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

19th
Jul 2016
Kim St. Dennis
@kimstdennis_twitter
Jul 19 2016 00:35
:fingerscrossed:
Paul Martin
@Spazholio
Jul 19 2016 00:47
Oh, right. Should have updated. So, I decided to use the Mac again. It worked this time. In something like 5-10 mins, rather than 30-45. So…yeah, whatever that was about...
Kim St. Dennis
@kimstdennis_twitter
Jul 19 2016 01:03
awesome
Julie
@julieraines
Jul 19 2016 01:04
Woohoo!
sarahspins
@sarahspins
Jul 19 2016 01:12
@Spazholio whoa, so it DOES take 10 minutes ;)
either that or your briefly warped to another universe where time moves faster than it does here
Paul Martin
@Spazholio
Jul 19 2016 01:19
Well, I was an extra in “Interstellar”...
John Dodds
@johners97
Jul 19 2016 01:23
Ok so i have realised how PuTTY works, this is where code is entered?
Martin Haeberli
@mhaeberli
Jul 19 2016 01:40
@johners97 - not sure what you mean by “where code is entered”? Assuming you have an RPi (or a working Edison), you should be able to connect to your <thingie> by saying “ssh pi@1.2.3.4" (or whatever the IP address is) and entering the password
I’m not a putty user, but maybe you have to say putty pi@1.2.3.4 instead
@johners97 anyway, once you log in to the RPi, then you can follow the steps to install various OpenAPS stuff on the RPi.
mariano42
@mariano42
Jul 19 2016 03:25
hi all... i am seeing something strange that I can't wrap my head around. I am still just running my loop from the command line, and the last two times i fired up the PI and ran the loop for the first time I encountered insufficient privileges errors and it was unable to do its thing.. .(stderr: 'error: insufficient permission for adding an object to repository database .git/objects
error: monitor/temp_basal.json: failed to insert into database
error: unable to index file monitor/temp_basal.json
fatal: updating files failed'
PREFLIGHT FAIL). This loop has ran plenty of times before, so I know the loop is fine. I then run the same command using sudo and it runs fine and does what I expect it to do. After that I can run the loop normally without error. That seems odd to me, but granted I am still learning linux. Anyone have any ideas?
mariano42
@mariano42
Jul 19 2016 03:36
or is that normal and expected?
Scott Leibrand
@scottleibrand
Jul 19 2016 04:00
That is actually cause by running openaps commands with sudo.
That creates files in .git owned by root that the normal user account can't edit.
I believe there is a section in the docs on how to use chown to fix it.
mariano42
@mariano42
Jul 19 2016 04:01
@scottleibrand oh.. ok, I will see if i can find that.
@scottleibrand thanks, I found it.
Scott Leibrand
@scottleibrand
Jul 19 2016 04:04
:+1:
Martin Haeberli
@mhaeberli
Jul 19 2016 04:17

still seeking advice

nightscout/recent-missing-entries.json

report fails … -

AttributeError: 'NoneType' object has no attribute 'read'
No JSON object could be decoded
Mitchell Slep
@moomoobloo
Jul 19 2016 04:41
@mhaeberli me too!
Martin Haeberli
@mhaeberli
Jul 19 2016 04:41
@moomoobloo :+1:

also:

openaps use ns shell upload treatments.json recently/combined-treatments.json

fails with

ValueError: No JSON object could be decoded
Dragan Maksimovic
@dm61
Jul 19 2016 04:46
@mhaeberli , @moomoobloo , do you have monitor/glucose.json and where is it coming from? (attached Dexcom CGM, or NS, or?)
Martin Haeberli
@mhaeberli
Jul 19 2016 04:46
@dm61 in my case, yes, from NS
head monitor/glucose.json
  {
    "direction": "Flat", 
    "trend": 4, 
    "dateString": "2016-07-19T04:38:38+00:00", 
    "sgv": 252, 
    "date": 1468903118000, 
    "device": "share2", 
    "_id": "578daf3693a430e19dee0f72", 
    "type": "sgv"
  }, 
  {
    "direction": "Flat", 
    "trend": 4, 
    "dateString": "2016-07-19T04:33:38+00:00", 
    "sgv": 252, 
    "date": 1468902818000, 
    "device": "share2", 
    "_id": "578dae0a93a430e19dee0f50", 
    "type": "sgv"
  },
Dragan Maksimovic
@dm61
Jul 19 2016 04:47
then, nightscout/recent-missing-entries.json should not be invoked. That report is supposed to fill in missing entries on NS from an attached Dex received. Since your glucose is from NS alerady, there is nothing to fill in
Martin Haeberli
@mhaeberli
Jul 19 2016 04:48
@dm61 ok, thx
@dm61 - it’s being invoked by “openaps report-nightscout” ...
Dragan Maksimovic
@dm61
Jul 19 2016 04:50
I think that report is constructed assuming you have Dex receiver attached.
I have two versions of report-nightscout: one when Dex received is attached, and another when I pull CGM data from NS. In the second version, I do not invoke that nightscout/recent-missing-entries.json report
sorry, I meant that alias (not report) is constructed assuming you have Dex received attached
Martin Haeberli
@mhaeberli
Jul 19 2016 04:59
@dm61 one more thing? advice on how to define report nightscout/openaps-status.json
Dragan Maksimovic
@dm61
Jul 19 2016 05:00
that one is complicated... give me a sec
Martin Haeberli
@mhaeberli
Jul 19 2016 05:02
k
Dragan Maksimovic
@dm61
Jul 19 2016 05:06
here is how I would do it manually as use: openaps use ns-status shell monitor/clock.json oref0-monitor/iob.json oref0-predict/oref0.json oref0-enacted/enacted-temp-basal.json monitor/battery.json monitor/reservoir.json monitor/status.json, you may try that from command line first, than if it works create nightscout/openaps-status.json
Martin Haeberli
@mhaeberli
Jul 19 2016 05:07
thx
usage: openaps-use [-h] [--format {text,json,base,stdout}] [--output OUTPUT]
                   [--version]
                   device ...
openaps-use: error: argument device: invalid choice: 'ns-status' (choose from u'calculate-iob', u'cgm', u'detect-sensitivity', u'determine-basal', u'get-profile', u'ns', u'oref0', u'pump', u'tz', u'units')
Dragan Maksimovic
@dm61
Jul 19 2016 05:10
yes, I recall I had to add ns-status device; later on I think I've seen someone say that ns device can do all, so I am not sure if ns-status is really required or not. I'll find how I created that ns-status device in a moment and will let you know
Martin Haeberli
@mhaeberli
Jul 19 2016 05:10
k thx
Dragan Maksimovic
@dm61
Jul 19 2016 05:12
here it is: oref0 device-helper ns-status 'ns-status $*' | openaps import
Martin Haeberli
@mhaeberli
Jul 19 2016 05:14
Looks promising! 5 minutes - just lost laptop power
Mitchell Slep
@moomoobloo
Jul 19 2016 05:15
are there any docs about Pebble integration?
Dragan Maksimovic
@dm61
Jul 19 2016 05:16
I have not done anything with Pebble from openaps, so I do not really know. I'd look at urchin repo - my understanding is it should be fairly straightforward
Mitchell Slep
@moomoobloo
Jul 19 2016 05:19
@dm61 thanks, didn't know about that!
Dragan Maksimovic
@dm61
Jul 19 2016 05:20
you are welcome. I am yet to give it a try myself
Martin Haeberli
@mhaeberli
Jul 19 2016 05:23

@dm61 brilliant!

openaps report add nightscout/openaps-status.json JSON ns-status shell monitor/clock.json oref0-monitor/iob.json oref0-predict/oref0.json oref0-enacted/enacted-temp-basal.json monitor/battery.json monitor/reservoir.json monitor/status.json

worked!

Dragan Maksimovic
@dm61
Jul 19 2016 05:24
:+1:
Mitchell Slep
@moomoobloo
Jul 19 2016 06:07
does nightscout have any alarms for OpenAPS, like if loop is down or manual correction bolus is needed?
Jason Calabrese
@jasoncalabrese
Jul 19 2016 06:08
it has alarms for when openaps isn't looping, when the pump data is stale, pump battery, and pump reservoir
Mitchell Slep
@moomoobloo
Jul 19 2016 06:09
how are those configured?
Jason Calabrese
@jasoncalabrese
Jul 19 2016 06:09
the bwp plugin uses the iob from openaps so it does a pretty good job auto snoozing high alarms when theres enough net iob
lots of options
Mitchell Slep
@moomoobloo
Jul 19 2016 06:09
any docs?
all the NS settings are defined in the readme
Mitchell Slep
@moomoobloo
Jul 19 2016 06:10
thanks!
Jason Calabrese
@jasoncalabrese
Jul 19 2016 06:10
it takes some trial and error to get it tuned to fit your needs
alarms work though the web app, pushover, and iftt maker
John Dodds
@johners97
Jul 19 2016 11:41
'code'
'''
var code = "formatted"
'''
Sorry ^ testing
*item
Jim is CrushingT1D
@CrushingT1D
Jul 19 2016 12:29
Is anyone using the dexcom g5 ios app downloaded from outside the us app store? I just downloaded the version from the australian app store and noticed the signup button does not work. Am a little nervous to use my us account in case some sort or problem comes up
But was wanting to test the mmol version of the app
Paul Martin
@Spazholio
Jul 19 2016 12:32
@cjo20 What format does my EdisonVoltage output need to be in in order for it to show properly in NS? After updating last night (and changing the variable from 2 to json) it's not showing up in my site anymore.
Paul Martin
@Spazholio
Jul 19 2016 12:39
Or maybe I should direct that to @jasoncalabrese? I think they were both working on that...
Jim is CrushingT1D
@CrushingT1D
Jul 19 2016 12:42
If anyone replies please @ mention me, hard to keep up :)
Julie
@julieraines
Jul 19 2016 13:07

I’m having trouble with max_iob.json (and I created preferences.json with max_iob 2 because 2 units seemed reasonable)
openaps report invoke settings/profile.json
get-profile://text/shell/settings/profile.json

module.js:340
throw err;
^
Error: Cannot find module '/home/pi/sasha/max_iob.json'
at Function.Module._resolveFilename (module.js:338:15)
at Function.Module._load (module.js:280:25)
at Module.require (module.js:364:17)
at require (module.js:380:17)
at Object.<anonymous> (/usr/local/lib/node_modules/oref0/bin/oref0-get-profile.js:60:23)
at Module._compile (module.js:456:26)
at Object.Module._extensions..js (module.js:474:10)
at Module.load (module.js:356:32)
at Function.Module._load (module.js:312:12)
at Function.Module.runMain (module.js:497:10)
reporting settings/profile.json

what am I doing wrong?

John Dodds
@johners97
Jul 19 2016 13:20
This message was deleted
Accidentally delted my own message...
Dragan Maksimovic
@dm61
Jul 19 2016 13:58
@julieraines looks like you created preferences.json, but you are invoking settings/profile.json that seems to be looking for max_iob.json. Maybe just rename preferences.json to max_iob.json, or modify settings/profile.json to look for preferences.json
Paul Martin
@Spazholio
Jul 19 2016 13:58
(which, IIRC, contains the max_iob value)
John Dodds
@johners97
Jul 19 2016 14:11
Is it wise to assign a static ip address to the raspberry pi?
Julie
@julieraines
Jul 19 2016 14:48
Ok - that sounds good! Thanks!
so the command is voltage json batteryVoltage battery
Chris Oattes
@cjo20
Jul 19 2016 15:46
Very clever
Paul Martin
@Spazholio
Jul 19 2016 16:23
@jasoncalabrese What you have in your uoload-status.json is what I have, I believe. It's been working great, up until last night when I updated the code. So this is 100% on me. =)
Jason Calabrese
@jasoncalabrese
Jul 19 2016 16:29
what code did you update?
Paul Martin
@Spazholio
Jul 19 2016 16:29
EdisonVoltage
Jason Calabrese
@jasoncalabrese
Jul 19 2016 16:30
does voltage json batteryVoltage battery work?
Paul Martin
@Spazholio
Jul 19 2016 16:30
Unsure. I made the changes that you recommended, and then realized that my rig hasn't run properly for the last 30m. Fixing that now, and I'll report back.
Jason Calabrese
@jasoncalabrese
Jul 19 2016 16:30
I did update oref0 so that it will handle a number or object for uploader battery
Paul Martin
@Spazholio
Jul 19 2016 16:33
Networking went all screwy. Rebooting.
Dave Ewall
@daveewall
Jul 19 2016 16:34
@jasoncalabrese Does this EdisonVoltage program require anything different to be hardwired to the board? I remember reading a while back about @cjo20 working on something... is that the same program? I'd love to see my battery voltage on the NS status too.
Paul Martin
@Spazholio
Jul 19 2016 16:35
I swear, I don't know why, but I periodically need to clean out my /var/lib/dhcp dir of all leases, otherwise the device just will NOT get online.
@jasoncalabrese No, it doesn't appear that it works. Here is the contents of my rig-battery.json: {"batteryVoltage":3807, "battery":63}.
Jason Calabrese
@jasoncalabrese
Jul 19 2016 16:39
maybe your oref0 isn't current with dev
Paul Martin
@Spazholio
Jul 19 2016 16:40
Incidentally, when I'm using screen everything's fine until I go to use vim. Once I do that, I'm constrained to what seems to be an 80 column layout. If anyone has a fix for that, I'd be suuuuper appreciative.
I don't know what version oref0 is on. I could give it an update though, just to be on the safe side.
Dave Ewall
@daveewall
Jul 19 2016 16:42
@jasoncalabrese Ignore me, I RTFM'd.
Paul Martin
@Spazholio
Jul 19 2016 16:45
HAH! Upgrading oref0 seems to have done the trick. Gonna let the loop run a few more times before I'm sold though. Did all the upgrades to oref0 include the json -e to json -E thing I've been seeing mentioned?
Hmmm...I don't think the BWP takes temp targets into account.
Paul Martin
@Spazholio
Jul 19 2016 16:52
Nor does it take current temp basal rates into account either. That's odd.
Paul Martin
@Spazholio
Jul 19 2016 16:58
@jasoncalabrese Ever since upgrading oref0, I'm getting the following. I haven't upgraded mmeowlink in a while, so this could all be answered by doing so, but I wanted to see if you've ever seen this before:
reporting enact/suggested.json
{"temp":"absolute","bg":131,"tick":"+3","eventualBG":148,"snoozeBG":148,"predBGs":{"IOB":[131,134,136,139,141,143,145,146,147,148,149,150,150,150,150,150,150,150,149,149,148,148,148,147,147,146,146,146,145,145,144,144,144,143,143,142,142,141,141,140,140,139,139,138,138,137,137,136]},"COB":0,"IOB":-0.111,"reason":"COB: 0, Dev: 15, BGI: 0.56, ISF: 25.3, Target: 80; Eventual BG 148 >= 80, temp 3.35<6.8U/hr","duration":30,"rate":6.8}
enact/enacted.json  raised  byte must be in range(0, 256)
pump://JSON/set_temp_basal/enact/enacted.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 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/medtronic.py", line 431, in main
    results = self.upload_program(program)
  File "/usr/local/lib/python2.7/dist-packages/openaps/vendors/medtronic.py", line 524, in upload_program
    return self.pump.model.set_temp_basal(**program)
  File "/usr/local/lib/python2.7/dist-packages/decocare/models/__init__.py", line 222, in set_temp_basal
    result = self._set_temp_basal(**basals)
  File "/usr/local/lib/python2.7/dist-packages/decocare/models/__init__.py", line 25, in __call__
    self.response = inst.session.query(self.msg, **kwds)
  File "/usr/local/lib/python2.7/dist-packages/decocare/session.py", line 101, in query
    self.execute(command)
  File "/home/edison/src/mmeowlink-source/mmeowlink/handlers/stick.py", line 225, in execute
    return sender(command)
  File "/home/edison/src/mmeowlink-source/mmeowlink/handlers/stick.py", line 140, in __call__
    self.upload()
  File "/home/edison/src/mmeowlink-source/mmeowlink/handlers/stick.py", line 127, in upload
    self.send_params( )
  File "/home/edison/src/mmeowlink-source/mmeowlink/handlers/stick.py", line 28, in send_params
    payload = bytearray([len(params)]) + bytearray(params)
ValueError: byte must be in range(0, 256)
No action required
Jason Calabrese
@jasoncalabrese
Jul 19 2016 17:04
I get some errors with interference, but don't recognize byte must be in range(0, 256)
bwp hasn't been updated to use the temp basals or temp targets
Paul Martin
@Spazholio
Jul 19 2016 17:05
It's now happening with every single iteration of the loop. It's not pushing any new temps to the pump either. =/
re: BMP - gotcha
Jason Calabrese
@jasoncalabrese
Jul 19 2016 17:05
it should
it should use COB too
but I'm not sure it's still bwp then
Paul Martin
@Spazholio
Jul 19 2016 17:06
heh
So...any ideas on what I can do about this not-setting-temp-basals thing anymore? It's got me a little concerned, and it only started happening after the upgrade.
Jason Calabrese
@jasoncalabrese
Jul 19 2016 17:09
I don't see how that would be caused by updating oref0, you updated other things too?
Paul Martin
@Spazholio
Jul 19 2016 17:09
Haven't touched a thing.
This is my production rig, so I'm pretty hands-off once I get it running the way I like.
Oh, after it started having the issue, I upgraded mmeowlink, but that didn't do anything other than add something in there about RileyLink.
And...I didn't think that oref0 had anything to do with setting the temp basals, so I was somewhat perplexed about that as well.
Jason Calabrese
@jasoncalabrese
Jul 19 2016 17:11
maybe someone in mmeowlink has seen it
how did you update mmeowlink? maybe some old *.pyc files around
Paul Martin
@Spazholio
Jul 19 2016 17:12
Ok, here's a weird thing - it just magically started working. Like, as you typed that sentence. Took about 30 mins to work again though.
I did a git pull and a sudo pip install -e .
Seems like anytime something goes weird, it takes several iterations of the loop for whatever it was to "shake out". Like when I get the "BG data is too old" error and I correct it - the next loop run never works. It always takes 3-4 more before it fixes itself. Very odd.
I'm going to watch the loop a bit and see what it does for the next few runs. If I keep having the issue, I'll ask over in mmeowlink.
Jason Calabrese
@jasoncalabrese
Jul 19 2016 17:15
I don't do this often, but I've seen @oskarpearson recommend doing this
# IN THE mmeowlink source directory:
find . -name '*.pyc' -print0 | xargs -t -r -0 rm
I'd only do it if you keep having problems
Paul Martin
@Spazholio
Jul 19 2016 17:16
Well, there ARE a lot of them there...
Does it harm anything if they're removed? I mean, if they're not there, they just have to be re-created a runtime, no? I'm not all that familiar with Python, so...
Jason Calabrese
@jasoncalabrese
Jul 19 2016 17:17
I'd think pip would clean them up, but not sure when they get created
Martin Haeberli
@mhaeberli
Jul 19 2016 17:23
our initial intended rig is an RPi3 with a large external battery pack (about the same size as the RPi3) - a 9600mAh Tecknet battery. We’re kind of using the battery as if it were a Uninterruptible Power Supply for the time being. Question being - is there any (easy or not) way to monitor the battery level so as to let us shut down the rig cleanly if need be?
Eric
@ecc1
Jul 19 2016 17:33
Or just find . -name '*.pyc' -delete
Jason Calabrese
@jasoncalabrese
Jul 19 2016 17:34
:+1:
Paul Martin
@Spazholio
Jul 19 2016 17:36
I have a theory that find + xargs + grep is Turing-complete. Can't prove it, but it feels right.
Jason Wittmer
@jbwittmer
Jul 19 2016 17:38
Question about "Meal Assist" - I went ahead and enabled Meal Assist. Loop seems to be working. Took me way longer than it should have because I can't type and/or proof-read. Anyway, working now. Question: "Meal Assist" continues to report as off. Am I correct in assuming that it will only report as "On" if it calculates that carbs won't be covered by boluses or high temps? Thanks.
Dana Lewis
@danamlewis
Jul 19 2016 17:52
@jbwittmer you must enter carbs for meal assist. also, we recommend moving on and turning on advanced meal assist, which is even better (and includes auto-sensitivity)
@mhaeberli I recommend also having a spare SD card loaded and ready to go, in case something does break/fry on the fly due to battery or otherwise, it’s great to just pop in the next SD card and you can repair the other one later.
Martin Haeberli
@mhaeberli
Jul 19 2016 17:58
@danamlewis thx
Dana Lewis
@danamlewis
Jul 19 2016 18:03
(clean shutdowns don’t happen for me every time)
Bog Dan
@bogdangorescu
Jul 19 2016 18:39
Anyone else having this error, when trying to read the pump? It seems like this is something about not having enought rights to read the serial port:
openaps use pump model
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/pkg_resources/__init__.py", line 719, in run_script
  File "build/bdist.linux-armv7l/egg/pkg_resources/__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 44, in __call__
    self.before_main(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/openaps/vendors/medtronic.py", line 61, in before_main
    self.setup_medtronic( )
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/openaps/vendors/medtronic.py", line 159, in setup_medtronic
    self.uart = stick.Stick(link.Link(self.scanner( )))
  File "/usr/local/lib/python2.7/dist-packages/decocare-0.0.29-py2.7.egg/decocare/link.py", line 24, in __init__
    self.open( port, dsrdtr=True, rtscts=True )
  File "/usr/local/lib/python2.7/dist-packages/decocare-0.0.29-py2.7.egg/decocare/link.py", line 35, in open
    self.serial = serial.Serial( self.port, **kwds )
  File "/usr/lib/python2.7/dist-packages/serial/serialutil.py", line 261, in __init__
    self.open()
  File "/usr/lib/python2.7/dist-packages/serial/serialposix.py", line 278, in open
    raise SerialException("could not open port %s: %s" % (self._port, msg))
serial.serialutil.SerialException: could not open port /dev/serial/by-id/usb-0a21_8001-if00-port0: 
[Errno 13] Permission denied: '/dev/serial/by-id/usb-0a21_8001-if00-port0'
Chris Oattes
@cjo20
Jul 19 2016 18:45
You need to add yourself to a group. Possibly dialup or dialout v
-v
Bog Dan
@bogdangorescu
Jul 19 2016 18:45
OK...thanks
Martin Haeberli
@mhaeberli
Jul 19 2016 18:47
any views on the relative range of slice of radio vs ti stick vs carelink … ?
Dana Lewis
@danamlewis
Jul 19 2016 18:47
TI > slice of radio > carelink
Chris Oattes
@cjo20
Jul 19 2016 18:48
@bogdangorescu
Martin Haeberli
@mhaeberli
Jul 19 2016 18:48
@danamlewis thx
Bog Dan
@bogdangorescu
Jul 19 2016 18:49
The group is dialout. Dialin doesn't exist...
let's see if this works.
Thanks
Chris Oattes
@cjo20
Jul 19 2016 18:49
adduser edison dialout
Bog Dan
@bogdangorescu
Jul 19 2016 18:50
nope...adduser adds a new user. I need to modify my existing user.
sudo usermod -a -G dialout userName
sarahspins
@sarahspins
Jul 19 2016 18:50
@danamlewis does RL have about the same range as the TI stick?
Bog Dan
@bogdangorescu
Jul 19 2016 18:50
then userName must logout and login again
Dana Lewis
@danamlewis
Jul 19 2016 18:51
I’d say TI > Rileylink ~= slide of radio > Carelink
Bog Dan
@bogdangorescu
Jul 19 2016 18:51
@cjo20 ---and no, it doesn't work....
Dana Lewis
@danamlewis
Jul 19 2016 18:51
of course, Rileylink is a little more convenient than slice of radio, so although I found range to be similar in general, usability is easier on a RL than slice.
Chris Oattes
@cjo20
Jul 19 2016 18:53
@bogdangorescu ls -l /serial/path/to/device
Jason Wittmer
@jbwittmer
Jul 19 2016 18:54
@danamlewis OK I'll look at AMA soon. If I'm entering carbs in pump for boluses, that's good enough for Meal Assist (or AMA for that matter?)
Dana Lewis
@danamlewis
Jul 19 2016 18:54
@jbwittmer yep, entering into pump is fine, you just need to do it consistently (i.e. always enter carbs when there are carbs)
Bog Dan
@bogdangorescu
Jul 19 2016 18:55
@cjo20 --- > openaps$ ls -l /dev/serial/by-id/usb-0a21_8001-if00-port0 openaps$ ../../ttyUSB0
Chris Oattes
@cjo20
Jul 19 2016 18:57
What about if you do ls -l /dev/serial/by-id
Martin Haeberli
@mhaeberli
Jul 19 2016 18:58
@danamlewis so how do you use RileyLink in that context - do you flow data pump -> RileyLink -> iOS app-> Nightscout -> OpenAPS ? or pump -> RileyLink -> bluetooth -> OpenAPS ?
Bog Dan
@bogdangorescu
Jul 19 2016 18:58
openaps@raspberrypi:~/Documents/stefan $ ls -l /dev/serial/by-id
total 0
lrwxrwxrwx 1 root root 13 iul 17 22:17 usb-0a21_8001-if00-port0 -> ../../ttyUSB0
lrwxrwxrwx 1 root root 13 iul 17 22:17 usb-DexCom_DexCom_Gen4_USB_Serial-if00 -> ../../ttyACM0
openaps@raspberrypi:~/Documents/stefan $
Dana Lewis
@danamlewis
Jul 19 2016 19:00
@mhaeberli I had an Edison/rileylink as my main mobile rig for a while, works same as TI to pull pump data into local pump history file (used locally), and which is also uploaded to NS.
Martin Haeberli
@mhaeberli
Jul 19 2016 19:00
@danamlewis as to AMA - suppose you add (bolus-free) carbs, for a low, for example - do you still enter in pump and have it calculate “0” insulin - for completeness ?
@danamlewis thx - you say “had” - what is your current preferred mobile rig?
@danamlewis and on the Edison, how did SGV / CGM data flow in? xDrip electronics ? or Share -> NS (assuming you are using Dexcom)
Bog Dan
@bogdangorescu
Jul 19 2016 19:01
@cjo20 --- I have to add sudo in front of all the commands I issue to devices connected to usb ports.
Dana Lewis
@danamlewis
Jul 19 2016 19:02
if it’s <5 grams of carbs, I don’t put it in (personal choice, I’m lazy!), but if it’s 10 or more, I put it in. You can also set it up to take the carb entry via NS if you prefer. some of the pumps require a .1 bolus to enter the carbs, some will still take the carb entry if you back out without giving the bolus, I can’t remember which version does which. but definitely worth putting big carb corrections in, then AMA does a good job jumping in sooner with any rebound BGs.
@mhaeberli my other main mobile rig has been Edison/TI, which has gotten flaky recently so might be switching back to Edison/rileylink again to compare.
I have openxshareble and I also have data coming in from Share/NS when I’m hotspotted. (I carry two receivers).
Martin Haeberli
@mhaeberli
Jul 19 2016 19:03
@danamlewis - so two Dexcom G4 platinum receivers … ? we are staying away from G5 here for the time being
Dana Lewis
@danamlewis
Jul 19 2016 19:04
yes, two g4 shares.
Martin Haeberli
@mhaeberli
Jul 19 2016 19:09
@danamlewis thx 4 all the info - btw, are you using any kind of 3d printed or other case for the edison, or shrinkwrap or something? Also, what you are doing sounds a lot like what I’d like to work with Eden on doing … As you can probably tell, we’re still not across the “enact” bridge, but I’m trying to gather as much data in real time as possible, flow it in to Nightscout, begin to track recommendations (so track, predict, recommend) then try to get comfortable with some looping …
Chris Oattes
@cjo20
Jul 19 2016 19:10
@bogdangorescu relog
sarahspins
@sarahspins
Jul 19 2016 19:10
@mhaeberli it looks like Dana's is stuffed in her dexcom case (@danamlewis is that a Tallygear, or something else?)
Dana Lewis
@danamlewis
Jul 19 2016 19:11
A lot of folks have nice 3d printed cases, I actually toss my Edison rigs in a soft Dexcom case that I use for running, because it’s got a clear plastic front and a loop so I can clip it to my pocket.
@sarahspins yep, it’s a tallygear dexcom case. I’m actually going to reach out and ask her to create a custom one, same idea but an inch or so shorter in length, doesn’t need to be that long for an Edison rig
Martin Haeberli
@mhaeberli
Jul 19 2016 19:12
@danamlewis cool thx
Dana Lewis
@danamlewis
Jul 19 2016 19:12
also worth noting, there’s an awesome designer who’s working on some neat 3d printed ideas to go with some of the custom Edison boards being created, so more to come in a few weeks re: those
Martin Haeberli
@mhaeberli
Jul 19 2016 19:12
:+1:
sarahspins
@sarahspins
Jul 19 2016 19:12
very very cool :+1:
Paul Martin
@Spazholio
Jul 19 2016 19:15
I can't recommend @eluhrs (where has he been, anyway?) 3D printed case enough: https://www.thingiverse.com/thing:1525618
Dana Lewis
@danamlewis
Jul 19 2016 19:17
:+1:
someone want to PR some links to cases in the usability page in phase 6 of the docs? :)
David Cintron
@loudestnoise
Jul 19 2016 19:19
woot woot
Here comes :100:
Dana Lewis
@danamlewis
Jul 19 2016 19:21
As a friendly reminder, these counts go up in the order based on when people submit the form (so some people submit two weeks after they start looping, some people do it when they wake up after the third night, etc.), so don’t get too jazzed about which particular number you think you are, because it’s “wrong” in the order of who started looping when! (And of course, don’t rush it because you think you’ll be a certain number, because you probably won’t be.) Be jazzed instead about the fact that you have, or are approaching having, a closed loop AP that works!! :)
sarahspins
@sarahspins
Jul 19 2016 19:28
I want to be #100 ;)
but really I'm just thrilled to be doing this :+1:
Dana Lewis
@danamlewis
Jul 19 2016 19:29
:)
Hannah Crabtree
@lollydaggle
Jul 19 2016 19:31
Also thrilled to have a (growing) community of people who understand - both diabetes and the urge to creatively solve problems. I haven't really come out about this project yet (someone at work will eventually notice the pouch pancreas) - what reactions have people gotten from friends, family and colleagues?
My friends thought I was "more diabetic" after getting a CGM. Pretty sure this takes it to a whole new level.
Dana Lewis
@danamlewis
Jul 19 2016 19:33
:)
sarahspins
@sarahspins
Jul 19 2016 19:33
@lollydaggle now they'll just think you're a nerdy diabetic :+1:
Funny aside... I volunteered at a Maker Faire info booth a few months back at SXSW and I got more comments on my Dexcom than anything else we had on display :)
I probably would have been doing OpenAPS back then if it wasn't for finding a compatible pump :+1:
savelik
@savelik
Jul 19 2016 19:39
Hello! I'm starting the buildout anyday now. Dexcom G5 + minimed 722 pump. So I decided to start with raspberry, took raspberry 3 to go BLE for internet. I see that everyone is switching to Edisons - is that a logical 2nd step to try to minimize the equipment? This is for my 9-year old daughter. She has been on MDI since diagnosis and FreeStyle Libre with a nightscan module that feeds into nightscout. I want to give OpenAPS a good try to be fully ready for the beginning of the school year.. Any suggestions/ideas/...? I read all the OpenAPS docs and after the 2nd reading they seem very clear.
Dana Lewis
@danamlewis
Jul 19 2016 19:40
welcome @savelik!
Yes, Edison is often a minimizer step, but I’d recommend getting the Pi rig up first, before buildling the second rig (Edison or otherwise)
John Dodds
@johners97
Jul 19 2016 19:40
have an idea. could the pi/edison be programmed to respond to low blood glucose by not only the loop but also by playing classical music to speakers via bluetooth to help stay calm?
Just a thought...
Dana Lewis
@danamlewis
Jul 19 2016 19:40
@johners97 you might be able to do that, if you wanted to :) using an IFTT integration
John Dodds
@johners97
Jul 19 2016 19:41
@danamlewis Would be better than having my pump bleep me it makes me freak out when im low! haha
savelik
@savelik
Jul 19 2016 19:41
thank you :) btw I love FreeStyle Libre because my daugher swims a lot and we can check her blood sugar with her just lifting her arm out of the water... Dexcom is a pain in the ass to re-synch in the same scenario and fingerpricking is not precise because of the wet fingers...
John Dodds
@johners97
Jul 19 2016 19:41
although it would still beep
Dana Lewis
@danamlewis
Jul 19 2016 19:42
@johners97 have you tried putting your pump to vibrate mode instead?
savelik
@savelik
Jul 19 2016 19:42

@danamlewis

Yes, Edison is often a minimizer step, but I’d recommend getting the Pi rig up first, before buildling the second rig (Edison or otherwise)

Got it, thanks a lot!

Dana Lewis
@danamlewis
Jul 19 2016 19:43
:+1: and then Edison can be school rig and Pi can be bedside/home rig, etc.
savelik
@savelik
Jul 19 2016 19:43
I want a 24/7 rig that I just set to recharge when she is back home with a backup rig just in case :)
Martin Haeberli
@mhaeberli
Jul 19 2016 19:43
@danamlewis in the multi-rig scenario, I assume you can flow all data through the same nightscout instance if you like ?
Dana Lewis
@danamlewis
Jul 19 2016 19:43
yes. backups are very, very good :)
@mhaeberli exactly!
John Dodds
@johners97
Jul 19 2016 19:44
@danamlewis yeah, imagine programming the pi or edison to shake dog discuits in some device to attract a cute dog to run around the room whilst low rofl
savelik
@savelik
Jul 19 2016 19:44
yes, I bought 2 pumps of the same model (722) just to have a backup
Dana Lewis
@danamlewis
Jul 19 2016 19:44
nice @savelik :)
sarahspins
@sarahspins
Jul 19 2016 19:45
@savelik it took me 6+ months to find my 723... I'm still hunting for a backup
Martin Haeberli
@mhaeberli
Jul 19 2016 19:45
@sarahspins :+1:
savelik
@savelik
Jul 19 2016 19:45
yep, difficult task, that is why I saw 2 in good shape for sale - I went for it :)
Last question - G5 - I am thinking to connect it via USB to raspberry to grab the data (it seems that only G4 has raw, and G5 doesn't - am I correct here?) and also I'm thinking to setup a dexcom app on the phone that will be a part of the rig to read and upload data as well. Is that a normal flow? :)
so in case the receiver for some reason breaks down or faulters - there's the data that I'm uploading with G5->iphone->Dexcom Cloud and I can turn bridge to nightscout and DL to the rig...
Dana Lewis
@danamlewis
Jul 19 2016 19:48
Yup. You can have data come to the rig from multiple places, and it’s good to design an offline way (i.e. plugging in receiver to rig) as well as online (i.e. share to Nightscout)
exactly
or vice versa
savelik
@savelik
Jul 19 2016 19:49
then I just switch the rig setup this way or that way... yep
Dana Lewis
@danamlewis
Jul 19 2016 19:49
you won’t have to switch the rig setup ;)
you can design it to check for one or the other, in priority order, so if you’re not hotspotted and not getting cloud data, it checks for the receiver plugged in, or vice versa
savelik
@savelik
Jul 19 2016 19:49
how will the rig know which data to grab?
got it
Dana Lewis
@danamlewis
Jul 19 2016 19:50
:+1:
good questions!
savelik
@savelik
Jul 19 2016 19:50
thank you Dana!! :)
Paul Martin
@Spazholio
Jul 19 2016 19:50
I'd honestly love it if the reports could include conditional logic at some point. =)
Dana Lewis
@danamlewis
Jul 19 2016 19:50
you bet!
::coughs file an issue cough::
savelik
@savelik
Jul 19 2016 19:50
I will check back by the end of the week with hopefully a working rig ;)
Dana Lewis
@danamlewis
Jul 19 2016 19:50
(@spazholio ^)
:) @savelik or with more questions!
savelik
@savelik
Jul 19 2016 19:51
hopefully with a working rig lol :))))) I do not want to bother you guys too much, you are amazing!!!
Dana Lewis
@danamlewis
Jul 19 2016 19:51
that’s what this channel is here for!
savelik
@savelik
Jul 19 2016 19:52
awesome!!!! (h)
Paul Martin
@Spazholio
Jul 19 2016 19:52
@danamlewis Well, I don't know if anyone would even be interested in something like that, or if the way people are using it right now is sufficient. Besides, someone ::cough:: @bewest @scottleibrand ::cough:: would have to code in all new functionality. =)
Dana Lewis
@danamlewis
Jul 19 2016 19:52
yes, but if you don’t put in an issue to log the idea, it doesn’t go on The List for consideration
so, if you’re serious about it, file an issue, or if not, not :)
Paul Martin
@Spazholio
Jul 19 2016 19:53
Do you think that would be an oref0 issue? Or...which project? =)
Dana Lewis
@danamlewis
Jul 19 2016 19:54
stick it somewhere and if it doesn’t end up in the right place, you’ll find out ;)
sarahspins
@sarahspins
Jul 19 2016 19:55
ahh, the beauty of crowdsourcing :+1:
Dana Lewis
@danamlewis
Jul 19 2016 19:56
:+1:
Chris Oattes
@cjo20
Jul 19 2016 20:05
@Spazholio you can do conditionals in aliases
Paul Martin
@Spazholio
Jul 19 2016 20:05
How so? By invoking bash or something similar?
Chris Oattes
@cjo20
Jul 19 2016 20:06
Yes
Paul Martin
@Spazholio
Jul 19 2016 20:06
(I love how you tell me that the exact second I hit "SUBMIT" on the issue) =)
Chris Oattes
@cjo20
Jul 19 2016 20:07
Sorry, I was eating dinner and enjoying the view from the balcony
Paul Martin
@Spazholio
Jul 19 2016 20:07
So if I wanted to "if CGM is connected, then don't bother looking for NS data, else go to NS to get BG data" - how would you recommend approaching that?
(no worries, man)
Dana Lewis
@danamlewis
Jul 19 2016 20:08
@Spazholio it's actually good, because you can go back and log any info you learn in the issue, whether or not you decide to close it or keep the discussing going about best ways to facilitate.
That way someone with a similar idea or Q in future will more easily find it
Paul Martin
@Spazholio
Jul 19 2016 20:09
I'm not naming names, but I can almost guarantee you there are some people who see my name on stuff like this and think, "Oh FFS, what now?" =)
Dana Lewis
@danamlewis
Jul 19 2016 20:10
That's not how this works ;)
Chris Oattes
@cjo20
Jul 19 2016 20:10
(The view from the balcony: http://i.imgur.com/O3YWQCb.png )
sarahspins
@sarahspins
Jul 19 2016 20:10
@cjo20 :+1:
Chris Oattes
@cjo20
Jul 19 2016 20:11
@Spazholio I'd try and pull data from the pump, then check if the file you expect it to produce still exists
sarahspins
@sarahspins
Jul 19 2016 20:11
@Spazholio I think you earned whatever help you need now after helping me with my Edison :+1:
I really do mean that :)
Dragan Maksimovic
@dm61
Jul 19 2016 20:12
@cjo20 wow :+1:
Paul Martin
@Spazholio
Jul 19 2016 20:12
HAH! I know I can be a pain in the ass (in a mostly good way) so in order to ameliorate my guilt, I try to help out wherever I can. And in YOUR case....well, I refuse to let technology beat me. Especially when it should just freaking WORK. =)
sarahspins
@sarahspins
Jul 19 2016 20:13
I know, right? I'm not exactly an idiot with these things but I was just way past my own personal limit for "WTF is going wrong with this" :)
Chris Oattes
@cjo20
Jul 19 2016 20:13
@Spazholio if you look at the link I sent in PM there are a few cases where I check for a file existing
sarahspins
@sarahspins
Jul 19 2016 20:13
and now it's just sitting on my desk, I'm still mad at it :P
Paul Martin
@Spazholio
Jul 19 2016 20:13
Just wait until you try to solder the damned thing.
sarahspins
@sarahspins
Jul 19 2016 20:14
nah, that will be a piece 'o cake :)
I got my 10-pin cable to tear apart in the mail today.. I might still tackle it
Jason Wittmer
@jbwittmer
Jul 19 2016 20:14
@cjo20 If you could send me the example for where you check for file existing I'd appreciate it. I've been wondering how to do this as well.
Scott Leibrand
@scottleibrand
Jul 19 2016 20:14
@Spazholio there are several ways to do that. https://github.com/scottleibrand/openaps-sh/blob/refresh-loops/setup.sh (and some older branches) show a couple approaches, and https://github.com/openaps/docs/wiki/CGM-Loop-with-openxshareble-scratchpad outlines some similar tools
Paul Martin
@Spazholio
Jul 19 2016 20:15
Be careful with those 10-pin cables. Those cables are so THIN that once you get 'em soldered, you really don't want to touch them anymore.
sarahspins
@sarahspins
Jul 19 2016 20:15
@Spazholio I've done it before :+1:
thats where everyone's favorite strain relief comes into play (aka hot glue)
Paul Martin
@Spazholio
Jul 19 2016 20:16
No kidding? Well...neat. I still want to re-do mine with some better cables, but that means buying a new breakout board. =)
Scott Leibrand
@scottleibrand
Jul 19 2016 20:16
@Spazholio but neither of those approaches is really as simple as you're looking for in openaps/oref0#173
Paul Martin
@Spazholio
Jul 19 2016 20:17
Yeah, I've done some of the "run the report, check to see if the file exists, then..." but that's more bash, and I know that's not really all too fun for some.
sarahspins
@sarahspins
Jul 19 2016 20:18
I've used flat cables to connect relays to controllers in multiple Arduino projects.. and all of those basically involve soldering those stupid 26g wires since crimp connectors aren't super reliable depending on the application
Chris Oattes
@cjo20
Jul 19 2016 20:19
check-success = ! bash -c "if [ ! -s enact/suggested.json ]; then touch retune; exit 1; fi;" that's how I check whether the loop completed, and then queue up an mmtune if it didn't
Scott Leibrand
@scottleibrand
Jul 19 2016 20:20
I am wondering if maybe we shouldn't put together some templates/scripts to set everything up to do a full cgm and read-only ns-uploader loop (of both cgm and pump data), so people can go ahead and get a consistent known-working setup with a single command (maybe after they've done a bunch of self-built aliases and have an understanding of how things fit together, so they can troubleshoot later)
Chris Oattes
@cjo20
Jul 19 2016 20:21
What would be interesting is an oref0 command that, once you've provided it with a sensible pump history, glucose history and model provided you with the set of aliases you need to get that consistent setup
All of the steps to get that far are (separately) documented, so if someone gets that far on their own, they're probably OK, if they can't work it out then they'll come here for help
Scott Leibrand
@scottleibrand
Jul 19 2016 20:23
yeah. that was kinda my original idea with setup.sh, but it is now somewhat inconsistent with a lot of the new stuff in the docs (templates, etc.)
Kim St. Dennis
@kimstdennis_twitter
Jul 19 2016 20:41
upkeep of a script like that (unless more abstracted) can be really tough.
@Spazholio are you thinking that working with a function within oref0 is easier for people that working with bash?
sarahspins
@sarahspins
Jul 19 2016 20:43
@danamlewis is using RL with Loop technically part of OpenAPS, or is it considered something else?
Dana Lewis
@danamlewis
Jul 19 2016 20:43
let me ask you a question before I answer that: why are you asking that? :)
sarahspins
@sarahspins
Jul 19 2016 20:44
curiosity - I am not real clear on where that "fits" in the scheme of things
Scott Leibrand
@scottleibrand
Jul 19 2016 20:45
@kimstdennis_twitter everything I do to my loop I do in setup.sh first and then deploy to my rigs, so I think I could manage that part. will just be a question of managing deployment of new versions and keeping us down to a single master and a single dev, not lots of feature branches.
Dana Lewis
@danamlewis
Jul 19 2016 20:46
:) it’s not technically part of OpenAPS, but Nate created an algorithm that he used early on as part of his OpenAPS implementation and then switched it into Loop when he built it. so it’s similar in a lot of ways, but different.
Kim St. Dennis
@kimstdennis_twitter
Jul 19 2016 20:46
I'd love the versioning in that as well @scottleibrand, I think even the docs could use seme versioning.
Dana Lewis
@danamlewis
Jul 19 2016 20:46
there’s talk of porting oref0 into LoopKit, but if we did that we’d also need to do a lot of the documentation and approach we take with OpenAPS in terms of safety steps and build process, because as you noted in the other channel, that’s not yet part of LoopKit
Scott Leibrand
@scottleibrand
Jul 19 2016 20:47
and importantly, Nate's dosing algorithm follows the original OpenAPS reference design pretty closely, so it has a lot of (but not all) the same safety design constraints.
sarahspins
@sarahspins
Jul 19 2016 20:47
Gotcha... I was mainly curious because the attitude towards an app solution seems somewhat mixed...based on a couple of videos of talks I've watched and the google doc that Scott linked me to the other day it almost seems like maybe/maybe not?
Scott Leibrand
@scottleibrand
Jul 19 2016 20:48
I'm fairly conflicted myself. :)
Dana Lewis
@danamlewis
Jul 19 2016 20:48
it’s not about the app, but it’s about having the framework to help guiding someone to self-build a closed loop setup safely, and so they know how it works and that they’re taking responsibility for what they’re building
(from my perspective)
Scott Leibrand
@scottleibrand
Jul 19 2016 20:48
:+1: that ^^
Dana Lewis
@danamlewis
Jul 19 2016 20:48
so whether that’s an app or that’s a hardware setup, that’s what I think is needed before pointing people to it
sarahspins
@sarahspins
Jul 19 2016 20:48
I mean, as a G5 user it isn't really a "shortcut" to the end product as much as a convenience of not having to rely on the internet or another device to get my CGM data in my loop
Hannah Crabtree
@lollydaggle
Jul 19 2016 20:50
RileyLink capabilities confuse me. It connects phone to pump via BLE. So, does that mean I could pull my G5 Share app data into my loop via RL and also port what my loop is doing to Nightscout without needing wifi?
sarahspins
@sarahspins
Jul 19 2016 20:50
Because that's where I was beginning to REALLY find myself frustrated before I found out about RL and ordered one.. yes it could be done, but only by carrying my receiver around, which I didn't really "want" to do... I already carry a device that does all the things I needed to do
@lollydaggle it reads from the G5 directly, but with some limitations
it's not for example, a replacement for the G5 app in the same way that the traditional xDrip app on android can replace the need for a receiver
@lollydaggle basically the app uses the phone's BT to read the G5, and RL provides the connectivity with the pump
Hannah Crabtree
@lollydaggle
Jul 19 2016 20:53
That makes sense! was just about to ask if it can connect with the pump too
sarahspins
@sarahspins
Jul 19 2016 20:53
you can build the app and have it reading G5 data without RL, there's just not a whole lot it does for you without pump data ;)
Hannah Crabtree
@lollydaggle
Jul 19 2016 20:54
so Edison + RL + Battery is the holy grail for portability?
Dana Lewis
@danamlewis
Jul 19 2016 20:54
Or Edison/Ti/battery
(for a second rig)
sarahspins
@sarahspins
Jul 19 2016 20:55
no? not as a G5 user..... because it would probably still mean carrying around the receiver, or always having internet
Dana Lewis
@danamlewis
Jul 19 2016 20:55
Still suggest a Pi as first rig/walking through the docs with the least amount of frustration to getting a working loop going and with understanding
Scott Leibrand
@scottleibrand
Jul 19 2016 20:55
RL doesn't add much portability for an Edison setup: it's more useful if you want a phone-based loop rather than a standalone one.
Dana Lewis
@danamlewis
Jul 19 2016 20:55
Not if you use openxshareble, @sarahspins
you don’t require connectivity if you use that
sarahspins
@sarahspins
Jul 19 2016 20:55
@danamlewis you can't do that with the G5
Scott Leibrand
@scottleibrand
Jul 19 2016 20:55
yeah, openxshareble is G4 only
Dana Lewis
@danamlewis
Jul 19 2016 20:55
true
sarahspins
@sarahspins
Jul 19 2016 20:55
the G5 is different
Dana Lewis
@danamlewis
Jul 19 2016 20:56
@lollydaggle are you G5 or G4?
sarahspins
@sarahspins
Jul 19 2016 20:56
G4 seems like there are lots of options... and you're already carrying the receiver so it's not a big deal to keep carrying it
with G5 I don't carry it, and don't really want to
Dana Lewis
@danamlewis
Jul 19 2016 20:57
for reliability purposes, I probably would still
Hannah Crabtree
@lollydaggle
Jul 19 2016 20:57
@danamlewis @sarahspins I'm G5. Honestly what I have now is portable enough but always looking for efficiencies
sarahspins
@sarahspins
Jul 19 2016 20:57
I mean, I would have, if that was the only reliable way to do it, but I don't tend to have a lot of problems with my phone getting G5 data
Dana Lewis
@danamlewis
Jul 19 2016 20:57
(but, I am still on G4 for a number of reasons, including that I have seen many many cases of G5 not being reliable even with the phone on the bedside table)
sarahspins
@sarahspins
Jul 19 2016 20:57
getting it to the cloud, sure, but I don't really experience the "dropouts" that a lot of G5 users talk about
like what @danamlewis just mentioned
Hannah Crabtree
@lollydaggle
Jul 19 2016 20:58
I don't seem to have G5 issues yet but it's only been 3 months
Dana Lewis
@danamlewis
Jul 19 2016 20:58
nice
sarahspins
@sarahspins
Jul 19 2016 20:58
I've actually had more random dropouts with my receiver... it doesn't seem to always reconnect when it should
I've been on G5 since December, I'm on my 4th transmitter (I've had two failures, one aged out)
so that may be a sticking point for some with the G5... the 3 month life of the transmitter, however, that is a software limit, and I think there will be a way around it soonish
Hannah Crabtree
@lollydaggle
Jul 19 2016 21:00
Yeah! My receiver was having issues last night. I also swear I don't have any setting options to disable or silence the alarms
my transmitter is about to age out (I'm on 93 days!)
sarahspins
@sarahspins
Jul 19 2016 21:01
but, I do have to say that as a dexcom user since the "egg" receiver, with the 7+ not having an extra "thing" to use it now is worth all of the minor (to me) downsides
I am on day 80 with this transmitter.. I think it will hold out until the very end like the other one I had "time out" did :+1:
I did get 112 days from that one
Scott Leibrand
@scottleibrand
Jul 19 2016 21:03
another thing to test with the G5 112-day limit: I've heard that's a runtime limit, not an elapsed-calendar-time limit. if so, you could alternate your two sensors (with two transmitters) with a day or so overlap to get the benefit of watching the two sensors and deciding when to switch between them
Dana Lewis
@danamlewis
Jul 19 2016 21:03
that might be something @channemann can sleuth out… ^
Scott Leibrand
@scottleibrand
Jul 19 2016 21:04
would be fairly easy to test with a 90-day-old transmitter if you keep close track of runtime
activate the new transmitter after 90 days and start alternating them each week, and see how many more weeks you get out of the old one
Chris Hannemann
@channemann
Jul 19 2016 21:05
I'll put it on my list; unsure as to whether that will be in my purview, but happy to queue up questions
Dana Lewis
@danamlewis
Jul 19 2016 21:05
:+1:
sarahspins
@sarahspins
Jul 19 2016 21:05
@scottleibrand with my 112 days, there was a 5 day stretch where I didn't have a sensor on (was out of town, didn't bring a spare), and it didn't matter
it still expired 112 days from when it was activated
Scott Leibrand
@scottleibrand
Jul 19 2016 21:05
ah, bummer
sarahspins
@sarahspins
Jul 19 2016 21:06
so I don't think it's a runtime limit... Dexcom knows when you started it and won't let you use it beyond
Scott Leibrand
@scottleibrand
Jul 19 2016 21:06
ok, nevermind that idea then
sarahspins
@sarahspins
Jul 19 2016 21:07
I know several people have reported though, that they keep getting readings in the app after it says it's expired, even though you can't calibrate or do anything else with it, that data still seems to go to share until the session ends
I didn't experience that myself though, my session ended and it just said pair new transmitter and stopped reading just like it would at the end of any other session
Jason Calabrese
@jasoncalabrese
Jul 19 2016 21:45
I'm pretty sure I heard that the x d r i p g5 android app was able to continue getting data after the share said it was dead
Paul Martin
@Spazholio
Jul 19 2016 21:46
@kimstdennis_twitter Yeah, that’s what I was thinking.
Paul Martin
@Spazholio
Jul 19 2016 22:01
@cjo20 Remember when you were helping me (and others) getting our wpa_supplicant.conf and /e/n/i files set up properly? Well, I just set up a new Edison and set everything up, but it appears as though my Edison is now renegotiating a wifi conncetion every 5-10 seconds. Checking /var/log/syslog shows it running wpa_action and dhclient one after another, a LOT. Any ideas why? I’m sitting at home, not in range of multiple APs or anything.
Chris Oattes
@cjo20
Jul 19 2016 22:12
Not sure why it would do that. Try rebooting.
Pete Schwamb
@ps2
Jul 19 2016 22:16
This message was deleted
Scott Leibrand
@scottleibrand
Jul 19 2016 22:18
@ps2 mischan?
Paul Martin
@Spazholio
Jul 19 2016 22:18
I did reboot, and that seems to have cleared it up. But I took it from work to home (three separate APs involved in that) so I’m a little worried it’s gonna happen again.
Pete Schwamb
@ps2
Jul 19 2016 22:18
gah. Yep, meant for RL chan.
mariano42
@mariano42
Jul 19 2016 22:18
Can anyone offer any advice on ways to ssh into the PI when you cant connect it to wifi? Unfortunately at work the wifi needs a windows certificate to connect to and the guest network needs you to open a browser to log in. My only option would be to hook it up to a keyboard/mouse/monitor everyday to set that up. I believe there is a way to get the PI to spin up an ad hoc network, but I am not sure if that will do exactly what I want. I am on day 2 of looping and missed plugging in my battery before it died today and after it restarted my loop never came back up and I have no way of knowing why right now! (lesson learned about keeping an eye on that battery)
Scott Leibrand
@scottleibrand
Jul 19 2016 22:19
@mariano42 most people use cellular hotspots if there is no "normal" wifi network available
mariano42
@mariano42
Jul 19 2016 22:19
@scottleibrand yea, unfortunately my verizon account doesnt allow it and its my phone through my company so I can't even get it added
Scott Leibrand
@scottleibrand
Jul 19 2016 22:20
maybe a mifi / freedompop type device?
or set up your own little local wifi network
if you have ethernet you could plug it in upstream, but even with just local wifi you can at least log in to the box that way
mariano42
@mariano42
Jul 19 2016 22:21
yea, that seemed to be the best suggestion i have found so far. I am far from a network expert so I will have to do some reading. I just wanted to make sure I wasn't missing something obvious before i went down that road
Paul Martin
@Spazholio
Jul 19 2016 22:21
I would kill for a way for the Edison/Pi to be able to send notifications to an iPhone or something.
Scott Leibrand
@scottleibrand
Jul 19 2016 22:22
@Spazholio what's wrong with Pushover?
Paul Martin
@Spazholio
Jul 19 2016 22:22
Sans internet connection. And possible two-way communication.
And for my Pebble to be able to send BT comms to my Edison.
And a pony.
Scott Leibrand
@scottleibrand
Jul 19 2016 22:23
@danamlewis can help you with the latter. ;-)
Paul Martin
@Spazholio
Jul 19 2016 22:23
Actually, the Pebble -> rig thing would be nice. So I can have it mapped to the setup-my-reverse-ssh-tunnel.sh script that I don’t have written. =)
Dana Lewis
@danamlewis
Jul 19 2016 22:24
No, pony has been on my wish list for a long time and that's never happened...
mariano42
@mariano42
Jul 19 2016 22:24
@scottleibrand thanks as always..
Scott Leibrand
@scottleibrand
Jul 19 2016 22:25
I keep getting her stuff like http://mylittlepony.hasbro.com/images/spring2016/ponies/char_twilightsparkle.png and she's never happy. ;-)
Dana Lewis
@danamlewis
Jul 19 2016 22:27
:)
Paul Martin
@Spazholio
Jul 19 2016 22:29
hehehe
Jason Wittmer
@jbwittmer
Jul 19 2016 22:45
@danamlewis Followup to earlier question about Meal Assist. I'm not sure I'm ready to implement Auto Sensitivity yet. Am I reading correctly that I can implement AMA by switching the latest AMA branch of oref0 tools? Thanks
Dana Lewis
@danamlewis
Jul 19 2016 22:45
It's actually in dev of oref0 now :) and we need to update the docs accordingly!
Ben West
@bewest
Jul 19 2016 22:46
@Spazholio if you learn python, setting that up should be relatively simple
Paul Martin
@Spazholio
Jul 19 2016 22:46
@bewest Prove it. Show me just how simple it can be. :)
Ben West
@bewest
Jul 19 2016 22:46
I already did that with previous examples
the bluez source dir has a test folder full of python scripts that offer some hints on this particular task though :-)
Jason Wittmer
@jbwittmer
Jul 19 2016 22:48
@danamlewis OK, thanks. Can you point me in the direction of instructions as to how to change my setup to the dev version of oref0? Sorry for all the questions.
Ben West
@bewest
Jul 19 2016 22:48
sudo npm install -g git://github.com/openaps/oref0.git#dev
Paul Martin
@Spazholio
Jul 19 2016 22:49
Wait, which thing are we talking about?
Jason Wittmer
@jbwittmer
Jul 19 2016 22:50
@bewest Thanks
Paul Martin
@Spazholio
Jul 19 2016 23:02

@jasoncalabrese I’m following the docs here: https://github.com/openaps/docs/wiki/CGM-Loop-with-openxshareble-scratchpad. When I get to the openaps use cgm scan bit is where it all falls down. I get this:

Traceback (most recent call last):
  File "/usr/local/bin/openaps-use", line 63, in <module>
    app( )
  File "/usr/local/lib/python2.7/dist-packages/openaps/cli/__init__.py", line 51, in __call__
    self.run(self.args)
  File "/usr/local/bin/openaps-use", line 57, in run
    output = app(args, self)
  File "/usr/local/lib/python2.7/dist-packages/openaps/uses/__init__.py", line 92, in __call__
    return self.method.selected(args)(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps/uses/__init__.py", line 31, in __call__
    return self.method(args, app)
  File "/home/edison/src/openxshareble/openxshareble/__init__.py", line 97, in __call__
    res = self.ble.run_mainloop_with(run, quit_with_loop=False)
  File "/home/edison/src/Adafruit_Python_BluefruitLE/Adafruit_BluefruitLE/bluez_dbus/provider.py", line 121, in _user_thread_main
    self._return_code = target()
  File "/home/edison/src/openxshareble/openxshareble/__init__.py", line 92, in run
    self.before_main(args, app)
  File "/home/edison/src/openxshareble/openxshareble/__init__.py", line 75, in before_main
    self.prolog(mac=mac)
  File "/home/edison/src/openxshareble/openxshareble/app.py", line 93, in prolog
    self.remote.connect()  # Will time out after 60 seconds, specify timeout_sec parameter
  File "/home/edison/src/Adafruit_Python_BluefruitLE/Adafruit_BluefruitLE/bluez_dbus/device.py", line 102, in connect
    self._device.Connect()
  File "/usr/lib/python2.7/dist-packages/dbus/proxies.py", line 145, in __call__
    **keywords)
  File "/usr/lib/python2.7/dist-packages/dbus/connection.py", line 651, in call_blocking
    message, timeout)
dbus.exceptions.DBusException: org.bluez.Error.Failed: Protocol not available

I’ve followed the openxshareble install docs in the way I’ve done it before, and bluetoothd is running with the —experimental flag. Any thoughts?

Jason Calabrese
@jasoncalabrese
Jul 19 2016 23:04
did you update /etc/dbus-1/system.d/bluetooth.conf? you're using BlueZ 5.37?
Paul Martin
@Spazholio
Jul 19 2016 23:18
Yes and yes.
Paul Martin
@Spazholio
Jul 19 2016 23:25
Running bluetoothctl finds the DEXCOMRX device almost immediately.
Oh, and every once in a while, the same command returns
Traceback (most recent call last):
  File "/usr/local/bin/openaps-use", line 63, in <module>
    app( )
  File "/usr/local/lib/python2.7/dist-packages/openaps/cli/__init__.py", line 51, in __call__
    self.run(self.args)
  File "/usr/local/bin/openaps-use", line 57, in run
    output = app(args, self)
  File "/usr/local/lib/python2.7/dist-packages/openaps/uses/__init__.py", line 92, in __call__
    return self.method.selected(args)(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps/uses/__init__.py", line 31, in __call__
    return self.method(args, app)
  File "/home/edison/src/openxshareble/openxshareble/__init__.py", line 97, in __call__
    res = self.ble.run_mainloop_with(run, quit_with_loop=False)
  File "/home/edison/src/Adafruit_Python_BluefruitLE/Adafruit_BluefruitLE/bluez_dbus/provider.py", line 121, in _user_thread_main
    self._return_code = target()
  File "/home/edison/src/openxshareble/openxshareble/__init__.py", line 93, in run
    output = self.main(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps/vendors/dexcom.py", line 44, in main
    return self.port or ''
AttributeError: 'scan' object has no attribute 'port’
Lynne Butts
@flutelynne_twitter
Jul 19 2016 23:30
Has anyone used something like an AT&T velocity for having wifi access?
Scott Leibrand
@scottleibrand
Jul 19 2016 23:31
is that like a Mifi?
Paul Martin
@Spazholio
Jul 19 2016 23:35
It is. I priced it out a while ago.
Lynne Butts
@flutelynne_twitter
Jul 19 2016 23:36
I don't know. My iphone hotspot never works, and I want to find a way to connect when away.
Scott Leibrand
@scottleibrand
Jul 19 2016 23:37
the main question will be whether it is "always on" or if you have to turn it on to connect
Paul Martin
@Spazholio
Jul 19 2016 23:38
I’m assuming you’ve got the iPhone hotspot screen open until the first device connects, right? Do you have a password with special characters in it?
@scottleibrand My main question was, “Is this worth another $20 a month just to use the data I already pay for?"
Lynne Butts
@flutelynne_twitter
Jul 19 2016 23:39
I will have to look into it further. It shows up on the choices and I set it up for the pi, but, right, the extra $20/month...
Possible long time away in coming months, and don't want to be without connectivity.
jaylagorio
@jaylagorio
Jul 19 2016 23:40
@Spazholio I use a FreedomPop device
Lynne Butts
@flutelynne_twitter
Jul 19 2016 23:40
@jaylagorio Does it work well?
jaylagorio
@jaylagorio
Jul 19 2016 23:40
When I turn it on and the Pi doesn't already have wifi access it connects up in less than 30 seconds
Paul Martin
@Spazholio
Jul 19 2016 23:40
I’ve got my iPhone hotspot working with all three of my devices seamelessly. I’d be happy to see if I can help at some point if you’d like.
@jaylagorio Cost?
jaylagorio
@jaylagorio
Jul 19 2016 23:40
The device works well, the speed and reception will vary depending on your area
The hotspot was $49 referb, the monthly cost is something like $10/mo
Lynne Butts
@flutelynne_twitter
Jul 19 2016 23:41
@Spazholio Thanks. I have to give a lesson now, but will come back maybe tomorrow or something to see about it.
@jaylagorio That doesn't sound too bad.
@Spazholio I think because I have the G5 app and the Pebble app going, there may be some interference.
jaylagorio
@jaylagorio
Jul 19 2016 23:43
I'm attempting to finagle it down to $0/mo and a 500MB cap but I haven't put a lot of effort into it yet
On the front page: "100% Free Mobile Phone & Internet Service / No Contract • No Commitment • Cancel Anytime"
I just haven't dialed it down yet
freedompop.com
Paul Martin
@Spazholio
Jul 19 2016 23:43
Well, I’m with AT&T (they give a discount for it at work) and AT&T says “$20 per additional device on the plan”.
Jason Calabrese
@jasoncalabrese
Jul 19 2016 23:44
@Spazholio I think it's looking at another device around you that doesn't have a port attribute
jaylagorio
@jaylagorio
Jul 19 2016 23:44
I'm on T-Mobile, the Freedompop thing is totally seperate
Jason Calabrese
@jasoncalabrese
Jul 19 2016 23:44
maybe not, but I saw some things like that
Paul Martin
@Spazholio
Jul 19 2016 23:44
@jasoncalabrese So if I give the MAC/serial (I know them) then I should be ok? I should be able to skip the scanning bit?
Jason Calabrese
@jasoncalabrese
Jul 19 2016 23:44
if you know the mac that will help a lot
almost imposible to get it to work without setting the mac
Lynne Butts
@flutelynne_twitter
Jul 19 2016 23:45
Same here. AT&T. Same quote.
Paul Martin
@Spazholio
Jul 19 2016 23:54
That “no port” thing seems to ring a faint bell.
Hang on, re-pairing with the phone for a few minutes to get some data into NS. Just got done with a new sensor session. I’ll try again in a couple.