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

3rd
May 2017
Meghan Rutledge
@mar5477
May 03 2017 00:00
Tried to go back and reinstall things, now I'm getting this error
Your branch is behind 'origin/dev' by 5 commits, and can be fast-forwarded.
(use "git pull" to update your local branch)
Updating 5088bd8..0fc8151
error: Your local changes to the following files would be overwritten by merge:
bin/oref0_init_pump_comms.py
bin/oref0_subg_ww_radio_parameters.py
Please, commit your changes or stash them before you can merge.
rsilvers129
@rsilvers129
May 03 2017 00:02
Any idea how I can upload .json glucose files to my new Nightscout install? I asked on Facebook and no responses.
Sgoldberg89
@Sgoldberg89
May 03 2017 00:07
@danamlewis @geoffro10 it stopped again
dana, getting closer to the grenade option...
Dana Lewis
@danamlewis
May 03 2017 00:08
don’t do it! ;p (remember your loop is still working, right? so really no reason to do that other than impatience)
what are your devicestatus from the last “it worked” until now?
Scott Leibrand
@scottleibrand
May 03 2017 00:12
@mar5477 you can do what it suggests and stash them: just git stash and then pull dev again
@rsilvers129 this is probably a better place than FB for that kind of discussion. have you built your openaps rig? If so, what CGM option did you choose?
rsilvers129
@rsilvers129
May 03 2017 00:13
My rig is built and loaded. I chose G5 from iPhone.
Scott Leibrand
@scottleibrand
May 03 2017 00:14
ah, k. if you had chosen G4-upload, MDT, or xdrip (I believe) it will have created a myopenaps-cgm-loop directory, which would have everything you'd need to upload to NS
might be worth re-running oref0-setup on another directory to install those bits (just don't schedule it in cron)
rsilvers129
@rsilvers129
May 03 2017 00:14
ok
Scott Leibrand
@scottleibrand
May 03 2017 00:15
then you can drill down into openaps alias show upload and see how it has the upload bits configured. there's a section in the docs for running reports manually that shows how to do most of that
you'll eventually get to a report that uploads something like nightscout/glucose.json - you might be able to just place your file there and run that openaps report invoke command to do the upload
or you can drill down further to figure out the syntax (with the nightscout command to do it manually
lots of complexity there, but a ton of flexibility too
rsilvers129
@rsilvers129
May 03 2017 00:22
Updating 5088bd8..0fc8151
error: Your local changes to the following files would be overwritten by merge:
bin/oref0_init_pump_comms.py
bin/oref0_subg_ww_radio_parameters.py
Please, commit your changes or stash them before you can merge.
Aborting
Couldn't pull latest oref0
Scott Leibrand
@scottleibrand
May 03 2017 00:36
:point_up: May 2, 2017 5:12 PM
Meghan Rutledge
@mar5477
May 03 2017 00:41
@scottleibrand root@MYAPS:~/src# git stash
Saved working directory and index state WIP on master: d5ea015 initializing openaps 0.2.0-dev
HEAD is now at d5ea015 initializing openaps 0.2.0-dev
root@MYAPS:~/src# mkdir -p ~/src; cd ~/src && git clone -b dev git://github.com/openaps/oref0.git || (cd oref0 && git checkout dev && git pull)
fatal: destination path 'oref0' already exists and is not an empty directory.
M bin/oref0_init_pump_comms.py
M bin/oref0_subg_ww_radio_parameters.py
Already on 'dev'
Your branch is behind 'origin/dev' by 5 commits, and can be fast-forwarded.
(use "git pull" to update your local branch)
Updating 5088bd8..0fc8151
error: Your local changes to the following files would be overwritten by merge:
bin/oref0_init_pump_comms.py
bin/oref0_subg_ww_radio_parameters.py
Please, commit your changes or stash them before you can merge.
Scott Leibrand
@scottleibrand
May 03 2017 00:43
you have to be inside the directory it's complaining about first
so cd ~/src/oref0
rsilvers129
@rsilvers129
May 03 2017 00:45
Updating 5088bd8..0fc8151
error: Your local changes to the following files would be overwritten by merge:
bin/oref0_init_pump_comms.py
bin/oref0_subg_ww_radio_parameters.py
Please, commit your changes or stash them before you can merge.
Aborting
Couldn't pull latest oref0
root@pancreas0:~/src/oref0#
Meghan Rutledge
@mar5477
May 03 2017 00:46
@rsilvers129 cd ~/src/oref0
git stash
then run again
rsilvers129
@rsilvers129
May 03 2017 00:46
ok
Meghan Rutledge
@mar5477
May 03 2017 00:46
*run oref0-setup again
rsilvers129
@rsilvers129
May 03 2017 00:47
that worked
Attempting to communicate with pump:
bash: /usr/local/bin/oref0_init_pump_comms.py: Permission denied
Meghan Rutledge
@mar5477
May 03 2017 00:50
did you run cd ~/src/oref0 && npm run global-install too/
cd ~/src/oref0 && npm run global-install
rsilvers129
@rsilvers129
May 03 2017 00:51
Doing it now
rsilvers129
@rsilvers129
May 03 2017 01:03
I got through it all. "916.588", 5, -46
Scott Leibrand
@scottleibrand
May 03 2017 01:04
:+1:
Meghan Rutledge
@mar5477
May 03 2017 01:11
@danamlewis I went back through my devicestatus entries from right before I stopped uploading. Looks like a couple minutes before there was a duplicate entry (everything was the same except for the "created_at" timestamp) and then loop pulled an older bg, and the next entry after that was the last successful one.
Not sure what's the best way to share since it's so long.
Scott Leibrand
@scottleibrand
May 03 2017 01:16
copy and paste it into https://gist.github.com/ ?
and then try deleting that duplicate entry and see if it causes the loop to start uploading again?
the entries I deleted have "DELETED" before them. I also added my bgs to the bottom from Dexcom Clarity so you can see what the actual order of my bg was
I changed a lot of other things too, so I'm still not back up yet, but thought it would be interesting to see what might have caused it
rsilvers129
@rsilvers129
May 03 2017 01:40
New 2500 enclosure is done. http://www.thingiverse.com/thing:2291125
jmcrawford
@jmcrawford
May 03 2017 02:02
Nice. How much did Intel pay you to leave the logo exposed?
rsilvers129
@rsilvers129
May 03 2017 02:06
I know. This would be perfect for one of their commercials.
Meghan Rutledge
@mar5477
May 03 2017 02:06
Welp. Not sure about anything at all anymore. But the gist above is what happened right before I stopped uploading to mlab from xdrip+.Deleting those entries did not get me back looping. I was able to change my REST API to send bg directly to my edison, but they weren't showing up in NS (my basal changes were). The whole time I was still getting new devicestatus entries, just nothing in my "entries" collection. Finally, changed back to MongoDB upload, created a new "entries1" collection in mlab, changed Collection name field in xdrip to "entries1" and changed mongo_collection in config vars to "entries1" now I'm looping and can see basal and bg in NS. Still no idea why it stopped in the first place or what about my "entries" collection it doesn't like.
Andy Sharrow
@dramageek
May 03 2017 02:06
I'd be curious to see a signal comparison with your cases. I thought it worked better with the radio chip out.
Dana Lewis
@danamlewis
May 03 2017 02:20
FYI - just merged SMB and oref1 and all of the related glorifulness to dev. Please see https://github.com/openaps/oref0/pull/424#issuecomment-298807656 and the note above and below it for how to test and specific things we’d like people to turn on one at a time (if you choose to test that thing). Please DO comment on that PR if you test something and give it a thumbs up; or let us know there/here if something breaks for you. Plus, usual commentary about using a second rig or second pump to test things on, be watchful as you observe new behaviors, etc. (No oref1 stuff will get turned on without you enabling it; so even testing dev with your current setup parameters & confirming everything works as-is is helpful too)
cameronrenwick
@cameronrenwick
May 03 2017 02:29
@rsilvers129 nice job!! I've got a new build I'm doing right now with a 2500 and I think your case will be just what I'm looking for. Would you consider building a case for a battery like this one https://elmwoodelectronics.ca/products/lithium-ion-battery-pack-3-7v-4400mah? I've got a "long day" rig that I'm building that I intend to use the 4400 battery. A bit clunkier for sure but.. any thoughts?
rsilvers129
@rsilvers129
May 03 2017 02:31
The thing is, you can stack two 2500s together and make a 5000 and it will be 14.3mm thick and 60mm long for the battery. That pack is 18mm thick and 65mm long.
I think Tesla uses them because they can run cooling in the space between them, but that doesn't matter to us due to the slow draw.
cameronrenwick
@cameronrenwick
May 03 2017 02:33
absolutely true.. sadly the battery acquiring in Canada is the hard part and the 2500 I've got is the only 1 I have. Seems the 4400s are easier to get a hold of
rsilvers129
@rsilvers129
May 03 2017 02:34
18650 is very common. I have lots of them.
Great for flashlights
(torches for you)
I tried to upload my .jsons to Nightscout, and it says:
{"status":401,"message":"Unauthorized","description":"Invalid/Missing"}
so I need a description field?
cameronrenwick
@cameronrenwick
May 03 2017 02:35
ya I think it's a supply demand thing... and lol I call them flashlights too.. my new samsung s7 has a "torch" button on it... told my son its' for cutting steel lol
rsilvers129
@rsilvers129
May 03 2017 02:37
You should just use a USB battery pack to keep your 2000 charged.
RedCyclist1
@RedCyclist1
May 03 2017 03:08
HI Gang, quick question, I am setting up my second rig, only this one will plug into my dexcom. Can I run 2 rigs at the same time? Is there a way to take advantage of the autotunes from my first rig to the new one or will it have to establish the 30 days from today? @danamlewis
;) yes
Yes re Autotune - you can just copy the first rig Autotune files over to second rig
RedCyclist1
@RedCyclist1
May 03 2017 03:31
@danamlewis got it thanks
Dana Lewis
@danamlewis
May 03 2017 03:40
:+1:
Enjoy the start of your fleet! ;)
RedCyclist1
@RedCyclist1
May 03 2017 03:41
Looking forward to offline looping...
Not sure if it was mentioned here, but there is an OpenAPS meet up in Phoenix/Tempe/Mesa area on May 11th!!!! sigh up if interested. http://whoozin.com/RUX-CT4-NVJJ...
Dana Lewis
@danamlewis
May 03 2017 03:44
:+1:
Aaron Michelson
@Aaybob
May 03 2017 03:51
@RedCyclist1 I'd like to hear about it. I did have offline looping. But nothing for about six weeks.
Dana Lewis
@danamlewis
May 03 2017 04:12
What kind of pump do you have? And where is it in relation to your rig?
Dana Lewis
@danamlewis
May 03 2017 04:18
Did you create the profile files and tweak the aliases?
Side note: continues to wish that the folks w 512 who do these steps would PR in better instructions for making these tweaks
Dana Lewis
@danamlewis
May 03 2017 04:23
(I feel like a broken record saying it, but if no one PRs it, it doesn't get better, and every person with a 512 or 712 ends up asking the same questions)
:+1: welcome
RedCyclist1
@RedCyclist1
May 03 2017 04:30
@Aaybob will let you know once up and running... right now we can't get out of the nano editor... none of the instructions are working... Oy
sanderlong
@sanderlong
May 03 2017 04:33
I feel really stupid asking this, but I'm stuck on step 2 run oref0-setup. I'm not clear on what I should be inputting in terminal
Dana Lewis
@danamlewis
May 03 2017 04:34
Have you successfully completed the other steps prior?
sanderlong
@sanderlong
May 03 2017 04:34
Yes
Is it this: cd && ~/src/oref0/bin/oref0-setup.sh ?
Dana Lewis
@danamlewis
May 03 2017 04:35
Step 2 is oref0-setup.sh script
cd && ~/src/oref0/bin/oref0-setup.sh
Yup! A bit buried because there's some key things to read prior
sanderlong
@sanderlong
May 03 2017 04:36
Thanks for confirming - I was thinking that was the next step and I was missing something
Dana Lewis
@danamlewis
May 03 2017 04:36
:+1:
sanderlong
@sanderlong
May 03 2017 04:36
Another question - do I need to have the basals and ratios programmed in the pump
I have an Animas now, so nothing is set on the Medtronic yet
There is a note in the instructions that lead me to believe I need to set up the Medtronic first: The very first time may take a while (15-20 minutes) for it to successfully read and pull a full history from your pump.
Dana Lewis
@danamlewis
May 03 2017 04:38
Correct: i would go ahead and program pump with everything
Even if you don't put a reservoir in it
sanderlong
@sanderlong
May 03 2017 04:39
Thank you, Dana!
Dana Lewis
@danamlewis
May 03 2017 04:39
Welcome! You can finish this step without it; it just won't loop successfully until it reads from the pump the proper profile and can find basals to evaluate, etc.
sanderlong
@sanderlong
May 03 2017 04:41
Thanks, I really appreciate your help!
Dana Lewis
@danamlewis
May 03 2017 04:42
:smile:
Scott Leibrand
@scottleibrand
May 03 2017 04:54
Also, if you're switching pumps, do that before you start looping. There are often differences you need to figure out before adding more variables into the mix. (And someone should PR in that advice if it's not already in the docs.)
tynbendad
@tynbendad
May 03 2017 06:20
are there any upcoming changes planned for explorer board? am about to order one for a 2nd rig... anyone know a good place to buy the low power edison board also? anyone measure a battery life difference with it?
Dana Lewis
@danamlewis
May 03 2017 06:26
No one has found any actual difference in battery for low power vs not on edison
And I haven't heard of any rev changes planned on Explorer lately
tynbendad
@tynbendad
May 03 2017 06:38
thx, that makes it easy :+1:
they ought to sell a version with the trace cut already, taking a knife to your new board is scary ;)
Scott Leibrand
@scottleibrand
May 03 2017 06:49
Maybe email ERD and suggest that?
They might have enough sales volume now to justify it.
Andrew Baugh
@baughaw
May 03 2017 13:03
@webwiztn I used ABS 100% fill and .3mm thickness
dmdfreak
@dmdfreak
May 03 2017 14:05
is there a way to change my phone BTmac in the rig without a total rerun of oref?
Geoff Aumaugher
@geoffro10
May 03 2017 14:45
@scottleibrand @danamlewis @mar5477 @Sgoldberg89 An update to the issue of pump and OpenAPS info not showing. I saw Meghan's post regarding her changing to entries1 and it working. I tried something different. I renamed entries to entries-old and created a new entries. I waited for a bit to see if it would straighten out. Nope... no go. So I copied the new entries back into entries-old, deleted entries, and renamed entries-old back to entries. I'll keep at it. Sigh....
katie disimone
@Kdisimone
May 03 2017 14:56
@dmdfreak crontab -e and new pairing
dmdfreak
@dmdfreak
May 03 2017 14:56
thank you @Kdisimone
ok I might need a little more than that @Kdisimone hehe. I was hoping it would be clearer so I would have to ask for more info. But I'm too newbie
katie disimone
@Kdisimone
May 03 2017 15:04

If you are logged into your rig, type crontab -e and look for the lines that have your phone's old BT MAC address on them. It's usually one of the first lines, for oref0-online. Change the BT mac. Then save the changes using Ctrl-x, answering yes, and then pressing return to use the save.

And then you will have to repeat BT pairing with the new phone using the instructions in the docs.

http://openaps.readthedocs.io/en/latest/docs/walkthrough/phase-4/bluetooth-tethering-edison.html

john36m
@john36m
May 03 2017 15:07
Regarding Fiasp insulin, I have read here that you can just go into Canadian pharmacy and get some, but I also read you need a prescription. Anyone know the definite answer. I will be in Canada in a week and would like to pick some up.
dmdfreak
@dmdfreak
May 03 2017 15:10
I guess thats my problem. The line isnt there @Kdisimone I guess I didnt install bluetooth the last time I ran oref setup because my last phone wouldnt work with it. I was trying to rerun it but its aborting. So I tried to update to the new dev version and went through the correct motions, but then it aborted when it reran oref0 again.
it says please commit your changes or stash them before you can merge.
katie disimone
@Kdisimone
May 03 2017 15:22
That same message was discussed yesterday, I believe. Scroll back up in Gitter and you should find help for that. 👍🏻
dmdfreak
@dmdfreak
May 03 2017 15:26
ok thank you
rsilvers129
@rsilvers129
May 03 2017 15:45
get stash
@john36m I think the Canadian phramacies will use their own doctor to provide the prescription. Still, I looked into it, and it was 10x the cost of getting Insulin from my insurance in the US.
Scott Leibrand
@scottleibrand
May 03 2017 15:47
@john36m Technically you need a prescription. Practically, my understanding is that they don't have to check if you have one.
Martin Haeberli
@mhaeberli
May 03 2017 15:47
is there magic to cause an Edison / Explorer to power on again automatically (after the battery is flat and it is charged up / charging) without pushing the power / boot button? Asking because, in some 3d cases, it’s not always convenient to get at that button.
rsilvers129
@rsilvers129
May 03 2017 15:47
It starts automatically.
Scott Leibrand
@scottleibrand
May 03 2017 15:47
@mhaeberli that should be the default behavior.
Martin Haeberli
@mhaeberli
May 03 2017 15:48
hmmm
Scott Leibrand
@scottleibrand
May 03 2017 15:48
You can test it by doing a halt and then plugging in.
rsilvers129
@rsilvers129
May 03 2017 15:48
I have never presed that button.
Martin Haeberli
@mhaeberli
May 03 2017 15:52
k i’ll try that in a few minutes - just now - charge light was happily on but main light was not. pressed button once, system looks to be booting.
ssh into system.
halt
but now system operating red light still on.
i’ll try that on my other two edisons (after I finish bringing them up to date)...
ah ok shut down now
plugging in causes charging light to come on
but not booting yet as far as i can tell
-> time for experiments w other two systems
kenkotch
@kenkotch
May 03 2017 16:12
has anyone tried to use ThisButton on a pebble watch recently? I keep getting "THISBUTTON added. Pebble not compatible" EDIT:nevermind- it seems that with the demise of pebble I am not able to install ANY apps through the pebble app. I'll see if I can side load it...
Eric
@ecc1
May 03 2017 16:14
Depending on how the board wires the DCIN signal (I haven't looked), the Edison will either boot as soon as the battery is charged to 3.5V, or as soon as it reaches 2.5V. See section 4.12 of the Edison module document for details.
Gina
@gclyon
May 03 2017 16:14
Any config needed to have edison read bg's from usb G5 for offline monitoring? I've got the pebble BT'd to edison working, and pancreabble working w Urchin watchface---all online tho.
Martin Haeberli
@mhaeberli
May 03 2017 16:15
@ecc1 thx that may be part of the issue - battery may well have been well below 3.5
Dana Lewis
@danamlewis
May 03 2017 16:17
@gclyon second battery plugged in, plus depending on which v of explorer you have: http://openaps.readthedocs.io/en/latest/docs/walkthrough/phase-1/offline-looping-and-monitoring.html#offline-looping
Gina
@gclyon
May 03 2017 16:20
I do have the version after March. ...So just adding another battery? should i be able to check this with it plugged into computer instead (and turn off wifi)?
Dana Lewis
@danamlewis
May 03 2017 16:22
Check this = check which? Offline? Ya by turning off wifi and making sure your hotspot or Bluetooth tether isn't on
JELCRAWFORD
@JELCRAWFORD
May 03 2017 16:37
@john36m - I just called my pharmacy to ask - they indicated they could do an 'emergency' prescription without a physician, but would prefer if the PWD brought either a prescription or a sample of what they were using - so if you had a doc at home write a prescription, I suspect they'd provide it - probably not useful to provide a sample of whatever you are using now. Some pharmacies in some provinces have 'prescribing pharmacists' - I suspect you'd be better if you were at one of those - the pharmicist has extra training and privileges. I don't know costs for Fiasp, but our most recent prescription costs for two vials of Novorapid in Alberta was $65CAD. I think past fills of Apidra were similar for a benchmark around insulin pricing up here. Hope that helps.
dmdfreak
@dmdfreak
May 03 2017 16:45
found the ref above and fixed that problem @Kdisimone I rerun setup using the new BTmac but when I try to do the bluetooth setup, it still shows the old phone mac. I tried the cronetab -e after that and I still cant see a listing for the btmac there
rsilvers129
@rsilvers129
May 03 2017 16:51
I don't know enough to write this, but while I am learning it would be great to have a list of the top 10 or so commands that one needs to know on the Edison to be an effective user. I would totally write it if I could. So: to download and upgrade to the latest version, type this... to run autotune, do this... it could be added to the "helpful tips" section. Or I could do it in five week or so.
Arlene Samowich
@webwiztn
May 03 2017 16:53
@baughaw Thanks for the info!
katie disimone
@Kdisimone
May 03 2017 16:53

@rsilvers129 I've already started something like that and extended it to aliases so that you don't have to write the longer versions of those commands.

http://openaps.readthedocs.io/en/latest/docs/walkthrough/phase-1/add-alias.html

Editing runagain
Editing wifi
Editing preferences.json
Looking at autotune profile
Etc
Logs that you want to see the most frequently.
dmdfreak
@dmdfreak
May 03 2017 16:55
ok I got it going. Thanks again @Kdisimone
katie disimone
@Kdisimone
May 03 2017 16:56
That list on the page includes all the commands I most frequently use for the last 4 or 5 months of OpenAPS
@dmdfreak awesome!!! Glad to hear!
dmdfreak
@dmdfreak
May 03 2017 16:56
is there a way to check if I'm on the oref1 now?
katie disimone
@Kdisimone
May 03 2017 16:58
cd src
cd oref0
git branch
dmdfreak
@dmdfreak
May 03 2017 17:01
ok. that say *dev in green and master in grey
Dana Lewis
@danamlewis
May 03 2017 17:01
@dmdfreak you won't be "on" oref1, but per the dev pr, if on dev you'll see notes about various oref1 features you can enable and test one at a time
katie disimone
@Kdisimone
May 03 2017 17:01
Yes. The one with the * and green is what you are on. And like dana says...it's not oref1 yet
Dana Lewis
@danamlewis
May 03 2017 17:02
It won't switch to being oref1
It builds upon oref0 but isn't a wholesale replacement - just additional features on top of oref0
dmdfreak
@dmdfreak
May 03 2017 17:03
oh ok. before I got sidetracked I remember seeing I need to make changes manually somewhere. I'll go check on that
katie disimone
@Kdisimone
May 03 2017 17:04
@dmdfreak yes, if you did a fresh script, don't forget preferences.json
Dana Lewis
@danamlewis
May 03 2017 17:04
Yup. Dev, then enable specific features and test them one at a time. The dev to master pr has a list in the comments. Please do let us know which scenarios you test and what works.
katie disimone
@Kdisimone
May 03 2017 17:04
@danamlewis so I can still always see Oreo cookies when I go to GitHub?! I appreciate that 😊
Because oref1 doesn't make me think of oreos
dmdfreak
@dmdfreak
May 03 2017 17:05
lol
Dana Lewis
@danamlewis
May 03 2017 17:05
Haha. Yep.
Dana Lewis
@danamlewis
May 03 2017 17:12
(There's a restaurant in Seattle called orfeo that bugs me greatly that they keep misspelling oref0 ;))
Luis Betancourt
@betluis
May 03 2017 17:48
@lukas-ondriga, @scottleibrand hello... question, the hot button/http push option for temp targets from an offline pebble is still not embeded in oref0 master correct? is it now a recommended approach already?
Martin Haeberli
@mhaeberli
May 03 2017 17:55
has anyone integrated pebble / fitbit / Basis … physical activity data as an input ? (some of these, I believe, have bluetooth input; some have web APIs…)
Scott Leibrand
@scottleibrand
May 03 2017 17:56
no, we haven't found any with a good easy way to get activity data in real time. even if we do, usefulness might be limited, because by the time the activity starts it's sorta too late to do a zero temp. you need the user to set an activity mode beforehand to avoid having to do a carb correction
if we could get < 5m detection of walking, though, that might help for being able to turn off basal / stop SMBing for less strenuous activity that is likely to massively slow down carb absorption.
if you can get a prototype integration going with one of those services, that's the kind of automatic activity mode is something I'd love to incorporate into the oref0 algorithms
Martin Haeberli
@mhaeberli
May 03 2017 17:58
@scottleibrand ok, thx; though at least in teens for example we have found persistent nighttime lows after an “active” day vs a “passive” day… anecdotal example - not our teen - kid plays soccer all day - tough management after bedtime unless bed time carbo loading or smart enough to set a much lower basal profile; our kid not as extreme - but a day of working at the barn with horses has a much different night time than a day at school
thx 4 suggestion re quick detection of walking; this also is a very relevant scenario here.
Scott Leibrand
@scottleibrand
May 03 2017 17:59
ah. so you're looking less at real-time reaction, and more as a way to inform autosens
Martin Haeberli
@mhaeberli
May 03 2017 17:59
@scottleibrand yes, indeed
Scott Leibrand
@scottleibrand
May 03 2017 17:59
it might be easier to update the autosens algorithm to, for example, compare 24h and 6h sensitivity, and go with whichever is more sensitive
Martin Haeberli
@mhaeberli
May 03 2017 18:00
:+1:
Geoff Aumaugher
@geoffro10
May 03 2017 18:00
@scottleibrand @danamlewis @mar5477 @Sgoldberg89 Another update. I renamed devicestatus to devicestatus-old and created a new devicestatus. IT'S WORKING!! So, obviously, there is something borked in the old stuff. I wonder if there is a way to export the whole thing and check it out in Excel or something like that.
Scott Leibrand
@scottleibrand
May 03 2017 18:01
once @sulkaharo is done porting @MilosKozak's autosens updates, we can re-test that and consider whether further improvements are warranted for more quickly detecting sudden-onset sensitivity
in the mean time, you can always set a long high temp target until autosens catches up
Luis Betancourt
@betluis
May 03 2017 18:25
@mhaeberli how do you manage the night after an active football day today?
I mean, with current open apps tools you see consistent lows right? I am interested in knowing how you have managed this situation
Martin Haeberli
@mhaeberli
May 03 2017 18:26
not football but horses - today we are not looping - moving towards looping. tight monitoring by parent(s); we often forget to stay ahead of the ball by looking at trends in the 3 hours before bed. soon we’ll set up better alarms that we have on the basic Dexcom … by looking for velocity
Luis Betancourt
@betluis
May 03 2017 18:27
I see
Martin Haeberli
@mhaeberli
May 03 2017 18:27
so plan to move towards autotune at least, probably even before looping regularly (have run Loop app before but not consistently)
Scott Leibrand
@scottleibrand
May 03 2017 18:28
if you're not looping yet, you'll probably find that autosens makes a huge difference there. it takes a few hours to get fully up to speed, but even after 3-4 hours it'll be detecting sensitivity. and in the mean time, the negative deviations will make it low temp even if it doesn't know about the increased sensitivity yet
Martin Haeberli
@mhaeberli
May 03 2017 18:28
:+1:
Scott Leibrand
@scottleibrand
May 03 2017 18:29
autotune doesn't help with day to day variations: that's more for getting your average basal rates and ratios dialed in, and then autosens can adjust from there
Martin Haeberli
@mhaeberli
May 03 2017 18:29
@scottleibrand understood - part of it, as you well know, is just consistently gathering not just BG data but pump data - temp basals, carbs, other treatments, etc (at least as I understand for autotune)
philosophically, I’d like to have “ambient autotune for dummies” - that is, set up an edison/explorer, have it follow her around for a week and gather the data, look at autotune’s suggestions, even no closed looping yet...
yes, a PR for me to write after I actually do it
Luis Betancourt
@betluis
May 03 2017 18:31
Scott if we have autosens would it mean temp target is not needed?
Martin Haeberli
@mhaeberli
May 03 2017 18:31
‘ambient’ because minimally invasive to current T1’s routine
Luis Betancourt
@betluis
May 03 2017 18:32
How do you visualize the use of temp targets with autosens together
Dana Lewis
@danamlewis
May 03 2017 18:32
@mhaeberli yep just set up the rig and let it read at night when she's in range. Leave maxiob at 0, etc.
Scott Leibrand
@scottleibrand
May 03 2017 18:32
@betluis temp targets are usually used for manually adjusting oref0's behavior for the next hour or so. autosens operates automatically over 24h worth of data
Dana Lewis
@danamlewis
May 03 2017 18:33
@betluis you may still want to use temp targets even if autosens is running
^ because that
Sgoldberg89
@Sgoldberg89
May 03 2017 18:35
@geoffro10 @scottleibrand @danamlewis quick update on the issue, I went back and deleted ~10 devicestatus entries and it's been working since. Sounds like Geoff had a good fix too, I'll try to next if it stops working again.
Scott Leibrand
@scottleibrand
May 03 2017 18:35
(And autotune makes little tweaks so that its actions are an average based on the last several weeks' worth of data.)
Did you make a copy of the bad devicestatus entries?
Martin Haeberli
@mhaeberli
May 03 2017 18:35
@scottleibrand @danamlewis so I assume it’s helpful not harmful to get broader time coverage than just overnights ...
with respect to autotune data gathering
Dana Lewis
@danamlewis
May 03 2017 18:36
@mhaeberli only if you also want it for low temping. Otherwise it'll pull full pump history when it gets into range so doesn't matter much
Geoff Aumaugher
@geoffro10
May 03 2017 18:36
@Sgoldberg89 How did you determine which to delete?
Dana Lewis
@danamlewis
May 03 2017 18:37
(Or in sense of getting used to remembering to carry rig, learning the logs, etc)
PieterGit
@PieterGit
May 03 2017 18:37
@geoffro10 have you tried robomongo to do mongo queries? i use that for debugging mongodb status sometimes
Sgoldberg89
@Sgoldberg89
May 03 2017 18:37
@geoffro10 Started from the last entry and just worked my way up the list until I felt I have proved my point
*had
Martin Haeberli
@mhaeberli
May 03 2017 18:37
@danamlewis thx
Geoff Aumaugher
@geoffro10
May 03 2017 18:38
@PieterGit No, haven't heard of it. I'll check it out. @Sgoldberg89 Ok, I'm going to switch them back and start deleting from the end of the old db.
PieterGit
@PieterGit
May 03 2017 18:40
Any volunteers with pi3 / ti usb / ww setup around? I'm looking for somebody to confirm/test my --ww_ti_usb_reset fix . I'm starting to believe it probably also can happen with non-WW pump btw . See openaps/oref0#456
i know @ceben80 , but don't know if there are any others willing to help
Luis Betancourt
@betluis
May 03 2017 18:41
Thanks all for the comments
Martin Haeberli
@mhaeberli
May 03 2017 18:48
quick dumb question - besides RTD - can i easily set zero-temp value from 0 to some de minimus - like 0.05 - our T1 likes to keep at least some basal running ...
Luis Betancourt
@betluis
May 03 2017 18:51
Back to the hot button application though, would it be possible to make it work directly from the pebble to the rig via Bluetooth without needing the android phone? The intention is to make it fully offline
Also can't be IOS?
Gina
@gclyon
May 03 2017 18:54
trying to take edison explorer bd (after march) offline with pebble... i don't have a 2nd battery (yet) but is there any way to tell if it is working before getting one? i plugged edison into pc via otg and G5 to uart (and disabled wifi on edison)... it does not pull BG's this way. Is this b/c the G5 needs the OTG port (that I'm using to ssh with)?
Scott Leibrand
@scottleibrand
May 03 2017 18:54
@mhaeberli as far as I can tell, there is no actual difference between a 0 and a 0.05 U/hr temporary basal rate. the pump actually delivers insulin as microboluses anyway, so a 0.05 U/hr rate just means "deliver tiny microboluses really infrequently". given that OpenAPS seldom sustains a zero temp basal for more than an hour, there's no real difference there AFAICT
Geoff Aumaugher
@geoffro10
May 03 2017 18:55
Hmmm.... so now I'm getting a quota error in Mongo. UGH... anyone know how to clean it up and/or to purge to get free space?
Scott Leibrand
@scottleibrand
May 03 2017 18:56
the only plausible argument I've heard for avoiding zero basal is to prevent sustained zero basal for 2h or more, because that might cause pump site occlusions. and in that case, the discussion ended up on the idea of actually limiting how long the zero basal could be run, rather than trying to keep a minimal basal level at all times.
but, in the end, the people arguing for that decided it wasn't necessary / worth the trouble, and never implemented it
Dana Lewis
@danamlewis
May 03 2017 19:01
@gclyon yes G5 needs to be in otg. Plug wall power if no second battery into uart and see if it sets temps for the next bit without wifi
JELCRAWFORD
@JELCRAWFORD
May 03 2017 19:03
@scottleibrand - in support of not doing anything, prior to OpenAPS we actually had scheduled, several times a day, a zero basal rate for a couple of hours each time - mid afternoon and long period overnight. No occlusions from that. We know of another family with a son that also had zero basal times when he was kindergarten through to grade 5 - again, no noted problems for them either. As always though, each will be different...
JELCRAWFORD
@JELCRAWFORD
May 03 2017 19:09
@scottleibrand ...and with Openaps we have temp basal of zero running periodically pretty often every day....
Martin Haeberli
@mhaeberli
May 03 2017 19:10
@scottleibrand thx ...
Scott Leibrand
@scottleibrand
May 03 2017 19:10
@JELCRAWFORD :+1: thx
Gina
@gclyon
May 03 2017 19:13
@danamlewis thx. it is indeed temping!
Dana Lewis
@danamlewis
May 03 2017 19:14
:+1: great! You can also set a higher or lower temp and watch it cancel, too, to confirm it's working offline
Gina
@gclyon
May 03 2017 19:18
so for the additional battery needed,.. same type necessary or would a smaller (phone charger battery) work?
well, i don't guess it'd really be any smaller in size. i'm just wanting to plug in a battery i have (phone) but don't want to screw anything up
Scott Leibrand
@scottleibrand
May 03 2017 19:20
it has to be a USB battery like you'd use to charge a phone. any size is fine, just a question of how long you want it to last. if you get a reasonably sized one, you can unplug the lipo from the rig and just use the external battery.
Gina
@gclyon
May 03 2017 19:20
ahh cool
Scott Leibrand
@scottleibrand
May 03 2017 19:21
if you're going to unplug the lipo, you'll probably want to get an external battery that supports pass-through charging, so it doesn't turn off your rig when you go to charge it
there are some recommendations in the docs for some brands that support that. most of those are larger ones big enough to power a Raspberry Pi rig: that's fine, or you can get a smaller one with the same capabilities.
Gina
@gclyon
May 03 2017 19:23
i do have one of those (from the pi days) Thx. Y'all are so great!
Dana Lewis
@danamlewis
May 03 2017 19:24
:+1: :)
garykidd
@garykidd
May 03 2017 19:53
If I was to switch from G5 Dex CGm to MM...do I have to do anything besides re-run setup script?
Dana Lewis
@danamlewis
May 03 2017 19:58
that should be all - and maybe run openaps first-upload if it doesn’t stick bgs into Nightscout
garykidd
@garykidd
May 03 2017 20:00
awaiting a replacement transmitter from Dex--tomorrow night but cant wait...
garykidd
@garykidd
May 03 2017 20:09
Understanding that Dex G5 is still more much more accurate than MM....in the real world I wouldn't be very surprised if it did a comparable job overall with looping....just pure speculation....
Scott Leibrand
@scottleibrand
May 03 2017 20:12
from what I've heard the MDT sensor is more sensitive to "calibration hygiene", but when well-calibrated it usually works fine for looping. definitely a YDMV thing, though.
garykidd
@garykidd
May 03 2017 20:21
calibration hygiene meaning that BG not moving rapidly while calibrating?
Scott Leibrand
@scottleibrand
May 03 2017 20:21
yeah, and calibrating every at least every 12 hours
garykidd
@garykidd
May 03 2017 20:23
Got it. I confess I've gone way longer with Dex and still been amazed at needing little to no adjustment once I've finally gotten around to it. Anyway---MM through tonight has got to better than flying blind while sleeping IMO.
I hadn't been aware that Dex G5 will now back fill missing BGs from up to 3hrs earlier upon reconnection...that's a serious improvement if its true of course.
dmdfreak
@dmdfreak
May 03 2017 20:35
So I'm on the latest Dev branch. Yay! But now I have another problem. I have the correct serial number in for the pump and it was working fine before I updated, but now the rig doesn't appear to be talking to the pump. No temp basels have been set. But Nightscout is show that the rig is communicating with the site. Anyone have any suggestions?
jmcrawford
@jmcrawford
May 03 2017 20:36
pump battery is ok?
dmdfreak
@dmdfreak
May 03 2017 20:36
Brand new. Had to change it last night.
jmcrawford
@jmcrawford
May 03 2017 20:37
can you run a manual command:
openaps use pump model
dmdfreak
@dmdfreak
May 03 2017 20:37
Tried that. Throws errors
jmcrawford
@jmcrawford
May 03 2017 20:37
what kind of errors?
dmdfreak
@dmdfreak
May 03 2017 20:40
That I can't say. I'm mobile but it was saying line errors in the code oddly enough. Maybe the install was corrupt?
Let me run it again. One sec
Scott Leibrand
@scottleibrand
May 03 2017 20:44
make sure you ran npm run global-install from ~/src/oref0 after updating
garykidd
@garykidd
May 03 2017 20:49
@scottleibrand if I rerun script will by BT tethering components remain
?
Scott Leibrand
@scottleibrand
May 03 2017 20:50
you'll need to provide the MAC address each time you run oref0-setup, but the pairing and trusting stuff should be unaffected by that
garykidd
@garykidd
May 03 2017 20:51
OK; cool. Thanks. Still waiting on the transmitter to complete charge---hope the thing still works.
dmdfreak
@dmdfreak
May 03 2017 20:57
myopenapsroot@timsedison:~/myopenaps# openaps use pump modelTraceback (most recent call last): File "/usr/local/bin/openaps-use", line 63, 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-use", line 57, in run output = app(args, self) File "/usr/local/lib/python2.7/dist-packages/openaps/uses/init.py", line 92, in call
return self.method.selected(args)(args, app) File "/usr/local/lib/python2.7/dist-packages/openaps/uses/init.py", line 31, in call
return self.method(args, app)
File "/usr/local/lib/python2.7/dist-packages/openaps/uses/use.py", line 44, in call
self.before_main(args, app) File "/usr/local/lib/python2.7/dist-packages/openaps/vendors/medtronic.py", line 62, in before_main self.setup_medtronic( ) File "/usr/local/lib/python2.7/dist-packages/mmeowlink/vendors/mmeowlink.py", line 115, in setup_medtronic setup_func(self) File "/usr/local/lib/python2.7/dist-packages/mmeowlink/vendors/mmeowlink.py", line 75, in setup_medtronic_link link = builder.build(radio_type, port) File "/usr/local/lib/python2.7/dist-packages/mmeowlink/link_builder.py", line 16, in build return SubgRfspyLink(port) File "/usr/local/lib/python2.7/dist-packages/mmeowlink/vendors/subg_rfspy_link.py", line 55, in init self.open() File "/usr/local/lib/python2.7/dist-packages/mmeowlink/vendors/serial_interface.py", line 23, in open
raise AlreadyInUseException("%s already in use" % self.device)mmeowlink.vendors.serial_interface.AlreadyInUseException: /dev/spidev5.1 already in use
Martin Haeberli
@mhaeberli
May 03 2017 21:02
that suggests to me that the cron jobs are already running, so you can’t talk in parallel to it over the command line
dmdfreak
@dmdfreak
May 03 2017 21:02
@jmcrawford
Martin Haeberli
@mhaeberli
May 03 2017 21:02
stop the cron jobs, try openaps pump use model, etc, then resume the cron jobs
dmdfreak
@dmdfreak
May 03 2017 21:09
Ok sorry. I was lagging with a bad connection. Just caught up in the chat feed
Martin Haeberli
@mhaeberli
May 03 2017 21:10
I got lost searching for instructions on how to flash the HamShield Explorer board from Edison (flash the CC1110 or whatever). tried searching RTD. Yes, I owe a PR to improve the instructions now that I know how they are supposed to work… but can’t even find them any more ...
Dana Lewis
@danamlewis
May 03 2017 21:11
openaps.readthedocs.io/en/latest/docs/Resources/troubleshooting.html?highlight=ccprog#could-not-get-subg-rfspy-state-or-version-have-you-got-the-right-port-device-and-radio-type
dmdfreak
@dmdfreak
May 03 2017 21:12
Ok. Did that and it returned 722 which is my pump model
garykidd
@garykidd
May 03 2017 21:19
Curious if anyone can share experience with expired sensors...specifically MM?
Martin Haeberli
@mhaeberli
May 03 2017 21:21
@danamlewis thx
PopcornPam
@PopcornPam_twitter
May 03 2017 21:23
@dmdfreak , we use a lot of expired sensors at my clinic in Haiti (mostly dexcom), and they seems to be fine, but we rarely try to get a full week, let alone multi weeks.
dmdfreak
@dmdfreak
May 03 2017 21:25
@garykidd ⬆️
garykidd
@garykidd
May 03 2017 21:26
copy that! thank you.
PieterGit
@PieterGit
May 03 2017 21:57
@scottleibrand @danamlewis : with a fresh rig i get: {"carbs":0,"reason":"not enough glucose data to calculate carb absorption"}. Is this caused by [report "monitor/glucose-oref0.json"] set to 30.0 minutes or will it disappear by itselves.
Scott Leibrand
@scottleibrand
May 03 2017 21:58
which setup was this again?
PieterGit
@PieterGit
May 03 2017 21:59
CGM=G4-upload, dev-branch, WW-pump, PI3, with G4 connected to USB and "adv_target_adjustments": true
dev-branch + openaps/oref0#456 to be exact.
Scott Leibrand
@scottleibrand
May 03 2017 22:01
the monitor/glucose-oref0.json contents get merged together as more data comes in
pi@pi2-erf ~/myopenaps-cgm-loop/monitor $ grep -c glucose glucose-zoned-merge.json
288
PieterGit
@PieterGit
May 03 2017 22:02
now on 47, so it's just a new rig problem i guess?
Scott Leibrand
@scottleibrand
May 03 2017 22:02
there is a command you can run to get a full 24h
openaps glucose-long maybe
no, openaps reset-glucose-merge
PieterGit
@PieterGit
May 03 2017 22:04
still 48
Scott Leibrand
@scottleibrand
May 03 2017 22:04
reset-glucose-merge ! bash -c "openaps glucose-long && rm -f monitor/glucose-zoned-merge.json && cp monitor/glucose-zoned-long.json monitor/glucose-zoned-merge.json && openaps report invoke monitor/glucose-raw-merge.json"
to find those I did: cd ~/myopenaps-cgm-loop and then openaps alias show
PieterGit
@PieterGit
May 03 2017 22:10
hitting calibration bug of eu dexcom:
pi@pi3:~/myopenaps-cgm-loop $ openaps reset-glucose-merge cgm://JSON/oref0_glucose/monitor/glucose-oref0-long.json reporting monitor/glucose-oref0-long.json tz://JSON/rezone/monitor/glucose-zoned-long.json reporting monitor/glucose-zoned-long.json raw://JSON/shell/monitor/glucose-raw-merge.json Could not require: /home/pi/myopenaps-cgm-loop/monitor/cal.json { [Error: Cannot find module '/home/pi/myopenaps-cgm-loop/monitor/cal.json'] code: 'MODULE_NOT_FOUND' } reporting monitor/glucose-raw-merge.json, but now 53
openaps extras fails with:
$ openaps extras
cgm://JSON/nightscout_calibrations/monitor/cal.json
monitor/cal.json  raised  Could not parse Calibration
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 40, in main
    output = task.method(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps/uses/use.py", line 45, in __call__
    output = self.main(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps/vendors/dexcom.py", line 830, in main
    results = super(iter_calibrations, self).main(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps/vendors/dexcom.py", line 420, in main
    for item in candidates:
  File "/usr/local/lib/python2.7/dist-packages/dexcom_reader/readdata.py", line 301, in iter_records
    records = list(self.ReadDatabasePage(record_type, x))
  File "/usr/local/lib/python2.7/dist-packages/dexcom_reader/readdata.py", line 267, in GenericRecordYielder
    yield record_type.Create(data, x)
  File "/usr/local/lib/python2.7/dist-packages/dexcom_reader/database_records.py", line 158, in Create
    return cls(unpacked_data, raw_data)
  File "/usr/local/lib/python2.7/dist-packages/dexcom_reader/database_records.py", line 180, in __init__
    self.check_crc()
  File "/usr/local/lib/python2.7/dist-packages/dexcom_reader/database_records.py", line 47, in check_crc
    raise constants.CrcError('Could not parse %s' % self.__class__.__name__)
dexcom_reader.constants.CrcError: Could not parse Calibration
garykidd
@garykidd
May 03 2017 22:15
an thoughts on why my NS site seems to have conked out.....cant see openaps or pump pills or some others??
Dana Lewis
@danamlewis
May 03 2017 22:16
@garykidd future treatments?
garykidd
@garykidd
May 03 2017 22:18
hmmm...not sure how to tell--I can see some data points but mostly it appears off scale of graph
garykidd
@garykidd
May 03 2017 22:23
@danamlewis I'm pretty sure this is almost it but NS admin section shows 0 future treatments....realizing now though that rig was down for some time recently...
Dana Lewis
@danamlewis
May 03 2017 22:24
time sync issues then?
garykidd
@garykidd
May 03 2017 22:24
where does that reveal itself...I suspect so.
katie disimone
@Kdisimone
May 03 2017 22:28
5.1 (with doubled the carbs) from last a1c of 5.7. Thanks OpenAPS.
Dana Lewis
@danamlewis
May 03 2017 22:28
@Kdisimone :fire: :confetti_ball: :heart: :heart_eyes: :raised_hands: go ya’ll!
garykidd
@garykidd
May 03 2017 22:34
congrats @Kdisimone
rerunning install script ...easy first step I suppose
Dana Lewis
@danamlewis
May 03 2017 22:35
@garykidd was your rig online? if it got back online, it probably would have fixed it’s time being out of sync if that was the case
JELCRAWFORD
@JELCRAWFORD
May 03 2017 22:37
Addtl info re: insulin purchase in Canada - Alberta - this may vary by province, but I had the opportunity to ask a parent at school who is a pharmacist - she was quite clear that a prescription is NOT required. They do sometimes refuse people who are very obviously using insulin as a growth hormone. So presuming you are rolling with pump attached, you will likely not have a problem getting Fiasp. She did not know, off hand, the pricing point. She indicated reimbursement by insurance might be a problem for those of you looking for it out of your home country, but access generally would not be a problem.
garykidd
@garykidd
May 03 2017 22:46
hoping that when I have some BGs actually rolling in from MDT CGM rig/NS will rectify itself...right now MDT retrieve errors only so I'm presuming that's what's supposed to happen in warmup state...
Dana Lewis
@danamlewis
May 03 2017 22:47
ohhh that’s why you’re having bugs
yea, can’t loop during warmup
might also have to run openaps first-upload once your sensor is awake
RedCyclist1
@RedCyclist1
May 03 2017 23:44
@danamlewis What cords are folks using to do offline looping between rig and dexcom? I am looking for something short and mini to mini. What are other folks using for this?
Actually that is an open question for everyone. :)
garykidd
@garykidd
May 03 2017 23:49
@danamlewis :smile: up and running....just had to have some BGs roll in...thx.
PieterGit
@PieterGit
May 03 2017 23:52
any packet guru's for dexcom G4 / dexcom_reader which are able to help me with openaps/dexcom_reader#16 (problem with calibration records with a Dexcom G4 connected by USB)