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

4th
Dec 2017
peterfulvi
@peterfulvi
Dec 04 2017 00:35
I'm having issues with my G5 pairing to a little posh mobile phone I have. I try to pair the dexcomsl service and it say inorrect pin number. Would this have anything to do with the operating system or is there a step I'm missing in the pairing process.
Cfomodz
@Cfomodz
Dec 04 2017 05:44
Thank you
Lorenzo Sandini
@lsandini
Dec 04 2017 06:21
IMG_2236.JPG
yay, assembled the Pi Zero W and a TI stick and voilà :)
PieterGit
@PieterGit
Dec 04 2017 08:08
@lsandini Nice. You could not get an Edison/Explorer board anymore?
Of course openaps/oref0 uses eventualBG and the insulin curve to determine what to do, but I wonder if it would be beneficial for us to implement an extra safety measure with Figure 3. That is: limit the max IOB to be lower than "Insulin Dose Adjustment" based on the Trend Arrow and ISF (which is called Correction Factor, CF in the article). Thoughts?
Paper also references a recent paper on "Exercise management in type 1", http://www.sciencedirect.com/science/article/pii/S2213858717300141 . Did anybody read that and nows if that can be useful for the we are not waiting community
Lorenzo Sandini
@lsandini
Dec 04 2017 08:45
@PieterGit I have been planning a small-scale study for a while and it seems this will be happening. Since medical journals care mostly for prospective studies, that's what I am trying to do. It will be prospective, longitudinal analytic, no cross-over design or controls, but an observation of the transition from CSII + CGM to openAPS. I have purchased enough Edisons for that (mostly as Arduino kits as standalone Edisons are hard to find), and I hope I'll get the explorer boards for all of them. Still no funding so it's out of my pocket, but I am pretty confident I'll get a grant soon :) I wanted to make sure I have a "plan B" in case I ran out of Edison/Explorer rigs, and the Pi 0 W was fun to build. I need to determine the capacity of the power bank needed for 12-24h, or put a 3.3V ->5V converter if I want to use LiPo batteries. All WIP :)
PieterGit
@PieterGit
Dec 04 2017 08:48
@lsandini : sounds nice. Join https://gitter.im/openaps/hardware-dev if you want to get involved in the next level after PI0 and TI USB... Feel free to ask questions here if you need help
Dana Lewis
@danamlewis
Dec 04 2017 08:52
@lsandini there's another batch of EBs coming so shouldn't be an issue
Scott Leibrand
@scottleibrand
Dec 04 2017 09:10
@PieterGit we do use recent BG trend information (calculated directly, not inferred from tend arrows) to help determine eventualBG.
Scott Leibrand
@scottleibrand
Dec 04 2017 09:16
If you come across any instances where we dose more insulin than would be allowed by that heuristic, I’d be happy to investigate whether we’re being too aggressive there. I suspect you won’t find any though.
Lorenzo Sandini
@lsandini
Dec 04 2017 10:23
@PieterGit The impact of aerobic/anaerobic exercise on glucose metabolism is well documented, but often misunderstood even by healthcare professionals. It would be really interesting to measure the intensity and duration of exercise and see if it would be useful for openaps calculations. Bringing that data in real-time into Nightscout from a variety of sensors could be an objective, but the issue is the precision of the sensors. Heart rate from wristband sensors is not accurate and who wants to wear a chest strap 24/7 ? Step counters are very unspecific (stationary bike, weight lifting, ice hockey...). Moreover transferring the chest strap data in real time would require reading the data from the strap and uploading it to NS... how ? There have been efforts to make receivers for Polar straps, maybe someone here has some experience in that field ?
Velibor Maric
@vebaba
Dec 04 2017 10:43
@lsandini I didn't experiment but Polar (and others) has BT HR senders. Some apps (like Strava or Sports tracker) has ability to read those. Also, I believe that they read all other BT HR straps.
ejoe132
@ejoe132
Dec 04 2017 13:21
I have an open network I need to connect to, but cant find in the docs the right entry for it, from memory is it enter the name then for password "=keymanagement_none"?
It says:

If you want to add open networks to your list, then add:

network={
key_mgmt=NONE
priority=-999
}

ejoe132
@ejoe132
Dec 04 2017 14:17
@sdneufer Thank you
Scott Leibrand
@scottleibrand
Dec 04 2017 14:57
@lsandini getting the data in real time is still the biggest challenge. After that it should be fairly straightforward to do something like activate an activity mode based on HR or steps, but not sure how you’d reliably detect anaerobic activity that raises BG.
Diabetic_Dude
@diabetic_dude_twitter
Dec 04 2017 16:20
anyone have a good plan for repairing the JST connection on batteries? It appears that the female end of my JST connection is coming undone, I would love to fix it because the battery is still in good working order aside from the lousy connection on the end. I have looked but can only find ones with the cables attached
https://www.adafruit.com/product/261 this is mainly all I can find, I would love to just by the white piece on the end and re crimp it if that can be done.
Velibor Maric
@vebaba
Dec 04 2017 16:49
@diabetic_dude_twitter I'm not sure which exactly JST standard you need but you might need something like this http://uk.farnell.com/jst-japan-solderless-terminals/sph-002t-p0-5s/contact-0-05-0-22mm-cut-reel/dp/1671245?MER=bn_level5_5NP_EngagementRecSingleItem_3# and new housing
What would I do is to get some unused pin from something else, solder it to old connection which become loose and then solder wire on pin and finally put thermo shrinking cover to isolate it...
Scott Leibrand
@scottleibrand
Dec 04 2017 16:58
lipo batteries degrade with age, so if you can just buy a new one I'd do that instead of trying to repair it (and risking shorting it out and causing a fire)
Doran Barton
@fozzmoo
Dec 04 2017 17:19
Is there a document that walks someone through what to do if/when Nightscout just suddenly stops displaying anything in the OpenAPS pill?
Doran Barton
@fozzmoo
Dec 04 2017 17:20
Excellent (and quick). Thx Dana!
Doran Barton
@fozzmoo
Dec 04 2017 17:43
Yep. devicestatus collection got too big.
I really should move my mongodb off mlab and to my linode box.
Dave Acklam
@dcacklam
Dec 04 2017 18:00
@diabetic_dude_twitter
To fix the plug on the battery, you can take the pins out, re-crimp them, and put them back in. In my case I also put some solder over the crimp.
@diabetic_dude_twitter
If you flip the plug over (so that the 'spine' is facing down) you will see 2 plastic tabs. Pry one of those up, and use a stick-pin or other similar tool to pull the metal contact-pin out the back... Repeat for the other side.
Once you get the pin out, just use a needle nose pliers to pull the crimp open, cut back the wire a little, and then re-crimp... Add solder to 'make sure' the pin is on good, and push it back into the plug... Then look at how your wires were set up & figure out a way to prevent strain/the-wire-cracking-off-again...
Dave Acklam
@dcacklam
Dec 04 2017 18:52
@scottleibrand
Popped back in to let you guys know I'm looking at the latest dev-branch in-re x12 compatibility...
Dave Acklam
@dcacklam
Dec 04 2017 18:57

There's an issue with the attempt to pull bg_settings_raw returning a reference error (ValueError: Circular reference detected), erroring out the loop.

This same error occurred in the past on my 712, but it didn't force the loop to bail out with an error. Now it does.

Dave Acklam
@dcacklam
Dec 04 2017 19:08
Also, mmtune on the x12s, with the explorer, appears broken to the point where it actually causes communication errors, which can only be fixed with a firmware reflash
Eg, the pump will be talking to the rig just fine, and if I walk out of range or do something that triggers an MMtune (Say, reboot or otherwise let pump_loop_completed get too old) then it will lose comms & a reset will be needed...
Also, mmtune seems incapable of accurately tuning for an x12 with the explorer - it always finds all frequencies at -99, and always tunes to the default freq.
I've tried replacing the explorer board & tried an external antenna, and the mmtunes do not change - even though the pump & rig actually DO communicate... mmtune thinks there is no signal & acts accordingly.
Scott Leibrand
@scottleibrand
Dec 04 2017 19:17
Can you flag which exact lines in oref0-pump-loop are erroring out for 512?
Dave Acklam
@dcacklam
Dec 04 2017 19:17
Yep
It's every one of them that calls bg_targets_raw.json or bg_targets.json
Notably, I just did a git-pull on oref0, and didn't re-run the setup script...
So it may be that if I ran setup again, it would adjust the script to not have that...
The other error is that 'function mmtune' actually makes comms worse for my pump, not better...
Scott Leibrand
@scottleibrand
Dec 04 2017 19:19
Worth a try. Maybe open an issue for me to look at from a computer tomorrow? Bed time here.
Dave Acklam
@dcacklam
Dec 04 2017 19:19
On mine, I've had it call a firmware flash instead - but that's not something I think should be put into the over-all code...
Got it... Sleep well...
Diabetic_Dude
@diabetic_dude_twitter
Dec 04 2017 19:36
@dcacklam thank you, spot on, following what you said to do allowed me to find my weak point (where the strain is at) and gave me a way to reinforce the connection in the future. @vebaba thank you too my backup plan it is to use an old JST connector I have on battery that expanded
I also have another battery on order that should come in hopefully by Friday
Diadon81
@Diadon81
Dec 04 2017 19:42
Hi, someone know how to solve this error in pump-looplog Old pumphistory-24h refreshValueError: month must be in 1..12
ah, that doesn't work... do search for 'month must be in 1..12' and post from Nov13
Diadon81
@Diadon81
Dec 04 2017 20:18
@vebaba yeah, I checked, it looks like problem remains unsolved and final recommendation just to wait or push lots of temp basal to move pumphistory page.
Diadon81
@Diadon81
Dec 04 2017 21:04
@scottleibrand are there any other solution for refreshValueError: month must be in 1..12 problem instead of pushing a bunch of temp basals or just wait, when the history start to be outdated?
Velibor Maric
@vebaba
Dec 04 2017 21:07
there was more tries tomorrow. they tried to drill it down
Diadon81
@Diadon81
Dec 04 2017 21:08
because I am getting this issue in pumphistory-24h, it means I should wait at least for 27 hours :(
@vebaba yesterday?
Velibor Maric
@vebaba
Dec 04 2017 21:09
on nov14th. scroll down on rosa<something> posts and read hers and Scotts
I believe that there were no result but she reinstalled later but there were some tries to drill it down
Diadon81
@Diadon81
Dec 04 2017 21:14
yeah, I checked all the history, unfortunatelly, final conclusion was as I said in previous message. The reason looks clear, there is some kind of wrong history record. Because it was made by a pump itself (as far as I understood), it is impossible to avoid this issue.
Velibor Maric
@vebaba
Dec 04 2017 21:37
:(
Enda Farrell
@endafarrell
Dec 04 2017 21:53
Just a quick note to say that we have another Looper - me :-) Open Looping during the day (I need a case ...), Closed Looping tonight for the first time - but my biggest concern is that my "CGM" is a Freestyle Libre > Downpatched Sony SmartWatch3 > Libre Alarm(+) > xDrip+ > Nightscout. I've been trying different Libre positions - now its on the inside of my arm facing my ribs - that seems a lot better than on the outside of my arm.
Velibor Maric
@vebaba
Dec 04 2017 22:42
I was always wondering whether SW3 stands for full day measurement duty? I use it as xdrip+ 'monitor'