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

18th
Jan 2017
JaysonEwer
@JaysonEwer
Jan 18 2017 00:18
@scottleibrand Ok, I guess they aren't "configs" but there are a few files that you manually modify, right? Do you have a way to keep all your rigs on the same settings or do you manually change those files on each of your rigs? Sorry for the late reply.
fjungman
@fjungman
Jan 18 2017 00:20
Has anyone looked at the Medtronic 751-NAS (or 551) pump? The S/W version is 1.1F 1.1 0B 0B , which sounds low enough, if it is based upon the same code base as the 722/522 line. Please let me know if you have heard of these working.
j4cbo
@j4cbo
Jan 18 2017 00:24
what year is it from? I wonder if they just reset the version numbers... my 551 came with 3.0-something
Scott Leibrand
@scottleibrand
Jan 18 2017 00:33
@JaysonEwer yeah, there are some customizations I have in my crontab, and personal Linux environment stuff, that I have to set up each time I set up a new rig. But I don't change those much: pretty much all the development etc. happens in the oref0 repo. That's actually one of the reasons I did oref0-setup: I was tired of having to manually keep all my different setups the same. :)
@fjungman do you have access to go into the pump's Utilities menu and look for the PC Connect option under Connect Devices? if that's present, it won't be able to loop, but if it's absent, it should be good.
fjungman
@fjungman
Jan 18 2017 00:49
Yep, it has PC Connect, so no-go. This is for a friend, who is now shopping for a 722... :)
Scott Leibrand
@scottleibrand
Jan 18 2017 00:50
bummer. glad it was easy to rule out though.
fjungman
@fjungman
Jan 18 2017 00:50
@j4cbo I don't know how to tell the year, but there is a number on the back "6012" which might suggest 2012...
Yeah, you have this down to a science, Scott!
Aaron Michelson
@Aaybob
Jan 18 2017 00:55
The only difference I see from home and away is that I'm using battery power. So I'm trying a new battery brand.
Aaron Michelson
@Aaybob
Jan 18 2017 01:07
how does one get this file ? is it an "only in dev" thing? monitor/edison-battery.json
Dana Lewis
@danamlewis
Jan 18 2017 01:10
I think it’s still in dev
related, that needs to be added to the docs (I think) (unless @cjo20 comes in and says otherwise, I don’t remember it being added but I do remember seeing gist’s floating around before it went into dev with setup)
Chris Oattes
@cjo20
Jan 18 2017 01:11
It's part of oref0-setup now
Dana Lewis
@danamlewis
Jan 18 2017 01:12
master, though? i thought it was still dev
Chris Oattes
@cjo20
Jan 18 2017 01:12
Looks like dev only at the moment
Aaron Michelson
@Aaybob
Jan 18 2017 01:13
:worried: Is there good reason not to go to dev?
Dana Lewis
@danamlewis
Jan 18 2017 01:14
there’s a lot of stuff like bluetooth and other things that are still needing a lot of love and testing…so depends on what you’re going to use. if just basics, it’s probably stable enough. but generally I wouldn’t do dev unless you have multiple rigs, and can keep at least one rig on master.
Aaron Michelson
@Aaybob
Jan 18 2017 01:16
I just booted on my third rig. My first one I burned so it can't be powered by battery but powered by serial port I just found it works.
Rigs haven't worked away from home using master.
Chris Oattes
@cjo20
Jan 18 2017 01:18
@Aaybob Do you run your loop as root?
Aaron Michelson
@Aaybob
Jan 18 2017 01:18
yes
or that would be in cron
/root/myopenaps is in cron.....everything is as root
Chris Oattes
@cjo20
Jan 18 2017 01:24

A simple way of getting it working

cd ~/src
git clone https://github.com/cjo20/EdisonVoltage
cd EdisonVoltage
make voltage

Then you need to add a line in cron that runs ~/src/EdisonVoltage/voltage batteryVoltage battery > /root/myopenaps/monitor/edison-battery.json every 5 minutes

Aaron Michelson
@Aaybob
Jan 18 2017 01:25
thnks
Chris Oattes
@cjo20
Jan 18 2017 01:25
That should generate the file anyway
Aaron Michelson
@Aaybob
Jan 18 2017 01:25
right i have the grep code to shutdown....
Chris Oattes
@cjo20
Jan 18 2017 01:27
I'd just get it generating the file reliably first
If you start playing around with shutdown before you get it working, it might end up shutting down immediately, which might make it difficult to remove :P
Aaron Michelson
@Aaybob
Jan 18 2017 01:31
What would be the most sure way to go mobile? dev-ble, dev-wifi, master wifi?
Chris Oattes
@cjo20
Jan 18 2017 01:31
Hope the battery stuff works
cjo20 @cjo20 -> bed
Aaron Michelson
@Aaybob
Jan 18 2017 01:32
g'night
j4cbo
@j4cbo
Jan 18 2017 02:11
today is a good day
my explorer board arrived and Dexcom processed my order
Dana Lewis
@danamlewis
Jan 18 2017 02:12
double woot! :) diabetes christmas!
;p
jmcrawford
@jmcrawford
Jan 18 2017 02:28
Would an Edison disable wifi/BT if the battery was running low, as a way of prolonging its life?
Scott Leibrand
@scottleibrand
Jan 18 2017 02:29
not by default
jmcrawford
@jmcrawford
Jan 18 2017 02:46
Ok, just trying to troubleshoot. Lost BT connectivity out of the blue, I noticed that the battery was reporting 10%. Might not be related. It picked up wifi as soon as we got home.
j4cbo
@j4cbo
Jan 18 2017 02:49
i'm getting... really weird network connectivity to the edison
64 bytes from 192.168.8.197: icmp_seq=29 ttl=64 time=11.831 ms 64 bytes from 192.168.8.197: icmp_seq=30 ttl=64 time=1192.005 ms 64 bytes from 192.168.8.197: icmp_seq=31 ttl=64 time=191.388 ms 64 bytes from 192.168.8.197: icmp_seq=32 ttl=64 time=1512.458 ms 64 bytes from 192.168.8.197: icmp_seq=33 ttl=64 time=507.345 ms
incredibly glitchy ssh
ah. fixed by iwconfig wlan0 power off
Scott Leibrand
@scottleibrand
Jan 18 2017 03:44
ifconfig?
j4cbo
@j4cbo
Jan 18 2017 03:45
iwconfig
Scott Leibrand
@scottleibrand
Jan 18 2017 03:45
huh, k.
j4cbo
@j4cbo
Jan 18 2017 03:46
i'm running into the same spiFromDesc issue as @tim2000s
Eric Luhrs
@eluhrs
Jan 18 2017 03:49

Hi, all. I made some progress getting the latest dev version setup on an Edison/explorer board rig last night. oref0-setup completes successfully with all advanced options enabled, and the loop is running in cron. However, I noticed that three of my report output files are empty: settings/autosens.json, settings/profile.json, and monitor/iob.json. Some comments above, between @C-Ville and @scottleibrand suggested running the profile.json report manually. Doing that yields the following error: ``` root@wopr-ee1:~/wopr# openaps-report invoke settings/profile.json
get-profile://text/shell/settings/profile.json
/usr/local/bin/oref0-get-profile pump_settings.json bg_targets.json insulin_sensitivities.json basal_profile.json [preferences.json] [<carb_ratios.json>] [<temptargets.json>] [--model model.json]

Options:
--model, -m Pump model response [default: false]
--help Show help [boolean]
--exportDefaults Show default preference values [default: false]
--updatePreferences Check for any keys missing from current prefs and add from defaults [default: false]

Unknown argument: autotune
reporting settings/profile.json
root@wopr-ee1:~/wopr#

Scott Leibrand
@scottleibrand
Jan 18 2017 03:49
@eluhrs you have an old oref0 somehow
npm run global-install in ~/src/oref0/ should update it
was this a completely fresh install, or did it have oref0 already?
Eric Luhrs
@eluhrs
Jan 18 2017 03:50
It was a new install last night. Entirely new OS and hardware.
Scott Leibrand
@scottleibrand
Jan 18 2017 03:50
hmm, that is concerning
what does git branch in ~/src/oref0/ show?
Eric Luhrs
@eluhrs
Jan 18 2017 03:51
My days and nights have been running together lately, so I suppose it is possible that it was the previous day. But I don't think so.
dev
Scott Leibrand
@scottleibrand
Jan 18 2017 03:52
how about npm list -g oref0 ?
Eric Luhrs
@eluhrs
Jan 18 2017 03:53
I already started the global-install (which reported oref0@0.3.6-dev).
Scott Leibrand
@scottleibrand
Jan 18 2017 03:54
k
I guess someone will need to do a complete reimage and install to see what is up with the install before we can merge to master
j4cbo
@j4cbo
Jan 18 2017 03:56
for posterity: i fixed the spiFromDesc issue by going into ~/src/mraa/build and re-doing make && make install, and then deleting /usr/local/lib/i386-linux-gnu/python2.7/site-packages/*mraa*
Eric Luhrs
@eluhrs
Jan 18 2017 03:56
Should I re-run oref0-setup again or will cron just pickup any updates that were applied?
j4cbo
@j4cbo
Jan 18 2017 03:56
(the correct mraa appears to be in /usr/local/lib/python2.7/dist-packages/)
Scott Leibrand
@scottleibrand
Jan 18 2017 03:56
@eluhrs I think that one will get picked up without re-running oref0-setup
thx @j4cbo
j4cbo
@j4cbo
Jan 18 2017 04:02
ok, now getting Could not get subg_rfspy state or version. Have you got the right port/device and radio_type?
Eric Luhrs
@eluhrs
Jan 18 2017 04:02
Thanks, @scottleibrand. Now I have non-empty profile.json, iob.json, and autosens.json reports, however running the profile.json report gives me this warning: Could not parse autotune_data. Does that just mean that I haven't generated enough data for analysis?
Scott Leibrand
@scottleibrand
Jan 18 2017 04:03
your nightly cron hasn't run yet
you can run it manually the first time
j4cbo
@j4cbo
Jan 18 2017 04:03
silly question, but the explorer board ships with subg_rfspy already installed on the cc1110... right?
Scott Leibrand
@scottleibrand
Jan 18 2017 04:03
it always has in the past, but I'm not 100% sure about the latest groupgets batch
if not, you can flash it from the Edison with ccprog
j4cbo
@j4cbo
Jan 18 2017 04:06
alright, i'll try reflashing it... assuming we want spi1_alt2? or is there a prebuilt binary around?
wait, never mind, found the binary (linked from https://github.com/EnhancedRadioDevices/915MHzEdisonExplorer_SW)
success!
Eric Luhrs
@eluhrs
Jan 18 2017 04:15
@scottleibrand Thanks. Running the nightly autotune manually worked. Is there any documentation on how to read the autotune.log file? Am I correct to believe that I need to make any changes manually to the pump? Or is there a way to remotely update pump settings?
Scott Leibrand
@scottleibrand
Jan 18 2017 04:21
the cron job automatically updates what the rig uses, so if you're looping 24/7 that means you don't need to change the pump until autotune starts bumping up against the 1.2x and 0.7x limits, unless you're going to go off the loop
@danamlewis and I have just now gotten to that point with a few hours of the basal schedule, so we'll probably be adjusting a few basals soon on the pump, and maybe raising ISF
Eric Luhrs
@eluhrs
Jan 18 2017 04:24
I don't follow the part about the 1.2x and 0.7x limits. What are they?
Scott Leibrand
@scottleibrand
Jan 18 2017 04:25
the autosens min and max in preferences.json are also used by autotune to determine how far it's allowed to adjust things from the pumpprofile.
those default to 0.7 (30% more sensitive) and 1.2 (20% more resistant)
j4cbo
@j4cbo
Jan 18 2017 04:26
that's interesting... after rebooting the edison it's gone back to Could not get subg_rfspy state or version. Have you got the right port/device and radio_type?
Eric Luhrs
@eluhrs
Jan 18 2017 04:26
Oh, right. I'll read up more now that I've got it working.
Scott Leibrand
@scottleibrand
Jan 18 2017 04:28
@j4cbo did the red lights blink and go out?
(they should blink whenever it runs reset_spi_serial.py, and stay off. dimly lit means a cc1110 firmware problem)
mmtune runs reset_spi_serial.py before doing anything else, so if you're watching pump-loop.log it should already be doing that
j4cbo
@j4cbo
Jan 18 2017 04:31
hm let me try again. i think they didn't blink at all
they went dim the first time around, before i had programmed the firmware at all
@scottleibrand fresh boot, no crontab, first command i run is openaps use pump mmtune: same error, no LED activity
Scott Leibrand
@scottleibrand
Jan 18 2017 04:36
sounds like it's time to re-run cc-prog. did you erase before writing?
j4cbo
@j4cbo
Jan 18 2017 04:37
i did
Scott Leibrand
@scottleibrand
Jan 18 2017 04:38
not sure then, but I would still try it again. which hex file did you use?
j4cbo
@j4cbo
Jan 18 2017 04:38
if i run ccprog again it starts working. i wonder if the edison's default gpio state is holding the cc1110 in reset or something?
i used spi1_alt2_EDISON_EXPLORER_US_STDLOC.hex per instructions at https://github.com/EnhancedRadioDevices/915MHzEdisonExplorer_SW
Eric Luhrs
@eluhrs
Jan 18 2017 04:42
Is it the OTG port that charges the Explorer? And does it do pass-through charging?
Scott Leibrand
@scottleibrand
Jan 18 2017 05:27
either port can be used for charging, and the lipo can charge without shutting down the Edison, yes
j4cbo
@j4cbo
Jan 18 2017 05:31
i haven't gotten around to getting a lipo yet, so i tried running it from a big USB charger brick
and... it draws too little power :P the brick does its auto-shutoff thing
good problems to have, i guess
Scott Leibrand
@scottleibrand
Jan 18 2017 05:37
heh ya
PieterGit
@PieterGit
Jan 18 2017 07:38
@scottleibrand @danamlewis . What are the features that need love and attention before releasing dev to master? It would be good to have a more recent master that is considered stable. Currently dev grows new features every week so the gap between dev and master grows. This will probably cause newbies to start with dev instead of master
Scott Leibrand
@scottleibrand
Jan 18 2017 07:39
yep, that seems to happen every release cycle
first thing would be to go through everything and look/test for regressions affecting people who don't enable the optional features. then we need to decide which optional features are well tested enough to make available in master, and which ones need to be removed from the interactive setup script and only available to people who know to look for them in the code and find the command line flag...
samihusseingit
@samihusseingit
Jan 18 2017 07:42
for AAPS ad OpenAPS, we must enter each carb grams taken?
Scott Leibrand
@scottleibrand
Jan 18 2017 07:42
advanced meal assist will only work properly if it knows about your carbs
samihusseingit
@samihusseingit
Jan 18 2017 07:42
ok
thanks
Ryan Dalisky
@rdalisky
Jan 18 2017 12:00
Has anyone replaced the battery inside a medtronic enlite transmitter?
OrlaWilson
@OrlaWilson
Jan 18 2017 12:04
Hi. Im on Day 3 using my new edison rig but keep running into a Nightscout Error 403, Im using Azure. CPU Day quota exceeded. Assume its becuase the edison/explorer board is pulling all the data from NS. Is this just an azure issue? Do I have to start paying to get a bigger daily quota?:(
Tim Street
@tim2000s
Jan 18 2017 12:07
@ceben80 @colinlennon I'm not seeing xDrip data on the edi. Just want to check that the xDrip settings are correct. I'm using: https://NS-passcode@hostname.local:5000/api/v1/
@OrlaWilson That's an issue with running OpenAPS or Loop.
Best thing is to set up a heroku Nightscout instance.
OrlaWilson
@OrlaWilson
Jan 18 2017 12:26
@tim2000s Thanks, I wondered if that would be the case. I guess thats what Ill be doing in my spare time today, then:)
Matthias
@ceben80
Jan 18 2017 12:30
@tim2000s are u using the hostname or IP to connect?
I was not able to establish a connection per hostname, but IP works
Tim Street
@tim2000s
Jan 18 2017 12:36
Thanks @ceben80 -just realised when I tried to ssh from android. Should be interesting now...
Tim Street
@tim2000s
Jan 18 2017 12:50
I can see inbound messages on the tcpdump!
Matthias
@ceben80
Jan 18 2017 13:16
@tim2000s for me very annoying with hardcoded IPs due to the fact my huwai sometimes DCHP different IPs and static definitions MAC to IP is not possible with this router :worried:
Tim Street
@tim2000s
Jan 18 2017 13:17
It's not updating the xDrip/glucose.json file. 🙁
Matthias
@ceben80
Jan 18 2017 13:20
wtf
maybe Colin @colinlennon has an idea :wink:
Brian Rabinovitz
@bdr1177
Jan 18 2017 14:27
Does COB matter for anything other than advanced meal assist? I'm having a problem with my NS site right now and the openaps log a reporting carbs as undefined.
Brian Rabinovitz
@bdr1177
Jan 18 2017 14:39
"is" reporting
Eric Luhrs
@eluhrs
Jan 18 2017 15:30
I am having some heat issues with my explorer board. It gets MUCH hotter than my old intel breakout board. Almost too hot to touch. While installing it, I would get console messages about CPU 0 and CPU 1 being throttled due to excessive heat. During heavy use, such as when doing dist-upgrade, it would periodically reboot. I'm not sure how to confirm, but I suspect it was due to heat. Now that I have 3.6-dev installed, it reboots almost repeatedly (which makes it tough to look at log files). I have tried two different explorer boards and two different compute modules with similar results. I see that @dougtakeuti and @scottleibrand had a conversation about heat issues back on Jan 2, but I couldn't find a resolution. Any suggestions gratefully accepted.
Chris Oattes
@cjo20
Jan 18 2017 15:33
@eluhrs: how hot does it actually get?
sudo apt-get install lm-sensors to install the software, sensors to report the values
Eric Luhrs
@eluhrs
Jan 18 2017 15:53
Thanks, @cjo20. Currently only 81/80C. But the only way I could get it to stop rebooting was to run from battery. Had been running from USB previously. I think it was hotter before, but I can log in to run sensors before it reboots....
"can't" log in
Starting to think the problem might be related to USB. I plugged back in to monitor temp until reboot. It had gone down to 72C while unplugged, but then steadily increased to 81C before it rebooted....
Eric Luhrs
@eluhrs
Jan 18 2017 16:02
I managed to login between reboots, and found that temp was 89C. The critical temp is set to 90, so I think it must be rebooting whenever it reaches 90C.
Eric
@ecc1
Jan 18 2017 16:06
That is way too hot . My module on an Intel mini-breakout is running at 43°C core temps
Does it follow the compute module (if it's on a Sparkfun or Intel breakout board)?
Chris Oattes
@cjo20
Jan 18 2017 16:18
My edison is @ 66C, plugged in to UART
Eric Luhrs
@eluhrs
Jan 18 2017 16:19
@cjo20 Is that on an explorer board or some other breakout?
@ecc1 I can't check now, but I have tried more than one compute module on this explorer.
Chris Oattes
@cjo20
Jan 18 2017 16:20
On an alpha explorer board
Eric Luhrs
@eluhrs
Jan 18 2017 16:21
Interesting, @cjo20. That's about where my board is when it is UNPLUGGED. Once I hook it up to either USB, the temp climbs.
I think I should try a different computer. I made a custom USB mini cable for my ReilyLink, but I think I caused a short when I forgot to wrap Kaptan around the soldered pins. That USB port not longer works on my iMac. I wonder if I caused damage to the others.....
jmcrawford
@jmcrawford
Jan 18 2017 16:27
@eluhrs Are you charging a battery? More current == more heat
Eric Luhrs
@eluhrs
Jan 18 2017 16:27
Yes, probably. How do I read battery voltage from the shell?
jmcrawford
@jmcrawford
Jan 18 2017 16:28
check monitor/edison-battery.json
Eric Luhrs
@eluhrs
Jan 18 2017 16:28
is there a shell command?
the json file shows 72 percent, so yes, it is pulling a charge.
Even unplugged, the temp has now passed 70c.
Matthias
@ceben80
Jan 18 2017 16:32
@eluhrs for display EdisonVoltage I have append monitor/mmtune.json --uploader monitor/edison-battery.jsonto format-ns-status
format-ns-status = ! bash -c "ns-status monitor/clock-zoned.json monitor/iob.json enact/suggested.json enact/enacted.json monitor/battery.json monitor/reservoir.json monitor/status.json monitor/mmtune.json --uploader monitor/edison-battery.json > upload/ns-status.json"
shown in NS as follows
blob
Eric Luhrs
@eluhrs
Jan 18 2017 16:34
Thanks, @ceben80. What happens if one value is low? Do they both appear red?
Actually, I guess they do. 158.4u doesn't seem too low. :)
Matthias
@ceben80
Jan 18 2017 16:36
it's the pump battery
needs to be replaced :smile:
EdisonVoltage is shown when mouse over the battery symbol (where you can see 44%)
back to your q @eluhrs noting happens when the voltage is low
I have build in a cron line to shutdown my Eddie proper when 3200 mV is reached
Tim Street
@tim2000s
Jan 18 2017 16:47
@eluhrs I was testing charging earlier. USB connected, I hit 66. Wireless rig uncased got to 68, wireless in the case got up from 52 to 75 in about 4 mins so I stopped.
Eric Luhrs
@eluhrs
Jan 18 2017 16:51
Hmmm. So that could be a problem. At least if you want it to be portable. :(
My temporary solution is to leave it on my window sill. It is a leaky old window, so now the temp is down to 40c.
Brian Rabinovitz
@bdr1177
Jan 18 2017 16:57
Sorry to ask again but don't think anyone answered. Does openaps use cob for anything other than meal assist?
Dana Lewis
@danamlewis
Jan 18 2017 16:58
@bdr1177 openaps will calculate the observed cob for advanced meal assist, rather than going with a pre-defined absorption rate as we had originally done.
but I think you’re asking for a reason - what kind of trouble are you having with your NS? Is it everything, or just COB reporting wonky? And, how do you usually enter carbs - via NS Care Portal, or pump bolus wizard?
Brian Rabinovitz
@bdr1177
Jan 18 2017 17:15
Yes, cob is not showing up on NS site, regardless of how the carbs are entered (care portal or pump bolus). The openaps basal cob is showing up fine. I'm troubleshooting the NS issue with the NS folks right now (Kate Farnsworth and Amy Cowen) who are both stumped. I'll probably just rebuild my NS site from scratch. I'm not planning on using meal assist right now and my understanding is that cob won't otherwise affect the openaps calculations and I just wanted to confirm that. We're still in testing phase right now (taking it slow) so I just want to get everything sorted out before trying to go live.
Dana Lewis
@danamlewis
Jan 18 2017 17:15
yea, it won’t effect you if you’re not using meal assist or advanced meal assist.
are you using azure or heroku?
Brian Rabinovitz
@bdr1177
Jan 18 2017 17:17
I switched over to heroku for openaps, but I still have an old azure site as well. The cob problem occurs with care portal on both sites. They share an mlab database, so I'm hoping nothing is corrupted there because I don't want to lose the NS history.
Dana Lewis
@danamlewis
Jan 18 2017 17:18
nah, that shouldn’t matter. (just checking because if you weren’t already on heroku and were going to mess with rebuilding sites, I would say kill 2 birds with one stone, but you’re already there so doesn’t matter)
Brian Rabinovitz
@bdr1177
Jan 18 2017 17:20
okay, thanks!
Dana Lewis
@danamlewis
Jan 18 2017 17:20
(also after you rebuild and if it’s still not working and want help troubleshooting, starting sharing screenshots & we can try to help, too.)
Brian Rabinovitz
@bdr1177
Jan 18 2017 17:31
@danamlewis Thanks so much. Will do.
j4cbo
@j4cbo
Jan 18 2017 17:34
first night with the loop running, woke up at 110 :)
Dana Lewis
@danamlewis
Jan 18 2017 17:35
:smile: :confetti_ball: :clap:
j4cbo
@j4cbo
Jan 18 2017 17:35
well, actually 130, but the enlite thought I was 110 :/
Tim Street
@tim2000s
Jan 18 2017 17:41
@j4cbo don't worry. I regularly wake up at what Dex tells me is 100 and blood says is 115...
JELCRAWFORD
@JELCRAWFORD
Jan 18 2017 17:51
@bdr1177 - just curious what COB issues you are seeing - we have periodically had some weird numbers on COB - have never been able to really figure it out (ie. suddenly showing 90 carbs at 11pm at night on my 10yo - he'd been asleep for 3 hours and supper had been 5 hrs before). New build on Edison doesn't seem to have the same issues - at least we've not seen any as yet...
Dana Lewis
@danamlewis
Jan 18 2017 17:52
@JELCRAWFORD on AMA, MA, or not?
and was that from NS COB pill, or as reported by OpenAPS?
JELCRAWFORD
@JELCRAWFORD
Jan 18 2017 17:54
Yes to both - carbs were consistent between OpenAPS and NS COB. When we had the really weird numbers, we turned off AMA as with Scott's help we realized it APS wouldn't low temp bolus b/c it saw so many carbs onboard. We have watched since and not seen it recur (on the Edison now). We do have a rather large spike in the evenings and we wondered if it was possible that the APS was assuming that the only way we could have such a massive increase was to assume big carb load. I think we have some screen shots around from that night... lemme see...
Dana Lewis
@danamlewis
Jan 18 2017 17:55
@JELCRAWFORD gotcha - was wondering if that might be what was happening or if it was another fluke…but it shouldn’t magically appear/assign carbs. (unless in the future w/ the floating carb feature, but that’s not relevant here). last q (I think) - was this old rig with setup scripts, or from before setup scripts?
Brian Rabinovitz
@bdr1177
Jan 18 2017 17:56
I think my issue affects both NS and openaps. The NS COB pill stays at zero no matter what. And the openaps log shows COB as "undefined," which I assume means that it can't get a number. The COB occasionally pops up in NS and I'm trying to figure out what triggers that. Also, at least some of the time, the NS BWP seems to be able to see the COB when she is low and it is recommending a carb correction. Doesn't matter if the carbs are entered in care portal or through the pump and everything is showing up correctly in the care portal treatments.
Dana Lewis
@danamlewis
Jan 18 2017 17:56
(NS COB is calculated separately, btw, but it’s helpful to ID the problem to know if it’s getting data or not)
wondering if ns cob only shows up when you do care portal
JELCRAWFORD
@JELCRAWFORD
Jan 18 2017 17:57
Mike says - All set-up scripts...
Dana Lewis
@danamlewis
Jan 18 2017 17:58
@JELCRAWFORD k, thanks
Brian Rabinovitz
@bdr1177
Jan 18 2017 17:58
@danamlewis Not quite sure what you mean, unless you mean that the openaps log showing cob as "undefined" is unrelated to what NS shows?
Dana Lewis
@danamlewis
Jan 18 2017 17:59
NS COB is calculated separately from openaps pill related cob
but if you sometimes see NS COB that’s probably from when you entered carbs in NS care portal directly
jmcrawford
@jmcrawford
Jan 18 2017 17:59
@bdr1177 When we had no COB shown in either NS or APS, it was due to not running AMA. Once we turned that back on, we had COB shown in both.
Dana Lewis
@danamlewis
Jan 18 2017 17:59
which means carbs aren’t coming in from the pump somehow, but if your loop is otherwise working, that doesn’t quite seem right, so there might be something going on
JELCRAWFORD
@JELCRAWFORD
Jan 18 2017 18:00
IMG_0288.PNG
Ok - here are a few shots of various problems. 1st image is of our night of 'magical' carb ( I swear this is why I never lose weight - it's the magical carbs...!)
Brian Rabinovitz
@bdr1177
Jan 18 2017 18:01
@danamlewis But the carbs from the pump do show up in the care portal treatment reports and on the NS graph itself. It's just the pill that doesn't show up.
@jmcrawford That's interesting. Maybe I'll try to run AMA during our testing and see if that changes things.
Dana Lewis
@danamlewis
Jan 18 2017 18:02
@bdr1177 may be as @jmcrawford suggested re: AMA missing
:+1:
JELCRAWFORD
@JELCRAWFORD
Jan 18 2017 18:02
Also had a few times where the carb total seemed weird - it seemed as though it was totalling the entire day... I can't say for 100% certainty on this one that it was consistent between OpenAPS and NS though - I'm pretty sure, but not totally sure..... next photo coming...
Brian Rabinovitz
@bdr1177
Jan 18 2017 18:03
@danamlewis So does the openaps log showing COB as "undefined" also indicate a problem? I understand that it won't actually affect calculations unless we're using AMA.
Dana Lewis
@danamlewis
Jan 18 2017 18:04
it won’t effect the basic loop running, no, but if you have AMA and getting undefined it usually means there’s a gap somewhere
JELCRAWFORD
@JELCRAWFORD
Jan 18 2017 18:04
IMG_0285.PNG
Brian Rabinovitz
@bdr1177
Jan 18 2017 18:05
Ok. I'll play around with the AMA tonight and see if that changes anything. Don't have the aps rig here with me.
JELCRAWFORD
@JELCRAWFORD
Jan 18 2017 18:05
IMG_0284.PNG
One more of magical carb - I swear we do not let him midnight snack and we'd know if he was... (note time 1:12 am) - less significant re: carb, but still appeared for some weird reason...
JELCRAWFORD
@JELCRAWFORD
Jan 18 2017 18:16
@danamlewis we do not enter any carb into NS that also goes through the pump. We enter carb corrections (which will not go through pump, obviously) into NS though so they are recorded. Very initially we had double entries for school time, when he wasn't on APS - we had son shift to entering carbs in a NS note rather than a 'carb entry form' so we would still have the details we wanted at home to watch, but so that we didn't have the double entry problem. When we had double entries, we would then delete them as soon as he came in from school, the APS picked him up and then entered in the pump data thusly duplicating the entries he had made into care portal. I do not believe that any of the screen shots above were double entry issues.
Scott Leibrand
@scottleibrand
Jan 18 2017 18:23
@bdr1177 COB unknown just means you're not running AMA.
(Usually. It can occur with AMA, if there is a problem with the carb data.)
Brian Rabinovitz
@bdr1177
Jan 18 2017 18:25
@scottleibrand Thanks for the clarification. I didn't know if it just pulled the COB info no matter what as part of its status check.
Tim Street
@tim2000s
Jan 18 2017 18:25
@ceben80 do you have an SQLite directory on you local xDrip setup. I've just noticed that I don't.
C-Ville
@C-Ville
Jan 18 2017 18:27
@scottleibrand @danamlewis will NS entered carb corrections be used for the loop or only the data entered using the Bolus wizzard in NS?
Dana Lewis
@danamlewis
Jan 18 2017 18:28
@C-Ville if your rig is online (internet connected), it can pull in info from NS (Carbs, temp targets)
C-Ville
@C-Ville
Jan 18 2017 18:30
@danamlewis temptargets I new, was just not sure if carb only entries would be included.
Dana Lewis
@danamlewis
Jan 18 2017 18:30
:+1:
Scott Leibrand
@scottleibrand
Jan 18 2017 18:30
@bdr1177 Yeah, that's not ideal, but everyone who would change it is using AMA and not affected. PRs enthusiastically welcomed. :smile:
Dana Lewis
@danamlewis
Jan 18 2017 18:31
so have to watch out for double entries, like described above
C-Ville
@C-Ville
Jan 18 2017 18:31
until last week I was running a fairly old installation without AMA, and I am just now trying to work myself into AMA for looping
Tim Street
@tim2000s
Jan 18 2017 18:31
@C-Ville you'll love it
Dana Lewis
@danamlewis
Jan 18 2017 18:32
:)
C-Ville
@C-Ville
Jan 18 2017 18:33
I tried @colinlennon s xdrip modification, but this essentially seems to prevent me from using the full NS functionality
Brian Rabinovitz
@bdr1177
Jan 18 2017 18:45
@danamlewis @JELCRAWFORD As a follow up to cob discussion, I just opened up NS on my phone and it showed what appeared to be correct cob. I refreshed the page and cob went back down to zero.
Matthias
@ceben80
Jan 18 2017 18:47
@tim2000s see here :wink:
edison@eddie:~$ sudo find / -iname sqlite3
/home/.rootfs/usr/bin/sqlite3
/home/.rootfs/usr/lib/python2.7/sqlite3
/home/.rootfs/usr/lib/python3.4/sqlite3
/home/.rootfs/usr/share/doc/sqlite3
JELCRAWFORD
@JELCRAWFORD
Jan 18 2017 18:49
@bdr1177 we certainly noted fluctuations as you are describing, but could not come up with an explanation. The zero COB was less concerning to us than the huge COB at night. As indicated though, we have not seen this since shifting to the Edison/Explorer board a couple of weeks ago - so if it continues to happen, it is not so obvious/evident. Weird. We did turn AMA on on the new rig.
Brian Rabinovitz
@bdr1177
Jan 18 2017 18:52
We're testing Edison/explorer board, which I just set up last weekend after testing Rpi with carelink for a few weeks. But haven't set up AMA yet.
JELCRAWFORD
@JELCRAWFORD
Jan 18 2017 18:55
We ran Rpi from mid Nov through to late Dec - not a great solution for school though, so we were motivated to get the Edison up and running. Even our son identifies that the APS usually means he has less corrections and less diabetes related stuff to 'think about' at school. We still watch pretty closely, but consider it a big win just with his comments (nevermind the greater likelihood of a full nights sleep and the better overall control). It seems like a project one will always be tweaking :)
Brian Rabinovitz
@bdr1177
Jan 18 2017 18:59
@JELCRAWFORD That's awesome so glad to hear it's working well for you. We just did testing with the Rpi/carelink and never actually went live, because we also saw that it wouldn't really be practical at school. We're excited to go live with the Edison/explorer soon.
JELCRAWFORD
@JELCRAWFORD
Jan 18 2017 19:01
@bdr1177 We ran the Rpi at home overnight and on weekends (and a good chunk of Christmas break)- was a good exercise to run it where we could watch and get a feel for it before he took it to school. Also still have it running for times when the Edison is getting an oil change or other work up:) I'd say having both to run is valuable. Particularly as Rpi has no battery needs, just a plug in, so if Edison not working well, we can always fall back to Rpi - even if only at night.
JELCRAWFORD
@JELCRAWFORD
Jan 18 2017 19:19
Screen Shot 2017-01-18 at 12.16.24 PM (2).png
@danamlewis @bdr1177 ...hm.... apparently I spoke too soon about the COB issues....
hmm.. now reading 37..... same thought as before - weird.
Tim Street
@tim2000s
Jan 18 2017 19:28
Cheers @ceben80 - looks like although I can see input via port 5000 on my xdrip set up, I am not getting any data into the entries table of xDripAPS.db, which explains why it isn't running from local.
Matthias
@ceben80
Jan 18 2017 19:30
@tim2000s do you have a xdrip.ini located in ~/myopenaps :question:
Tim Street
@tim2000s
Jan 18 2017 19:30
Yes, and it looks correct
Or does it? Should it have something in both fields = and args =?
Matthias
@ceben80
Jan 18 2017 19:32
btw Tim I finished my 1000mAh lipo test today with the EB...last nearly 7 hours
Tim Street
@tim2000s
Jan 18 2017 19:32
That's not too bad.
xdrip.ini
[device "xdrip"]
fields =
cmd = bash -c "curl -s http://localhost:5000/api/v1/entries?count=288 | json -e \"this.glucose = this.sgv\""
args =
Matthias
@ceben80
Jan 18 2017 19:33
that's my ini file
edison@eddie:~/myopenaps$ cat xdrip.ini
[device "xdrip"]
fields =
cmd = bash -c "curl -s http://localhost:5000/api/v1/entries?count=288 | json -e \"this.glucose = this.sgv\""
args =
haha :smile:
simultaneously posting
Tim Street
@tim2000s
Jan 18 2017 19:35
In xDrip: http://localhost:5000/api/v1/entries or http://localhost:5000/api/v1/
Matthias
@ceben80
Jan 18 2017 19:37
blob
Tim Street
@tim2000s
Jan 18 2017 19:38
Hmmm - same as I have
Matthias
@ceben80
Jan 18 2017 19:38
but localhost is your edison IP?
Tim Street
@tim2000s
Jan 18 2017 19:38
Yes
Matthias
@ceben80
Jan 18 2017 19:38
weird
Tim Street
@tim2000s
Jan 18 2017 19:39
Very
I can tell that the xDripAPS script has run due to the entries table in xDripAPS.db
Matthias
@ceben80
Jan 18 2017 19:40
@colinlennon Hi Colin, any idea here :question: thx in advance :wink:
edison@eddie:~/myopenaps$ ps -aux | grep -i xdrip
edison     362  0.0  0.0   2036   500 ?        Ss   17:03   0:00 /bin/sh -c python /home/edison/.xDripAPS/xDripAPS.py
edison     364  0.1  1.2  19556 12760 ?        S    17:03   0:20 python /home/edison/.xDripAPS/xDripAPS.py
edison   14571  0.0  0.0   4320   916 pts/2    S+   20:41   0:00 grep -i xdrip
and your ps @tim2000s ?
Tim Street
@tim2000s
Jan 18 2017 19:44
Exactly the same...!
But the content of my xdrip/glucose.json is:
[]
j4cbo
@j4cbo
Jan 18 2017 19:48
thinking i may move my nightscout install to the edison
camping mode all the time
Tim Street
@tim2000s
Jan 18 2017 19:49
I like the idea of that @j4cbo
Dana Lewis
@danamlewis
Jan 18 2017 19:53
:+1: needs some documenting ;)
j4cbo
@j4cbo
Jan 18 2017 19:56
hint hint :) I'm, uh, not a fan of how nightscout seems totally unauthenticated by default
Ryan anglani
@ryan_anglani_twitter
Jan 18 2017 20:02
Hello all: I am trying to go mobile with our openaps and we have the motoG with g4 share with ting. I am not able to get a hotspot with the motoG. the motoG gives a message that says unable to connect to data network has anyone come across this problem ? and how did you fix it? thank you
jburke41
@jburke41
Jan 18 2017 20:05
Just switched to HeroKu from Azure. However, the status on my OpenAps pill is unknown. I can't seem to figure out how to set "entries" as the collection name. Has anyone else had this issue?
Ryan anglani
@ryan_anglani_twitter
Jan 18 2017 20:06
it also says you may not be subscribed to the tethering and hotspot service
Tim Street
@tim2000s
Jan 18 2017 20:08
@ryan_anglani_twitter - I think you need to talk to your Cellphone service provider
Dana Lewis
@danamlewis
Jan 18 2017 20:20
@jburke41 in general, sometimes it takes 10-15 minutes for things to show up when you switch. but if it’s been after 15 minutes, it’s probably something else that didn’t get set right.
related, since we’re primarily recommending heroku, if someone wants to update the screenshots for a heroku example here in the docs, that would probably help some: http://openaps.readthedocs.io/en/latest/docs/walkthrough/phase-1/nightscout-setup.html#enable-these-plugins
jburke41
@jburke41
Jan 18 2017 20:22
Updating the screenshots would be fantastic bc very confusing with old ones
Dana Lewis
@danamlewis
Jan 18 2017 20:22
maybe take screenshots as you go, maybe you can make the PR :)
jburke41
@jburke41
Jan 18 2017 20:22
def will try!
Dana Lewis
@danamlewis
Jan 18 2017 20:22
:+1: :clap:
jburke41
@jburke41
Jan 18 2017 20:23
If anyone knows where you enter the collection name as "entries" in Heroku that would be great!
Ryan anglani
@ryan_anglani_twitter
Jan 18 2017 20:26
it looks like the motoG from boost mobile cannot provide a hotspot .....time to change things
fiberfan
@fiberfan
Jan 18 2017 20:47
@danamlewis I am filling out an application for a mini maker fair in SLC in April. The application wants contributors. Are there others in addition to you, Scott and Ben West who I should list as contributors? The attendees at a maker fair are ones more likely to be interested in both Nightscout and OpenAPS.
Dana Lewis
@danamlewis
Jan 18 2017 20:47
@fiberfan cool! you could just list “the #OpenAPS community” as contributors to catch us all
and the “#WeAreNotWaiting community” for Nightscout contributors
happy to review it if you’d like/want other input
also I bet @mccgm would be interested in helping, too
fiberfan
@fiberfan
Jan 18 2017 20:52
I will mention it to @mccgm on Saturday at a Utah OpenAPS meeting. If I am accepted I will definitely want input on what info I bring.
Dana Lewis
@danamlewis
Jan 18 2017 20:52
:+1:
fiberfan
@fiberfan
Jan 18 2017 20:54
Would you recommend Individual maker, Cause or mission-based organization or None of the above for Individual/Group type?
Sandra Keßler
@SandraK82
Jan 18 2017 20:54
a little update on the blueReader (nfc 2 ble transmitetr for the freestyle libre): https://unendlichkeit.net/wordpress/?p=421
Dana Lewis
@danamlewis
Jan 18 2017 20:55
err, not having seen the application instructions/form, probably “cause”.
Scott Leibrand
@scottleibrand
Jan 18 2017 20:56
@SandraK82 :+1:
Ryan anglani
@ryan_anglani_twitter
Jan 18 2017 20:57
``` * sudo wpa_cli scan
          • killall -g --older-than 15m openaps
          • cd /home/pi/gavinopenaps && oref0-reset-git
          • cd /home/pi/gavinopenaps && oref0-truncate-git-history
          • cd /home/pi/gavinopenaps && ps aux | grep -v grep | grep -q 'openaps g et-bg' || ( date; openaps get-bg ; cat cgm/glucose.json | json -a sgv dateString | head -1 ) | tee -a /var/log/openaps/cgm-loop.log
          • cd /home/pi/gavinopenaps && ps aux | grep -v grep | grep -q 'openaps n s-loop' || openaps ns-loop | tee -a /var/log/openaps/ns-loop.log
          • cd /home/pi/gavinopenaps && ( ps aux | grep -v grep | grep -q 'openaps pump-loop' || openaps pump-loop ) 2>&1 | tee -a /var/log/openaps/pump-loop.log
            NIGHTSCOUT_HOST=http://gavinscgm.herokuapp.com
          • cd /home/pi/gavinaps && oref0-reset-git
          • cd /home/pi/gavinaps && oref0-truncate-git-history
          • cd /home/pi/gavinaps && ps aux | grep -v grep | grep -q 'openaps get-b g' || ( date; openaps get-bg ; cat cgm/glucose.json | json -a sgv dateString | h ead -1 ) | tee -a /var/log/openaps/cgm-loop.log
          • cd /home/pi/gavinaps && ps aux | grep -v grep | grep -q 'openaps ns-lo op' || openaps ns-loop | tee -a /var/log/openaps/ns-loop.log
          • cd /home/pi/gavinaps && ( ps aux | grep -v grep | grep -q 'openaps pum p-loop' || openaps pump-loop ) 2>&1 | tee -a /var/log/openaps/pump-loop.log
          • cd /home/pi/gavins && oref0-reset-git
          • cd /home/pi/gavins && oref0-truncate-git-history
          • cd /home/pi/gavins && ps aux | grep -v grep | grep -q 'openaps get-bg' || ( date; openaps get-bg ; cat cgm/glucose.json | json -a sgv dateString | hea d -1 ) | tee -a /var/log/openaps/cgm-loop.log
          • cd /home/pi/gavins && ps aux | grep -v grep | grep -q 'openaps ns-loop ' || openaps ns-loop | tee -a /var/log/openaps/ns-loop.log
          • cd /home/pi/gavins && ( ps aux | grep -v grep | grep -q 'openaps pump- loop' || openaps pump-loop ) 2>&1 | tee -a /var/log/openaps/pump-loop.log
          • cd /home/pi/gavin && oref0-reset-git
          • cd /home/pi/gavin && oref0-truncate-git-history
          • cd /home/pi/gavin && ps aux | grep -v grep | grep -q 'openaps get-bg' || ( date; openaps get-bg ; cat cgm/glucose.json | json -a sgv dateString | head -1 ) | tee -a /var/log/openaps/cgm-loop.log
          • cd /home/pi/gavin && ps aux | grep -v grep | grep -q 'openaps ns-loop' ``` what does this mean?
Scott Leibrand
@scottleibrand
Jan 18 2017 20:58
you need to hit shift-enter after the last ` and again before and after the last set to get the formatting you want
Looks like you have a whole bunch of myopenaps instances of different names. you should delete the old ones and just keep one of them. or delete everything and re-run oref0-setup on the directory you want
fiberfan
@fiberfan
Jan 18 2017 20:59
@danamlewis The form is at http://slcmakerfaire.com/call-for-makers/
Tim Street
@tim2000s
Jan 18 2017 20:59
crontab -r will delete the contents of cron if you want to start again
Scott Leibrand
@scottleibrand
Jan 18 2017 21:00
be careful with that: it doesn't prompt you or anything, just deletes it
Tim Street
@tim2000s
Jan 18 2017 21:00
;)
Scott Leibrand
@scottleibrand
Jan 18 2017 21:00
it's apparently easy for some people to accidentally hit r when doing crontab -e
fortunately most everything in crontab comes from oref0-setup, so really easy to restore it
Ryan anglani
@ryan_anglani_twitter
Jan 18 2017 21:02
@scottleibrand how can I delete everything
fiberfan
@fiberfan
Jan 18 2017 21:03
If you want a line by line deletion, in Nano ctrl-k deletes the line the cursor is on.
katie disimone
@Kdisimone
Jan 18 2017 21:03
@jburke41 the collection name is under Settings, Reveal Config Vars and then the line “MONGO_COLLECTION"
annaloop_·_Settings___Heroku.jpg

@jburke41 however, if you migrated from Azure to Heroku…you may want to make sure that you actually used “entries” in your original mLab setup. If you intend to migrate and use your old mLab database from heroku for continuity then you would also want to copy and paste your MONGODB_URI from azure into heroku (also shown in the screenshot above) and use the same collection name as you had setup in mLab.

The “deploy to Heroku” button from cgm-remote-monitor will automatically setup a new MongoDB for you and use “entries” as the collection name.

Ryan anglani
@ryan_anglani_twitter
Jan 18 2017 21:11
``` * sudo wpa_cli scan
          • killall -g --older-than 15m openaps
          • cd /home/pi/gavinopenaps && oref0-reset-git
          • cd /home/pi/gavinopenaps && oref0-truncate-git-history
          • cd /home/pi/gavinopenaps && ps aux | grep -v grep | grep -q 'openaps get-bg' || ( date; openaps get-bg ; cat cgm/glucose.json | json -a sgv dateString | head -1 ) | tee -a /var/log/openaps/cgm-loop.log
          • cd /home/pi/gavinopenaps && ps aux | grep -v grep | grep -q 'openaps ns-loop' || openaps ns-loop | tee -a /var/log/openaps/ns-loop.log
          • cd /home/pi/gavinopenaps && ( ps aux | grep -v grep | grep -q 'openaps pump-loop' || openaps pump-loop ) 2>&1 | tee -a /var/log/openaps/pump-loop.log
            NIGHTSCOUT_HOST=http://gavinscgm.herokuapp.com
          • cd /home/pi/gavinaps && oref0-reset-git
          • cd /home/pi/gavinaps && oref0-truncate-git-history
          • cd /home/pi/gavinaps && ps aux | grep -v grep | grep -q 'openaps get-bg' || ( date; openaps get-bg ; cat cgm/glucose.json | json -a sgv dateString | head -1 ) | tee -a /var/log/openaps/cgm-loop.log
          • cd /home/pi/gavinaps && ps aux | grep -v grep | grep -q 'openaps ns-loop' || openaps ns-loop | tee -a /var/log/openaps/ns-loop.log
          • cd /home/pi/gavinaps && ( ps aux | grep -v grep | grep -q 'openaps pump-loop' || openaps pump-loop ) 2>&1 | tee -a /var/log/openaps/pump-loop.log
          • cd /home/pi/gavins && oref0-reset-git
          • cd /home/pi/gavins && oref0-truncate-git-history
          • cd /home/pi/gavins && ps aux | grep -v grep | grep -q 'openaps get-bg' || ( date; openaps get-bg ; cat cgm/glucose.json | json -a sgv dateString | head -1 ) | tee -a /var/log/openaps/cgm-loop.log
          • cd /home/pi/gavins && ps aux | grep -v grep | grep -q 'openaps ns-loop' || openaps ns-loop | tee -a /var/log/openaps/ns-loop.log
          • cd /home/pi/gavins && ( ps aux | grep -v grep | grep -q 'openaps pump-loop' || openaps pump-loop ) 2>&1 | tee -a /var/log/openaps/pump-loop.log
          • cd /home/pi/gavin && oref0-reset-git
          • cd /home/pi/gavin && oref0-truncate-git-history
          • cd /home/pi/gavin && ps aux | grep -v grep | grep -q 'openaps get-bg' || ( date; openaps get-bg ; cat cgm/glucose.json | json -a sgv dateString | head -1 ) | tee -a /var/log/openaps/cgm-loop.log
          • cd /home/pi/gavin && ps aux | grep -v grep | grep -q 'openaps ns-loop' || openaps ns-loop | tee -a /var/log/openaps/ns-loop.log
          • cd /home/pi/gavin && ( ps aux | grep -v grep | grep -q 'openaps pump-loop' || openaps pump-loop ) 2>&1 | tee -a /var/log/openaps/pump-loop.log
            ```
I'm keep pressing the ``` and its not coming out
the openaps is up and running nevermind :)
Scott Leibrand
@scottleibrand
Jan 18 2017 21:21
@ryan_anglani_twitter hit shift-enter after the first ``` to put them on their own line without any text afterward
in the web interface you can also use the up arrow to edit the last thing you posted, or hit edit from the ... menu to the right of the timestamp
PieterGit
@PieterGit
Jan 18 2017 22:05
@scottleibrand : sorry, no time for openaps/oref0#261 (autotune) the coming days. i did a first autotune run last week and we changed basals on the pump. i now did manual basal tuning based on autotune results, trying to mimic 'eating soon' bolus strategy a bit, and being cautious of increasing basals because the RPI3 mobile rig still has some issues. I'll review the new profile compared to the new autotune results this weekend. I hope I can do some PR'stuff, but getting the Edison with USB working is higher on my wish lists.
Scott Leibrand
@scottleibrand
Jan 18 2017 22:06
:+1:
C-Ville
@C-Ville
Jan 18 2017 22:46
@scottleibrand did you remove xdrip from the current oref0 dev? i tried to run the setupscript but it tells me xdrip is not supported
i tried 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
Jan 18 2017 23:01
nope
C-Ville
@C-Ville
Jan 18 2017 23:02
strange, i am currently not offered the xdrip choice anymore
i guess i screwed up somewhere, now it seems to be working
C-Ville
@C-Ville
Jan 18 2017 23:09
I do have another question regarding xdrip. does the loop using xdrip pull cgm data from nightscout if you do not get new data from xdrip?
scottleibrand @scottleibrand has never used it
Dana Lewis
@danamlewis
Jan 18 2017 23:10
(if someone wants to send me an xdrip for my menagerie we can do some testing with it, though)
C-Ville
@C-Ville
Jan 18 2017 23:11
sounds tempting :-)
Dana Lewis
@danamlewis
Jan 18 2017 23:11
lol
C-Ville
@C-Ville
Jan 18 2017 23:12
I will need to build another one soon, so I just make it 2
Aaron Michelson
@Aaybob
Jan 18 2017 23:59
left house got nothing from rig. returned to serial port tail -f loop-pump shows
temp_basal.json more than 5m old
Temp refreshError, retrying
Listening: .No pump comms detected from other rigs
Mmtune: "916.660", 5, -32 Starting p
……………..
Restarting ntp (via systemctl): ntp.service.
Waiting for ntpd to synchronize... No!
ntpd did not synchronize.