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

1st
Oct 2016
garykidd
@garykidd
Oct 01 2016 00:00
Is it possible that DIA varies a bit with magnitude of bolus...I know bolus wizards, for example, aren't setup as if they were..... but it almost seems like that in real-world practice/experience (though, admittedly it could be other factors contributing to the thought/question).
garykidd
@garykidd
Oct 01 2016 00:12
@scottleibrand was curious if you had any immediate thoughts to my post from Sep 29 23:24?
Scott Leibrand
@scottleibrand
Oct 01 2016 00:15
link? dunno what timezone you're in
garykidd
@garykidd
Oct 01 2016 00:16
ET
Scott Leibrand
@scottleibrand
Oct 01 2016 00:16
the one about glucose.json being empty?
garykidd
@garykidd
Oct 01 2016 00:16
yeah..
Scott Leibrand
@scottleibrand
Oct 01 2016 00:16
you can either plug in CGM or get it from NS
if your NS download isn't working, you should check NS host and API secret
garykidd
@garykidd
Oct 01 2016 00:17
neither seems to work..left cgm plugged in for a long time...
Scott Leibrand
@scottleibrand
Oct 01 2016 00:17
I'd take a look at /var/log/openaps/cgm-loop.log first
and then look at the contents of the cgm-loop alias and try the components manually
garykidd
@garykidd
Oct 01 2016 00:21
cgm://JSON/oref0_glucose/raw-cgm/raw-entries.json
Fri 30 Sep 00:03:01 EDT 2016
cgm://JSON/oref0_glucose/raw-cgm/raw-entries.json
Fri 30 Sep 00:04:01 EDT 2016
cgm://JSON/oref0_glucose/raw-cgm/raw-entries.json
Fri 30 Sep 00:05:01 EDT 2016
cgm://JSON/oref0_glucose/raw-cgm/raw-entries.json
Fri 30 Sep 00:06:01 EDT 2016
cgm://JSON/oref0_glucose/raw-cgm/raw-entries.json
Fri 30 Sep 00:07:02 EDT 2016
cgm://JSON/oref0_glucose/raw-cgm/raw-entries.json
Fri 30 Sep 00:08:01 EDT 2016
cgm://JSON/oref0_glucose/raw-cgm/raw-entries.json
Fri 30 Sep 00:09:01 EDT 2016
cgm://JSON/oref0_glucose/raw-cgm/raw-entries.json
Fri 30 Sep 00:10:01 EDT 2016
cgm://JSON/oref0_glucose/raw-cgm/raw-entries.json
Fri 30 Sep 00:11:01 EDT 2016
cgm://JSON/oref0_glucose/raw-cgm/raw-entries.json
Fri 30 Sep 00:12:01 EDT 2016
cgm://JSON/oref0_glucose/raw-cgm/raw-entries.json
Fri 30 Sep 00:13:01 EDT 2016
cgm://JSON/oref0_glucose/raw-cgm/raw-entries.json
Fri 30 Sep 00:14:01 EDT 2016
cgm://JSON/oref0_glucose/raw-cgm/raw-entries.json
that's /var/log/openaps/cgm-loop.log in case it wasn't obvious.
Scott Leibrand
@scottleibrand
Oct 01 2016 00:23
so just "not working". have to run component aliases to see why
garykidd
@garykidd
Oct 01 2016 00:24
to view the contents of the cgm-loop alias....is that best to just cat openaps.ini
Scott Leibrand
@scottleibrand
Oct 01 2016 00:24
no, openaps alias show cgm-loop
garykidd
@garykidd
Oct 01 2016 00:24
lot of things to read in the alias section......I have a hard time distinguishing one alias form another....would be huge to put a hard return between each alias actually
Scott Leibrand
@scottleibrand
Oct 01 2016 00:25
not sure what you're looking at
garykidd
@garykidd
Oct 01 2016 00:25
oh was saying of openaps.ini
Scott Leibrand
@scottleibrand
Oct 01 2016 00:25
you should never have to look at that
garykidd
@garykidd
Oct 01 2016 00:25
the alias section...
got it
Scott Leibrand
@scottleibrand
Oct 01 2016 00:26
you're welcome to, of course, but there are commands to do everything without it
garykidd
@garykidd
Oct 01 2016 00:27
cool
holy cow.....lot of sub aliases
stet that
hmmmm...is that def. the alias....
This message was deleted
Scott Leibrand
@scottleibrand
Oct 01 2016 00:30
maybe not. look at crontab -l
garykidd
@garykidd
Oct 01 2016 00:30
pi@raspberrypi:~/APS401 $ openaps alias show cgm-loop
usage: openaps-alias show [-h] [--ini] [--json]
                          [{autosens,bg-fresh-check,enact,format-latest-nightscout-treatments,format-ns-status,gather,get-bg,get-ns-bg,get-ns-glucose,get-settings,invoke,latest-ns-treatment-time,merge-pumphistory-long,mmtune,monitor-cgm,monitor-pump,ns-loop,ns-meal-carbs,ns-temptargets,pebble,prep-pumphistory-entries,pump-loop,refresh-old-profile,refresh-old-pumphistory,refresh-old-pumphistory-24h,refresh-profile,refresh-pumphistory-24h,refresh-pumphistory-and-enact,refresh-temp-and-enact,upload,upload-ns-status,upload-pumphistory-entries,upload-recent-treatments,wait-for-long-silence,wait-for-silence,*}]
openaps-alias show: error: argument name: invalid choice: 'cgm-loop' (choose from 'autosens', 'bg-fresh-check', 'enact', 'format-latest-nightscout-treatments', 'format-ns-status', 'gather', 'get-bg', 'get-ns-bg', 'get-ns-glucose', 'get-settings', 'invoke', 'latest-ns-treatment-time', 'merge-pumphistory-long', 'mmtune', 'monitor-cgm', 'monitor-pump', 'ns-loop', 'ns-meal-carbs', 'ns-temptargets', 'pebble', 'prep-pumphistory-entries', 'pump-loop', 'refresh-old-profile', 'refresh-old-pumphistory', 'refresh-old-pumphistory-24h', 'refresh-profile', 'refresh-pumphistory-24h', 'refresh-pumphistory-and-enact', 'refresh-temp-and-enact', 'upload', 'upload-ns-status', 'upload-pumphistory-entries', 'upload-recent-treatments', 'wait-for-long-silence', 'wait-for-silence', '*')
Scott Leibrand
@scottleibrand
Oct 01 2016 00:30
sorry, it's monitor-cgm
* * * * * cd /home/edison/cgm-loop && ( ps aux | grep -v grep | grep -q 'openaps monitor-cgm' || openaps monitor-cgm ) 2>& 1 | tee -a /var/log/openaps/cgm-loop.log
so openaps alias show monitor-cgm
garykidd
@garykidd
Oct 01 2016 00:31
on rp3
Scott Leibrand
@scottleibrand
Oct 01 2016 00:31
your crontab -l will show a different loop dir, but should show openaps monitor-cgm as well
Eric Jensen
@elnjensen
Oct 01 2016 00:32
@scottleibrand @garykidd (and of course anyone else who is interested). I'm pleased to report that I got wifi roaming working on the Edison. I was having problems with it - would connect to available network on boot, but then wouldn't switch when that became available.
garykidd
@garykidd
Oct 01 2016 00:32
I reverted to my previous cron setup already
Eric Jensen
@elnjensen
Oct 01 2016 00:32
Now works great, with no wpa stuff in cron at all.
garykidd
@garykidd
Oct 01 2016 00:33
@elnjensen nice....willl want to fully undertand that in short order
pi@raspberrypi:~/APS401 $ openaps alias show monitor-cgm
monitor-cgm report invoke raw-cgm/raw-entries.json cgm/cgm-glucose.json
pi@raspberrypi:~/APS401 $ cat raw-cgm/raw-entries.json
cat: raw-cgm/raw-entries.json: No such file or directory
pi@raspberrypi:~/APS401 $ cat cgm/cgm-glucose.json
cat: cgm/cgm-glucose.json: No such file or directory
pi@raspberrypi:~/APS401 $ ls
cgm  cgm.ini  detect-sensitivity.ini  determine-basal.ini  enact  get-profile.ini  iob.ini  meal.ini  monitor  ns-glucose.ini  ns.ini  openaps.ini  oref0.ini  pebble.ini  preferences.json  pump.ini  pump-session.json  raw-cgm  settings  tz.ini  units.ini  upload
pi@raspberrypi:~/APS401 $ cd cgm
pi@raspberrypi:~/APS401/cgm $ ls
ns-glucose.json
pi@raspberrypi:~/APS401/cgm $
Eric Jensen
@elnjensen
Oct 01 2016 00:33
Here's my /etc/network/interfaces:
auto lo iface lo inet loopback

# auto usb0 
iface usb0 inet static
    address 192.168.2.15
    netmask 255.255.255.0


# auto wlan0 
allow-hotplug wlan0
iface wlan0 inet manual
    wpa-roam /etc/wpa_supplicant/wpa_supplicant.conf

iface default inet dhcp
Scott Leibrand
@scottleibrand
Oct 01 2016 00:34
@garykidd so it has never successfully created those files. you can invoke them manually though
openaps report invoke raw-cgm/raw-entries.json etc
garykidd
@garykidd
Oct 01 2016 00:34
@scottleibrand still struggling to understand why it hasnt though
Eric Jensen
@elnjensen
Oct 01 2016 00:34
The two changes (from what I had initially) were to change iface wlan0 inet dhcp to iface wlan0 inet manual and in the next line, to change wpa-conf to wpa-roam.
Scott Leibrand
@scottleibrand
Oct 01 2016 00:34
@garykidd no way to know without trying it
@elnjensen odd, that should change precisely nothing
Eric Jensen
@elnjensen
Oct 01 2016 00:35
@scottleibrand I know you've said that you didn't think that wpa-roam mattered, but I verified that, at least for the above file, that switch makes the difference.
Scott Leibrand
@scottleibrand
Oct 01 2016 00:35
(since you have iface default inet dhcp)
jmcrawford
@jmcrawford
Oct 01 2016 00:35
@elnjensen What does your /etc/wpa_supplicant/wpa_supplicant.conf look like?
Hannah Crabtree
@lollydaggle
Oct 01 2016 00:36
this is a silly question - at what point in the setup script do I need to have my devices plugged in? can't get openaps use pump model or openaps use cgm iter_glucose to work
I'm suspicious RE: my carelink working but the CGM should communicate
Eric Jensen
@elnjensen
Oct 01 2016 00:36
Interestingly there's is essentially zero documentation on wpa-roam vs. wpa-conf. Lots of pages where people post their configs, but no mention of it in man pages at all.
Scott Leibrand
@scottleibrand
Oct 01 2016 00:37
yeah, I had to do some serious digging to find any info on it
Eric Jensen
@elnjensen
Oct 01 2016 00:37
So I think a lot of the declarations about what it does or doesn't do are, shall we say, not well documented...
@jmcrawford Give me a sec to sanitize the wpa_supplicant.conf file...
Scott Leibrand
@scottleibrand
Oct 01 2016 00:37
@elnjensen agreed
perhaps someone should read the code and update the man pages...
Eric Jensen
@elnjensen
Oct 01 2016 00:40
But which code? Is it a part of /etc/network/interfaces, or of wpa-supplicant? I guess must be the latter?
Scott Leibrand
@scottleibrand
Oct 01 2016 00:40
dunno, or I probably would've done it. :)
but I can't read core OS code effectively
Eric Jensen
@elnjensen
Oct 01 2016 00:41
@jmcrawford OK, here's my /etc/wpa_supplicant/wpa_supplicant.conf. Nothing special in here, just a few networks and the kind of setup you commonly see recommended:
country=US
ctrl_interface=DIR=/var/run/wpa_supplicant GROUP=netdev
update_config=1

network={
        ssid="aaa"
        psk="****"
        priority=1
        }

network={
        ssid="xxx"
        psk="****"
        scan_ssid=1
        priority=2
        }

network={
        ssid="yyy"
        psk="****"
        priority=3
        }
Scott Leibrand
@scottleibrand
Oct 01 2016 00:42
@lollydaggle I would plug everything in before I start, but it shouldn't matter
garykidd
@garykidd
Oct 01 2016 00:44
pi@raspberrypi:~ $ cd && ~/src/oref0/bin/oref0-setup.sh
Usage: oref0-setup.sh <--dir=directory> <--serial=pump_serial_#> [--tty=/dev/ttySOMETHING] [--max_iob=0] [--ns-host=https://mynightscout.azurewebsites.net] [--api-secret=myplaintextsecret] [--cgm=(G4|G5|MDT)] [--enable='autosens meal']
Start interactive setup? [Y]/n y
What would you like to call your loop directory? [myopenaps] APS401
Ok, APS401 it is.
What is your pump serial number? #########
Ok, ########## it is.
What Kind of CGM are you using? (i.e. G4, G5, MDT) G5
Ok, G5 it is.
Are you using mmeowlink? If not, press enter. If so, what TTY port (i.e. /dev/ttySOMETHING)? /dev/ttyACM0
Ok, TTY /dev/ttyACM0 it is.
Are you using Nightscout? If not, press enter.
If so, what is your Nightscout host? (i.e. https://mynightscout.azurewebsites.net)? https://######.azurewebsites.net
Ok, https://projectfeedback.azurewebsites.net it is.
And what is your Nightscout api secret (i.e. myplaintextsecret)? #############
Ok, ########## it is.
Do you need any advanced features? y/[N] y
Enable automatic sensitivity adjustment? y/[N] y
Enable advanced meal assist? y/[N] y
Setting up oref0 in /home/pi/APS401 for pump ######## with G5 CGM,
NS host https://##########.azurewebsites.net, TTY /dev/ttyACM0, advanced features  autosens  meal 
Continue? y/[N] y
Checking /home/pi/APS401: true
/home/pi/APS401 already exists
/home/pi/src/oref0/ already exists; pulling latest
fatal: unable to connect to github.com:
github.com[0: 192.30.253.112]: errno=No route to host

Couldn't pull latest oref0
Eric Jensen
@elnjensen
Oct 01 2016 00:45
I think all the magic happens in /etc/network/interfaces.
Scott Leibrand
@scottleibrand
Oct 01 2016 00:45
@garykidd you're not online?
Hannah Crabtree
@lollydaggle
Oct 01 2016 00:45
Hmm, I re-ran the setup script anyway. I'm getting my favorite python error when I run the CGM:
Scott Leibrand
@scottleibrand
Oct 01 2016 00:45
@garykidd also you missed blacking out one instance of your NS URL
Hannah Crabtree
@lollydaggle
Oct 01 2016 00:45
'NoneType' object has no attributes 'iter_records'
Scott Leibrand
@scottleibrand
Oct 01 2016 00:46
heh, that's the same one I'm fighting with trying to get openxshareble working
you can also do lsusb to see if the CGM is recognized
garykidd
@garykidd
Oct 01 2016 00:47
hmmmm...seems strange b/c the computer and my pi on same network i'm ssh'ing into pi ...and this computer has internet access.....
just rebooting pi to see if that'll work
Hannah Crabtree
@lollydaggle
Oct 01 2016 00:50
this is my pump/carelink error: serial.serialutil.SerialException: Attempting to use a port that is not open. Wondering if the carelink is fried at this point.
Scott Leibrand
@scottleibrand
Oct 01 2016 00:51
dunno, but it might be a good time to start thinking about something more capable like a TI stick
Hannah Crabtree
@lollydaggle
Oct 01 2016 00:52
@scottleibrand already on top of it! Ordered the TI and Edison this morning.
Scott Leibrand
@scottleibrand
Oct 01 2016 00:52
:-)
did you see the Explorer board post?
Hannah Crabtree
@lollydaggle
Oct 01 2016 00:53
Yeah, I put my name in. I know I'll eventually want to get there. I'm trying to take it one step at a time.
Scott Leibrand
@scottleibrand
Oct 01 2016 00:53
:+1:
Tim Mecklem
@tmecklem
Oct 01 2016 01:05
@danamlewis did you get my name for 2? It looks like the form has a quantity field now and I don’t remember it having that before.
Eric Jensen
@elnjensen
Oct 01 2016 01:08
@scottleibrand Downloaded the source code for wpa-supplicant, and the string wpa-roam doesn't appear in any file in the source tree. Don't think I'll chase this unicorn any farther - just happy to have it working! Any place in the docs that /etc/network/interfaces is described? I could take a look and see if a modest PR there would make sense. (And BTW, now that this is setup, it seems I don't need to every run wpa_cli scan - it just finds a new network pretty quickly on its own.)
Dana Lewis
@danamlewis
Oct 01 2016 01:09
@tmecklem yep! Added the field after a few of you quick twitch folks had already started inputting :)
Tim Mecklem
@tmecklem
Oct 01 2016 01:09
:D
Dana Lewis
@danamlewis
Oct 01 2016 01:11
And when you actually order/pay you'll be able to confirm or adjust how many you want. We're nearing 75 already (wow!) so we'll likely round the order up to have a few spares for people who discover this later.
Tim Mecklem
@tmecklem
Oct 01 2016 01:12
My wife and her friend are at their T1D meetup tonight and they’ve both been closed loop for a while now. I suspect from the interest there will be a hotspot of openaps fans here around Cincinnati soon.
garykidd
@garykidd
Oct 01 2016 01:12
@scottleibrand yup....have fun with that...oh well
Tim Street
@tim2000s
Oct 01 2016 01:12
👍👍
Dana Lewis
@danamlewis
Oct 01 2016 01:13
@tmecklem nice :)
Tim Mecklem
@tmecklem
Oct 01 2016 01:13
And judging from their BGs, they didn’t have a presentation tonight :)
garykidd
@garykidd
Oct 01 2016 01:13
@scottleibrand I re-installed everthing...same error...agian
Tim Street
@tim2000s
Oct 01 2016 01:13
You have any numbers for the uk @scottleibrand Or @danamlewis
garykidd
@garykidd
Oct 01 2016 01:13
MODULE NOT FOUND
Dana Lewis
@danamlewis
Oct 01 2016 01:14
@tim2000s numbers - you mean number of boards ordered or something else?
I didn't ask geo on the form so I just have a hunch based on names that I know
garykidd
@garykidd
Oct 01 2016 01:15
pi@raspberrypi:~/APS401 $ openaps report invoke raw-cgm/raw-entries.json
cgm://JSON/oref0_glucose/raw-cgm/raw-entries.json
raw-cgm/raw-entries.json  raised  'NoneType' object has no attribute 'iter_records'
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 532, in main
    iter_glucose = self.get_glucose_data(params, args)
  File "/usr/local/lib/python2.7/dist-packages/openaps/vendors/dexcom.py", line 513, in get_glucose_data
    return itertools.takewhile(self.comparison, self.dexcom.iter_records('EGV_DATA'))
AttributeError: 'NoneType' object has no attribute 'iter_records'
wait...I selected G5...but is there raw with G5.....maybe the report name and the data don't match?
Tim Street
@tim2000s
Oct 01 2016 01:20
Sorry @danamlewis I meant users. I requested a couple of boards.
I know of myself and three others in the U.K. Just wondering how many we had.
Dana Lewis
@danamlewis
Oct 01 2016 01:22
that sounds about right
garykidd
@garykidd
Oct 01 2016 01:31
@scottleibrand is it becuae there isnt a --no-raw added to openaps use oref0_glucose maybe?
in raw-cgm/raw-entries.json call/
David
@davidkeddydb
Oct 01 2016 01:39
Ok, so i made a mistake on the pump today that i see could cause others a problem. Had a high fat meal, meant to square wave some and didn't so I canceled the blous after .1 delivered. I then re-entered and squared as I wanted. COB sees it now as double the carbs. I should have just bolused the amount rather than entering carb amount.
garykidd
@garykidd
Oct 01 2016 01:40
pick it up tomorrow I guess.
I've seen double entries before in NS without even doing that...I think if you take a big bolus that straddles two NS uploads it might pick it up as two entries....as well.
Scott Leibrand
@scottleibrand
Oct 01 2016 01:42
@davidkeddydb yeah, unfortunately that behavior is different between different pumps too
@garykidd yeah, it's possible that you need to add --no-raw - not sure if anyone has thoroughly tested the --G5 support in oref0-setup yet
Tim Street
@tim2000s
Oct 01 2016 01:47
I'm running G5 but it's through mobile. I don't have a receiver.
garykidd
@garykidd
Oct 01 2016 01:58
that part seems easy but I would've thought the NS would kick in for me but nada
@tim2000s that means you're using share and/or NS then for BG?
@tim2000s did you need to add --no-raw to oref0_glucose calls in oref0-setup script?
Im actually thinking the continual sudo wpa cli scans are making my router conk out.....same with my damn * * * * * /home/pi/bin/bt-pan client XX:YY:ZZ:AA:BB:CC in cron causing BT issues in car.....so many links...so many possible broken links...
garykidd
@garykidd
Oct 01 2016 02:04
@scottleibrand might you be around tomorrow night...I thought I'd have energy tonight....but no dice.
Scott Leibrand
@scottleibrand
Oct 01 2016 02:04
perhaps. nothing scheduled yet, but we'll see
garykidd
@garykidd
Oct 01 2016 02:07
before I hit the hay...any ideas....maybe that NO MODULE FOUND error is OK and it'll start to set basals over time...i was wathcing the damn tail loop log mor ethan NS and now I am wondering if it might kick in on NS
even with that error
Scott Leibrand
@scottleibrand
Oct 01 2016 02:08
not sure what "no module found" thing you're talking about
garykidd
@garykidd
Oct 01 2016 02:09
pi@raspberrypi:~/APS401 $ tail -F /var/log/openaps/pump-loop.log
Listening: .No comms detected
Old pumphistory: 
Error, retrying
Starting pump-loop at Fri 30 Sep 21:06:28 EDT 2016:
Listening: .No comms detected
Old pumphistory: 
Error, retrying
Starting pump-loop at Fri 30 Sep 21:07:28 EDT 2016:
Listening: .Starting pump-loop at Fri 30 Sep 21:10:02 EDT 2016:
Listening: .No comms detected
Old pumphistory: Refreshed pumphistory
Could not parse input data:  { [Error: Cannot find module '/home/pi/APS401/monitor/glucose.json'] code: 'MODULE_NOT_FOUND' }
determine-basal://text/shell/enact/suggested.json
reporting enact/suggested.json
Old pumphistory-24h refreshed
Profile less than 60m old
temp_basal.json less than 5m old
Pumphistory less than 5m old
Settings refreshed
Pumphistory-24 less than 20m old
Completed pump-loop at Fri 30 Sep 21:11:28 EDT 2016
Scott Leibrand
@scottleibrand
Oct 01 2016 02:09
oh, you just mean your glucose.json being empty
garykidd
@garykidd
Oct 01 2016 02:09
sorry...almost.....;MODULE_NOT_FOUND'
yeah...I guess....I am so tired.
Scott Leibrand
@scottleibrand
Oct 01 2016 02:10
that won't fix itself.
tomorrow you can work on getting your reports fixed up for your G5, and/or figuring out what's wrong with your NS setup
garykidd
@garykidd
Oct 01 2016 02:10
happens after every time it Refreshed pumphistory
Scott Leibrand
@scottleibrand
Oct 01 2016 02:11
yeah, that's when it checks for updated BG data
but monitor/glucose.json is empty, so it never finds any there
the pump-loop doesn't actually try to pull the BG data: that's happening in the cgm-loop
so all the pump-loop is telling you is "nope, still empty"
garykidd
@garykidd
Oct 01 2016 02:11
one thing b4 I crash...what's the eaiest way to check NS BG calls within the contect of an oref0-setup instance
riger that
Scott Leibrand
@scottleibrand
Oct 01 2016 02:12
try openaps use ns shell preflight
garykidd
@garykidd
Oct 01 2016 02:12
roger
Scott Leibrand
@scottleibrand
Oct 01 2016 02:12
it will try to connect to NS and return true or false
garykidd
@garykidd
Oct 01 2016 02:16
see that shows true
so maybe it wold actually work if I let it go for awhile
LBehler
@LBehler
Oct 01 2016 02:17
how did you guys get your "ns" added as a device in openaps? That step isn't in my setup doc
@scottleibrand
Scott Leibrand
@scottleibrand
Oct 01 2016 02:18
@LBehler in this context, oref0-setup does it
@garykidd if you want to troubleshoot NS further, try openaps get-ns-glucose
and then head -20 cgm/ns-glucose.json
garykidd
@garykidd
Oct 01 2016 02:26
I stepped away from iphone/ receiver recently---would it just pull older 20 records?
pi@raspberrypi:~/APS401 $ openaps get-ns-glucose
ns://JSON/shell/cgm/ns-glucose.json
No JSON object could be decoded
Unknown request: oref0_glucose_since
followed by ...at the end...
Format potential entries (glucose values) for Nightscout.

    openaps use ns shell format-recent-type tz entries monitor/glucose.json  | json -a dateString | wc -l
    # Add it as a report
    openaps report add nightscout/recent-missing-entries.json JSON ns shell format-recent-type tz entries monitor/glucose.json
    # fetch data for first time
    openaps report invoke nightscout/recent-missing-entries.json

    # add report for uploading to NS
    openaps report add nightscout/uploaded-entries.json JSON  ns shell upload entries.json nightscout/recent-missing-entries.json
    # upload for fist time.
    openaps report invoke nightscout/uploaded-entries.json

Traceback (most recent call last):
  File "/usr/local/bin/openaps-report", line 82, in <module>
    app( )
  File "/usr/local/lib/python2.7/dist-packages/openaps/cli/__init__.py", line 51, in __call__
    self.run(self.args)
  File "/usr/local/bin/openaps-report", line 75, in run
    output = app(args, self)
  File "/usr/local/lib/python2.7/dist-packages/openaps/cli/subcommand.py", line 52, in __call__
    return self.method.main(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps/reports/invoke.py", line 48, in main
    reporters.Reporter(report, device, task)(output)
  File "/usr/local/lib/python2.7/dist-packages/openaps/reports/reporters/__init__.py", line 33, in __call__
    self.blob = self.serialize(data)
  File "/usr/local/lib/python2.7/dist-packages/openaps/reports/reporters/__init__.py", line 31, in serialize
    return self.method.serialize(render(data), self)
  File "/usr/local/lib/python2.7/dist-packages/openaps/vendors/process.py", line 56, in prerender_json
    return json.loads(data)
  File "/usr/lib/python2.7/json/__init__.py", line 338, in loads
    return _default_decoder.decode(s)
  File "/usr/lib/python2.7/json/decoder.py", line 366, in decode
    obj, end = self.raw_decode(s, idx=_w(s, 0).end())
  File "/usr/lib/python2.7/json/decoder.py", line 384, in raw_decode
    raise ValueError("No JSON object could be decoded")
ValueError: No JSON object could be decoded
pi@raspberrypi:~/APS401 $
could that all be cgm related stuff vs. NS?
Scott Leibrand
@scottleibrand
Oct 01 2016 02:29
oh, Unknown request: oref0_glucose_since is a clue
means you're using an old version, of oref0 I think
garykidd
@garykidd
Oct 01 2016 02:30
hmmmm...maybe I lost internet connection while downloading latest within script?
Scott Leibrand
@scottleibrand
Oct 01 2016 02:30
or maybe my script thought you were recent enough when you weren't
I think I saw this once before, and not sure we ever fixed it in the script, just had whoever it was download the latest oref0
garykidd
@garykidd
Oct 01 2016 02:32
hmmmm.....you'd know better than I....but thinking on it...I did try 4 or 5 times to get this running...so at least it;s consistent...so I think that means that ....it probably isn't what I thought above about losing internet connection mid-stream/setup.
sudo npm install -g oref0 ?
Scott Leibrand
@scottleibrand
Oct 01 2016 02:34
one sec, trying to figure out the proper fix in the script to have you test that
garykidd
@garykidd
Oct 01 2016 02:34
k
Scott Leibrand
@scottleibrand
Oct 01 2016 02:36
ok, do cd ~/src/oref0/ && git pull for me, and then re-run oref0-setup
this time it should say Installing latest oref0 dev and then do the npm spinner
This message was deleted
garykidd
@garykidd
Oct 01 2016 02:38
pi@raspberrypi:~/APS401 $ cd ~/src/oref0/ && git pull
remote: Counting objects: 4, done.
remote: Compressing objects: 100% (4/4), done.
remote: Total 4 (delta 0), reused 0 (delta 0), pack-reused 0
Unpacking objects: 100% (4/4), done.
From git://github.com/openaps/oref0
   d21a334..f1693b3  dev        -> origin/dev
Updating d21a334..f1693b3
Fast-forward
 bin/oref0-setup.sh | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
Scott Leibrand
@scottleibrand
Oct 01 2016 02:38
cool, that's my 1-line change, now re-run oref0-setup.sh
garykidd
@garykidd
Oct 01 2016 02:39
OK...gonna clear my crontab so that I can create a new new one from script and not just append the existing one.....brb
Scott Leibrand
@scottleibrand
Oct 01 2016 02:40
sorry you're running into all the bugs
thx for hanging in there.
garykidd
@garykidd
Oct 01 2016 02:40
no way man....sorry I'm hitting you up non stop.
but maybe it'll be useful in the end...hopefully
Scott Leibrand
@scottleibrand
Oct 01 2016 02:41
well you've already helped me fix one bug: that should help the next person in this situation
garykidd
@garykidd
Oct 01 2016 02:42
Well then...that's cool. there have been a few small moments when I've actually known something that someone else hasn't...and it's good to be able to contribute...even if its something small.
True for openaps
true for life, really
dougtakeuti
@dougtakeuti
Oct 01 2016 02:43
Hi @scottleibrand, I ran the script in a new openaps instance in the Edison rig. However, when I sent it to school running the old loop, it would not connect to the school wifi anymore. I've been using @TC2013 wifi hack successfuly for a few months. Is there anything in the script that could have caused it or is it just a coincidence?
Scott Leibrand
@scottleibrand
Oct 01 2016 02:44
maybe * * * * * sudo wpa_cli scan
if you have toby's hack going, you can comment that out
dougtakeuti
@dougtakeuti
Oct 01 2016 02:44
yes, I saw that
Scott Leibrand
@scottleibrand
Oct 01 2016 02:44
or try that instead of his thing, dunno
dougtakeuti
@dougtakeuti
Oct 01 2016 02:45
ok, will play around with that, thanks
garykidd
@garykidd
Oct 01 2016 02:47
hmmmm...don't think I saw this in any install attempt...
Checking /home/pi/APS401: true
/home/pi/APS401 already exists
/home/pi/src/oref0/ already exists; pulling latest
Already up-to-date.
Checking oref0 installation
Installing latest oref0 dev

> oref0@0.2.0 global-install /home/pi/src/oref0
> npm install && sudo npm install -g && sudo npm link && sudo npm link oref0

npm WARN package.json mocha@2.5.3 No README data
npm ERR! Error: EPERM, chmod '/home/pi/src/oref0/node_modules/oref0/bin/oref0-calculate-iob.js'
npm ERR!  { [Error: EPERM, chmod '/home/pi/src/oref0/node_modules/oref0/bin/oref0-calculate-iob.js']
npm ERR!   errno: 50,
npm ERR!   code: 'EPERM',
npm ERR!   path: '/home/pi/src/oref0/node_modules/oref0/bin/oref0-calculate-iob.js' }
npm ERR! 
npm ERR! Please try running this command again as root/Administrator.

npm ERR! System Linux 4.4.21-v7+
npm ERR! command "/usr/bin/nodejs" "/usr/bin/npm" "install"
npm ERR! cwd /home/pi/src/oref0
npm ERR! node -v v0.10.29
npm ERR! npm -v 1.4.21
npm ERR! path /home/pi/src/oref0/node_modules/oref0/bin/oref0-calculate-iob.js
npm ERR! code EPERM
npm ERR! errno 50
npm ERR! stack Error: EPERM, chmod '/home/pi/src/oref0/node_modules/oref0/bin/oref0-calculate-iob.js'
npm ERR! 
npm ERR! Additional logging details can be found in:
npm ERR!     /home/pi/src/oref0/npm-debug.log
npm ERR! not ok code 0

npm ERR! oref0@0.2.0 global-install: `npm install && sudo npm install -g && sudo npm link && sudo npm link oref0`
npm ERR! Exit status 50
npm ERR! 
npm ERR! Failed at the oref0@0.2.0 global-install script.
npm ERR! This is most likely a problem with the oref0 package,
npm ERR! not with npm itself.
npm ERR! Tell the author that this fails on your system:
npm ERR!     npm install && sudo npm install -g && sudo npm link && sudo npm link oref0
npm ERR! You can get their info via:
npm ERR!     npm owner ls oref0
npm ERR! There is likely additional logging output above.
npm ERR! System Linux 4.4.21-v7+
npm ERR! command "/usr/bin/nodejs" "/usr/bin/npm" "run" "global-install"
npm ERR! cwd /home/pi/src/oref0
npm ERR! node -v v0.10.29
npm ERR! npm -v 1.4.21
npm ERR! code ELIFECYCLE
npm WARN This failure might be due to the use of legacy binary "node"
npm WARN For further explanations, please read
/usr/share/doc/nodejs/README.Debian

npm ERR! 
npm ERR! Additional logging details can be found in:
npm ERR!     /home/pi/src/oref0/npm-debug.log
npm ERR! not ok code 0
Scott Leibrand
@scottleibrand
Oct 01 2016 02:47
yeah, this is the first time it's attempting to install oref0 for you
garykidd
@garykidd
Oct 01 2016 02:47
so that's OK
Scott Leibrand
@scottleibrand
Oct 01 2016 02:48
looks like you messed up perms on your ~/src/oref0 directory at some point by doing something with sudo maybe
garykidd
@garykidd
Oct 01 2016 02:48
sure
Scott Leibrand
@scottleibrand
Oct 01 2016 02:48
just sudo rm -rf ~/src/oref0/ and re-clone it
(re-clone = the step 1: mkdir -p ~/src; cd ~/src && git clone -b dev git://github.com/openaps/oref0.git || (cd oref0 && git checkout dev && git pull))
garykidd
@garykidd
Oct 01 2016 02:49
done.
re do script then
Scott Leibrand
@scottleibrand
Oct 01 2016 02:49
ya
garykidd
@garykidd
Oct 01 2016 02:52

getting some

> oref0@0.2.0 global-install /home/pi/src/oref0
> npm install && sudo npm install -g && sudo npm link && sudo npm link oref0

npm WARN deprecated lodash.assign@4.2.0: This package is deprecated. Use Object.assign.
npm WARN deprecated to-iso-string@0.0.2: to-iso-string has been deprecated, use @segment/to-iso-string instead.
npm WARN deprecated jade@0.26.3: Jade has been renamed to pug, please install the latest version of pug instead of jade
npm WARN engine escodegen@1.8.1: wanted: {"node":">=0.12.0"} (current: {"node":"0.10.29","npm":"1.4.21"})
npm WARN engine hawk@3.1.3: wanted: {"node":">=0.10.32"} (current: {"node":"0.10.29","npm":"1.4.21"})
npm WARN engine form-data@2.0.0: wanted: {"node":">= 0.12"} (current: {"node":"0.10.29","npm":"1.4.21"})

and still waiting for likely more....

Scott Leibrand
@scottleibrand
Oct 01 2016 02:53
cool. so my fix correctly detected that you needed a new oref0, and it's now installing it
hopefully that's all that's needed to get BGs from NS
and then we can figure out the --G5 stuff tomorrow
garykidd
@garykidd
Oct 01 2016 02:54
amen brother
so it just finished actually...almost scare (nahhh) to watch llp log
loop
Scott Leibrand
@scottleibrand
Oct 01 2016 02:58
maybe try it manually first so you don't have to wait
openaps get-ns-glucose
garykidd
@garykidd
Oct 01 2016 02:58
nope....I can already tell it seems to be working---at least a step further than prior attempts
Scott Leibrand
@scottleibrand
Oct 01 2016 02:59
:)
you can look at monitor/glucose.json too
LBehler
@LBehler
Oct 01 2016 02:59
Hey @scottleibrand I'm trying to follow along with your conversation with @garykidd but I'm stuck at the command to download the latest version of oref0 from git. When I run the cmd "cd ~/src/oref0/ && git pull" it barfs. I don't have a src directory
Scott Leibrand
@scottleibrand
Oct 01 2016 03:00
@lbehler that's for someone who has already cloned the oref0 repo. you'll want to do step 1 at http://openaps.readthedocs.io/en/oref0-setup/docs/walkthrough/phase-2/oref0-setup.html instead
garykidd
@garykidd
Oct 01 2016 03:01
yes....working
LBehler
@LBehler
Oct 01 2016 03:01
Dude thanks so much. We really appreciate your patience lol
garykidd
@garykidd
Oct 01 2016 03:02
it just set temp basal......and you know what else.....
my stupid change to mmtune.py is still in "effect" and knowing only one part of how that works...the one change I made..even though it should probably be changed.....it works.... range 916.707 to 916.709 default 916.708.....it works all the time
Scott Leibrand
@scottleibrand
Oct 01 2016 03:09
I'm thinking about doing something not too dissimilar for the Explorer board: saving the last known good mmtune frequency, and then setting that (without actually tuning) every loop when we reset the cc1110
right now when I reset the cc1110 radio chip every reboot, I have to do a full mmtune to get it to be able to do anything
garykidd
@garykidd
Oct 01 2016 03:13
I guess it would be easy enough to track what freq was selected....so that maybe you could potentially limit the range of the scan to only incorporate prior know "good" frequencies....or maybe even just hop from one to the other without any c=kind of call to mmtune whatsoever.....because my hunch is...that while the best freq may drift....they may likley fall within a fairly tigh range....much narrower than the full mmtune standard ranges
Scott Leibrand
@scottleibrand
Oct 01 2016 03:13
yep
we've had lots of ideas for how to do mmtune smarter, but it hasn't been a priority for anyone to actually code up
garykidd
@garykidd
Oct 01 2016 03:16
I actually think I have lots of ideas...just don't know how to code...I know I can learn..and I think I may try to...at least a little bit. I love the logic underlying it all. years ago I toyed with C...but other things cuaght my attention...even before C++ was catching on = a long time ago.
Scott Leibrand
@scottleibrand
Oct 01 2016 03:17
yeah, mmtune is probably an area where you could get up to speed and start contributing pretty quickly. it helps that it's of particular interest to you given your weird pump
garykidd
@garykidd
Oct 01 2016 03:17
gonna hit the hay. I do appreciate the help. APS401 = functional.
Scott Leibrand
@scottleibrand
Oct 01 2016 03:18
sweet! sleep well with a working loop. :)
garykidd
@garykidd
Oct 01 2016 03:18
Will pursue...
roger that.
and likewise
dougtakeuti
@dougtakeuti
Oct 01 2016 03:30

@scottleibrand another doubt: the script creates this alias:

upload-pumphistory-entries = ! bash -c "openaps prep-pumphistory-entries && ns-upload-entries upload/pumphistory-entries.json"

shouldn't it be something like this?

upload-pumphistory-entries = ! bash -c "openaps prep-pumphistory-entries && ns-upload $NIGHTSCOUT_HOST $API_SECRET pumphistory.json upload/pumphistory-entries.json"
Scott Leibrand
@scottleibrand
Oct 01 2016 03:35
don't waste too much time fixing that stuff. it all needs replaced per what I told @tim2000s earlier
I believe that particular one might be some legacy DIYPS stuff that we don't need at all
I think it actually was uploading the pumphistory to entries.json as well as treatments.json
that might be part of the source of @tim2000s and others' extra NS usage
Matthias
@ceben80
Oct 01 2016 03:41
Morning guys and greetings from Germany :wink: ....looping around 90 :tada:
Starting the day with 4.8 A1C :star2:
Scott Leibrand
@scottleibrand
Oct 01 2016 03:43
heh. that might be lower than mine (non-T1D) ;-)
Eric Jensen
@elnjensen
Oct 01 2016 03:48
One way to enter carbs but no bolus into the pump is to use bolus wizard, and then just zero out the bolus it suggests. Is there any easier way?
Just tried the "Capture Event" and it put a note in NS, but doesn't seem to have registered as actual carbs eaten.
Scott Leibrand
@scottleibrand
Oct 01 2016 03:49
@elnjensen "carbs but no bolus" method works on some pumps, but not others
Eric Jensen
@elnjensen
Oct 01 2016 03:50
Sorry, which of the two things I just mentioned is that? :worried:
Scott Leibrand
@scottleibrand
Oct 01 2016 03:50
some people consider it a bug ( :point_up: September 30, 2016 6:39 PM ) and others like you a feature
entering carbs into the bolus wizard and then canceling the bolus
Eric Jensen
@elnjensen
Oct 01 2016 03:51
Ah, got it. Is there any other way to achieve that carb-only entry? (Short of going to NS to enter there - I mean via pump.)
Scott Leibrand
@scottleibrand
Oct 01 2016 03:52
probably would have to do some development to get the event carbs captured
I don't know of any easier ways. we use off-pump carb entry exclusively, because @danamlewis hates the bolus wizard
Eric Jensen
@elnjensen
Oct 01 2016 03:53
OK, no worries - almost certainly not worth it (Or at least low enough priority that it would never get done. :smile: )
Scott Leibrand
@scottleibrand
Oct 01 2016 03:53
so we use DIYPS, pebble, alexa, and once in a while NS
Eric Jensen
@elnjensen
Oct 01 2016 03:54
Cool - CarePortal watch app is pretty straightforward, nice and quick to use.
Brian Bonner
@itsn0tm3
Oct 01 2016 03:54
@scottleibrand Interestingly out of the 4 rigs that I moved over to the new oref-setup script, 3 had the same problem
actually nevermind
I think this one is different
other 2 were abotu the six package
this one I think has to do with sudo
Scott Leibrand
@scottleibrand
Oct 01 2016 03:55
yeah, I saw the six thing on an upgrade. haven't seen it on any of the ones I've flashed fresh though
Eric Jensen
@elnjensen
Oct 01 2016 04:01
If I do a carb correction, is it useful to enter that? I guess it could be useful for AMA to know that there are carbs working to bring BG up.
Brian Bonner
@itsn0tm3
Oct 01 2016 04:01
yeah the newest rig I had did just fine
the upgrades were the ones that had problems
Scott Leibrand
@scottleibrand
Oct 01 2016 04:02
@elnjensen yes, it is often useful, unless you are running exceptionally sensitive (beyond what autosens can adjust for) and would rather ride high than risk another low
although even then I would recommend entering the carbs and setting a temp target
Eric Jensen
@elnjensen
Oct 01 2016 04:02
Great, thx.
Dana Lewis
@danamlewis
Oct 01 2016 04:06
@elnjensen definitely enter if it's a medium or large carb correction, then it will help a lot for any rebound.
Eric Jensen
@elnjensen
Oct 01 2016 04:06
Thanks, @danamlewis ! :candy:

@scottleibrand Found the docs for wpa-roam! It's in man wpa_action. Here's the money quote:

wpa_action is given as an argument to a wpa_cli daemon.

          wpa_cli -i eth1 -a /sbin/wpa_action -B

   This can be done by using the wpa-roam option in the interfaces file.  wpa-roam  takes  one  argument,  a  user  provided
   wpa_supplicant.conf file.

   The  inet  METHOD  must be 'manual' for this interface, as it will be configured according to wpa_cli action events. Also
   supply a 'default' interfaces stanza using the dhcp inet METHOD so that networks without an 'id_str' option can  fallback
   to  attempting  to  receive an ip via dhcp. If one or more networks requires additional network configuration, provide an
   unique 'id_str' for each network, and an interfaces stanza using the 'id_str' value as a LOGICAL interface.

So the combination of wpa-roam and manual does seem to be the key here.

Scott Leibrand
@scottleibrand
Oct 01 2016 04:09
yeah, I found that. we knew that wpa-roam needs manual and a default interface
but that doesn't explain whether such a basic wpa-roam config is identical to a wpa-conf config with dhcp specified on the interface
afaict wpa-roam is only useful for having the same IP address etc. on a wifi and an ethernet interface or similar, so you can plug in or use wifi and roam between those two
it has nothing to do, afaik, with switching between wifi networks
Eric Jensen
@elnjensen
Oct 01 2016 04:12
Have you tested that assumption, though? I found that that was exactly the change that made it work for me.
Scott Leibrand
@scottleibrand
Oct 01 2016 04:12
I have the wpa-conf config and it works for me, so dunno
Brian Bonner
@itsn0tm3
Oct 01 2016 04:13
@scottleibrand interesting thing for you, on this rig I just worked on, redoing npm install, npm link, npm link oref0 and npm run global-install fixes the error on the 2nd run and works just fine after that
Scott Leibrand
@scottleibrand
Oct 01 2016 04:14
:+1:
Brian Bonner
@itsn0tm3
Oct 01 2016 04:14
not sure why it doesn't work during the first run through the setup but the 2nd time works (shrugs)
dougtakeuti
@dougtakeuti
Oct 01 2016 04:31
@scottleibrand thanks for letting me know. I'll just get rid of that bit and see what happens.
What I'm really trying to figure out right is why upload/latest-treatments.json is empty.
When I run nightscout cull-latest-openaps-treatments monitor/pumphistory-zoned.json settings/model.json $(openaps latest-ns-treatment-time), I get []. Running openaps latest-ns-treatment-time, gives 2020-01-19T16:14:21-02:00. That doesn't seem right.
Matthias
@ceben80
Oct 01 2016 04:48
@scottleibrand from today on I am running with a Huwai mobile hotspot, will be delivered in the next 6 hours or so....not the smallest but 5200 mAh :star2:
Scott Leibrand
@scottleibrand
Oct 01 2016 05:19
@dougtakeuti what does your Nighscout admin page show for future treatments?
Rob Kresha
@rkresha
Oct 01 2016 06:30
@scottleibrand why use rsync -rtu when cp -pu would accomplish the same thing, especially why copying individual files? Just curious, I only ever think of rsync in a large copy across a network
when copying*
Scott Leibrand
@scottleibrand
Oct 01 2016 06:31
I don't have a -u flag listed in my man cp
but the meta-answer is "because I know that way works" :)
Rob Kresha
@rkresha
Oct 01 2016 06:32
ha
and you like being fancy
Scott Leibrand
@scottleibrand
Oct 01 2016 06:33
heh, more likely I just learned one way and stuck with it.
but looks like cp -pu would probably work on the debian systems we're working with
if there's any reason to switch, would just need to do some testing on all platforms
if not, I know of a lot of other things that actually need refactoring. ;-)
I tend to write fairly inelegant inefficient code that just barely works, so it's good to have someone coming along behind me and helping to refactor it and make it more robust
Rob Kresha
@rkresha
Oct 01 2016 06:35
sure make me out to be the petty bad guy :panda_face:
ha
Scott Leibrand
@scottleibrand
Oct 01 2016 06:36
haha not exactly. ;-)
in all seriousness, I didn't know cp -pu did the same thing, so I'll probably try that first in future
Rob Kresha
@rkresha
Oct 01 2016 06:39
rsync has some overhead in a "large # of files" scenario, where it does all of the comparing of timestamps up front instead of inline... so in this case there wouldn't be any extra overhead.... my backstory is I just ran the oref0-setup to compare it to what I've been running with, noticed the rsync commands immediately and starting wondering if there was some backstory there
Scott Leibrand
@scottleibrand
Oct 01 2016 06:39
that came from my knowing how to do it with rsync from doing network copies
of lots of TB of video content @ work :)
Rob Kresha
@rkresha
Oct 01 2016 06:40
that's how I got familiar with robocopy
anyways, my .02 cents for being simplistic would be to use cp -pu .. I've see -u in all of the flavors I've dealt with: ubuntu, suse, rhel, OEL, etc
Scott Leibrand
@scottleibrand
Oct 01 2016 06:42
apparently just not OS X
Rob Kresha
@rkresha
Oct 01 2016 06:42
whomp whomp
Scott Leibrand
@scottleibrand
Oct 01 2016 06:42
but we aren't running oref0 there, so NBD
Rob Kresha
@rkresha
Oct 01 2016 06:42
it was the part of the apple that was consumed
it wasn't funny, I know
Scott Leibrand
@scottleibrand
Oct 01 2016 06:43
feel free to put in a PR if you like. will just need you or someone to test it on Pi and Edison
Rob Kresha
@rkresha
Oct 01 2016 06:43
consider PI done
I have a external shell script running my old loop and was using it... will work on the PR
Scott Leibrand
@scottleibrand
Oct 01 2016 06:44
yay, openxshareble works with oref0-setup now! :) openaps/oref0#229
surprised I got that working so fast. turns out my main problem was a receiver that stopped doing BLE reliably
wonder if it would be better after a hard paperclip reset
working well on a new G4 share receiver though
Rob Kresha
@rkresha
Oct 01 2016 06:47
my dilemma is I don't want to pay for a mifi or hotspot, but also don't want to have my share receiver tethered to my rig... and since I can't connect two bluetooths device to the share receiver, I'm stuck... #diabeticproblems
Scott Leibrand
@scottleibrand
Oct 01 2016 06:47
we work around that by having two receivers
which is also good for avoiding 2h restart gaps
Rob Kresha
@rkresha
Oct 01 2016 06:48
so I have a start G4 receiver as well, but the algorithm on it is different so my numbers are slightly different and it doesn't have bluetooth... is dexcom still selling g4 w/ share?
standartd*
ugh
Scott Leibrand
@scottleibrand
Oct 01 2016 06:49
ya, you can still get G4 with share
they'll be selling G4 transmitters for awhile yet, until all the pump makers switch to G5
so G4 w/ Share is their standard G4 receiver now in the US
Rob Kresha
@rkresha
Oct 01 2016 06:49
you remember cost on the g4 w/ share?
I've git my deductible for the year so now would be the time
Scott Leibrand
@scottleibrand
Oct 01 2016 06:50
nope. want to say $600, but who knows
Rob Kresha
@rkresha
Oct 01 2016 06:50
git = hit
smh
Scott Leibrand
@scottleibrand
Oct 01 2016 06:50
or would that be git HEAD? ;-)
git diff HEAD desk ;-p
Rob Kresha
@rkresha
Oct 01 2016 06:52
my mind just exploded
anyone else ever notice, that sometimes it takes a couple runs of git merge for it to see the difference in branches/forks?
Scott Leibrand
@scottleibrand
Oct 01 2016 06:53
nope
you might need to go pull the branch first though
git checkout master && git pull && git checkout yourbranch && git merge master
Rob Kresha
@rkresha
Oct 01 2016 06:54
what about between forks
so when I'm in my dev and I do git merge openaps/dev
Scott Leibrand
@scottleibrand
Oct 01 2016 06:54
are you talking GUI stuff? dunno about that
maybe git fetch openaps first?
Rob Kresha
@rkresha
Oct 01 2016 06:55
what's the diff between merge and fetch.. please don't send me a letmegooglethatforyoulink
Scott Leibrand
@scottleibrand
Oct 01 2016 06:55
lol. pull is fetch+merge essentially
fetch just updates the behind the scenes cache. merge actually updates the current branch
Rob Kresha
@rkresha
Oct 01 2016 06:56
but pull and push is relative to your local copy vs github, yes?
Scott Leibrand
@scottleibrand
Oct 01 2016 06:56
usually
git remote -v shows your pull and push destinations
Rob Kresha
@rkresha
Oct 01 2016 06:56
fetch is more metadata?
Scott Leibrand
@scottleibrand
Oct 01 2016 06:56
it's loading the current state of the remotes so you can do things with them
downloads all the branches, diffs, etc.
Rob Kresha
@rkresha
Oct 01 2016 06:57
yep, that's probably the key in my situation
thanks@
Rob Kresha
@rkresha
Oct 01 2016 07:04
@scottleibrand PR to oref0-setup or dev ?
Scott Leibrand
@scottleibrand
Oct 01 2016 07:04
dev please
I'm trying to delete oref0-setup so I can re-branch it from dev
Rob Kresha
@rkresha
Oct 01 2016 07:04
figured that was the case, just double-checking
Scott Leibrand
@scottleibrand
Oct 01 2016 07:04
have to close all open PRs against it first. :)
Rob Kresha
@rkresha
Oct 01 2016 07:14
Alright Scott, nice chatting with you. Thanks for all that you do! :thumbsup:
Scott Leibrand
@scottleibrand
Oct 01 2016 07:14
have a good night!
Tim Street
@tim2000s
Oct 01 2016 11:05
I gave AMA the ultimate test last night. Late night pizza. I'm still amazed at the result. Went to bed at 6.2 and woke up in the fives. Wow.
PieterGit
@PieterGit
Oct 01 2016 11:47
@scottleibrand : just tried wip/radio_ww branch. can confirm that it works
i noticed you use --cgm=shareble, would it not be better to use --cgm=g4 --glucose_source=shareble. That way you could also use: --cgm=g4 --glucose_source=wired (use G4 directly from USB) or --cgm=g4 --glucose_source=ns (use Nightscout to get glucose readings to openaps/oref0)
PieterGit
@PieterGit
Oct 01 2016 11:55
alternative names: --glucose-transport, usb instead of wired, etc.
Peter Miller
@ochenmiller
Oct 01 2016 12:09
One year anniversary! A year ago today we'd been looping for three nights. Haven't looked back - except at those awesome charts!
PieterGit
@PieterGit
Oct 01 2016 12:22
@scottleibrand : is there a reason why oref0-setup.sh uses bluetooth/bluez 5.37 instead of the latest 5.42?
PieterGit
@PieterGit
Oct 01 2016 12:43
@scottleibrand: found a bug in the shareble script. the if ! python -c "import openxshareble" 2>/dev/null; then must be removed, because the oref0-setup.sh depends on $HOME/src/openxshareble/bluetoothd.conf . I had openxshareble installed, but with pip, so ~/src/openxshareble did not exist
kenstack
@kenstack
Oct 01 2016 13:20
@elnjensen hereis one reference to the walsh curves with graphs http://jdst.org/May2009/Articles/VOL-3-3-ORG12-ELLINGSEN.pdf
Tim Street
@tim2000s
Oct 01 2016 13:21
Anyone know of any tools I can use on Windows to flash my Slice of Radio card for Pi. Still having issues with boost libraries on my Linux install
kenstack
@kenstack
Oct 01 2016 13:26
@bustavo will post the equations behind glucodyn today sorry not sure where that pdf went …if you have any questons just ping me - the curves we use in there are just curve fits of the walsh equations, but imho the walsh curves we have found are a bit too agressive in terms of prediction of used insulin. the MDT curves are based on someone elses curves … its in the old 722 manuals
Tim Street
@tim2000s
Oct 01 2016 14:33
Right - if you're having problems installing the cc-tool for mmeow link, there is a flash programmer at this adddress that works from windows: http://www.ti.com/tool/cc-debugger#technicaldocuments
I've just used it effectively.
Gustavo
@bustavo
Oct 01 2016 15:15
Just added the basic_math.pdf to Glucodyn Github repo: https://github.com/perceptus/glucodyn ( cc @kenstack )
Tim Street
@tim2000s
Oct 01 2016 16:21
In pump-loop, what's the purpose of sleep $[ ( $RANDOM / 2048 ) ]s ? Why not simply sleep 5m?
dougtakeuti
@dougtakeuti
Oct 01 2016 16:34
@scottleibrand not sure what you meant by checking future events in the Nightscout admin page. Did you mean in azure?
Brett
@burgertime14
Oct 01 2016 16:36
Been looking for about 30 min, anyone have a link to nightscout set up on Heroku?
dougtakeuti
@dougtakeuti
Oct 01 2016 16:40
@burgertime14 is this what you're looking for? http://www.nightscout.info/wiki/labs/how-to-heroku
Brett
@burgertime14
Oct 01 2016 16:42
@dougtakeuti Thanks. Not sure how I was able to find it, may have something to do with my current state of mind after Azure shut me out :angry:
*wasn't
dougtakeuti
@dougtakeuti
Oct 01 2016 16:44
same situation here regarding azure...
Brett
@burgertime14
Oct 01 2016 16:45
@dougtakeuti -- @rkresha is having issues as well, he believes, related to the mlabs upgrade
dougtakeuti
@dougtakeuti
Oct 01 2016 16:46
But heroku is working fine, and I have the impression it is faster
Brett
@burgertime14
Oct 01 2016 16:47
Would it require the $7/mo plan or does free still have enough juice to cover a NS install?
dougtakeuti
@dougtakeuti
Oct 01 2016 16:49
i've been running free
Brett
@burgertime14
Oct 01 2016 16:49
Awesome, thanks --- signing up for Heroku right now
dougtakeuti
@dougtakeuti
Oct 01 2016 16:51
actually, let me correct that. - running free with my old loop, without autosens. It was my current attempt to implement the advanced features that made azure go beyond the free limit.
Brett
@burgertime14
Oct 01 2016 16:52
I am completely ok with $7/mo, I may choose that, just for piece of mind. I am not ok with $55/mo for Azure
Scott Leibrand
@scottleibrand
Oct 01 2016 17:23
I would recommend the Heroku $7/mo plan for openaps at this point. Maybe should update the docs to recommend that over Azure.
@tim2000s the random sleep is to avoid synchronization of multiple rigs.
And that random interval is a few seconds, not minutes. We run pump-loop every minute and check if anything has changed that requires pulling new data.
Scott Leibrand
@scottleibrand
Oct 01 2016 17:33
@tim2000s awesome re: pizza test! :)
@PieterGit bluez 5.37 is just because that's what we know works. do you know if there's a more recent version that can be installed via package instead of compiled on the edison? would love to speed up that part of the install.
@PieterGit thanks for the bug report. I'll clone openxshareble regardless to get that file, but only install if no packaged version is already present.
Scott Leibrand
@scottleibrand
Oct 01 2016 17:40
Tim Street
@tim2000s
Oct 01 2016 17:43
@scottleibrand That's what i thought - I've dialed it back to five minutes, as my sgv is only updating on a 5 minute basis, 5 mins for a bolus isn't really an issue, and you get the correct time for that from the pump, and in all fairness, it reduces the load on Azure so stops it locking me out ;)
Scott Leibrand
@scottleibrand
Oct 01 2016 17:45
that seems like a decent workaround until we fix NS uploads to stop overloading Azure. (although I would try reducing the frequency of ns-loop rather than pump-loop there). but once that is fixed, you'll probably find it works better to let it refresh other stuff more frequently, so it's ready to go once a new BG comes in. that becomes particularly important if you start using multiple rigs
Brett
@burgertime14
Oct 01 2016 17:52
@scottleibrand @dougtakeuti I have Heroku up and running. It is so much faster than Azure, it will be difficult to make me go back
dougtakeuti
@dougtakeuti
Oct 01 2016 17:53
:+1:
Dana Lewis
@danamlewis
Oct 01 2016 18:04
@ochenmiller happy looping-versary :) :confetti_ball:
Peter Miller
@ochenmiller
Oct 01 2016 18:50
We're calling it the "Hankiversary" in our house, as our rig has been called "Hank the Pank" since day one! :smile:
Dana Lewis
@danamlewis
Oct 01 2016 18:51
Heh :)
Sandra Keßler
@SandraK82
Oct 01 2016 19:09
since today part time Looping :D so excited
Dana Lewis
@danamlewis
Oct 01 2016 19:12
Woot!
Sandra Keßler
@SandraK82
Oct 01 2016 19:13
part time because my setup is not yet transportable, 3d-Printer is working hard :)
thanks to the great setup script it was so easy, only problem was missing hint for configuring for eu. maybe i find the time to write a complete from Zero to hero guide in the next days (raspi + ti usb + xdrip + openaps and all done locally on raspi + xdrip handy)
Scott Leibrand
@scottleibrand
Oct 01 2016 19:37
@SandraK82 what parts did you have to modify for an EU pump?
Sandra Keßler
@SandraK82
Oct 01 2016 19:57
@scottleibrand mmtune.py has to be changed:
class MMTune:
FREQ_RANGES = {
'US': { 'start': 916.5, 'end': 916.9, 'default': 916.630 },
'WW': { 'start': 868.4, 'end': 868.6, 'default': 868.525 }
}
Scott Leibrand
@scottleibrand
Oct 01 2016 19:57
Sandra Keßler
@SandraK82
Oct 01 2016 19:57
the ww frequentes have to be modified to this values
Scott Leibrand
@scottleibrand
Oct 01 2016 19:58
no, your range is narrower
why is that needed?
is mmtune doing usedDefault: true for you?
Sandra Keßler
@SandraK82
Oct 01 2016 19:58
no
the tip was from germanLoop group
eespecially a doc which is passed on
Scott Leibrand
@scottleibrand
Oct 01 2016 19:59
the 'start': 868.150, 'end': 868.750, means it will scan frequencies within that range and pick the best one. your frequencies are all inside the range it's scanning
Tim Street
@tim2000s
Oct 01 2016 19:59
Hmmm. I'm using a pump that originated in Germany and it's working fine with the defualt WW settins
Sandra Keßler
@SandraK82
Oct 01 2016 20:00
with the default ones i did not get my pump (522 ww)
Tim Street
@tim2000s
Oct 01 2016 20:00
Mine is 522 ww too
Sandra Keßler
@SandraK82
Oct 01 2016 20:00
with the changed ones it was easy
Scott Leibrand
@scottleibrand
Oct 01 2016 20:00
was that the only change? or did you do https://github.com/oskarpearson/mmeowlink/wiki/Non-USA-pump-settings too?
Sandra Keßler
@SandraK82
Oct 01 2016 20:00
however pump reception is a huge problem currently for me
i used a modified firmware, so the https://github.com/oskarpearson/mmeowlink/wiki/Non-USA-pump-settings is not needed
Scott Leibrand
@scottleibrand
Oct 01 2016 20:02
wonder if your firmware is still not picking ideal values for the other parameters, so you have to get exactly the right frequency on tuning, rather than just "close enough"
Sandra Keßler
@SandraK82
Oct 01 2016 20:03
if i use my other equipment and look at the bands here... it is very crowded
Scott Leibrand
@scottleibrand
Oct 01 2016 20:04
I think we need to get mmtune to write its last known good settings, and then have it do a search outward from there (starting with very narrow adjustments) until it sees a local maximum
Sandra Keßler
@SandraK82
Oct 01 2016 20:04
living in plain sight to the central station of kassel, more or less on top of the rails in an old building with a handytower 10m from my seat....
Scott Leibrand
@scottleibrand
Oct 01 2016 20:05
that would allow mmtune to adaptively narrow in on exactly the right frequency, without everyone having to tweak their settings
Sandra Keßler
@SandraK82
Oct 01 2016 20:05
the ti is less then two meter away from me and the pump at my bra
Scott Leibrand
@scottleibrand
Oct 01 2016 20:06
we could also have mmtune try both the US and WW settings if it doesn't yet know which one your pump is
Sandra Keßler
@SandraK82
Oct 01 2016 20:07
we use home automatation equipment in our flat which is supposed to work over 100m indoors.. it mails regulary within 10m
Scott Leibrand
@scottleibrand
Oct 01 2016 20:08
another thing you could try is having mmtune try narrower frequency spacing
not reduce the range, but increase the number it tries
Sandra Keßler
@SandraK82
Oct 01 2016 20:10
currently mmtune is called evry 5 or so minutes and fails sometimes to find the pump
Scott Leibrand
@scottleibrand
Oct 01 2016 20:11
you might have better luck only calling it on pump-loop failure as oref0-setup does?
once you have tuning good, no reason to change it until comms start failing
Sandra Keßler
@SandraK82
Oct 01 2016 20:12
no, i dont call it by myself, ist called by oref0
Scott Leibrand
@scottleibrand
Oct 01 2016 20:12
oh, so you see failures that often? yeah, that sucks.
thought you were saying you put it in cron to run every 5m
I think I'll be doing some work on mmtune soon. see also recent convo in https://gitter.im/oskarpearson/mmeowlink
if anyone wants to help, please LMK. :)
Sandra Keßler
@SandraK82
Oct 01 2016 20:14
we are getting used to it, currently ns+loop tv runs on most devices here :D evry room has a tail -F and ns page open all the time
Scott Leibrand
@scottleibrand
Oct 01 2016 20:14
heh nice
Sandra Keßler
@SandraK82
Oct 01 2016 20:15
the benefit of working professionally in sw engineering.. enough devices to use are here
Scott Leibrand
@scottleibrand
Oct 01 2016 20:15
we have 6 rigs spaced around our apartment, all taking turns. :)
Sandra Keßler
@SandraK82
Oct 01 2016 20:16
wow, all rigs working at the same time?
Scott Leibrand
@scottleibrand
Oct 01 2016 20:18
yep. somme in range, some not, but all playing nicely
Sandra Keßler
@SandraK82
Oct 01 2016 20:19
nice :+1:
i guess i will try around a bit more. current target is to change the firmware to talk with one of my ble boards, like rileylink
ble has pretty good range here ( i dont get why, the band is scrambled as hell and i found ~200 devices nearby on peak)
Scott Leibrand
@scottleibrand
Oct 01 2016 20:22
the MDT 900MHz protocol is really simplistic, so probably doesn't take a lot to be better than that at dealing with a noisy RF environment
Pete Schwamb
@ps2
Oct 01 2016 20:27
ble works well because of channel hopping, and because it’s gfsk. mdt is basically morse code at high speed. :)
Scott Leibrand
@scottleibrand
Oct 01 2016 20:28
heh
Pete Schwamb
@ps2
Oct 01 2016 20:32
And I’m not speaking metaphorically. It really is done the exact same way you’d send morse code.
Takes up a lot of bandwidth, and is very susceptible to noise.
Well, the coding isn’t the same, but the modulation is.
Sandra Keßler
@SandraK82
Oct 01 2016 20:34
as soon as my loop is portable, i will try at my parents home.. very small village sourrouded by nothing then woods
garykidd
@garykidd
Oct 01 2016 21:51
Apologies...i actually did search fo this but couldnt seem to find the variable to enable with azure portal to enable purple predict BG lines associated with AMA....
Scott Leibrand
@scottleibrand
Oct 01 2016 21:51
env SHOW_FORECAST="openaps ar2"
and then you have to hit the ... next to 24HR and check the box
garykidd
@garykidd
Oct 01 2016 22:02
blob
isnt there supposed to be n line though?
one
Jason Calabrese
@jasoncalabrese
Oct 01 2016 22:10
@rkresha @scottleibrand I think in most places we can use oref0 copy-fresher or a jq merge and then use a report invoke instead of naked cp/rsync
Scott Leibrand
@scottleibrand
Oct 01 2016 22:27
I think jq merge adds too many edge failure cases
what does oref0 copy-fresher do that rsync -tuv or cp -up doesn't?
garykidd
@garykidd
Oct 01 2016 22:55
@scottleibrand something wrong with purple lines right?
Scott Leibrand
@scottleibrand
Oct 01 2016 22:56
@PieterGit we don't need --cgm=g4 --glucose_source=ns because oref0-setup actually sets up for both locally connected CGM (G4 USB, G5 USB or G4 Share BLE) and NS. if the CGM is connected, it uses that; if not, it uses NS.
@garykidd there are supposed to be multiple lines when you have COB
but the lines are supposed to be smooth: that sudden uptick means something is not quite right
garykidd
@garykidd
Oct 01 2016 22:57
got even crazier but maybe becuase I lost comm with rig for a bit
Scott Leibrand
@scottleibrand
Oct 01 2016 22:57
@PieterGit as there are only those three mutually exclusive options for local CGM data, I don't think it makes sense to have two switches
garykidd
@garykidd
Oct 01 2016 22:58
blob
Scott Leibrand
@scottleibrand
Oct 01 2016 23:03
can you paste the determine-basal output from your pump-loop.log? should look like this:
{"temp":"absolute","bg":172,"tick":-10,"eventualBG":59,"snoozeBG":72,"predBGs":{"IOB":[172,162,151,140,129,119,109,99,90,82,74,68,61,56,51,47,44,41,39,39,39,39,39,39,39,39,39,39,39,40,40,41,41,41,42],"aCOB":[172,170,165,158,149,140,129,119,110,102,94,87,81,76,71,67,64,61,59,58,56,56,56,56,56,57,58,58,59,60,60,61,61,61,62],"COB":[172,162,151,140,130,119,109,100,92,84,77,70,65,60,56,52,49,47,46,44,44,44,44,45,46,47,48,49,50,51,52,53,54,54,55,56,56,57,57,58,58,58,59]},"COB":7,"IOB":3.455,"reason":"COB: 7, Dev: 3, BGI: -10.66, ISF: 40, Target: 100; Eventual BG 59 < 100, setting -1.9U/hr, but 30m left and 0 ~ req 0U/hr: no action required"}
the part right after reporting enact/suggested.json
garykidd
@garykidd
Oct 01 2016 23:19
{"temp":"absolute","bg":156,"tick":-1,"eventualBG":221,"snoozeBG":141,"predBGs":{"IOB":[156,154,150,147,143,137,130,122,113,103,92,80,67,55,43,39],"aCOB":[156,150,143,139,134,128,122,117,112,107,103,99,94,91,87,84,80,77,75,72,70,68,67,65,64,63,62,61,61,61,61,61,61,67,74,80,86,92,98,104,109,115,120,125,130,135,140,145,149,153,158,162,166,169,173],"COB":[156,155,153,153,153,152,150,149,148,147,146,145,145,144,143,143,142,142,142,141,141,141,141,142,142,142,142,143,143,143,144,144,145,151,157,163,169,174,179,184,189,193,197,200,204,207,210,212,214,216,218,219,220,220,221]},"COB":74,"IOB":6.663,"reason":"COB: 74, Dev: 91, BGI: -16.19, ISF: 40, Target: 146; 221-141 in range: no temp required; setting current basal of 0.85 as temp","duration":30,"rate":0.85}
Scott Leibrand
@scottleibrand
Oct 01 2016 23:21
what is your DIA set to?
garykidd
@garykidd
Oct 01 2016 23:21
@scottleibrand not sure where to pu it but when selecting G5 cgm option in oref0-setup..... it does appear that you need to amend raw cgm report...I had an error...but this has seemed to work....
openaps report add raw-cgm/raw-entries.json JSON cgm oref0_glucose --no-raw --hours 25 --threshold 100
Scott Leibrand
@scottleibrand
Oct 01 2016 23:21
what was it before you changed it?
garykidd
@garykidd
Oct 01 2016 23:22
it appeared that the base report didn't have --no-raw as we had speculated last night.
Scott Leibrand
@scottleibrand
Oct 01 2016 23:22
odd: mine does
root@edison-eb2 ~/myopenaps # openaps report show raw-cgm/raw-entries.json --cli
openaps use cgm oref0_glucose --hours "25" --threshold "100" --no-raw
and that's on a G4
garykidd
@garykidd
Oct 01 2016 23:22
before chagning it it was...
openaps report show raw-cgm/raw-entries.json --cli openaps use cgm oref0_glucose --hours "25" --threshold "100"
hmmmm...I wonder if somehow G4 election is different in this respect....it isn't case sensitive is it?
incedentally it appears that the quotes are needed...not does the order matter for the hours, threshold and no-raw elections ...is that correct?
DIA is 4
Scott Leibrand
@scottleibrand
Oct 01 2016 23:25
order shouldn't matter
this is odd:
root@edison-eb2 ~/myopenaps # openaps report show raw-cgm/raw-entries.json --cli
openaps use cgm oref0_glucose --hours "25" --threshold "100" --no-raw
root@edison-eb2 ~/myopenaps # openaps report show raw-cgm/raw-entries.json --json
{"type": "report", "name": "raw-cgm/raw-entries.json", "raw-cgm/raw-entries.json": {"count": "", "use": "oref0_glucose", "no_raw": "False", "reporter": "JSON", "seconds": "", "sensor": "", "hours": "25", "device": "cgm", "gaps": "", "microseconds": "", "threshold": "100", "minutes": "", "date": "display_time", "glucose": ""}}
notice that in --cli it show --no-raw, but in --json it says "no_raw": "False"
what do you get with --json ?
garykidd
@garykidd
Oct 01 2016 23:30
hmmmm...I wonder if this could be related to the issue I had awhile back where G5 wasnt working for my setup....despite election of no-raw....it didnt really take effect....can look back...
Scott Leibrand
@scottleibrand
Oct 01 2016 23:30
try just running openaps report show raw-cgm/raw-entries.json --json
garykidd
@garykidd
Oct 01 2016 23:31
pi@raspberrypi:~/APS401 $ openaps report show raw-cgm/raw-entries.json --json
{"type": "report", "name": "raw-cgm/raw-entries.json", "raw-cgm/raw-entries.json": {"count": "", "use": "oref0_glucose", "no_raw": "True", "reporter": "JSON", "seconds": "", "sensor": "", "hours": "25.0", "device": "cgm", "gaps": "", "microseconds": "", "threshold": "100", "minutes": "", "date": "display_time", "glucose": ""}}
Scott Leibrand
@scottleibrand
Oct 01 2016 23:32
and openaps report show raw-cgm/raw-entries.json --cli ?
garykidd
@garykidd
Oct 01 2016 23:33
well keep in mind...my last entry is after I changed the report---in case that wasn't clear
one sec
Scott Leibrand
@scottleibrand
Oct 01 2016 23:33
ya, that's what I'm trying to compare
garykidd
@garykidd
Oct 01 2016 23:34
openaps report show raw-cgm/raw-entries.json --cli
openaps use cgm oref0_glucose --hours "25.0" --no-raw --threshold "100"
Scott Leibrand
@scottleibrand
Oct 01 2016 23:34
huh. so --cli lies and shows --no-raw regardless?
scottleibrand @scottleibrand tests that
garykidd
@garykidd
Oct 01 2016 23:35
wait---are you saying in your or mine
Oh...in yours right?
Scott Leibrand
@scottleibrand
Oct 01 2016 23:35
I'm saying there's a bug in openaps toolkit that displays --no-raw on the --cli output regardless whether "no_raw" is "True" or "False"
which was confusing me, obviously
garykidd
@garykidd
Oct 01 2016 23:37
openaps/openaps#104
could that be related to the above?
different context but...
Scott Leibrand
@scottleibrand
Oct 01 2016 23:40
huh, weird. if I try to re-apply openaps report add raw-cgm/raw-entries.json JSON cgm oref0_glucose --hours 25 --threshold 100 it keeps "no_raw": "True". but if I do an openaps report remove raw-cgm/raw-entries.json first it removes it, and then openaps report show raw-cgm/raw-entries.json --cli looks fine
looks like the bug is only if you have "no_raw": "False", which you can only get from a json import (exported from before they changed the behavior)
garykidd
@garykidd
Oct 01 2016 23:41
the #104 above was referring to thispart...

```
Here is the breaking commit: 6fee307 (cc @bewest)

The fix seems straightforward enough (fields['no_raw'] was always being set to False), but please let me know if I'm missing something or there's more that I'm not seeing.

Also, if someone could give this a quick sanity test on a G4, it would be appreciated! Thanks!

Scott Leibrand
@scottleibrand
Oct 01 2016 23:41
so, I should probably remove "no_raw": "False" from the oref0-setup json import file, and also update the --cgm=G5 option to re-add the report with --no-raw
garykidd
@garykidd
Oct 01 2016 23:42
Based on these data points...that seems exactly right