimone thanks, I messaged you.
cameronrenwick
@cameronrenwick
May 14 2017 18:48
when I use bolus wizard and enter current BG, carbs and amount of bolus it shows up in the NS pills as expected. It doesn't show up on my BG green line. If I bolus without the wizard it does. I'm ok with this but as I intend to start SMB in the next little while I'm wondering if I should be also entering values in the care portal? looking for advice here...
Dana Lewis
@danamlewis
May 14 2017 18:56
No need to double enter
cameronrenwick
@cameronrenwick
May 14 2017 18:59
perfect thx..
Helene Brashear
@sillystring13
May 14 2017 19:58
should ns-dedupe-treatments be getting called automatically somewhere? I'm getting lots of dupes in the ns data - I think that every time I switch networks, it's uploading the data again
rsilvers129
@rsilvers129
May 14 2017 20:05
18485727_10156113971948662_3834048266354982655_n.jpg
I used printer filament to rivet it together instead of screws.
cameronrenwick
@cameronrenwick
May 14 2017 20:56
is that "take-apart-able" then?
PopcornPam
@PopcornPam_twitter
May 14 2017 21:26
possibly dumb question: I can't seem to get the rig to be powered when plugged in through usb, but not connected to battery. I tried both ports. What am I missing?
@rsilvers129 ... a thought for your cases... maybe make a small loop or hole that you could loop a keychain, string, something if you wanted to clip it on something or attach something to it (in my case a gps tracker)
Dana Lewis
@danamlewis
May 14 2017 22:11
Pam - bad cable?
rsilvers129
@rsilvers129
May 14 2017 22:17
I built a second rig today and cable issues caused me some delays. The other that confused me was myopenaps directory ended up in /root instead of /root/src. Not sure how that happens.
PopcornPam
@PopcornPam_twitter
May 14 2017 22:40
@danamlewis , I have tried a few different times in different plugs with different cables. Is there a particular port that works better than the other?
rsilvers129
@rsilvers129
May 14 2017 22:42
Some PCs don't have powered USB ports.
Dana Lewis
@danamlewis
May 14 2017 22:42
Shouldn't matter - might be something else going on. What power source are you using?
PopcornPam
@PopcornPam_twitter
May 14 2017 22:45
powerstrips that have microusb plugs
Also, if I want to wipe and edison and reflash it, how do I wipe it?
cameronrenwick
@cameronrenwick
May 14 2017 22:46
the one thing you do want to confirm is that your micro usb is in fact a data cable. Some are only charge cables (and thus only have 2 wires inside despite having the same micro and USB (on the other end) connector. Easiest way to confirm you have a data cab;e @PopcornPam_twitter is to connect your phone with your computer and if it comes up as a component that you can transfer files to it is a data cable. I've had mixed data and charge cables that have lead me to all sorts of problems/
PopcornPam
@PopcornPam_twitter
May 14 2017 22:47
no, definately data.. they work to plug the rig into the computer, and fine when battery is in, just no go without a battery
Also.. not sure why I am getting this (root@edisonhost:~/src/oref0# cd myopenaps
-bash: cd: myopenaps: No such file or directory)
oh, nevermind.. i forgot to go up to regular cd
RedCyclist1
@RedCyclist1
May 14 2017 23:17
I am going to move to dev with smb. Is this the correct command?
mkdir -p ~/src; cd ~/src && git clone -b dev git://github.com/openaps/oref0.git || (cd oref0 && git checkout dev && git pull)
Scott Leibrand
@scottleibrand
May 14 2017 23:40
Yes, that will work. It will end up just doing the part in parentheses because you already have oref0 cloned.