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

22nd
May 2017
Martin Haeberli
@mhaeberli
May 22 2017 00:02
@scottleibrand I like the custom Loop app idea, although I suspect it may be downstream of Share working... and I assume this custom loop thing can be found in RTD.
Martin Haeberli
@mhaeberli
May 22 2017 01:04
@laurathiessen pushover is for notifications on your phone if you have pushover installed
Bridge is for getting dexcom CGM to NightScout via the internet
BWP is basal wizard preview
Scott Leibrand
@scottleibrand
May 22 2017 01:33
I would disable the cob pill in NS: it presents misleading data vs. what is in the OpenAPS pill.
John Benjamin
@Bender1061
May 22 2017 01:55
SO I've been getting
MDT CGM configured; not waiting
Radio ok. Listening: .No pump comms detected from other rigs
Preflight OK. get-bg = ! bash -c " (echo -n MDT cgm data retrieve && openaps monitor-cgm 2>/dev/null >/dev/null && grep -q glucose cgm/cgm-glucose.json && echo d) && cp -pu cgm/cgm-glucose.json cgm/glucose.json && cp -pu cgm/glucose.json monitor/glucose-unzoned.json && (echo -n MDT cgm data reformat && openaps report invoke monitor/glucose.json 2>/dev/null >/dev/null && echo ted)"
MDT cgm data retrieved
MDT cgm data reformatted
Profile less than 60m old. jq: error: null and number cannot be divided
date: invalid date ‘@’
pump_loop_completed less than 5m ago. Refresh
a lot, and while the loop will keep happening working and the rig is running, I'm wondering if this is why for some strange reason my upload of the cgm data will sometimes take over an hour to update on night scout
the nightscout/glucose.json updates in a timely fashion. but the actual uploading to nightscout does not
David
@davidkeddydb
May 22 2017 02:00
Screenshot 2017-05-21 21.57.03.png
rsilvers129
@rsilvers129
May 22 2017 02:02
The author of the custom version of Loop that allowed location G5 data to pass to the Edison - he set up a web page with instructions. I lost the link to it. Does anyone have it?
David
@davidkeddydb
May 22 2017 02:02
I am getting this error ed
Could not parse input data: [SyntaxError: /root/dke6/monitor/iob.json: Unexpected end of input] should I redo set up or is there something else I should be looking at
Scott Leibrand
@scottleibrand
May 22 2017 02:15
I would run the settings refresh manually. That last line of its error output isn't enough to see what is failing.
rsilvers129
@rsilvers129
May 22 2017 02:53
I just started my loop for the first time to test it. I am not wearing the pump yet. The logs show that it is working. So great. But I have a second rig, and that one is having a problem seeing the radio. I reflashed the radio. I am on dev. Any ideas? Here is a screen capture.
bad.png
Scott Leibrand
@scottleibrand
May 22 2017 03:13
Try swapping the Edisons / Explorer boards and see which one the problem follows?
If it follows the Explorer board it's likely a cc1110 firmware or hardware issue. If it follows the Edison it's likely a config issue.
DaisyHK
@laurathiessen
May 22 2017 03:27
@scottleibrand ok thanks i will not use cob
@mhaeberli what would pushover be insalled on the phone? i have an android with xdrip+ right now
Martin Haeberli
@mhaeberli
May 22 2017 03:30
@laurathiessen as I understand it, you can install pushover https://pushover.net/ on one or more phones that you want alerted when certain events take place. I haven’t played with this in depth - I was just trying to answer your question - ‘what is pushover’. I do know that, in openaps, I can install and write python code that sends alerts when certain things happen, via pushover. But I also understand there may be some built in pushover stuff in Nightscout proper
DaisyHK
@laurathiessen
May 22 2017 03:33
@mhaeberli oh ok thanks. i will try the fb page maybe someone know there
John Benjamin
@Bender1061
May 22 2017 03:45
so I found a problem with oref0-pump-loop.sh the function function refresh_smb_temp_and_enact {
has touch -d "$(date -R -d @$(jq .[0].date/1000 monitor/glucose.json))" monitor/glucose.json Problem is on the MDT monitor/glucose.json, we don't have a .date that jq can read so it fails that routine. now after it's formatted in nightscout/glucose.json it's got one.
DaisyHK
@laurathiessen
May 22 2017 03:48
@mhaeberli @scottleibrand Bridge stuff is in regards to iphone mobile app? i have android and xdrip + so guessing leave these blank?
Martin Haeberli
@mhaeberli
May 22 2017 03:49
ok, you could say it that way - as i understand, it is in regard to ‘bridging’ data from Dexcom into Nightscout.
sounds like you shouldn’t need them
because xdrip+ feeds nightscout already
John Benjamin
@Bender1061
May 22 2017 03:50
I'll have to look to see if I can change that to the one in nightscout, or something else. but that is most definitely going to cause a problem. gut on the eastern zone it's bed time, so it will have to be later. but if anyone is curious, this is what our monitor/glucose.json looks like
 {
    "trend_arrow": "FLAT", 
    "direction": "Flat", 
    "name": "GlucoseSensorData", 
    "date_type": "relative", 
    "_tell": 2, 
    "display_time": "2017-05-21T23:41:00-04:00", 
    "packet_size": 0, 
    "op": 60, 
    "glucose": 120
  },
Scott Leibrand
@scottleibrand
May 22 2017 03:50
@Bender1061 thanks, I've been trying to get someone with MDT CGM to figure that one out for awhile
John Benjamin
@Bender1061
May 22 2017 03:52
i'm thinking make the nightscout/glucose.json happen during the reformatting of the cgm data, and just use that
but that might not work for everyone
DaisyHK
@laurathiessen
May 22 2017 03:53
@mhaeberli ok thanks! SHOW_RAWBG is set to the default never, but there is nothing on the site about it... i am guessing this is raw data but what is bad about seeing the raw data?
SHOW_RAWBG
Browser default raw display mode vaild settings are always never or noise
Martin Haeberli
@mhaeberli
May 22 2017 03:55
no recent direct experience w/show_rawbg - but in my current view, i can float over / click on a BG point and it will tell me the value; I assume with show-rawbg acquired via xdrip you can also get the underlying raw values
DaisyHK
@laurathiessen
May 22 2017 03:57
ok so i will leave it as the defult for now then. thanks!
it is now asking for payment info to verify the account.... guessing this is just for in case it goes over but I wont be automatically charged? I remember with azure you needed to change it to pay as you go in order to not get charged but i did not see anything in the instructions on this.
oh never mind
Martin Haeberli
@mhaeberli
May 22 2017 04:00
correct in my experience. it charged me zero - I then hit the limit
DaisyHK
@laurathiessen
May 22 2017 04:00
sorry kept reading
Martin Haeberli
@mhaeberli
May 22 2017 04:00
with three sites
after three weeks
DaisyHK
@laurathiessen
May 22 2017 04:00
3 sites?
Martin Haeberli
@mhaeberli
May 22 2017 04:00
but fixing it was just $7 /month
three heroku sites for a diversity of tests
DaisyHK
@laurathiessen
May 22 2017 04:01
wow you got a lot of T1D to follow lol
Martin Haeberli
@mhaeberli
May 22 2017 04:01
7 ;; mo / site
just 1 t1d
but different scenarios - openaps tune; openaps oref0; Loop
tec
etc
DaisyHK
@laurathiessen
May 22 2017 04:01
lol yes just teasing ;)
Martin Haeberli
@mhaeberli
May 22 2017 04:02
:+1:
DaisyHK
@laurathiessen
May 22 2017 04:10
tune vs oref0 ? thought there was only the oref0
Martin Haeberli
@mhaeberli
May 22 2017 04:13
oref0 with autotune vs without
for example
DaisyHK
@laurathiessen
May 22 2017 04:19
oh ok thanks for that clarfication
DaisyHK
@laurathiessen
May 22 2017 04:26
@mhaeberli this step indicates it is for LOOP but I am building OPENAPS so should i follow these?
image.png
Martin Haeberli
@mhaeberli
May 22 2017 04:31
looks good to me at first glance - it is in the openaps ‘read the documents’ collection, for starters. ; for doing openaps, you do not need to or want to enable the ‘loop’ plugin ; that plugin is used with iphone users who are running the iphone Loop app and who have rileylinks to talk to the pump
@scottleibrand fyi - information for your regression; the latest dev doesn’t seem to consistently solve the problem I have seen where the BG s are showing stale in the openaps logs but they are showing current on Nightscout (and are coming in through Nightscout); my own ‘solution’ is , by hand, to log in to the openaps / edison, fetch the most recent bg via the curl api, just to get it all confirmed, and then IF that is more current than the last 10 minutes, run openaps get-ns-bg
(running this afternoon’s dev as far as i can tell)
DaisyHK
@laurathiessen
May 22 2017 04:44
ok... think there is a problem my new site does not display all the plugins I set up in heroku
image.png
image.png
Scott Leibrand
@scottleibrand
May 22 2017 04:49
@mhaeberli it might be worth disabling your cron and running the stuff cron would run manually to step through things and see what's not happening automatically for you. it's strange that you're the only one seeing this.
Martin Haeberli
@mhaeberli
May 22 2017 04:51
@scottleibrand ok, thx. i’ll have to do that. fwiw, I thought I had seen some other users over the last few weeks seeing similar problems.
based on my own interpretation of discussions here ...
thanks for your patience with my whining ...
Scott Leibrand
@scottleibrand
May 22 2017 04:53
@mhaeberli none of the others I've seen are using a simple nightscout-only setup: that should be pretty bulletproof. all the other problems I've seen are with xdrip or MDT CGM
Martin Haeberli
@mhaeberli
May 22 2017 04:53
ok thx
Scott Leibrand
@scottleibrand
May 22 2017 04:54
you might also double-check that only one Nightscout is involved in this particular rig's setup
Martin Haeberli
@mhaeberli
May 22 2017 04:55
@scottleibrand you and @danamlewis do so much service for our community already; BUT I could pm you a papertrail extract in case anything stands out;
Scott Leibrand
@scottleibrand
May 22 2017 04:56
I doubt papertrail will be much help. nightscout glucose downloads don't log much, so you'll need to step through that part manually while looking at the glucose.json contents
Martin Haeberli
@mhaeberli
May 22 2017 04:56
k
rsilvers129
@rsilvers129
May 22 2017 05:01
drat. I installed jubilinix 0.0.0 by mistake. Will the update commands make it turn into the latest, or should I reflash?
Martin Haeberli
@mhaeberli
May 22 2017 05:04
@rsilvers129 per @esialb I was able to update from 1.0.0 to 1.1.0 (or whatever) by downloading the .zip kernel into root on the edison, expanding, and rebooting. but your mileage may vary
@scottleibrand

May 21 21:00:50 etghopenaps0 pump-loop.log: Settings less than 10m old
May 21 21:00:50 etghopenaps0 pump-loop.log: Pumphistory-24 less than 20m old
May 21 21:00:50 etghopenaps0 pump-loop.log: Completed pump-loop at Sun May 21 20:56:45 PDT 2017
May 21 21:00:50 etghopenaps0 network.log: Error, connecting BT to 10:41:7F:C7:58:E4, getting bnep0 IP
May 21 21:00:50 etghopenaps0 network.log: At Sun May 21 20:56:46 PDT 2017, my wifi network name is AirportNetwork, and my public IP is: 2601:0646:c001:be3e:fec2:deff:fe32:f019
May 21 21:00:50 etghopenaps0 network.log: Finished oref0-online.
May 21 21:00:50 etghopenaps0 network.log: Starting oref0-online.
May 21 21:00:50 etghopenaps0 network.log: At Sun May 21 20:57:02 PDT 2017 my local IP is: 10.0.0.217
May 21 21:00:56 etghopenaps0 ns-loop.log: Starting ns-loop at Sun May 21 20:57:03 PDT 2017:
May 21 21:00:56 etghopenaps0 network.log: At Sun May 21 20:57:02 PDT 2017, my wifi network name is AirportNetwork, and my public IP is: 
May 21 21:00:56 etghopenaps0 network.log: Error, cycling networking 
May 21 21:00:56 etghopenaps0 pump-loop.log: Starting pump-loop at Sun May 21 20:57:06 PDT 2017:
May 21 21:00:56 etghopenaps0 network.log: Stopping networking (via systemctl): networking.service.
May 21 21:00:56 etghopenaps0 ns-loop.log: ns://JSON/shell/cgm/ns-glucose.json
May 21 21:00:56 etghopenaps0 ns-loop.log: No JSON object could be decoded
May 21 21:00:56 etghopenaps0 ns-loop.log: Refreshed temptargets and meal-carbs
May 21 21:00:56 etghopenaps0 ns-loop.log: {"batteryVoltage":3638, "battery":44}
May 21 21:00:56 etghopenaps0 network.log: Starting networking (via systemctl): networking.service.
May 21 21:00:56 etghopenaps0 network.log: and getting new wlan0 IP
May 21 21:00:56 etghopenaps0 pump-loop.log: Radio ok, Listening: .No pump comms detected from other rigs
May 21 21:00:56 etghopenaps0 pump-loop.log: Profile less than 60m old
May 21 21:00:56 etghopenaps0 pump-loop.log: temp_basal.json less than 5m old
May 21 21:00:56 etghopenaps0 pump-loop.log: Pumphistory less than 5m old
May 21 21:00:56 etghopenaps0 pump-loop.log: Settings less than 10m old
May 21 21:00:56 etghopenaps0 pump-loop.log: Pumphistory-24 less than 20m old
May 21 21:00:56 etghopenaps0 pump-loop.log: Completed pump-loop at Sun May 21 20:57:43 PDT 2017
May 21 21:00:56 etghopenaps0 ns-loop.log: Uploaded
May 21 21:00:56 etghopenaps0 pump-loop.log: Starting pump-loop at Sun May 21 20:58:13 PDT 2017:
May 21 21:00:56 etghopenaps0 network.log: At Sun May 21 20:58:16 PDT 2017, my wifi network name is , and my public IP is: 
May 21 21:00:56 etghopenaps0 network.log: At Sun May 21 20:58:16 PDT 2017 my public IP is: 
May 21 21:00:56 etghopenaps0 network.log: Error, connecting BT to 10:41:7F:C7:58:E4, getting bnep0 IP
May 21 21:03:38 etghopenaps0 network.log: At Sun May 21 20:58:22 PDT 2017, my wifi network name is , and my public IP is: Finished oref0-online.
May 21 21:03:38 etghopenaps0 pump-loop.log: Radio ok, Listening: .No pump comms detected from other rigs
May 21 21:03:38 etghopenaps0 pump-loop.log: Profile less than 60m old
May 21 21:03:38 etghopenaps0 pump-loop.log: temp_basal.json less than 5m old
May 21 21:03:38 etghopenaps0 pump-loop.log: Pumphistory less than 5m old
May 21 21:03:38 etghopenaps0 pump-loop.log: Settings less than 10m old
May 21 21:03:38 etghopenaps0 pump-loop.log: Pumphistory-24 less than 20m old
May 21 21:03:38 etghopenaps0 pump-loop.log: Completed pump-loop at Sun May 21 20:58:49 PDT 2017
May 21 21:03:38 etghopenaps0 network.log: Starting oref0-online.
May 21 21:03:38 etghopenaps0 network.log: At Sun May 21 20:59:01 PDT 2017 my local IP is: 
May 21 21:03:38 etghopenaps0 network.log: At Sun May 21 20:59:02 PDT 2017, my wifi network name is , and my public IP is: 
May 21 21:03:38 etghopenaps0 network.log: Error, cycling networking 
May 21 21:03:38 etghopenaps0 ns-loop.log: Starting ns-loop at Sun May 21 20:59:02 PDT 2017:
May 21 21:03:38 etghopenaps0 network.log: Stopping networking (via systemctl): networking.service.
(so a network flap led to an unparsable json …)
but i’ll turn off cron and single step ...
Scott Leibrand
@scottleibrand
May 22 2017 05:17
actually network issues are probably more straightforward to troubleshoot
you certainly shouldn't be seeing the network.log erroring out and cycling networking that often
and oref0-online is a lot easier to step through manually than the nightscout BG download stuff :)
Martin Haeberli
@mhaeberli
May 22 2017 05:20
ok so i should look at troubleshooting?
Scott Leibrand
@scottleibrand
May 22 2017 05:22
yeah. I would comment out the oref0-online cron, and then manually run each thing it runs to see what's erroring and why
Martin Haeberli
@mhaeberli
May 22 2017 05:24
context - i have a comcast cable box with a 2.4 and 5 g wifi router, the edison is configured for both of them and bluetooth; the comcast box seems to offer lossy service off and on, which may be driving the flaps.
but yeah
DaisyHK
@laurathiessen
May 22 2017 05:48
I have set p my heroku nightscout site but it does not display the OpenAPS plugin, I have it in the settings of the app
image.png
Martin Haeberli
@mhaeberli
May 22 2017 05:58
@laurathiessen when you go to the web page to view the app proper (where it shows the bg graph) at the top right, there are three bars (kind of like an equals sign); click on that; scroll to the bottom, authenticate with your secret, save
then go back, and turn on all the plugins that are offered, including openaps
i hope that helps
openaps ns-loop fails with
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 48, in main
    reporters.Reporter(report, device, task)(output)
  File "/usr/local/lib/python2.7/dist-packages/openaps/reports/reporters/__init__.py", line 33, in __call__
    self.blob = self.serialize(data)
  File "/usr/local/lib/python2.7/dist-packages/openaps/reports/reporters/__init__.py", line 31, in serialize
    return self.method.serialize(render(data), self)
  File "/usr/local/lib/python2.7/dist-packages/openaps/vendors/process.py", line 56, in prerender_json
    return json.loads(data)
  File "/usr/lib/python2.7/json/__init__.py", line 338, in loads
    return _default_decoder.decode(s)
  File "/usr/lib/python2.7/json/decoder.py", line 366, in decode
    obj, end = self.raw_decode(s, idx=_w(s, 0).end())
  File "/usr/lib/python2.7/json/decoder.py", line 384, in raw_decode
    raise ValueError("No JSON object could be decoded")
ValueError: No JSON object could be decoded
DaisyHK
@laurathiessen
May 22 2017 05:59
@mhaeberli i have done all that already, the plug in just does not apper as an option
Martin Haeberli
@mhaeberli
May 22 2017 05:59
going to try to break it down ...
@laurathiessen ok, sorry, thx
DaisyHK
@laurathiessen
May 22 2017 06:00
@mhaeberli oh no worries... i am just stumped on how to fix it, thanks for trying will see if anyone else can help. I am asking katie on fb as well
Martin Haeberli
@mhaeberli
May 22 2017 06:01
@laurathiessen I tried to look at your image .png, but failed to see it at this end…; i suggest you make sure you are running the right branch etc of nightscout
Scott Leibrand
@scottleibrand
May 22 2017 06:01
make sure you have the plugin enabled in your heroku settings, and that you've restarted dynos
DaisyHK
@laurathiessen
May 22 2017 06:02
@mhaeberli right branch?
Martin Haeberli
@mhaeberli
May 22 2017 06:02
@laurathiessen follow @scottleibrand first - he definitely knows a lot more than i do
DaisyHK
@laurathiessen
May 22 2017 06:02
@scottleibrand it is enabled, restarted dynos?
Martin Haeberli
@mhaeberli
May 22 2017 06:02
but i’m referring to git branch
DaisyHK
@laurathiessen
May 22 2017 06:03
pic of what would show you if I was right or not for tht?
@mhaeberli
Martin Haeberli
@mhaeberli
May 22 2017 06:11
@laurathiessen i think you sent a screen shot of your settings; gitter wouldn’t let me zoom into it; no worries; @scottleibrand I think is a better guide ...
Scott Leibrand
@scottleibrand
May 22 2017 06:12
actually I've pretty much tapped out my knowledge of how to fix pill problems in heroku
Martin Haeberli
@mhaeberli
May 22 2017 06:12
but on that note, @scottleibrand , openaps get-bg is in my cron file, but of course it fails
with
# openaps get-bg
AttributeError: 'NoneType' object has no attribute 'iter_records'
grep: cgm/cgm-glucose.json: No such file or directory
not the root cause, though
Martin Haeberli
@mhaeberli
May 22 2017 06:18
need to break it down, but fixed it temporarily again on the way to diagnosis: openaps get-ns-bg failed (because cgm/ns-glucose.json empty), but when I tested the API via curl first, then it succeeded. so when it’s broken again, I’ll deconstruct get-ns-bg...
rsilvers129
@rsilvers129
May 22 2017 06:21
So on my problem where one rig works, and one can’t find the Edison radio, I didn’t want to unplug the Edisons and swap them on Explorers if I didn’t need to, because I didn’t want to risk bending a pin or something on the socket. So I reflashed the Edison and did a fresh install. Same problem. Most likely it is the hardware. This is a brand new Explorer latest version. How unusual is an Explorer radio not working issue?
What is carbsReq? Is this a way to get a notification that I need to eat X grams of carb to offset an oncoming low?
Scott Leibrand
@scottleibrand
May 22 2017 06:27
@rsilvers129 I've never heard of any problems with bent pins or similar on the Edison's bus, except when Tim completely crushed his entire rig. so swap it carefully and you should be fine.
Martin Haeberli
@mhaeberli
May 22 2017 06:42
@scottleibrand so the get-ns-glucose report is failing with unparseable json object.
now trying to deconstruct that (note that it seems to work fine if I preflight with the api request)
Martin Haeberli
@mhaeberli
May 22 2017 06:51
ok, preflighting doesn’t ALWAYS work ...
Dana Lewis
@danamlewis
May 22 2017 06:58
@mhaeberli ps - since you are having issues w NS data, are you in the habit of plugging in the receiver every night? Good habit to be in regardless
Martin Haeberli
@mhaeberli
May 22 2017 07:05
@danamlewis yes - generally plugged in at night. problems arise at a few points - sometimes share app just dies and has to be kicked. but i believe the framing root cause is networking here at night - i need to follow scott’s advice re the oref0-online step through (too tired tonight), but my direct console logs show
 [ 3426.626009] link down if wlan0 may call cfg80211_disconnected. event : 16, reason=2 from 88:ad:43:05:f8:88
[ 3426.659304] wl_android_wifi_off in
[ 3426.669714] wifi_platform_set_power = 0
[ 3433.117546] 
[ 3433.117546] Dongle Host Driver, version 1.141.59 (r)
[ 3433.117546] Compiled in /home/robin/src/edison-linux-helper/edison-bcm43340 on Apr  4 2017 at 22:42:52
[ 3433.117724] wl_android_wifi_on in
[ 3433.137712] wifi_platform_set_power = 1
[ 3433.688819] F1 signature OK, socitype:0x1 chip:0xa94c rev:0x2 pkg:0x0
[ 3433.697992] DHD: dongle ram size is set to 524288(orig 524288) at 0x0
[ 3433.707086] dhdsdio_download_firmware: firmware path=/lib/firmware/fw_bcmdhd.bin, nvram path=/lib/firmware/bcmdhd.cal
[ 3433.848990] sdioh_request_buffer: [1] doing memory copy buf=f4dc6000, len=2037
[ 3433.861116] dhdsdio_write_vars: Download, Upload and compare of NVRAM succeeded.
[ 3434.029635] dhd_bus_init: enable 0x06, ready 0x06 (waited 0us)
[ 3434.039988] wifi_platform_get_mac_addr
[ 3434.046927] wifi_get_mac_addr_intel: unable to open /config/wifi/mac.txt
[ 3434.058689] Firmware up: op_mode=0x0015, MAC=fc:c2:de:32:f0:19
[ 3434.091234] Firmware version = wl0: Aug 15 2014 15:09:17 version 6.20.190.2 (r497061) FWID 01-bb9bbd07
[ 3434.102682] dhd_preinit_ioctls wl ampdu_hostreorder failed -23
[ 3438.848248] wl_bss_connect_done succeeded with 88:ad:43:05:f8:88
[ 3438.878543] wl_bss_connect_done succeeded with 88:ad:43:05:f8:88
iterating every few minutes or so
obviously i have to be tethered directly to edison to debug this part (the networking part)
BUT when the network ‘flaps’ in this way, it seems sometimes to break the download from nightscout
nightscout indeed has fresh data, visually and per the api, but openaps can’t seem to get it sometimes, absent (maybe) a reboot, or this not completely clear set of operations - check api with curl, try openaps get-ns-bg (sometimes also a swing at ns-loop helps…)
Not today, but soon I’ll try this on a second Edison brick to see whether I can tell the difference ...
PieterGit
@PieterGit
May 22 2017 07:27
do others also sometimes hear a high frequent noise when a adafruit lipo (2500 mah in my case, https://www.adafruit.com/product/328) is charging?
David Goelzhaeuser
@idoodler
May 22 2017 07:40
@PieterGit Are you using the Explorer board?
PieterGit
@PieterGit
May 22 2017 07:40
yes
David Goelzhaeuser
@idoodler
May 22 2017 07:42
@PieterGit Mine is silent. I am using an LiPo charger from Adafruite in another project, it is also making high frequent noise. I guess it depends on the quality of the components.
PieterGit
@PieterGit
May 22 2017 07:47
ok, thanks, i'll try the other explorer board tonight to see if that makes a difference
David Goelzhaeuser
@idoodler
May 22 2017 07:50
Oh, btw. I am not yet using this battery, it should come today. It may also depend on the battery, but right now I am using a realy big 4400 mAh battery without any sound.
samihusseingit
@samihusseingit
May 22 2017 09:21
for setting a temp target, if you set the time to lets say 30 minutes, vs 60 minutes
does it actually consider the time( so he should be at this target within 30 minutes), vs (he should be at this target within 60 minutes)
meaning it would be more aggressive/faster to get within the time range provided?
or not?
Mike
@libxmike
May 22 2017 09:55
@/all if anyone in EU needs Medtronic 522 please PM me, I have one with compatible FW for sale as I use just my 754
Tim Street
@tim2000s
May 22 2017 11:12
@libxmike I know someone who wants one. Can you PM me please?
fidoman
@efidoman
May 22 2017 11:24

@mhaeberli - I had a similar problem with MDT on the dev build with microbolus enabled. I ended up making a custom loop where get-bg was done no matter what and the wait for bg function was made to where it will not wait. Here's the code I changed. It made mine work but it definitely is not the correct fix, just a hack for now ....

function wait_for_bg {
# EYF call get-bg here
openaps get-bg

echo -n "Waiting up to 4 minutes for new BG: "
for i in `seq 1 24`; do
    \# set mtime of monitor/glucose.json to the time of its most recent glucose value

# EYF -comment out the touch which is failing on my system for some reason and causing constant 4 minute waits
# touch -d "$(date -R -d @$(jq .[0].date/1000 monitor/glucose.json))" monitor/glucose.json
if (! ls monitor/pump_loop_completed >/dev/null ); then

Brian Rabinovitz
@bdr1177
May 22 2017 11:36
The docs say to use a 2000 mAh battery. Am I correct in recalling that that folks have also successfully used the 2500 mAh adafruit battery?
Savek-CC
@savek-cc
May 22 2017 12:03
@bdr1177 only voltage and connector are relevant - the rest is up to you and your case
rsilvers129
@rsilvers129
May 22 2017 12:42
@bdr1177 2500 is common. Really large sizes would take too long to charge such that you could not charge them with the Explorer within say eight hours. So if someone wanted say a 10,000, they would probably want another way to charge it.
peterfulvi
@peterfulvi
May 22 2017 13:51
Starting supermicrobolus pump-loop at Mon May 22 09:48:02 EDT 2017 with 3 second wait_for_silence:
get-bg = ! bash -c " (echo -n MDT cgm data retrieve && openaps monitor-cgm 2>/dev/null >/dev/null && grep -q glucose cgm/cgm-glucose.json && echo d) && cp -pu cgm/cgm-glucose.json c
gm/glucose.json && cp -pu cgm/glucose.json monitor/glucose-unzoned.json && (echo -n MDT cgm data reformat && openaps report invoke monitor/glucose.json 2>/dev/null >/dev/null && echo ted)"
MDT CGM configured; not waiting
Radio ok. Listening: .No pump comms detected from other rigs
Preflight OK. get-bg = ! bash -c " (echo -n MDT cgm data retrieve && openaps monitor-cgm 2>/dev/null >/dev/null && grep -q glucose cgm/cgm-glucose.json && echo d) && cp -pu cgm/cgm-glucose.json cgm/glucose.json && cp -pu cgm/glucose.json monitor/glucose-unzoned.json && (echo -n MDT cgm data reformat && openaps report invoke monitor/glucose.json 2>/dev/null >/dev/null && echo ted)"
MDT cgm data retrieved
MDT cgm data reformatted
Profile less than 60m old. jq: error: null and number cannot be divided
date: invalid date ‘@’
pump_loop_completed less than 5m ago. RefreshWarning: clock input Invalid Date is unzoned; please pass clock-zoned.json instead
ed pumphistory
pumphistory.json: "Bolus 2017-05-22T09:47:27 head[4], body[0] op[0x01]"
Checking pump clock: "2017-05-22T09:48:39-04:00" is within 1m of current time: Mon May 22 09:49:01 EDT 2017
and that pumphistory is less than 1m old.  Temp refreshed
{"carbs":13,"boluses":2,"mealCOB":0,"currentDeviation":-1.33,"maxDeviation":6.88,"minDeviationSlope":-0.94}
{"iob":2.473,"activity":0.0295,"bolussnooze":0,"basaliob":1.912,"netbasalinsulin":4.1,"hightempinsulin":4.1,"microBolusInsulin":0.6,"microBolusIOB":0.444,"time":"2017-05-22T13:49:05.000Z","lastBolusTime":1495460847000}
{"delta":-2,"glucose":176,"short_avgdelta":-1.51,"long_avgdelta":-1.26}
{"duration":29,"rate":2.2,"temp":"absolute"}
Autosens adjusting basal from 1.2 to 1.3; target_bg from 100 to 96; sens from 35 to 31.8
Carb Impact: 2.7 mg/dL per 5m; CI Duration: 0 hours; remaining 4h+ CI: 0 mg/dL per 5m
Accel. Carb Impact: 10 mg/dL per 5m; ACI Duration: 0 hours
UAM Impact: 3.2 mg/dL per 5m; UAM Duration: 0.3 hours
minPredBG: 102 minIOBPredBG: 107 minUAMPredBG: 96 avgPredBG: 103 COB: 0 carbs: 13
bgUndershoot: -17 zeroTempEffect: 19 carbsReq: -8
Checking deliverAt: 2017-05-22T13:49:14.000Z is within 1m of current time: Mon May 22 09:49:15 EDT 2017
and that smb-suggested.json is less than 1m old
enact/smb-suggested.json: {"insulinReq":0,"bg":176,"reservoir":"203.5","temp":"absolute","snoozeBG":113,"predBGs":{"UAM":[176,174,170,166,162,157,153,149,145,142,138,134,131,128,125,122,120,117,115,113,111,109,108,106,105,104,103,102,101,100,100,99,99,98,98,97,97,97,97,96],"IOB":[176,174,171,169,166,164,161,158,155,152,149,145,142,139,136,133,130,128,126,124,122,120,119,117,116,115,114,113,112,111,111,110,110,109,109,108,108,108,107]},"minPredBG":999,"IOB":2.473,"reason":"COB: 0, Dev: 16, BGI: -4.69, ISF: 32, Target: 96, minPredBG 102, IOBpredBG 107, UAMpredBG 96; 113 > 105: no temp required (letting high temp of 2.2 run). ","COB":0,"eventualBG":113,"tick":-2,"deliverAt":"2017-05-22T13:49:14.000Z"}
No smb_enact needed. Temp refreshed: monitor/temp_basal.json: {"duration":29,"rate":2.2,"temp":"absolute"}
No bolus needed (yet). Settings less than 10m old
Edison on battery: 71%. Pumphistory-24 < 90m old
Completed supermicrobolus pump-loop at Mon May 22 09:49:19 EDT 2017:
I got it to work but there seems to be different errors in my loop could someone help me sort it out
So now all i have to do is enter the carb amounts to my meals without boulsing?
Diabetic_Dude
@diabetic_dude_twitter
May 22 2017 14:05
could someone tell me what they know about the remainingCarbsCap option in the preferences.json file ?
Martin Haeberli
@mhaeberli
May 22 2017 14:07
@ericfeibelman thx
Diabetic_Dude
@diabetic_dude_twitter
May 22 2017 14:24
thanks :)
KensingtonB
@KensingtonB
May 22 2017 14:46
Hi -- hoping for some advice on dealing with "no space left on device" error messages. I am trying to update to latest oref 0 or even oref1, but cannot seem to get past this initial error.
Lynne Butts
@flutelynne_twitter
May 22 2017 15:18
You're welcome
katie disimone
@Kdisimone
May 22 2017 15:18
@KensingtonB try rebooting. If that doesn't work, you may need to clear out space in your tmp directory afaik
KensingtonB
@KensingtonB
May 22 2017 15:20
Thanks Katie. How exactly do I clear out that space?
peterfulvi
@peterfulvi
May 22 2017 15:26
Another question. I'm about to eat lunch, I'm trying the oref1 algorithm out. How do I enter my carbs into my pump without initiating a boulus. Would I just cancel the bolus after entering the carb amount
That is through bolus wizard
Scott Leibrand
@scottleibrand
May 22 2017 15:29
No, you need to do a 0.1U bolus with it to ensure it creates the bolus wizard entry.
@samihusseingit no, it's not any more aggressive for a shorter temp target. The duration just determines for how long it applies the modified target behavior. Once it expires, it goes back to normal.
peterfulvi
@peterfulvi
May 22 2017 15:42
Great thanks
peterfulvi
@peterfulvi
May 22 2017 16:18
@scottleibrand I tried that and it gave me a full boulus do I have to change any of my settings in my pump. It didn't let me adjust it before it gave me the insulin. Thanks
Scott Leibrand
@scottleibrand
May 22 2017 16:19
you should be able to just hold down the down button when it brings up the recommended full bolus amount to reduce that to 0.1U
Martin Haeberli
@mhaeberli
May 22 2017 16:24
@KensingtonB I had the running out of space issue with jubilinux 0.1.0 (logs were filling up fast); going to 0.1.1 helped. Clearing out logs as an interim helped, but I don’t have the magic memorized; google is your friend
until I upgraded, I had good luck changing from daily to hourly logs, and keeping only 4 of them or so...
AND compressing them
Martin Haeberli
@mhaeberli
May 22 2017 17:13
peterfulvi
@peterfulvi
May 22 2017 18:01
@scottleibrand it just scrolls down the menu when I do that
@scottleibrand just got it​ thanks you have to initiate it
katie disimone
@Kdisimone
May 22 2017 18:18
@scottleibrand I take it UAM can help make smbs kick in a little faster (and/or stronger) in the case of a rapidly rising BG?
Scott Leibrand
@scottleibrand
May 22 2017 18:24
immediately after a meal, that only applies if you don't enter (or way underestimate) carbs. but even if you enter carbs accurately, UAM is more helpful on the tail end if you end up rising more than expected from them, or if absorption stops prematurely.
UAM is quite useful on unexpected rises in BG though, regardless of the reason.
Jason Wittmer
@jbwittmer
May 22 2017 18:27
So, my son just entered "double" carbs at school (he cancelled a bolus after giving minimal insulin then re-entered it). Nurse was "out of the office". Ugh. Fortunately his remainingCarbCap is set near his usual lunch carbs. So I'm correct that oref1 should wait to bolus "extra" insulin until it sees carb impact? I set a higher temp-target just as a safety and we'll monitor closely. Will be interesting to see what happens. Anyone else w/ similar experience?
katie disimone
@Kdisimone
May 22 2017 18:35
yup, we’ve had double carbs periodically. A high temp target will help tremendously
Jason Wittmer
@jbwittmer
May 22 2017 18:36
Thanks @Kdisimone
Scott Leibrand
@scottleibrand
May 22 2017 18:42
remainingCarbsCap will prevent it from bolusing too much up front, before carb absorption is observed. However, even with that cap, it will still assume that any observed carb absorption is going to continue twice as long as it would otherwise, so there's a risk of SMB doing too much insulin. Setting a high temp target will disable SMB for the duration, which will limit any extra insulin to what can be delivered via high temp basal (subject to the high temp target of course). I believe you can set high temp targets up to 200 mg/dL. Although that probably won't necessary in this case, you should keep an eye on IOB and be prepared to set a higher temp target if needed.
katie disimone
@Kdisimone
May 22 2017 18:42
@scottleibrand thanks. I turned UAM on this morning, and Anna ate a “mini mcmuffin” and had to totally guess on carbs. Saw a pretty swift BG rise, and a pretty swift response today too. Wasn’t sure how much was the BG rise vs the UAM seeing the bg rise. I think UAM may help her with these pesky “nutrition breaks” where they serve up doritos and mcmuffins and donuts. She’s been having to split bolus on those (and she’s been upping carb counts…sometimes too far). Today’s looking like UAM maybe help pick up the split portion a little more and perhaps keep her foot off the gas on upping the carb counts (which would be nice)
Scott Leibrand
@scottleibrand
May 22 2017 18:43
yep, that's exactly what UAM is designed to help with. we've found that with oref1 SMB+UAM, a really rough guess on carb count, and an initial meal bolus sized accordingly, is generally enough for oref1 to take care of the rest.
katie disimone
@Kdisimone
May 22 2017 18:46
@scottleibrand do you want me to file an issue for this issue? Twice now, coincidentally, Anna has eaten a meal just prior to restarting her dex session. Both times, after 2-3 hours when she remembers to enter the calibration points and start up again, all the carbs are remaining and none have decayed. SMBs have started up immediately after she comes back online because of the undecayed carbs. It hasn’t been a huge deal, we just set a temp target…but can there be some logic added to it that if BGs aren’t there for 120 minutes that carbs decay or SMB is suspended for 30 minutes after BGs start coming in again? Or something? I’m just throwing ideas around in my head. This kind of stuff usually has a lot of other implications that I’m not imagining for the one scenario I’m specifically encountering.
Scott Leibrand
@scottleibrand
May 22 2017 18:48
There is an issue open for that already.
Might be worth commenting on it
katie disimone
@Kdisimone
May 22 2017 18:48
IMG_3249.PNG
Oh cool…I’ll go find it
thanks!
ericmtx
@ericmtx
May 22 2017 18:48
I've also just started my daughter (7th grade) on the SMB+UAM testing. So far so good - Katie thanks for your blog also. I shared that with my daughter so she feels like part of the experimental girl club now.
Scott Leibrand
@scottleibrand
May 22 2017 18:49
@Kdisimone as a workaround, you can also set your min_5m_carbimpact higher: if you're running oref1, that will just make it decay old carbs faster when no carb absorption is observed (for example because there's no BG data)
and of course, if you're using G4, you can do the two receiver staggered restart trick
katie disimone
@Kdisimone
May 22 2017 18:50
Yeah, I have unfortunately been away from home both times the restarts have happened these last two weekends and she’s opted to not stagger because…teenager.
@ericmtx Well, I’ll be adding UAM to it shortly. Thanks! Preview on the mcmuffin is looking very very good
Scott Leibrand
@scottleibrand
May 22 2017 18:52
@Kdisimone heh. I think that's as much of a PWD "yeah yeah, I can handle it manually" thing as a teenager thing. (and they're right, they usually can, and it often makes things easier for the PWD, which is good. but as a side effect it also makes things a bit harder for the automated system and those of us trying to observe / help remotely)
katie disimone
@Kdisimone
May 22 2017 18:52
Screen Shot 2017-05-22 at 11.51.51 AM.png
@scottleibrand SOOOO true!
ericmtx
@ericmtx
May 22 2017 18:56
For our testing so far our daughter is still dialing in her full carbs on the pump, along with BG, and then accepting the usual bolus. The SMBs today are then picking up at little over an hour after her full bolus to round out her BG rise.
Scott Leibrand
@scottleibrand
May 22 2017 18:56
:+1:
katie disimone
@Kdisimone
May 22 2017 18:56
And…yesterday was a movie with her friends while I was out of town. She sent me a test saying “i’m going to the movies with Zoey. I’m bolusing for popcorn, but no texting unless you really need to please.” Happy to say that smb kept me from having to text.
Screen Shot 2017-05-21 at 8.11.33 PM.png
Scott Leibrand
@scottleibrand
May 22 2017 18:58
<3 seeing that we're helping give kids back some of the freedom and independence T1D has stolen away. :-D
ericmtx
@ericmtx
May 22 2017 18:59
Yep - daughter just turned 13 this weekend. So yesterday and today have been big chunks of cookie cake. SMB is handling that pretty well!
katie disimone
@Kdisimone
May 22 2017 19:01
Oh, it’s the best part of looping! So nice to have that lighter touch with kid’s autonomy.
IMG_3399.PNG
Scott Leibrand
@scottleibrand
May 22 2017 19:03
:-D :+1: <3
Jason Wittmer
@jbwittmer
May 22 2017 19:05
+1 to what everyone else just said.
Dana Lewis
@danamlewis
May 22 2017 19:09
/me ’s heart is so full of this channel’s awesomeness.
thanks everyone for sharing how this is helping real-world :) <3 :confetti_ball:
Martin Haeberli
@mhaeberli
May 22 2017 19:09
:+1:
katie disimone
@Kdisimone
May 22 2017 19:32
IMG_3404.PNG
and…this is why you have safety features like long suspensions in SMB. So if your loop fails…the pump does what it needs to do still (her rig has come apart, screw came loose. sigh. almost had them tightened again this weekend and forget.)
but also…wow. it pre-planned that temp basal suspension ending time just about perfectly from a pretty far away time frame. would’ve been nice to take this experiment all the way to finish with rig working…but even in the rig failure, there was some good insight.
Screen Shot 2017-05-22 at 12.33.42 PM.png
Scott Leibrand
@scottleibrand
May 22 2017 19:41
:+1:
David Goelzhaeuser
@idoodler
May 22 2017 19:48
see-battery-life-connected-devices-iphone-450x800.jpg
Does anyone know how to enable Bluetooth Battery Service to see the Edisons Battery level on the iPhone?
Scott Leibrand
@scottleibrand
May 22 2017 19:51
That'd be cool: I did not even know such a thing was possible. We do report edison-voltage to NS, though, so you can see that with Glimpse.
David Goelzhaeuser
@idoodler
May 22 2017 19:53
Ok, I will take a look on Bluetooth Battery Service then:)
Scott Leibrand
@scottleibrand
May 22 2017 19:57
would be doubly cool if it could report both rig and pump battery levels
David Goelzhaeuser
@idoodler
May 22 2017 19:57
Yeah, that was my second tought:)
Is the Edison using BLE?
Scott Leibrand
@scottleibrand
May 22 2017 20:00
only if you install openxshareble (which most people don't use)
if you could figure out how to properly daemonize that and get it reliable (without all the connects and disconnects) that would be an even bigger win
rsilvers129
@rsilvers129
May 22 2017 20:23
Yesterday I saw in some docs a way to flash just the radio of the Explorer board. Now now matter how I search, I can’t find that page again. Does anyone know where that doc is?
Tim Street
@tim2000s
May 22 2017 20:27
If possible (and people don't mind) there's a possibility I'll be presenting this at the UKs biggest professional diabetes care conference in March. I'd like to share stories and the benefits people get.
Dana Lewis
@danamlewis
May 22 2017 20:27
:clap:
David Goelzhaeuser
@idoodler
May 22 2017 20:28
Sounds great!
rsilvers129
@rsilvers129
May 22 2017 20:47
@danamlewis Thank you. The program name helped me find what I had seen before. It was this: http://openaps.readthedocs.io/en/latest/docs/Resources/troubleshooting.html
rsilvers129
@rsilvers129
May 22 2017 21:07
It didn’t help. I tried everything I can think of, and can’t make my second Explorer board radio work. I tried 18 different combinations.
Screen Shot 2017-05-22 at 4.17.35 PM.png
Dana Lewis
@danamlewis
May 22 2017 21:08
Did you swap your existing Edison to that board to test?
rsilvers129
@rsilvers129
May 22 2017 21:08
@danamlewis Yes. I named my boards and Edisons 0 and 1. In the chart I tried every combination.
Either Edison will work on my first Explorer.
Neither will work on the new Explorer.
Curiously though, Master worked with some combination where Dev did not.
But then after running Master once, Dev started to work (but only on the original Explorer).
aabramowicz
@aabramowicz
May 22 2017 21:15
Open.JPG
Got this error today...so rig stopped. What happened?
aabramowicz
@aabramowicz
May 22 2017 21:31
The problem corrected itself but I'm wondering what happened...did we cause it by resetting the rig?
Tim Street
@tim2000s
May 22 2017 21:31
@SandraK82 :+1: :+1: :+1:
tynbendad
@tynbendad
May 22 2017 21:32
does openaps 'sleep' itself if its not getting a connection or something?
Starting pump-loop at Mon May 22 11:41:25 MDT 2017:
Starting pump-loop at Mon May 22 11:54:10 MDT 2017:
"916.636", 0, -99 Starting pump-loop at Mon May 22 11:56:39 MDT 2017:
"916.540", 5, -80 Starting pump-loop at Mon May 22 11:58:58 MDT 2017:
Starting pump-loop at Mon May 22 13:08:05 MDT 2017:
Starting pump-loop at Mon May 22 14:18:14 MDT 2017:
(that is from a "grep Starting" - note the times)

it stopped working while he was at school, about 10AM, i've just now rebooted it to see if it will come back, still waiting...

its back. but what hap'd?

PieterGit
@PieterGit
May 22 2017 21:35
did anybody else noticed quite a lot of time sets with the oref1 code? See openaps/oref0#477
Scott Leibrand
@scottleibrand
May 22 2017 21:40
@rsilvers129 did you try ccprog yet? did you verify you're flashing the correct .hex file (the one from ERD for SPI)?
@aabramowicz there are lots of ways that a process can be killed while it still has a lockfile open. the oref0-reset-git script will clean up the old logfile after a few minutes if it doesn't get fixed some other way first
@tynbendad yes, it will wait-for-silence longer the worse the signal strength is, up to about 2 minutes. look at the lines you filtered out to see the logic
that wouldn't account for the hour-long gap, though
aabramowicz
@aabramowicz
May 22 2017 21:46
@scottleibrand ok...so since its working I can ignore it.
Scott Leibrand
@scottleibrand
May 22 2017 21:48
yep. unless you're motivated enough to go figure out how to disable git entirely in the openaps toolkit, that's the best workaround we've found so far.
tynbendad
@tynbendad
May 22 2017 21:55
@scottleibrand it looks to me like there are regular git errors, even starting a few days ago, but when it went silent it seems like maybe it couldn't recover from one (this happened several days ago as well, but that time even a reboot couldn't fix it).
Scott Leibrand
@scottleibrand
May 22 2017 21:57
if you can figure out what kind of error it was that it couldn't fix (or what git command fails when that occurs), we can add a check for that to oref0-reset-git
tynbendad
@tynbendad
May 22 2017 21:57
i've already rebooted, and its working - if it helps i could zip up the pump-loop and put it somewhere, but i don't think it shows enough detail (for me)
here are the last few loops - over an hour apart, and no indication other than BG data old (NS had great BG data but the rig stopped connecting to bt tether)
Completed pump-loop at Mon May 22 12:00:17 MDT 2017

Starting pump-loop at Mon May 22 13:08:05 MDT 2017:
Radio ok, Listening: .No pump comms detected from other rigs
Old pumphistory: Refreshed pumphistory
{"carbs":198,"boluses":10.925,"mealCOB":198}
{"ratio":1.01}
BG data is too old (it's probably this), or clock set incorrectly.  The last BG data was read at Mon May 22 2017 10:06:07 GMT-0600 (MDT) but your system time currently is Mon May 22 2017 13:09:20 GMT-0600 (MDT)
determine-basal://text/shell/enact/suggested.json
reporting enact/suggested.json
{"reason":"BG data is too old (it's probably this), or clock set incorrectly.  The last BG data was read at Mon May 22 2017 10:06:07 GMT-0600 (MDT) but your system time currently is Mon May 22 2017 13:09:20 GMT-0600 (MDT)"}
{"reason":"BG data is too old (it's probably this), or clock set incorrectly.  The last BG data was read at Mon May 22 2017 10:06:07 GMT-0600 (MDT) but your system time currently is Mon May 22 2017 13:09:20 GMT-0600 (MDT)"}
Waiting for ntpd to synchronize... OK!
{"bg":150,"temp":"absolute","snoozeBG":354,"recieved":true,"predBGs":{"COB":[150,150,149,147,144,140,136,130,124,116,108,99,90,79,68,55,42,39],"aCOB":[150,154,158,160,161,162,161,159,157,154,149,144,138,131,123,114,104,92,80,67,54,41,39],"IOB":[150,149,147,144,140,134,127,118,108,97,85,71,56,41,39]},"rate":0.9,"reason":"COB: 130, Dev: 33, BGI: -5.63, ISF: 67, Target: 95; Eventual BG 39 < 90, bolus snooze: eventual BG range 39-354; setting current basal of 0.9 as temp","COB":130,"eventualBG":39,"timestamp":"2017-05-22T10:07:15.522921","duration":30,"tick":"+0","IOB":8.034}
Old settings refreshed
temp_basal.json less than 5m old
Pumphistory less than 5m old
Settings less than 10m old
pumphistory-24h refreshed
Completed pump-loop at Mon May 22 13:09:59 MDT 2017

Starting pump-loop at Mon May 22 14:18:14 MDT 2017:
Radio ok, Listening: .No pump comms detected from other rigs
Old pumphistory: Refreshed pumphistory
{"carbs":198,"boluses":10.925,"mealCOB":198}
{"ratio":1.01}
BG data is too old (it's probably this), or clock set incorrectly.  The last BG data was read at Mon May 22 2017 10:06:07 GMT-0600 (MDT) but your system time currently is Mon May 22 2017 14:19:30 GMT-0600 (MDT)
determine-basal://text/shell/enact/suggested.json
reporting enact/suggested.json
{"reason":"BG data is too old (it's probably this), or clock set incorrectly.  The last BG data was read at Mon May 22 2017 10:06:07 GMT-0600 (MDT) but your system time currently is Mon May 22 2017 14:19:30 GMT-0600 (MDT)"}
{"reason":"BG data is too old (it's probably this), or clock set incorrectly.  The last BG data was read at Mon May 22 2017 10:06:07 GMT-0600 (MDT) but your system time currently is Mon May 22 2017 14:19:30 GMT-0600 (MDT)"}
Waiting for ntpd to synchronize... OK!
{"bg":150,"temp":"absolute","snoozeBG":354,"recieved":true,"predBGs":{"COB":[150,150,149,147,144,140,136,130,124,116,108,99,90,79,68,55,42,39],"aCOB":[150,154,158,160,161,162,161,159,157,154,149,144,138,131,123,114,104,92,80,67,54,41,39],"IOB":[150,149,147,144,140,134,127,118,108,97,85,71,56,41,39]},"rate":0.9,"reason":"COB: 130, Dev: 33, BGI: -5.63, ISF: 67, Target: 95; Eventual BG 39 < 90, bolus snooze: eventual BG range 39-354; setting current basal of 0.9 as temp","COB":130,"eventualBG":39,"timestamp":"2017-05-22T10:07:15.522921","duration":30,"tick":"+0","IOB":8.034}
Old settings refreshed
temp_basal.json less than 5m old
Pumphistory less than 5m old
Settings less than 10m old
pumphistory-24h refreshed
Completed pump-loop at Mon May 22 14:20:09 MDT 2017
DaisyHK
@laurathiessen
May 22 2017 22:02
ok I have completed the phase 1 stuff of setting up the edison and it was working fine last night. I plugged it in this morning as I guessed the battery would be drained. Now when I plug it in the red light is on only for a few minutes then shuts off.... how can I figure out what is wrong?
Scott Leibrand
@scottleibrand
May 22 2017 22:02
which red light?
the one in the corner is the battery charging light. the one between the two microUSB ports is the power light for the Edison
tynbendad
@tynbendad
May 22 2017 22:04
i think maybe a case with reset button accessible to him is best course of action right now
Scott Leibrand
@scottleibrand
May 22 2017 22:05
yeah, being able to reboot if it's not working is helpful
tynbendad
@tynbendad
May 22 2017 22:10
looks like @danimaniac's case has accessible reset button, will try that
DaisyHK
@laurathiessen
May 22 2017 22:12
@scottleibrand ok i just disconnected the battery and reattached it and I think it is fine now.... if i push in that black pin that is the power bar will it d the same "reset" as I hate disconnecting the battery all the time
tynbendad
@tynbendad
May 22 2017 22:14
@laurathiessen i think you may have to hold it down for 10 seconds to power it off, then press it again for a sec to turn it on. not sure if it does anything when its just pressed once.
DaisyHK
@laurathiessen
May 22 2017 22:15
@tynbendad ok thanks will try that next time...
actually the device is pulled into the computer and both lights were on now the battery charging light is off
does it only light up when it needs charging?
tynbendad
@tynbendad
May 22 2017 22:19
its likely, i've never watched it that much ;)
DaisyHK
@laurathiessen
May 22 2017 22:20
also i accidentially did not sign in correctly, used PW edison as username and I cant figure out how to go back, ESC:Q does not undo it
tynbendad
@tynbendad
May 22 2017 22:20
exit
DaisyHK
@laurathiessen
May 22 2017 22:20
image.png
Martin Burke
@burmart_twitter
May 22 2017 22:20
Hi. New to this diabetes thing and I am definitely not at the point where one of these systems makes sense for me. I'm just wondering if there is a place for pictures that give a good estimate of how big an OPENAPS rig is. I see lots of pictures of Raspberry Pi ones, but don't think I've seen any for ones using edison. Are they about the same size?
DaisyHK
@laurathiessen
May 22 2017 22:21
@tynbendad ok... then how do you know it is working when it is not connected to anything? what light shows the rig is working?
tynbendad
@tynbendad
May 22 2017 22:21
they are small - are you familiar with rileylink size?
Dana Lewis
@danamlewis
May 22 2017 22:21
@burmart_twitter hi Martin! How about chapstick size? :D https://twitter.com/danamlewis/status/776248916077522944 (a lot smaller than Pi’s)
tynbendad
@tynbendad
May 22 2017 22:22
there ya go ;)
Martin Burke
@burmart_twitter
May 22 2017 22:22
GEEZ!
That looks a lot more appealing than the Pi versions I've seen pictures of.
tynbendad
@tynbendad
May 22 2017 22:23
@laurathiessen one light stays on when power is on, but i only know if its really working if i see it changing temp basal on nightscout
Martin Burke
@burmart_twitter
May 22 2017 22:23
Awesome.
DaisyHK
@laurathiessen
May 22 2017 22:25
@tynbendad ok then i have an issue with my rig as the light does not stay on :( how do i figure out what is wrong?
tynbendad
@tynbendad
May 22 2017 22:25
the middle light doesn't stay on?
DaisyHK
@laurathiessen
May 22 2017 22:27
oh never mind thought the middle red light only stayed on when it was showig that edison was being worked on ... nevermind if I disconnect it it stays light
the left one turns off though even when its plugged in but like you said that is not an issue
Scott Leibrand
@scottleibrand
May 22 2017 22:28
the charging LED in the corner only stays lit until the battery is fully charged
DaisyHK
@laurathiessen
May 22 2017 22:29
ok sorry thanks for the clarification
ok back t my original question about my putty and logging in incorrectly
Scott Leibrand
@scottleibrand
May 22 2017 22:30
I think someone answered that one: exit to log out
Martin Haeberli
@mhaeberli
May 22 2017 22:30
@tybendad - your stale BG data looks a bit familiar to me but I haven’t figured out mine, either
DaisyHK
@laurathiessen
May 22 2017 22:30
@scottleibrand oh sorry did not know that menat to input that into putty
ok thanks
Martin Haeberli
@mhaeberli
May 22 2017 22:32
@laurathiessen also note - depending on which explorer board version you have (1.2 works, but original is subtle), you may have to restart your edison from the power button on the explorer board if it shuts down all the way when its battery goes low. I understand that with 1.2 (that I have some of, but haven’t verified), it will turn on automatically when plugged back into power
DaisyHK
@laurathiessen
May 22 2017 22:33
if i only need to select a different time zone I can just input this command and follow the prmpts? dpkg-reconfigure tzdata
Scott Leibrand
@scottleibrand
May 22 2017 22:35
correct
tynbendad
@tynbendad
May 22 2017 22:40
@mhaeberli our rig was up on wifi fine all w/e, only seems to konk out occasionally when its on bt tether, so likely has something to do with that.
DaisyHK
@laurathiessen
May 22 2017 22:41
@mhaeberli how do i know which version i have?
Martin Haeberli
@mhaeberli
May 22 2017 22:41
@tynbendad great - we are having errors sometimes near our wifi, which can lead to stale bg data even when nightscout is live
@laurathiessen there is writing printed on one end of the board
@laurathiessen it says “1.2” or maybe more, on the 1.2 boards
when did you buy your board?
tynbendad
@tynbendad
May 22 2017 22:43
yeah, it either says "Explorer Block GMD 2016" or "... 2017 V1.2"
i just got a few of the older ones (using the new one), so i guess all the more reason to have power button accessible
Martin Haeberli
@mhaeberli
May 22 2017 22:43
:+1:
tynbendad
@tynbendad
May 22 2017 22:44
@danimaniac how do you hit reset button in this case, poke something thru a vent hole? https://github.com/danimaniac/OpenAPS-Explorer-Board-Edison-vented-case/tree/w/2500mAh-battery
DaisyHK
@laurathiessen
May 22 2017 22:44
ok i just checked 1.2, so with that one i need to rebooth from the power button when in qusestion?
Martin Haeberli
@mhaeberli
May 22 2017 22:45
@laurathiessen no - 1.2 should generally be fine, unless you need to force a reboot
tynbendad
@tynbendad
May 22 2017 22:45
i may need my kid to reset it, so would be great if it was easier (specific/clear hole), or a plastic piece built in that could be depressed... @rsilvers129
Martin Haeberli
@mhaeberli
May 22 2017 22:45
which you generally should be able to do, worst case, from the USB tether
i think i remember one of the vented case implementations had a way to push the button without opening it
tynbendad
@tynbendad
May 22 2017 22:46
yes, @danimaniac's cases have that
rsilvers129
@rsilvers129
May 22 2017 23:00
All of my enclosures have the Intel logo on the outside for cooling and to not heat the battery. That makes the reset button facing the battery. It would be nice if the next board revision had pins for a remote reset button.
garykidd
@garykidd
May 22 2017 23:01
curious if anyone may have heard anything about ways to view BGs on androidwear 2.0 devices on iOS? Really anxious to figure out a solution for Huawei Watch 2 on iOS.
DaisyHK
@laurathiessen
May 22 2017 23:01
i dont have OPENAPS all set up yet but the next steps in the documents are offline looping and monitoring. should i set this up with my pebble watch? I am going to use a medtronic 754 with a 2.7A, android galaxy s4, xdrip+ , dexcon g5, and my pebble watch. I dont think I can set up the xDripAPS until i have a working OpenAPS rig up and running...
Dana Lewis
@danamlewis
May 22 2017 23:08
correct. come back to that when you have basic rig running
glenrob1940
@glenrob1940
May 22 2017 23:10
Fired up a new explorer board and have an error 299 problem (excessive operator ignorance). I've got a time issue. I configured tzdata, and checked that nightscout was set correct (MDT) but I get the following error/warning about time
DaisyHK
@laurathiessen
May 22 2017 23:10
@danamlewis ok so then there is nohing to do until I want to actually set up my Open APS?
OpenAPS rig
glenrob1940
@glenrob1940
May 22 2017 23:11
''
Starting pump-loop at Mon May 22 17:03:02 MDT 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
BG data is too old (it's probably this), or clock set incorrectly. The last BG data was read at Mon May 22 2017 15:44:24 GMT-0600 (MDT) but your system time currently is Mon May 22 2017 17:04:07 GMT-0600 (MDT)
determine-basal://text/shell/enact/suggested.json
reporting enact/suggested.json
{"reason":"BG data is too old (it's probably this), or clock set incorrectly. The last BG data was read at Mon May 22 2017 15:44:24 GMT-0600 (MDT) but your system time currently is Mon May 22 2017 17:04:07 GMT-0600 (MDT)"}
{"reason":"BG data is too old (it's probably this), or clock set incorrectly. The last BG data was read at Mon May 22 2017 15:44:24 GMT-0600 (MDT) but your system time currently is Mon May 22 2017 17:04:07 GMT-0600 (MDT)"}
Waiting for ntpd to synchronize... OK!
{"bg":171,"temp":"absolute","snoozeBG":153,"recieved":true,"predBGs":{"IOB":[171,171,170,168,166,163,160,156,151,146,141,136,131,126,121,116,111,107,103,99,96,92,89,86,83,81,79,77,75,73,72,71,70,69]},"rate":1.85,"reason":"COB: undefined, Dev: 30, BGI: -5.04, ISF: 40, Target: 103; Eventual BG 77 < 100, bolus snooze: eventual BG range 77-153; setting current basal of 1.85 as temp. Setting neutral temp basal of 1.85U/hr","eventualBG":77,"timestamp":"2017-05-22T15:50:35.957073","duration":30,"tick":"+1","IOB":3.097}
pumphistory more than 5m old: Refresh^C
root@edison1:~/test63# date
Mon May 22 17:06:47 MDT 2017
root@edison1:~/test63#
"
Dana Lewis
@danamlewis
May 22 2017 23:11
@laurathiessen yes, you’ll do xdripAPS after you get the basic online rig setup so you know it’s working.
rsilvers129
@rsilvers129
May 22 2017 23:12
This may be a clue as to why my radio is not working on one rig. One one, the TTY is in a different color when using ls than the other.
Dana Lewis
@danamlewis
May 22 2017 23:12
@rsilvers129 did you read the logs and see those different time stamps?
rsilvers129
@rsilvers129
May 22 2017 23:12
Screen Shot 2017-05-22 at 7.10.57 PM.png
Screen Shot 2017-05-22 at 7.10.52 PM.png
DaisyHK
@laurathiessen
May 22 2017 23:13
@danamlewis ok i will be back on here in a few weeks then... cause I am off on a work trip for a week. thanks so much!
rsilvers129
@rsilvers129
May 22 2017 23:13
@danamlewis I am not sure what you mean.
Dana Lewis
@danamlewis
May 22 2017 23:14
@rsilvers129 the logs you posted show it’s not getting recent BG data : {"reason":"BG data is too old (it's probably this), or clock set incorrectly. The last BG data was read at Mon May 22 2017 15:44:24 GMT-0600 (MDT) but your system time currently is Mon May 22 2017 17:04:07 GMT-0600 (MDT)"}
ah, wrong person’s logs :D sorry @rsilvers129
ya’ll have the same color no-picture avatar here :D
glenrob1940
@glenrob1940
May 22 2017 23:15
Nightscout is showing the same time 17:04 so I don't know where the old data is coming from and/or lack of new data
Scott Leibrand
@scottleibrand
May 22 2017 23:15
@glenrob1940 lol @ error 299 btw ;-)
@glenrob1940 in addition to NS, what kind of CGM setup did you configure?
for some reason, your glucose.json isn't getting updated, and still has stale data from 15:44 MDT.
glenrob1940
@glenrob1940
May 22 2017 23:18
It's a dexcom G5 with the dexcom ios App,
Scott Leibrand
@scottleibrand
May 22 2017 23:18
have you ever plugged in the G5 receiver?
glenrob1940
@glenrob1940
May 22 2017 23:19
Yes, I have
Both the G5 receiver and nightscout are showing the same BG value
The ios App is showing the identical value also
Should I try blowing away glucose.json?
Scott Leibrand
@scottleibrand
May 22 2017 23:23
I would try plugging in or unplugging the receiver and see if that kicks it into gear
glenrob1940
@glenrob1940
May 22 2017 23:24
I'm not sure what you mean,
Should I cycle power on the receiver?
Scott Leibrand
@scottleibrand
May 22 2017 23:25
no, I'm talking about plugging it into your rig via USB OTG
have you ever done that, or are you just doing it via NS?
glenrob1940
@glenrob1940
May 22 2017 23:25
Oh, okay and with that info, no I have never plugged in via OTG - I will try that now
Scott Leibrand
@scottleibrand
May 22 2017 23:25
you'll need a special cable or adapter you probably don't have yet
so in that case, we should focus on why the ns-loop isn't downloading CGM data from NS. is the rig full online? can you ping google.com?
DaisyHK
@laurathiessen
May 22 2017 23:27
@danamlewis i thought there was a way to have OPENAPS running in the background but for you to still decide what to do.... is there more info on this somewhere?
glenrob1940
@glenrob1940
May 22 2017 23:28
Yes ping works fine. I'll go scrounge around cables - My old explorer board OTG never worked reliably so it's been awhile since I had cables in one spot.
garykidd
@garykidd
May 22 2017 23:30
@scottleibrand was wondering what the approach would be if community discovered a serious error in code. I know checking back in here for one would surface these types of issues. Was wondering if an email could automatically be generated or some other pushover notification or the like?
Dana Lewis
@danamlewis
May 22 2017 23:31
@garykidd we have a safety-critical google group that people sign up for that would be used to email out, in addition to posting here, to openaps-dev, on Facebook, and on Twitter
garykidd
@garykidd
May 22 2017 23:32
that's quite reassuring.
Dana Lewis
@danamlewis
May 22 2017 23:32
This group, FYI, in case anyone somehow skipped adding themselves when they filled out the “I’m now looping” form: https://groups.google.com/forum/#!forum/openaps-loopers (There are no messages that have been sent out; so you should just be able to add yourself.).
(This page in the docs http://openaps.readthedocs.io/en/dev/docs/walkthrough/phase-4/keeping-up-to-date.html links to the now-looping form, which has the link for the safety-critical notification group, and a reminder also for those who are using Loopkit/Loop to also sign up for that safety notification list if they are using both systems)
garykidd
@garykidd
May 22 2017 23:35
I actually wasnt member of fist but I'm 99% sure I filled out the second link form...thx
Dana Lewis
@danamlewis
May 22 2017 23:37
you joined Nov. 2, 2016
garykidd
@garykidd
May 22 2017 23:38
It introduces a lot of potential issues but if there was a clear error other known major issue...I'd be open to it updating on it's own....but recognizing the issues here...it still might not be a bad idea to automate a response to alert users via required settings upon running oref0 or 1 at inception...just an idea...
@danamlewis thx
Dana Lewis
@danamlewis
May 22 2017 23:38
we don’t have a way to do that, so email and all the other communication channels is what we plan to use
Scott Leibrand
@scottleibrand
May 22 2017 23:44
We set up the safety-critical Google group email list after we discovered the y2k16 bug (where a date field didn't have enough bits, so it broke on the rollover to Jan 1 2016). That bug caused OpenAPS to fail to calculate IOB, so it ended up looping solely based on the CGM readings without regard to how much insulin had already been delivered. While the resulting oscillations were suboptimal, it turns out you can actually keep overnight BG in range even if all you have is the CGM data: it's just not quite as flat as with the proper IOB calculations to dampen things. Since we've done that, we haven't had any other serious bugs that we've needed to alert people about.
garykidd
@garykidd
May 22 2017 23:47
@danamlewis understand that there's no way to do so now; was just offering a thought on a potential feature. There's a lot to think about and it seems that most of the work done here is concentrated among a pretty small group of folks to whom I remain incredibly grateful. It's far easier to offer up ideas without considering all the ramifications and constraints I realize. Thanks much.
glenrob1940
@glenrob1940
May 22 2017 23:49
I tried plugging G5 receiver, with a couple of cable combinations - none of which appeared to work as nothing showed up in lsusb. A listing of /monitor follows and clearly glucose.json is about half an hour behind.
''
root@edison1:~/test63/monitor# ls -al
total 72
drwxr-xr-x 2 root root 4096 May 22 15:44 .
drwxr-xr-x 9 root root 4096 Apr 28 21:56 ..
-rw-r--r-- 1 root root 43 May 22 17:41 battery.json
-rw-r--r-- 1 root root 0 May 22 17:45 carbhistory.json
-rw-r--r-- 1 root root 21 May 22 17:41 clock.json
-rw-r--r-- 1 root root 27 May 22 17:41 clock-zoned.json
-rw-r--r-- 1 root root 37 May 22 17:45 edison-battery.json
-rw-r--r-- 1 root root 6501 May 22 17:28 glucose.json
-rw-r--r-- 1 root root 7709 May 22 17:41 iob.json
-rw-r--r-- 1 root root 81 May 22 17:41 meal.json
-rw-r--r-- 1 root root 81 May 22 17:45 meal.json.new
-rw-r--r-- 1 root root 1370 May 22 16:51 mmtune.json
-rw-r--r-- 1 root root 3501 May 22 17:41 pumphistory.json
-rw-r--r-- 1 root root 3561 May 22 17:41 pumphistory-zoned.json
-rw-r--r-- 1 root root 4 May 22 17:41 reservoir.json
-rw-r--r-- 1 root root 69 May 22 17:41 status.json
-rw-r--r-- 1 root root 61 May 22 17:41 temp_basal.json
root@edison1:~/test63/monitor#
So, .. should I blow away glucose.json next?
garykidd
@garykidd
May 22 2017 23:50
@scottleibrand How much is known on how MM 670 algorithm works...realize this is rather open ended but I recall you sharing certain information on how it works...any high level thoughts you could share?
Scott Leibrand
@scottleibrand
May 22 2017 23:51
@glenrob1940 if there's nothing obvious in the ns-loop.log, I would comment out the ns-loop line in cron and try running it (and its components) manually to see if you can find the issue.
glenrob1940
@glenrob1940
May 22 2017 23:51
Okay, will try that
Scott Leibrand
@scottleibrand
May 22 2017 23:52
@garykidd not a lot, at least compared to what we'd like to know. are you trying to figure out something specific?
garykidd
@garykidd
May 22 2017 23:57
I seem to recall (a long while ago) that you may have commented about the MM formula is less responsive to BG rising in certain circumstances...
maybe that it wouldnt react to COB impact from a meal when BG was lower than target--maybe it just decreases basal no matter what?