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

7th
Jun 2016
Paul Martin
@Spazholio
Jun 07 2016 00:00
@JaysonEwer Your app suggestion worked PERFECTLY. I’m connecting to the internal IP instead of bouncing through my public site. I’m still gonna set up the autossh option as well, but this is a great fallback.
ryanchen26
@ryanchen26
Jun 07 2016 00:00
@scottleibrand Perhaps in the near future. Once I feel more confident about all this I'll double back to contribute to the docs. Still planning to make a flow chart...once understanding is 100%.
Scott Leibrand
@scottleibrand
Jun 07 2016 00:02
@ryanchen26 for this kind of thing I don't think you need a full understanding to usefully contribute. Just calling out something that other people are likely to get stuck on, and a fix or workaround, is useful.
Paul Martin
@Spazholio
Jun 07 2016 00:07
As someone fairly new to this, I’ll second that. In some respects, I’m getting a clue, but in others, I am completely lacking. And I figure if I’M getting tripped up, someone else probably will too. Doing some PRs has helped me get a better grasp on what I’m doing. And besides, if you got something wrong, @danamlewis and @scottleibrand will helpfully tell you how to fix it before they merge. =)
Jason Calabrese
@jasoncalabrese
Jun 07 2016 00:07
instead of /api/v1/entries.json you can use /api/v1/entries/sgv.json (same result as the find[type]=sgv
Ben West
@bewest
Jun 07 2016 00:07
ah forgot about that feature :-)
openaps use ns shell get entries/sgv.json count=288 for that
Jason Calabrese
@jasoncalabrese
Jun 07 2016 00:09
last night when updating 1 of the nighttime rigs while running auto-sens failed, then the empty json made the loop fail, think we really need to pull out those new errors
Paul Martin
@Spazholio
Jun 07 2016 00:09
I imagine it’s the same on the backend, right? Meaning one isn’t really more effective than the other?
s/effective/efficient
Jason Calabrese
@jasoncalabrese
Jun 07 2016 00:09
overall they make the loop less reliable
yeah, it's the same on the backend, the mongo query will be the same
Dana Lewis
@danamlewis
Jun 07 2016 00:10
:+1: @Spazholio (speaking of which, I know I owe you a comment on your usability PR :))
Paul Martin
@Spazholio
Jun 07 2016 00:10
No rush. =)
Ben West
@bewest
Jun 07 2016 00:10
@Spazholio it's just a matter of one being easier to type because the designers love you
Paul Martin
@Spazholio
Jun 07 2016 00:11
Unrelated to the errors @jasoncalabrese is talking about - if my pumphistory-24h.json report fails, should the entire loop fail? Because mine does. I imagine it’s because it can’t run auto-sens.json but should that be a loop-killer?
Ben West
@bewest
Jun 07 2016 00:13
opinions tend to vary
my feeling was yes, I want to expose the edge cases and error conditions so we can find them and stomp them out rather than allow them to potentially cause undefined behavior
Paul Martin
@Spazholio
Jun 07 2016 00:14
Maybe something that shows something up on NS? An “error pill” or something?
Dana Lewis
@danamlewis
Jun 07 2016 00:19
Stomp :)
Paul Martin
@Spazholio
Jun 07 2016 00:19
Is that “squashing a bug of some sort” or “that’s a hard ‘no’”?
Ben West
@bewest
Jun 07 2016 00:23
there are ways to influence the aliases to either blow up or not
sure, different people are expressing different opinions on when and where and how to blow up
for meals and autosens, if these are optional features, it makes sense to have like a --meals meals.json
I can add this very easily
fjungman
@fjungman
Jun 07 2016 00:30
Ok, this one is killing me. When I run the individual reports that make up this alias, they work. When I run the alias (as this command), it fails out. What hurts is that it says the report is invalid and to choose from the list, which lists the exact report . Doh!
pi@raspberrypi:~/vmax_aps $ openaps format-latest-nightscout-treatments && test $(json -f upload/latest-treatments.json -a created_at eventType | wc -l ) -gt 0 && ( openaps report invoke nightscout/recent-treatments.json && openaps report invoke nightcout/uploaded-missing-treatments.json ) || echo \"No recent treatments to upload\"
ns://JSON/shell/nightscout/recent-treatments.json
reporting nightscout/recent-treatments.json
usage: openaps-report invoke [-h]
{enact/enacted.json,enact/suggested.json,last_four_cgm_hours.json,last_four_pump_hours.json,monitor/battery.json,monitor/clock-zoned.json,monitor/clock.json,monitor/glucose.json,monitor/iob.json,monitor/ns-glucose.json,monitor/pumphistory-zoned.json,monitor/pumphistory.json,monitor/reservoir.json,monitor/status.json,monitor/temp_basal.json,monitor/upload-status.json,nightscout/recent-treatments.json,nightscout/uploaded-missing-treatments.json,nightscout/uploaded-status.json,settings/basal_profile.json,settings/bg_targets_raw.json,settings/insulin_sensitivities_raw.json,settings/model.json,settings/profile.json,settings/settings.json}
[{enact/enacted.json,enact/suggested.json,last_four_cgm_hours.json,last_four_pump_hours.json,monitor/battery.json,monitor/clock-zoned.json,monitor/clock.json,monitor/glucose.json,monitor/iob.json,monitor/ns-glucose.json,monitor/pumphistory-zoned.json,monitor/pumphistory.json,monitor/reservoir.json,monitor/status.json,monitor/temp_basal.json,monitor/upload-status.json,nightscout/recent-treatments.json,nightscout/uploaded-missing-treatments.json,nightscout/uploaded-status.json,settings/basal_profile.json,settings/bg_targets_raw.json,settings/insulin_sensitivities_raw.json,settings/model.json,settings/profile.json,settings/settings.json} ...]
openaps-report invoke: error: argument report: invalid choice: 'nightcout/uploaded-missing-treatments.json' (choose from u'enact/enacted.json', u'enact/suggested.json', u'last_four_cgm_hours.json', u'last_four_pump_hours.json', u'monitor/battery.json', u'monitor/clock-zoned.json', u'monitor/clock.json', u'monitor/glucose.json', u'monitor/iob.json', u'monitor/ns-glucose.json', u'monitor/pumphistory-zoned.json', u'monitor/pumphistory.json', u'monitor/reservoir.json', u'monitor/status.json', u'monitor/temp_basal.json', u'monitor/upload-status.json', u'nightscout/recent-treatments.json', u'nightscout/uploaded-missing-treatments.json', u'nightscout/uploaded-status.json', u'settings/basal_profile.json', u'settings/bg_targets_raw.json', u'settings/insulin_sensitivities_raw.json', u'settings/model.json', u'settings/profile.json', u'settings/settings.json')
"No recent treatments to upload"
This line:
openaps-report invoke: error: argument report: invalid choice: 'nightcout/uploaded-missing-treatments.json'
then it lists that exact option...
Scott Leibrand
@scottleibrand
Jun 07 2016 00:32
@jasoncalabrese want to put together a PR to fix the errors?
Paul Martin
@Spazholio
Jun 07 2016 00:34
@bewest This is a question that I doubt you can answer right off, but ever since changing that report, my loop fails. It can’t find monitor/glucose.json which is weird since I can run the report manually, as all of the raw files are there to create it. Do you have ANY idea why that simple change might have thrown things off?
Paul Martin
@Spazholio
Jun 07 2016 00:39
I…nevermind? I did literally nothing different, but my loop failed 8-9 times, and then magically started working.
:angry:
Paul Martin
@Spazholio
Jun 07 2016 00:47
It “feels” less reliable now though. That’s really, REALLY weird.
Paul Martin
@Spazholio
Jun 07 2016 01:02
@bewest I think I found a bug! I know you’re so happy to hear that! =)

Here’s my monitor/glucose.json:

[report "monitor/glucose.json"]
use = select                                                                                                                                                                                                                                                                                                          
reporter = JSON
no_timezone = False
current = now 
date = dateString
gaps =
input = raw-ns/ns-glucose-zoned.json raw-cgm/glucose-zoned.json
device = tz
timezone =
prev =

However, I’m getting down-to-the-second dupes in my monitor/glucose.json file.

Jason Calabrese
@jasoncalabrese
Jun 07 2016 01:07
@scottleibrand I just removed the lines from my fork, but I think @bewest has other ideas
Paul Martin
@Spazholio
Jun 07 2016 01:07
@bewest Belay that. There’s still oddness, but I don’t think it’s dupes...
Jason Calabrese
@jasoncalabrese
Jun 07 2016 01:08
jasoncalabrese/oref0@49eaf8c
I'd like to go further and be smarter about the failing when BG data is old, instead of doing nothing it should cancel any high temps
we need a better way to report errors so they don't get buried in the logs
Brooke Pazoles
@PazolesBrooke_twitter
Jun 07 2016 01:12
When APS is in bolus snooze, is it running a 0% basal or the neutral/regular basal?
Jason Calabrese
@jasoncalabrese
Jun 07 2016 01:13
the bolus snooze prevents some low temping, you can think about a bolus wizard subtracting the bolus snooze
without bolus snooze the loop would 0 temp after any large meal bolus
Brooke Pazoles
@PazolesBrooke_twitter
Jun 07 2016 01:14
Even if the BG levels are low?
Example - tonight I overdid my dinner bolus just a bit and was dropping in the 60's while still in bolus snooze. From what I understood, APS wanted me to keep a neutral temp, but it seems like I should have gone to a 0 basal
Jason Calabrese
@jasoncalabrese
Jun 07 2016 01:15
no, at some point no amount of bolus snooze will prevent a low temp
Paul Martin
@Spazholio
Jun 07 2016 01:15

Ok, so I took the output of the auto-sens calculation with the debug output that @scottleibrand suggested. Doing that I see three entries for 8:27pm:

-Mon Jun 06 2016 20:27:23 GMT-0400 (EDT)
-Mon Jun 06 2016 20:27:06 GMT-0400 (EDT)
>Sun Jun 05 2016 20:27:10 GMT-0400 (EDT)

I think I see why. My raw NS files have a dateString that looks like "dateString": "2016-06-06T20:27:23+00:00” and my raw CGM files look like this "dateString": "2016-06-06T20:27:06-04:00”. So the select is seeing ALL of them as discrete entities.

Jason Calabrese
@jasoncalabrese
Jun 07 2016 01:16
in the 60's is should have 0 temped no matter what
Brooke Pazoles
@PazolesBrooke_twitter
Jun 07 2016 01:16
Whats the threshold for 0 temping?
We were testing out a wifi hotspot when this happened - APS was delayed I think
Jason Calabrese
@jasoncalabrese
Jun 07 2016 01:17
sometimes when rising it will cancel a 0 temp in the 60's for us, but not when dropping
Brooke Pazoles
@PazolesBrooke_twitter
Jun 07 2016 01:17
Got it. Thanks Jason!
Jason Calabrese
@jasoncalabrese
Jun 07 2016 01:17
the threshold is based on your targets but it's about 80
Brooke Pazoles
@PazolesBrooke_twitter
Jun 07 2016 01:18
Thanks - that all makes sense
Matt Pazoles
@Pazoles
Jun 07 2016 01:19
Should it be printing a basal recommendation in the APS pill when it's under that threshold?
Jason Calabrese
@jasoncalabrese
Jun 07 2016 01:19
yes
and you'll see the threshold then
Matt Pazoles
@Pazoles
Jun 07 2016 01:19
Screen Shot 2016-06-06 at 9.18.35 PM.png
Jason Calabrese
@jasoncalabrese
Jun 07 2016 01:20
yeah, 64 < 70
should have low temped there
unless it thought there was already a low temp
Matt Pazoles
@Pazoles
Jun 07 2016 01:21
Cool, suggested.json has rate: 0 duration: 30, just wanted to make sure that there wasn't supposed to be something explicit in the APS pill.
pi@dougstamper:~ $ cat aps/enact/suggested.json
{"temp":"absolute","bg":64,"tick":"+0","eventualBG":88,"snoozeBG":59,"predBGs":{"IOB":[64,63,62,61,60,58,57,56,55,54,53,52,51,50,49,49,48,48,48,48,48,48,49,49,49,50,50,50,50,50,51],"aCOB":[64,71,77,83,89,94,99,103,107,110,113,116,118,120,121,122,122,122,122,122,122,122,123,123,123,124,124,124,124,125],"COB":[64,63,62,61,61,60,60,59,59,59,59,59,60,60,61,62,62,63,65,66,67,69,70,72,73,75,76,77,79,80,81,83,84,85,86,87,88]},"reason":"COB: 16, Dev: 9, BGI: -2.46, ISF: 60, Target: 95; BG 64<70","duration":30,"rate":0}
Thanks
fjungman
@fjungman
Jun 07 2016 01:22
found my issue! missing an "s" in nightscout. Really, really hard to see...
Jason Calabrese
@jasoncalabrese
Jun 07 2016 01:22
should have show enacted at that point
Paul Martin
@Spazholio
Jun 07 2016 01:22
@fjungman Good catch. Those can be the hardest to find sometimes. =)
Jason Calabrese
@jasoncalabrese
Jun 07 2016 01:22
maybe the suggest worked, but it didn't get to enact?
Ben West
@bewest
Jun 07 2016 01:24
@Spazholio use tz rezone -a missing --astimezone should fix that
Paul Martin
@Spazholio
Jun 07 2016 01:25
They both appear to have that already?
[report "raw-cgm/glucose-zoned.json"]
use = rezone
reporter = JSON
astimezone = True
date = timestamp dateString start_at end_at created_at                                                                                                                                                                                                                                                                
adjust = missing
input = raw-cgm/glucose.json
device = tz
timezone =
[report "raw-ns/ns-glucose-zoned.json"]
use = rezone                                                                                                                                                                                                                                                                                                          
reporter = JSON
astimezone = True
date = timestamp dateString start_at end_at created_at
adjust = misssing
input = raw-ns/ns-glucose.json
device = tz
timezone =
Ben West
@bewest
Jun 07 2016 01:26
for me that adjusts the hours and the zone, both
Paul Martin
@Spazholio
Jun 07 2016 01:28
GRRRRRRRRRRRRRRRRR
Thank you, @fjungman.
It helps to spell “misssing” correctly.
Ben West
@bewest
Jun 07 2016 01:28
oh, nice
using the tools would have highlighted the issue:
Paul Martin
@Spazholio
Jun 07 2016 01:29
That’s certainly one word for it. I’m currently saying quite a few others...
Ben West
@bewest
Jun 07 2016 01:29
openaps report show --cli raw-ns/ns-glucose-zoned.json
and then running what it prints would probably produce an error
Paul Martin
@Spazholio
Jun 07 2016 01:30
Yep, it does.
I swear, I’m about to create a sanity checker to find stupid things like that...
(because I know me - it’s gonna happen again)
Ben West
@bewest
Jun 07 2016 01:31
the sanity checker is to use the tools rather than editing the file :-)
report add would not have allowed misssing
Paul Martin
@Spazholio
Jun 07 2016 01:31
Get outta here with your “rationality” and “logic”.
Ben West
@bewest
Jun 07 2016 01:31
lol
Paul Martin
@Spazholio
Jun 07 2016 01:31
Well, I already added the report. I just had to tweak it due to something you suggested when things were wonky.
Wait…this is YOUR FAULT!
(somehow)
Ben West
@bewest
Jun 07 2016 01:32
my advice was to always try the use until it's right
then copy paste that into report add
Paul Martin
@Spazholio
Jun 07 2016 01:33
Yeah, yeah. You’re absolutely right, I don’t disagree. Sometimes it’s just quicker to do things manually. But…it leads to issues like this.
Lesson learned.
Paul Martin
@Spazholio
Jun 07 2016 01:39
Still having the issue, but I think I might see why. openaps report show --cli raw-ns/ns-glucose-zoned.json shows openaps use tz rezone raw-ns/ns-glucose.json. Note how the other fields aren’t in there? Is that because they’re the defaults, or because of something else?
Do I need to re-add the report? Is it actually using what the report show —cli is showing rather than what’s in openaps.ini?
Ben West
@bewest
Jun 07 2016 01:44
thinking...
actually, try sudo pip install -U openaps-contrib :-)
think there was a bug I fixed yesterday
Paul Martin
@Spazholio
Jun 07 2016 01:45
Hey, that shows the right stuff now…almost: openaps use tz rezone --date timestamp --date dateString --date start_at --date end_at --date created_at raw-ns/ns-glucose.json. I don’t see the astimezone thing now. Is “missing” the default, so it’s not shown in the report show —cli?
Not astimezone (which I also don’t see) - I meant the adjust.
Ben West
@bewest
Jun 07 2016 01:48
if you add --astimezone to the use, does it change the results?
Dan Piessens
@dpiessens
Jun 07 2016 01:48
@bewest Have you tried your gap filler on data that's getting uploaded to NS by other source (in my case xdrip) and then pushing the polyfill back?
Paul Martin
@Spazholio
Jun 07 2016 01:48
It does, thank you.
Should I worry about the “missing” bit being….well, missing?
Ben West
@bewest
Jun 07 2016 01:49
not if the use is giving you right results
need to add that use as report
Paul Martin
@Spazholio
Jun 07 2016 01:50
I was more concerned that the report show —cli isn’t reflecting something that’s in the openaps.ini but I agree - if it’s working, quit messing. =)
Ben West
@bewest
Jun 07 2016 01:50
yes, I'm using lsgaps to select my openxshareble records to upload to NS
Paul Martin
@Spazholio
Jun 07 2016 01:51
I’ve re-added the report, and waiting for the loop to run again.
Ben West
@bewest
Jun 07 2016 01:51
the --cli is intended to show the command needed to run what you put in, or what would get you the equivalent use as what's in the report
so if there's a default option, I think it won't repeat it if that's the one used
Dan Piessens
@dpiessens
Jun 07 2016 01:52
ok I think the order is raw -> tz rezone -> gaps -> select, is that correct?
Paul Martin
@Spazholio
Jun 07 2016 01:52
Ahhhh, ok. So if I put something like garbage = True then I shouldn’t expect to magically see a —garbage True appear in the report show —cli?
Ben West
@bewest
Jun 07 2016 01:52
yeah, I think gaps is usually run on eg NS + dexcom
so you get the gap from NS and apply it to the dexcom source to get the diff
it's actually built into the ns shell upload-non-empty-treatments thing
er, upload-non-empty-type
      openaps use ns shell lsgaps ${ZONE} ${TYPE} \                   
        |  openaps use ${ZONE} select --current now --gaps - ${FILE}      | json
this snippet is built into upload-non-empty-type
Dan Piessens
@dpiessens
Jun 07 2016 01:54
ok nice
Ben West
@bewest
Jun 07 2016 01:55
it fetches recent data from NS, runs it through lsgaps, selects the found gaps against the provided input file
Paul Martin
@Spazholio
Jun 07 2016 01:57
Argh.
openaps use tz rezone --date timestamp --date dateString --date start_at --date end_at --date created_at raw-ns/ns-glucose.json
That’s what the —cli shows. Still. I re-added the report.
Dan Piessens
@dpiessens
Jun 07 2016 01:58
that makes sense, I'll tinker :)
Paul Martin
@Spazholio
Jun 07 2016 01:59
openaps report add raw-ns/ns-glucose-zoned.json JSON tz rezone --astimezone --date timestamp --date dateString --date start_at --date end_at --date created_at raw-ns/ns-glucose.json <— that’s it, right?
Ben West
@bewest
Jun 07 2016 01:59
looks good
Paul Martin
@Spazholio
Jun 07 2016 01:59
openaps report show --cli raw-ns/ns-glucose-zoned.json still shows openaps use tz rezone --date timestamp --date dateString --date start_at --date end_at --date created_at raw-ns/ns-glucose.json
Ben West
@bewest
Jun 07 2016 01:59
hmm
Paul Martin
@Spazholio
Jun 07 2016 02:00
I mean, I can go into the openaps.ini and manually delete the report and then re-add, if you think it would help.
Ben West
@bewest
Jun 07 2016 02:01
ok
well you found yet another bug, is all
Paul Martin
@Spazholio
Jun 07 2016 02:02
Same result.
Deleted, report added, still no —astimezone showing.
Dan Piessens
@dpiessens
Jun 07 2016 02:02
other note: I noticed you can upload mmtune to ns... is there value in that data?
Ben West
@bewest
Jun 07 2016 02:09
ok, sudo pip install -U openaps-contrib
Paul Martin
@Spazholio
Jun 07 2016 02:09
Hey, there we go. Things are looking better now. Thanks.
Ben West
@bewest
Jun 07 2016 02:12
:-)
Paul Martin
@Spazholio
Jun 07 2016 02:13
I mean, everything’s rezoned properly, but there’s still somethiing wrong with the auto-sens.json or some reason, but I’m going to dig into that a bit...
I now have 1174 sgv entries in monitor/glucose.json. But at least they’re zoned properly! =)
It looks like the select is just adding both of the files together, rather than merging them intelligently. Could that be happening?
Here’s the —cli for monitor/glucose.json: openaps use tz select raw-ns/ns-glucose-zoned.json raw-cgm/glucose-zoned.json --date “dateString”. Does anything jump out at you?
Paul Martin
@Spazholio
Jun 07 2016 02:21
brb
Ben West
@bewest
Jun 07 2016 02:29
that's what it should do
Paul Martin
@Spazholio
Jun 07 2016 02:41
It’s supposed to mix ‘em? I thought it was supposed to merge them, but not merge dupes?
Paul Martin
@Spazholio
Jun 07 2016 02:49
If it’s NOT supposed to merge them intelligently, is there anything that we have that can take both sources and merge them together so that there aren’t any dupes?
Dana Lewis
@danamlewis
Jun 07 2016 02:58
Any loopers in NOLA area want to meet up Saturday or Sunday? Ditto for any NYC area loopers on Wednesday night?
*next Wed night, June 15
Dominick Gallegos
@dominickgallegos
Jun 07 2016 03:00
@danamlewis do you guys ever meet up in Boston?!
Dana Lewis
@danamlewis
Jun 07 2016 03:01
We did a few months ago when we were in town! There are a few Boston area folks who might be open to ya'lls own meetup
Dominick Gallegos
@dominickgallegos
Jun 07 2016 03:03
Any idea who they are? If not I'll search them out on FB - CGM on the Cloud.
Dana Lewis
@danamlewis
Jun 07 2016 03:05
@ochenmiller is one who spoke up last time :) and a few others spoke up about being willing to drive to Boston with enough notice!
Lynne Butts
@flutelynne_twitter
Jun 07 2016 03:08
I started looping live this evening. If I want to make some pump setting changes, what is the best method? sudo shutdown -h now, then unplug the battery or the carelink stick?
Dan Piessens
@dpiessens
Jun 07 2016 03:08
21:29
that's what it should dodnjttvihkthvnftvirtcgijhvejuccr
Brian Bonner
@itsn0tm3
Jun 07 2016 03:09
@billstackpole Hey Bill- Did you get your slice of radio working on a Pi3 or was it a Pi2?
Dana Lewis
@danamlewis
Jun 07 2016 03:09
@flutelynne_twitter you can just change them on the pump, then your next settings read should pull them in
Lynne Butts
@flutelynne_twitter
Jun 07 2016 03:09
@danamlewis even if it is looping?
Dana Lewis
@danamlewis
Jun 07 2016 03:09
Yep.
Lynne Butts
@flutelynne_twitter
Jun 07 2016 03:10
Great! I have to make my targets higher for now
Dana Lewis
@danamlewis
Jun 07 2016 03:10
Just know the settings changes won't be part of the loop until it next reads the full settings, depending on how you've built your loop
:) good plan
David Cintron
@loudestnoise
Jun 07 2016 03:17
@danamlewis Ah too bad not last weekend in NOLA! I was just there!
Dana Lewis
@danamlewis
Jun 07 2016 03:17
:)
Lynne Butts
@flutelynne_twitter
Jun 07 2016 03:19
My brain is dumb right now because I am 59 and dropping (I ate some carbs, just waiting). If I stop my ssh (sudo shutdown, etc.) is my loop going to keep running? Or should i just let my computer shut it off after it sleeps?
Paul Dickens
@thebookins
Jun 07 2016 03:20
Hi, New to this and just setting out. Wondering if there are any loopers in Australia happy to chat about their experiences?
Dana Lewis
@danamlewis
Jun 07 2016 03:20
If your loop is built to run offline, it will keep going even if you are not watching it through ssh. If you shut down the Pi, the loop will not work.
@thebookins quite a few Australians who have closed the loop, and several recently who I'm sure will chime in here :) @johnmales & others
Paul Dickens
@thebookins
Jun 07 2016 03:23
@danamlewis Great! Thanks!
Lynne Butts
@flutelynne_twitter
Jun 07 2016 03:23
I will just let the ssh time out on its own then?
Dana Lewis
@danamlewis
Jun 07 2016 03:23
based on the privacy settings people told me, not sure which of them mind being pointed out publicly in a case like this. If you're an AU looper who doesn't mind, let me know for situations like this ^ so we can tag you/connect you with others
Brian Bonner
@itsn0tm3
Jun 07 2016 03:26
ha if you get another deaf looper, please do feel free to point me out ;)
Ben West
@bewest
Jun 07 2016 03:31
you can exit ssh
it's fine
or close the window
the pi will run on it's own
Lynne Butts
@flutelynne_twitter
Jun 07 2016 03:32
@bewest Thanks! Closing the window is easiest for now.
Paul Martin
@Spazholio
Jun 07 2016 03:34
So @bewest - did I misunderstand what the tz select command does? It just seems to append one file onto the next. Or at least, it’s not deduping. =)
Lynne Butts
@flutelynne_twitter
Jun 07 2016 03:36
Nightscout keeps not being there with data. Went for a walk and hotspotted with my phone, all was fine. Came home, turned off hotspot, but no data on nightscout for 2 hours. I unplugged and replugged carelink first. Then I did openaps do-everything. And openaps report-nightscout. Nothing changed until I restarted the dynos in heroku. But now it is off again (21 mins), and the temp basals that showed before are not there. Is there a good way to refresh this?
Jarred Yaw
@jyaw
Jun 07 2016 03:42
Upgraded to autosens and AMA tonight! I see the purple lines! Not to bad of a process. The configure autosens and AMA doc out there was a great help in making the change.
Matt Walker
@walker0
Jun 07 2016 03:42
@thebookins I just started looping recently and extra bonus i'm Australian!
Paul Martin
@Spazholio
Jun 07 2016 03:43
Congrats to both of you!
Ben West
@bewest
Jun 07 2016 03:45
@Spazholio tz select is expecting to cull data from multiple files matching some kind of --gap specification
so if there were --gap file fed, it would select only the data matching the gaps
we had it like that earlier but you seemed unhappy it didn't include any data
Dana Lewis
@danamlewis
Jun 07 2016 03:46
@jyaw woohoo! PRs still welcome though to improve the process :)
Paul Martin
@Spazholio
Jun 07 2016 03:46
Yeah, that was part of the problem. Well, let me ask this - how do YOU generate your monitor/glucose.json?
My intention was to create a whole picture of 288 entries from whatever devices I might or might not have connected/downloaded, so I can generate the most accurate auto-sens.json possible.
And then I could upload whatever to NS - which would likely be a very small amount, since the Dexcom has share throwing things into NS, and I don’t do a lot on the pump itself.
Lynne Butts
@flutelynne_twitter
Jun 07 2016 03:49
@bewest Restarted dynos again, and that brought everything back. Thank you. I will see how the rest of the night goes.
Jarred Yaw
@jyaw
Jun 07 2016 03:51
@danamlewis ill take a look in the morning to see if anything I did was different but I pretty much used https://github.com/openaps/docs/blob/wip/ama-docs/docs/docs/walkthrough/phase-6/Configure-Automatic-Sensitivity-Mode-and-Advanced-Meal-Assist.md, i did cheat a bit with Scott's setup.sh for AMA but I still had to make edits for my carelink setup. I'm not exactly sure if the descriptors of the visualization or autosens output are right but they seemed reasonable :)
My first "eating soon" mode shall be tomorrow!
Dana Lewis
@danamlewis
Jun 07 2016 03:55
Still, any improvements would be great, we need as much English as possible so converting some of the setup.sh to English for the main docs is what we are hoping someone will do :)
Yay for eating soon! :)
Ben West
@bewest
Jun 07 2016 03:56
mine is very simple, I'm not mixing and matching data sources, and I'm not using auto-sens or AMA yet
$ openaps report show monitor/glucose.json  --cli
openaps use tz rezone --date timestamp --date dateString --date start_at --date end_at --date created_at --date system_time --date display_time --astimezone raw-cgm/glucose-raw.json
Dana Lewis
@danamlewis
Jun 07 2016 03:56
)(when we say English, we mean plain language instead of script files :))
Paul Martin
@Spazholio
Jun 07 2016 03:56
Ok, so you ALWAYS have your Dexcom connected?
Ben West
@bewest
Jun 07 2016 03:56
this is using ble/g4+share
the battery actually just died a few hours ago
Paul Martin
@Spazholio
Jun 07 2016 03:57
And that’s using openxshareble?
Ben West
@bewest
Jun 07 2016 03:58
yeah
Paul Martin
@Spazholio
Jun 07 2016 03:58
But then I can’t pair to my phone. So does openxshareble also post it to share? Is there any downside that you can think of?
Dana Lewis
@danamlewis
Jun 07 2016 03:59
It only pairs to one device
So you can pair with share, or you can pair with ble. (Or carry two receivers)
But you can't stuff data into share if you're doing ble
Ben West
@bewest
Jun 07 2016 04:00
I don't use dexcom share at all
Paul Martin
@Spazholio
Jun 07 2016 04:00
Right. But pairing my phone to my receiver, the iPhone app then posts my data to share...
Dana Lewis
@danamlewis
Jun 07 2016 04:00
Yes. I'm saying you have to choose
Paul Martin
@Spazholio
Jun 07 2016 04:00
Ahhhhh, ok that’s a bit of a drawback for how I currently do stuff. My wife likes to see my numbers in the app. =)
Ben West
@bewest
Jun 07 2016 04:00
not really, we could fairly easily upload to share also
Paul Martin
@Spazholio
Jun 07 2016 04:00
Are there any tools that upload to share?
Ben West
@bewest
Jun 07 2016 04:00
not sure, actually
Dana Lewis
@danamlewis
Jun 07 2016 04:01
@Spazholio yep trade offs. But you could do ble and then set up Nightscout so she could watch in. The NS app instead of share
Paul Dickens
@thebookins
Jun 07 2016 04:01
@walker0 thanks Matt!
Dana Lewis
@danamlewis
Jun 07 2016 04:02
She'd also then have insight to your loop activity, which might be helpful in case you're dropping/rising quickly
Paul Martin
@Spazholio
Jun 07 2016 04:02
True.
Does openxshareble prompt somewhere for the “mandatory" twice-daily calibrations?
Ben West
@bewest
Jun 07 2016 04:03
yeah, NS is showing everything that matters whereas share is showing a slice
the receiver itself does
it doesn't change using the receiver at all
and it's wireless, so that's nice
ah, there might be alarms for that
Paul Martin
@Spazholio
Jun 07 2016 04:05
Sorry for the barrage of questions, but does openxshareble also show more info than share? The calibration stuff, or the “noisy” entries?
Ben West
@bewest
Jun 07 2016 04:05
yeah, the uses are identical to usb
so it does raw and everything assuming g4
only difference from builtin dexcom is that it's wireless
Paul Martin
@Spazholio
Jun 07 2016 04:06
...
sigh
I suppose I know what I’m trying out tomorrow...
That also answers some questions I had about switching to the Edison. =)
Dana Lewis
@danamlewis
Jun 07 2016 04:07
;)
Paul Martin
@Spazholio
Jun 07 2016 04:34
Alas, I had to disable BT on the Pi3, so I might be switching sooner than I thought. =)
John Males
@johnmales
Jun 07 2016 04:57
@thebookins Hi Paul, yep my son (aka "the dude" as kids are usually referred to here) looping for 6 months. I’m in Sydney. Happy to chat here or privately :)
Jim is CrushingT1D
@CrushingT1D
Jun 07 2016 06:15
Now we are talking! Great to hear AU loopers are on the rise!! I am arriving in Sydney in a couple weeks :) @thebookins @johnmales
Hope to hear from the others too :crocodile:
Christer Jensen
@christerjensen
Jun 07 2016 06:37
FYI: I uploaded the Omnipod PDM ARM Firmware to omniaps.slack.com - also a IDB database of the firmware. Anyone good at analyzing these data?
Jason Calabrese
@jasoncalabrese
Jun 07 2016 07:11
@danamlewis I'll be in NYC next week, but not sure what the work plans are yet
live4sw
@live4sw
Jun 07 2016 12:14
@danamlewis I’d like to try to make it to the NYC meetup too, please keep us posted!
Dana Lewis
@danamlewis
Jun 07 2016 15:26
@live4sw awesome! Will do. @jasoncalabrese :+1:
Jason Calabrese
@jasoncalabrese
Jun 07 2016 15:29
I'm sure I'll have dinner with the group from work, but after that I could meet somewhere
David Cintron
@loudestnoise
Jun 07 2016 15:57
What’s the best AAA Lithium batteries to use in MedT pumps?
Ah, well looks like @channemann has done the research. Just found this
Dana Lewis
@danamlewis
Jun 07 2016 16:01
:+1:
::ponders suggesting a PR to the usability docs page::
Chris Hannemann
@channemann
Jun 07 2016 16:02
Feels like so long ago ...
Dana Lewis
@danamlewis
Jun 07 2016 16:02
lol @channemann
Chris Hannemann
@channemann
Jun 07 2016 16:03
I think there was a link to that gist in the hardware section of the docs at some point
Dana Lewis
@danamlewis
Jun 07 2016 16:11
yea. i think there are a few nuggest like that sprinkled in hardware/early on that I’m thinking should also be linked in Usability, because most people don’t look back once they get their hardware
Scott Leibrand
@scottleibrand
Jun 07 2016 16:35
@bewest is it documented anywhere how to use pip to install the latest dev version of everything? I don't see it in docs or gitter search, so I'm going back to quick-src for now
Jason Calabrese
@jasoncalabrese
Jun 07 2016 16:49
can't we use pip with local source like we do for mmeowlink?
I think I'll always want the local source for most things
Dana Lewis
@danamlewis
Jun 07 2016 17:04
re: Australia, and geography questions: I roughed out a page in the wiki in the OpenAPS docs with a rough list of where people are looping from, to help answer those questions. If (and that’s a big if, it’s not required or strongly suggested, it’s totally at your comfort level) you’d like to add yourself (by name, or handle, however you’re comfortable) to your geographic area for others to reach out to, feel free to. but this will never be a comprehensive loopers list, it’s purely intended to help show where there’s groupings of loopers and allow a little bit easier connection amongst local groups.
Feel free to add yourself, or combine cities into groups (i.e. I know the SF area blob that I grouped out of all the cities, but there may be other areas where a grouping rather than the burb listings makes sense - feel free to adjust those) where it makes sense
note that it’s a wiki page, so you just go in and edit, so I don’t have to approve every minute change (like a PR).
Paul Martin
@Spazholio
Jun 07 2016 17:05
@bewest So I’m embarking on my openxshareble adventure, and I’ve run into a hitch: I’ve disabled Share, and installed the vendor/device in OpenAPS, but when I run openaps use share iter_glucose_hours 3 I’m getting AttributeError: 'NoneType' object has no attribute ‘iter_records’ after the sending auth code bit. Have I missed something?
Dana Lewis
@danamlewis
Jun 07 2016 17:05
but, please only add yourself, don’t add anyone else without their explicit permission.
Paul Martin
@Spazholio
Jun 07 2016 17:06
There’s three of us in Michigan? That’s surprising. =)
Sarah Hodson Grady
@shgrady
Jun 07 2016 17:27
@Spazholio xDrip can load to Share, I just don't, my family prefers NS.
Paul Martin
@Spazholio
Jun 07 2016 17:27
I may give that a shot once I get openxshareble working. Just to get a feel for all the options. Thanks!
Jarred Yaw
@jyaw
Jun 07 2016 17:36
@danamlewis thanks for putting this together! At least I'm pretty certain there's not some local group I'm missing out on probably, since there's no other southeastern US locations that I saw.
Dana Lewis
@danamlewis
Jun 07 2016 17:54
@jyaw there’s an Atlanta person I left off. hold on :)
there we go.
but yea, not many SE that I know of, either
but no one has wanted to have a Hsv meetup the other times we’ve been back :) so if we do one while we’re there, maybe others will pop out of the woodwork this time :)
Ben West
@bewest
Jun 07 2016 18:17
@scottleibrand it's mentioned in the issues I'ev been working on
use pip install -U git+git://github.com/openaps/openaps.git'#dev'
I really recommending using master though
dev should not be needed unless your a dev wanting to edit sources
Dan Piessens
@dpiessens
Jun 07 2016 18:24
With AMA, how do temporary targets get into the calulation? I'm guessing it has to do with carbhistory.json, but I don't know how those events come from NS?
Also is it a problem if I'm using my bolus wizard with it?
I'd like to to the bolus wizard but use eating soon with it
Dana Lewis
@danamlewis
Jun 07 2016 18:29
there’s no problem with using your bolus wizard, but know that the bolus wizard won’t have an adjusted sensitivity number in it.
eating soon just changes the target to 80 (default is for 60 minutes). it shouldn’t change your meal bolus activity or thus in anyway impact what you do on bolus wizard
Dan Piessens
@dpiessens
Jun 07 2016 18:37
ok, but my loop should need to be pulling that temp target data at some point and I know I don't have a command pulling treatment events at this point
Scott Leibrand
@scottleibrand
Jun 07 2016 18:51
temp targets only come from nightscout currently
@bewest thanks. it was the raw calibration stuff that was missing when I just did pip install -U openaps
Dan Piessens
@dpiessens
Jun 07 2016 18:52
@scottleibrand OK, makes sense, my question was poorly phrased: where in the data collection process should the loop be pulling that data?
i.e. which report
but that is ugly and hacky
you should be able to set up a device and report to do it more cleanly
Dan Piessens
@dpiessens
Jun 07 2016 18:57
Ah that makes sense, perfect, thanks!
Dan Piessens
@dpiessens
Jun 07 2016 19:21
@scottleibrand how do I skip over a field on a device? for instance oref0 profile ends with "max-iob carb-ratios temp-targets"
if I don't have/need carb ratios, how do I skip it?
Paul Martin
@Spazholio
Jun 07 2016 19:26
@bewest Any ideas as to why I’m not quite able to use openxshareble as per the docs? Did I miss a step?
Dana Lewis
@danamlewis
Jun 07 2016 19:27
@Spazholio we had a lot of trouble getting it setup at first, too, for context
Paul Martin
@Spazholio
Jun 07 2016 19:28
Any suggestions? When running the list_dexcom command, I’m not even seeing the device. I saw it with bluetoothctl earlier, but I think that’s when I had it in discover mode.
Dan Piessens
@dpiessens
Jun 07 2016 19:37
@Spazholio it's pretty flaky on most devices, I've had issues based on interference
Paul Martin
@Spazholio
Jun 07 2016 19:38
Really? That’s surprising to hear, but good to know. I tend to keep the devices pretty close together. Hopefully that would mitigate some of the issues.
Dan Piessens
@dpiessens
Jun 07 2016 19:38
no wireless mouses/keyboard, phones, laptops and try to quiet the TI stick as that's blocked it for me
also watch out for cats, raindrops, solar flares, alignment of mars and any rare birds flying overhead :)
Paul Martin
@Spazholio
Jun 07 2016 19:40
Stupid Mars...
Thacher
@thacherT1D
Jun 07 2016 19:43
@danamlewis thanks!
has anyone used anything other than insulin in a pump before switching over? I want to test the functionality of the medtronic without necessarily removing my omniPod right away
Dana Lewis
@danamlewis
Jun 07 2016 19:45
you don’t need to put a reservoir in it, you can just rewind, start priming, and stick your finger in. once it senses the resistance, it’ll stop.
and then you don’t have to bother sticking a reservoir in.
Thacher
@thacherT1D
Jun 07 2016 19:45
thanks :)
Dana Lewis
@danamlewis
Jun 07 2016 19:45
:+1:
Scott Leibrand
@scottleibrand
Jun 07 2016 19:45
@dpiessens for optional arguments, you can provide a nonexistent filename
Dan Piessens
@dpiessens
Jun 07 2016 19:46
ok I'll try that
Ben West
@bewest
Jun 07 2016 20:24
oy
positional arguments are not optional
if there are optional arguments, let's use yargs to make an optional switch
David Cintron
@loudestnoise
Jun 07 2016 21:57
@thacherT1D I am transitioning from that same situation. I’m an OmniPodder, but most likely am gonna go full MedT with my 722 & OpenAPS. While it’s a little tricky for some parts, most you can do without connecting the pump or using insulin in it. I just did my first set with live insulin this weekend :smile:
and by connecting I mean to your person
Paul Martin
@Spazholio
Jun 07 2016 22:10
@bewest @danamlewis Do either of you recall having to do anything specific with the receiver prior to getting openxshareble working properly? I’ve tried severing all hints of my iPhone’s connectivity to the receiver, but still no love.
Paul Martin
@Spazholio
Jun 07 2016 22:16
Oooh, it needs a higher version of bluez? Or do you know if it just needs the —experimental flag? I’ve already got that running, if so.
Scott Leibrand
@scottleibrand
Jun 07 2016 22:17
I never did figure out what it needs, just finally managed to get it working that way
Dana Lewis
@danamlewis
Jun 07 2016 22:17
I glared at it really, really hard.
Paul Martin
@Spazholio
Jun 07 2016 22:17
heh, fair enough.
Scott Leibrand
@scottleibrand
Jun 07 2016 22:19
oh, I thought those glares were at me. :-p
Dana Lewis
@danamlewis
Jun 07 2016 22:19
glaring at you doesn’t make my devices magically work. ;) ok maybe sometimes it does, but this wasn’t one of those times
Scott Leibrand
@scottleibrand
Jun 07 2016 22:20
;-)
Andrew Dodd
@Entropy512
Jun 07 2016 22:20
glaring at the device usually works better. :P
Dana Lewis
@danamlewis
Jun 07 2016 22:21
yep :)
Scott Leibrand
@scottleibrand
Jun 07 2016 22:21
I like percussive maintenance myself
Andrew Dodd
@Entropy512
Jun 07 2016 22:21
Had a question, since I'm thinking of FINALLY making the switch from MDI to pump this summer - I saw that timomer had made some progress with the Accu-Chek Combo at https://groups.google.com/forum/#!topic/openaps-dev/hXDMbpMD3XI - Did they succeed in figuring out any of the protocol?
Looking at the discussion, directly interfacing from a phone might be problematic due to the MAC address hackery the meter side seems to do
Dana Lewis
@danamlewis
Jun 07 2016 22:22
@Entropy512 I think it’s still in progress
@timomer ^
Andrew Dodd
@Entropy512
Jun 07 2016 22:22
I noticed people were trying to sniff the CSR bluetooth module's I/O on the pump side - did anyone try doing the same on the meter side?
Scott Leibrand
@scottleibrand
Jun 07 2016 22:24
these sound like questions to ask the folks who hang out at https://omniapsslack.azurewebsites.net/ as well
Thacher
@thacherT1D
Jun 07 2016 22:25
I'm having trouble getting the dev branch working with nightscout -- from my azure site dashboard, when I make the switch from using the master branch to using the dev branch my site doesn't pull anything up... white screen... I've tried disconnecting/reconnecting the source, reseting the page, different browser... suggestions on what to try next?
Ben West
@bewest
Jun 07 2016 22:31
what version of bluez are you using?
bluetoothd -v
I'm usually able to get it to work with at least 5.33
Paul Martin
@Spazholio
Jun 07 2016 22:35
5.23
Thacher
@thacherT1D
Jun 07 2016 22:41
@loudestnoise that's great! I'm just getting going -- I've been thinking about it forever, but only just feeling up to the challenge ;)
David Cintron
@loudestnoise
Jun 07 2016 22:42
That was me two weeks ago
Thacher
@thacherT1D
Jun 07 2016 22:44
cool -- I'll definitely stay in touch then, hopefully I'll be there in two weeks too
also, I have no idea why, but the dev branch is working for me now... I just did all the same things I said I tried, but over again...
Paul Martin
@Spazholio
Jun 07 2016 22:49
Ok, compiled and installed the new version of bluez. I still have the AttributeError: 'NoneType' object has no attribute ‘iter_records’ issue though.
I’ve got the serial put in there, followed the directions (don’t think I missed anything), and I have the receiver in share mode.
Paul Martin
@Spazholio
Jun 07 2016 22:55

It seems like it’s never quite pairing:

<Adafruit_BluefruitLE.bluez_dbus.gatt.BluezGattService object at 0x75dfb230>
paired? 0
pairing...
paired
[UUID('00001800-0000-1000-8000-00805f9b34fb'), UUID('00001801-0000-1000-8000-00805f9b34fb'), UUID('00001804-0000-1000-8000-00805f9b34fb'), UUID('0000180a-0000-1000-8000-00805f9b34fb'), UUID('f0aca0b1-ebfa-f96f-28da-076c35a521db')]
finding service
SERVICE <Adafruit_BluefruitLE.bluez_dbus.gatt.BluezGattService object at 0x75584490>

This happens each time I run the command.

Lynne Butts
@flutelynne_twitter
Jun 07 2016 22:55
@thacherT1D I was a Ping user, but using MDI for the last month or so, taking a break so my sites would have a fresh start. I put my finger inside the reservoir compartment, as suggested, and was able to see how the settings would work. My husbands suggestion was to imitate the increases (couldn't do decreases) suggested, to see if it were reliable. I did that and compared it to the BWP on Nightscout, since I had been using that for a long time. I don't know if it would work to input the TBR's suggested with the MedT setup to make sure that your MedT pump is working correctly.
@loudestnoise Just started my live loop 24.5 hours ago. A whole new life! Used the phone to hotspot, walked a few miles, and waited it out while BG was dropping to see if the TBR worked! Didn't need to take any extra carbs! Felt great. I am not sure how to come out of the hotspot, though. We were going to bike ride, so I kept the hotspot going, but it must have tried to switch to the house wifi. So I ssh'd from my phone, shutdown, unplugged the battery, waited, then plugged in the battery again. House wifi picked it right up.
David Cintron
@loudestnoise
Jun 07 2016 23:01
@flutelynne_twitter Awesome. You can just turn hotspot off. Won't hurt for your Pi to lose connection. Especially if you're trying to coax it to join your home wifi.
Scott Leibrand
@scottleibrand
Jun 07 2016 23:02
on an iPhone, you just hit the blue bar, and then switch it off there. there should be something similar if you're on android.
Lynne Butts
@flutelynne_twitter
Jun 07 2016 23:03
@scottleibrand and @loudestnoise Thanks!
Dana Lewis
@danamlewis
Jun 07 2016 23:04
@flutelynne_twitter congrats on your awesome progress! :)
Lynne Butts
@flutelynne_twitter
Jun 07 2016 23:04
Also, If I lock the keypad on the pump, does that mess up with it doing the TBR's?
Dana Lewis
@danamlewis
Jun 07 2016 23:05
it shouldn’t, but why lock the keypad on the pump?
Lynne Butts
@flutelynne_twitter
Jun 07 2016 23:06
I used to lock it on the Animas. I guess I don't have to worry about it because it takes a few presses to actually complete any boluses. Thanks, @danamlewis
Scott Leibrand
@scottleibrand
Jun 07 2016 23:06
I didn't even realize you could lock it :)
Lynne Butts
@flutelynne_twitter
Jun 07 2016 23:07
@danamlewis @scottleibrand @bewest Thank you for making this Open Source! This is an awesome day!
Dana Lewis
@danamlewis
Jun 07 2016 23:09
Yea, it requires enough confirmation that I wouldn’t bother with it. :)
:sparkles: :clap: :raised_hands: :confetti_ball:
Scott Leibrand
@scottleibrand
Jun 07 2016 23:10
so great to see how anyone can build an openaps rig with enough willingness and determination to learn what they need to know. :)
Lynne Butts
@flutelynne_twitter
Jun 07 2016 23:19
Well, I couldn't have done it alone. That's for sure! I do have tons of determination, though.
Scott Leibrand
@scottleibrand
Jun 07 2016 23:19
There's not a single one of us who could've done it alone.
Andrew Dodd
@Entropy512
Jun 07 2016 23:20
That's the miracle of open source. :)
Dana Lewis
@danamlewis
Jun 07 2016 23:26
:)
Paul Martin
@Spazholio
Jun 07 2016 23:32
:thumbsup:
Brooke Pazoles
@PazolesBrooke_twitter
Jun 07 2016 23:50
Wish me luck guys! I'll be testing my open loop all week while Matt's in Italy. Hoping to officially close that sucker when he comes home on Monday! So far so GREAT!! Staying in range (65-130) about 90% of the time and getting better and better.
Dana Lewis
@danamlewis
Jun 07 2016 23:51
wooohoo! :)
Paul Martin
@Spazholio
Jun 07 2016 23:51
Congrats! :thumbsup:
Brooke Pazoles
@PazolesBrooke_twitter
Jun 07 2016 23:52
A quick question - when the APS rounds it's recommendation to fit the pump's increments, does it round up or down? If the rec is .88 will it do .9 or .85?
lmoynihan73
@lmoynihan73
Jun 07 2016 23:55
Newbie here. Trying to confirm that I can communicate with Dexcom and getting this error:
Traceback (most recent call last):
File "/usr/local/bin/openaps-use", line 4, in <module>
import('pkg_resources').run_script('openaps==0.1.5', 'openaps-use')
File "build/bdist.linux-armv7l/egg/pkgresources/_init.py", line 719, in run_script
File "build/bdist.linux-armv7l/egg/pkgresources/_init.py", line 1504, in run_script
File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/EGG-INFO/scripts/openaps-use", line 63, in <module>
app( )
File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/openaps/cli/init.py", line 51, in call
self.run(self.args)
File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/EGG-INFO/scripts/openaps-use", line 57, in run
output = app(args, self)
File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/openaps/uses/init.py", line 92, in call
return self.method.selected(args)(args, app)
File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/openaps/uses/init.py", line 31, in call
return self.method(args, app)
File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/openaps/uses/use.py", line 45, in call
output = self.main(args, app)
File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/openaps/vendors/dexcom.py", line 297, in main
for item in self.dexcom.iter_records(self.RECORD_TYPE):
File "/usr/local/lib/python2.7/dist-packages/dexcom_reader-0.1.9-py2.7.egg/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-0.1.9-py2.7.egg/dexcom_reader/readdata.py", line 267, in GenericRecordYielder
yield record_type.Create(data, x)
File "/usr/local/lib/python2.7/dist-packages/dexcom_reader-0.1.9-py2.7.egg/dexcom_reader/database_records.py", line 60, in Create
return cls(unpacked_data, raw_data)
File "/usr/local/lib/python2.7/dist-packages/dexcomreader-0.1.9-py2.7.egg/dexcomreader/database_records.py", line 42, in __init
self.check_crc()
File "/usr/local/lib/python2.7/dist-packages/dexcom_reader-0.1.9-py2.7.egg/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 EGVRecord