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

16th
Sep 2017
Craig Brenner
@cbrenner
Sep 16 2017 00:53
@danamlewis Congrats on the effort!
Scott Leibrand
@scottleibrand
Sep 16 2017 01:26
@cbrenner I'm not seeing why you think the zero temping in the first situation was a problem: it looks like it helped mitigate a low (from about 12:20-1:20), and then something else caused BG to start rising around 1:30. (the zero-temping was complete by about 11:30, so was unlikely to be the cause of a sharp spike 2h later)
how confident are you in the 82g carb count? often that pattern of a late high followed by a swift drop is caused by a carb overestimate, or a too-aggressive carb ratio / ISF
have you run autotune yet? does it suggest any changes in basals, ISF, or CR?
Craig Brenner
@cbrenner
Sep 16 2017 01:31
@scottleibrand Pretty confident with the 82g of carbs . I usually see the effect of basal changes with my son 1 - 2 hours later. So, my assumption was if the basal was zero temping for almost 2 hours form 9:30 - 11:30am then the cumulative impact would start to show itself around 1p - 1:30p and that would have been the rise. This has been the typical pattern when OpenAPS has been 0 temping for 1-2 hours after his meals
We run autotune everynight...ISF is generally pretty close +/- 5; CR is close. There have been basal recommendations some less some more and i just made those changes yesterday but autosense seems to have implemented these over the past several weeks.
Jacob H
@jdhigh
Sep 16 2017 02:50
Any thoughts in regards to the intel discontinuance scheduled for today and product order halt scheduled for tomorrow? should I order another edison board and chip before it's too late?
Dana Lewis
@danamlewis
Sep 16 2017 03:57
If you want a backup, yes order another edison. But suppliers should still have a good supply for a bit longer
thisistheusername
@thisistheusername
Sep 16 2017 12:57
Darrell Wright
@beached
Sep 16 2017 16:03
@scottleibrand this may be of interest to you https://blog.filippo.io/captive-browser/ with OpenAPS.
Scott Leibrand
@scottleibrand
Sep 16 2017 16:23
Looks like that relies on Chrome, so couldn't run on a rig.
@cbrenner if the impact of a temp starting at 9:30am didn't start until 1pm, that would indicate a peak insulin action time of 3.5h. It's usually more like 1-1.5h, but even if the delay is more like 2h the effect of 9:30-11:30 zero temping should be most visible from 11:30-1:30, when you had the (mitigated) low, not after 1:30 (when you had the rise).
Darrell Wright
@beached
Sep 16 2017 17:13
@scottleibrand I suspect something in python/node could be done with similar principals in the wii connected scripts. connect to http://captive.apple.com/hotspot-detect.html and detect the success. If not, click on all links, or post all forms(with any checks checked), and then after they return try again.
tynbendad
@tynbendad
Sep 16 2017 17:49
fwiw, in case anyone else was interested i ran an easy test - my 2500mah explorer edison rig ran without a loop for 35 hours before losing power, where with a loop running it gets about 16 hours. so, probably not a whole lot of room there to improve anyway.
kallnap
@kallnap
Sep 16 2017 18:43
@tynbendad your post reminds me that I need to hook up my 3500 mAh lipo instead
Dana Lewis
@danamlewis
Sep 16 2017 18:44
if anyone's using older batteries, don't forget to try a newer battery. I had an old 2k+ that was down to ~7 hours, put a newer 2k+, and I'm getting 12-14 hours out of it. same rig.
(year old battery)
Scott Leibrand
@scottleibrand
Sep 16 2017 20:37
@beached yeah, just a small matter of typing. ;-) LMK if you come across anyone who's done it for a headless Linux box.
Dana Lewis
@danamlewis
Sep 16 2017 20:39
@weedsDIY - easiest will be to change your NS URL and API secret in your "runagain" file, then do bash runagain
Craig Brenner
@cbrenner
Sep 16 2017 20:58
I am having some problems with the G5 data connection when our 2nd rig leaves a set WiFi connection and BT tethers to my son's iPhone to use his hotspot. It connects to his hotspot fine but then the G5 shows 'no data' on NS and on the Dexcom share apps. Locally my son is seeing the real-time BG data on his G5 app and the G5 app also shows that the internet connection is fine (green check) but no data is being transmitted to the cloud. Oddly, we don't have this problem with his primary rig just the 2nd rig. I'm not 100% sure but it seems that this problem started to happen when I upgraded to ver 0.53. My thinking was to uninstall the BT on the 2nd rig and then reinstall but not sure how to do that and do it so it is a fresh install. Any help would be appreciated.
If it matters, I am also seeing this error often on the 2nd rig: "localhost avahi-daemon: server.c: Packet too short or invalid while reading response record. (Maybe a UTF-8 problem?)"
WeedsDIY
@WeedsDIY
Sep 16 2017 22:00
@danamlewis
@danamlewis Thanks for the instruction. Back in business. Now back to learning more.
Dana Lewis
@danamlewis
Sep 16 2017 22:11
@weedsdiy awesome! :+1:
Dana Lewis
@danamlewis
Sep 16 2017 22:17
@cbrenner if bg data isn't getting to Dexcom app, need to troubleshoot that. Out of range of phone?
Craig Brenner
@cbrenner
Sep 16 2017 22:22
@danamlewis BG data is getting to Dexcom app (he sees real-time data locally but it doesn't send to the cloud). His phone is in his pocket so not out of range. Like I said, this only happens when his second rig (not his primary) BT tethers to his iPhone to use the hotspot. Seems like there is something odd going on when the BT connects to this phone and messes up something with his radio comm. To get things working again he has to restart his Dexcom app everytime
.....My guess is that there is something not right or corrupted with the 2nd rigs BT installation or supporting files. so trying to figure how to best to get a fresh install to validate this assumption