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

31st
May 2016
Eric Luhrs
@eluhrs
May 31 2016 00:06
@Spazholio Sorry, but I've never used a vender to print. I work at a college too. We have printers in our library, engineering department, and also art department. Also, you may want to check for a soldering lab if you have engineering.
Paul Martin
@Spazholio
May 31 2016 00:07
I think the one I used for my RileyLink case is widely considered to the the “least good” one we have. I’ll ask around. Thanks!
Eric Luhrs
@eluhrs
May 31 2016 00:08
we have a few Utilimakers and a Fabricator. The Fabricator is much, much better.
Paul Martin
@Spazholio
May 31 2016 00:11
I’ll keep that in mind. I honestly don’t know what we’ve got, but I know we have several scattered across a few departments.
I’ve found places that’ll print for you for something like $20. It’s not horrible, but….free is better, y’know?
Eric Luhrs
@eluhrs
May 31 2016 00:13
$20 is nuts for .50 cents worth of filament, although you should print at the highest quality, which will take 8-12 hours depending on the printer.
let me know if you cant get it done for free. Id rather mail you one than have you pay that much.
Paul Martin
@Spazholio
May 31 2016 00:14
That’s true, but you have to actually have the $XXX 3D printer. Gotta recoup that investment. I don’t blame them for charging $20, I just don’t wanna pay it. =)
Eric Luhrs
@eluhrs
May 31 2016 00:14
get another 1000c instead :)
Paul Martin
@Spazholio
May 31 2016 00:15
Does the one you have at your place print at a good enough quality for your liking? If so, I wouldn’t be averse to just getting one from you and paying you a few bucks for your trouble. And shipping, of course. =)
Eric Luhrs
@eluhrs
May 31 2016 00:17
Yes, the fabricator does an excellent job. I avoid the utilimakers.
pm your address and ill send one sometime this week.
Paul Martin
@Spazholio
May 31 2016 00:23
@bewest Is there a proper oref0 method to extract pump history? My current method uses iter_pump_history and I think that’s “old code".
Ben West
@bewest
May 31 2016 00:25
that's fine
Paul Martin
@Spazholio
May 31 2016 00:25
Ok, cool.
Ben West
@bewest
May 31 2016 00:26
for treatments, use openaps use ns shell format-recent-history-treatments monitor/zoned-pump-history.json model.json
Paul Martin
@Spazholio
May 31 2016 00:27
I…think I have that now. That looks familiar.
Ben West
@bewest
May 31 2016 00:27
that preps it, I believe, then I openaps use ns shell upload-non-empty-treatments the results from there
Paul Martin
@Spazholio
May 31 2016 00:28
I’m keeping my openaps.ini with some of the stuff that was originally set up when I started this. I figure in a year, it’ll be worth something to the Smithsonian. =)
Paul Martin
@Spazholio
May 31 2016 00:52
I’m noticing that my pumphistory-24h.json report dies often, and I’m pretty sure that’s just a fun side effect of radio communications with the pump. Would it be prudent to try to trap that, and then retry a few times until it runs successfully, or is it just better to let the loop die, and try again during the next iteration?
ddaniels1
@ddaniels1
May 31 2016 00:57
@bewest similar issues I think. Trying to add pump device (I think pump.ini is kept out of the git repo because of serial number). error:
edison@doubled:~/Documents/try5$ openaps device add pump mmeowlink subg_rfspy /dev/ttyMFD1 xxxxxx

Traceback (most recent call last):
  File "/usr/local/bin/openaps-device", line 6, in <module>
    exec(compile(open(__file__).read(), __file__, 'exec'))
  File "/home/edison/src/openaps/bin/openaps-device", line 35, in <module>
    app( )
  File "/home/edison/src/openaps/openaps/cli/__init__.py", line 47, in __call__
    self.configure_parser(self.parser)
  File "/home/edison/src/openaps/bin/openaps-device", line 21, in configure_parser
    self.commands = devices.configure_commands(parser, parent=self)
  File "/home/edison/src/openaps/openaps/devices/__init__.py", line 36, in configure_commands
    parent.devices = get_device_map(parent.config)
  File "/home/edison/src/openaps/openaps/devices/__init__.py", line 21, in get_device_map
    for device in Device.FromConfig(vendors, conf):
  File "/home/edison/src/openaps/openaps/devices/device.py", line 79, in FromConfig
    device.read(config=config)
  File "/home/edison/src/openaps/openaps/devices/device.py", line 36, in read
    self.extra.fields.update(dict(extra.items(self.section_name( ))))
  File "/usr/lib/python2.7/ConfigParser.py", line 642, in items
    raise NoSectionError(section)
ConfigParser.NoSectionError: No section: u'device "pump"'
Paul Martin
@Spazholio
May 31 2016 01:00
During my enact/suggested.json, I’m still getting problems with the CGM date being set wrong:
oref0-determine-basal://text/shell/enact/suggested.json
Optional feature Meal Assist enabled, but could not read required meal data.
{"ratio":0.89}
BG data is too old, or clock set incorrectly.  Your CGM time is Mon May 30 2016 19:56:07 GMT-0400 (EDT) but your system time is Mon May 30 2016 20:59:39 GMT-0400 (EDT)
reporting enact/suggested.json
The enact/suggested.json is set up as such: openaps use oref0-determine-basal shell monitor/iob.json monitor/temp_basal.json monitor/glucose.json settings/profile.json settings/auto-sens.json monitor/meal.json []. Is the remainder causing issues, possibly?
ddaniels1
@ddaniels1
May 31 2016 01:13
anyone have insight on the above error? everything I try to do the error I get is: ConfigParser.NoSectionError: No section: u'device "pump"'
wonder if I should uninstall openaps and reinstall.
Paul Martin
@Spazholio
May 31 2016 01:14
Hang on a second...
Isn’t the format for that openaps device add <my_pump_name> medtronic <my_serial_number>?
It’s been a while since I’ve done it though...
ddaniels1
@ddaniels1
May 31 2016 01:18
yeah, that's it. That's not the problem
Paul Martin
@Spazholio
May 31 2016 01:21
Ok. What’s going on?
Dan Piessens
@dpiessens
May 31 2016 01:32
@ddaniels1 It's saying there's no pump device, did you not source control your pump.ini file?
Sorry that first part was too obvious :)
ddaniels1
@ddaniels1
May 31 2016 01:33
Pump.ini didn't come over, I think because of the serial number so it is not in the git clone. Problem is I can't add any devices. wonder if I need to init a new openaps instance, then clone into there
Dan Piessens
@dpiessens
May 31 2016 01:34
what's the error when you add the device?
ddaniels1
@ddaniels1
May 31 2016 01:34
same error above, no matter what I do
ConfigParser.NoSectionError: No section: u'device "pump"'
Dan Piessens
@dpiessens
May 31 2016 01:34
can you nano into your config and remove any existing pump device section?
ddaniels1
@ddaniels1
May 31 2016 01:38
There isn't an existing pump there. I just tried initialize a new openaps instance. There all the commands work. I need to do that then clone the directory I had on the old edison there
OK, i got it, you were right, there was an existing pump section!
Thanks @dpiessens
:)
Ben West
@bewest
May 31 2016 01:42
openaps vendor add mmeowlink.vendors.mmeowlink
openaps device add pump mmeowlink subg_rfspy /dev/serial/by-id/usb-nightscout* should be enough
then to enter the serial number,
openaps use pump config --serial 123456 (this one is new as of this weekend)
made it so serial could be left out when addin
hmmmm if that doesn't work, let me know, looks like you had it right and it still errored
Paul Martin
@Spazholio
May 31 2016 01:45
@bewest It looks like I’ve got to use that monitor/auto-sens-glucose.json in a number of places, instead of just the one. Seems the lack of sgv entries is breaking several things...
Ben West
@bewest
May 31 2016 01:47
hmmm, yeah
well, there's plenty of sgv's in theres
but there's also mbgs and other things
basic idea was that entries would be for machine data, and treatments would be for people-entered data
Paul Martin
@Spazholio
May 31 2016 01:48
Yeah, but you miss one, and things are suddenly going ka-flooey. I’m keeping the monitor/glucose.json for things that need it, but the reports that are dying, I’m switching over to monitor/auto-sens-glucose.json and it after a few hours, the loop is finally working again.
I’ve opened an issue on openaps/oref0, but I didn’t capture the full scope. I worry that I’m doing something wrong, but hopefully someone will be able to tell me if so.
Ben West
@bewest
May 31 2016 01:50
:-)
Paul Martin
@Spazholio
May 31 2016 01:50
Another oddity - if I invoke monitor/meal.json on its own, it zips through, but during the loop, it takes 30-45s. Nothinig major, just peculiar.
(not worth troubleshooting, just odd)
Dan Piessens
@dpiessens
May 31 2016 01:57
@Spazholio file lock from another process / git maybe?
btw mind dumping the report sequence for CGM+NS with gap filler?
Paul Martin
@Spazholio
May 31 2016 02:00
How do you mean?
And I’m not using the gap filter just yet. But I have the command to add the report. =)
Dan Piessens
@dpiessens
May 31 2016 02:03
How do I mean on the git lock?
Paul Martin
@Spazholio
May 31 2016 02:03
Correct.
And I misspoke earlier.
If I openaps invoke settings/auto-sens.json then it takes forever. If I invoke monitor/meal.json it seems to run the same calculations, but goes very fast. Looks the same though. I don’t recall it doing that before...
I must have mucked up somwhere.
Dan Piessens
@dpiessens
May 31 2016 02:05
so what I've noted on some linux systems is git completes operations async from process terminiation. This means that if you call git add on a bunch of files it will say "done!" and still be doing things
Paul Martin
@Spazholio
May 31 2016 02:05
Here’s my report for monitor/glucose.json - openaps use tz select raw-ns/ns-glucose-zoned.json raw-cgm/glucose-zoned.json --date "dateString" --current “now”. This joins my raw NS and my raw CGM files into one.
My monitor/gaps.json is openaps use tz lsgaps raw-ns/ns-glucose-zoned.json raw-cgm/glucose-zoned.json --minutes "10" --date "dateString" --before "1 day ago" --after “now"
Dan Piessens
@dpiessens
May 31 2016 02:06
is one leveraging previous data versus starting from scratch?
Paul Martin
@Spazholio
May 31 2016 02:06
I think one is using pumphistory.json versus the other one using pumphistory-24h.json.
Dan Piessens
@dpiessens
May 31 2016 02:06
ah that could do it
Paul Martin
@Spazholio
May 31 2016 02:07
I think the one using the 24h is the right one. auto-sens is supposed to do a comparison against the last day. Unsure why meal.json is re-running the report though. Hmm.
Dan Piessens
@dpiessens
May 31 2016 02:07
hmm that is interesting
Ben West
@bewest
May 31 2016 02:08
doesn't sound like same report, but rather different report that maybe is sharing some code
Paul Martin
@Spazholio
May 31 2016 02:09
Yeah, it’s a different report, but it’s invoking the same --+++--+—>> stuff.
I don’t recall meal.json doing that previously, which is the part that has me confused.
Ben West
@bewest
May 31 2016 02:09
not using the meal stuff yet
been working on making the basics cleaner/easier
Paul Martin
@Spazholio
May 31 2016 02:10
Yeah, I know I’m going all bleeding edge here. But I really like the automatic ISF changes.
Ben West
@bewest
May 31 2016 02:10
:-D
running c-r-m locally on one edison at least :-D
Paul Martin
@Spazholio
May 31 2016 02:11
CRM?
Ben West
@bewest
May 31 2016 02:11
think that'll be nice, offering nightscout hosted locally
next piece is daemons
oacids already exists, but need a daemon for each device in addition
Dan Piessens
@dpiessens
May 31 2016 02:11
that would be cool
Paul Martin
@Spazholio
May 31 2016 02:12
oacids? New acronyms! Woo!
(what’s “oacids”?)
Ben West
@bewest
May 31 2016 02:12
plus one more daemon to do ble services, so mobile app can control/monitor openaps itself
in my google, I'm number one result
it's a daemon that can replace cron
so instead of setting up cron, you'll add openaps trigger {show,add,remove,until,emit} or openaps schedule {show,add,remove}, and maybe openaps do instead
oacids provides trigger and schedule, but they are linux-specific
most of openaps/mmeowlink can be run on mac/windows even
I'm somewhat open to more preferable names
Dan Piessens
@dpiessens
May 31 2016 02:17
well if you're daring like me, you can run windows 10 fast-ring and run this via bash
Ben West
@bewest
May 31 2016 02:17
yeah, should work
Dan Piessens
@dpiessens
May 31 2016 02:17
does for the most part, they haven't implemented all the usb device support in the kernel yet
Ben West
@bewest
May 31 2016 02:18
:-(
Dan Piessens
@dpiessens
May 31 2016 02:18
but it works for command testing, help and pipelines
Ben West
@bewest
May 31 2016 02:18
well, most of openaps should work almost anywhere
turns out windows/dos supports piping and stuff
yeah, definitely encourage testing things like that on a nice computer
the tab completion is fast and it helps a lot
Dan Piessens
@dpiessens
May 31 2016 02:19
yeah, agreed.
Ben West
@bewest
May 31 2016 02:20
anyway, that's why oacids is separate, it's linux-only
it's not just a matter of bash/syntax, it actually works with a system service that is typically only available on linux
open to better names, thinking currently it's something like Open access continuous insulin delivery system
or maybe open artificial continuous information data system
Dan Piessens
@dpiessens
May 31 2016 02:23
Looks neat, if people want to do more with this on windows I can help write the service for it
Rob Kresha
@rkresha
May 31 2016 02:31
Is there any way around the warmup period on the CGM, meaning can we keep openaps looping and making decisions or do we have to deal with 2 hrs of manual mode?
Ben West
@bewest
May 31 2016 02:31
oh, I was just working on that
Rob Kresha
@rkresha
May 31 2016 02:32
have it go off of raw data?
Ben West
@bewest
May 31 2016 02:32
someone created an oref0 raw think @jasoncalabrese
I just created new use for dexcom called calibrations, iter_calibrations $count and also iter_calibrations_hours $hours
so eg openaps use cgm calibrations
Rob Kresha
@rkresha
May 31 2016 02:35
to find calibration times
Ben West
@bewest
May 31 2016 02:37
looks like oref0 raw needs glucose and the calibrations
sorry
raw and calibrations
and some kind of factor
Rob Kresha
@rkresha
May 31 2016 02:37
ah... makes sense now, I fed it a dummy file for calibrations and it still worked, not sure what I'm losing out on
Ben West
@bewest
May 31 2016 02:40
oh?
I didn't think to try that
Rob Kresha
@rkresha
May 31 2016 02:40
maybe that is used as a filter or something
Paul Martin
@Spazholio
May 31 2016 02:41
Yeah, I’ve not been able to figure out how to actually utilize oref0 raw...
Ben West
@bewest
May 31 2016 02:41
very interesting, I'm interested now in pulling boluses with bgs and adding them as calibrations
Paul Martin
@Spazholio
May 31 2016 02:42
@bewest Do you know offhand who the “primary” is on auto-sens? Is it @scottleibrand?
Ben West
@bewest
May 31 2016 02:47
yes
Paul Martin
@Spazholio
May 31 2016 02:50
Ok, cool. I’ll wait for him to get back from vacation before I bombard him with questions. =)
Ben West
@bewest
May 31 2016 03:09
hmmm
I have some more changes that fix the remaining --cli issues
Paul Martin
@Spazholio
May 31 2016 03:13
Need something tested? I didn’t notice any issues on my end with the —cli reports, but I’d be happy to run a before and after if you’d like.
Ben West
@bewest
May 31 2016 03:14
I noticed when you pasted ;-)
there were a few things that were incorrect
Paul Martin
@Spazholio
May 31 2016 03:14
I figured. =)
Ben West
@bewest
May 31 2016 03:15
hmm
I just released openaps again
sudo pip install -U openaps should create a nice openaps --version with 0.1.5
Paul Martin
@Spazholio
May 31 2016 03:15
Tell me what you need me to run?
Ben West
@bewest
May 31 2016 03:15
would love to see the results of openaps report show --cli
Paul Martin
@Spazholio
May 31 2016 03:15
Which report?
Ben West
@bewest
May 31 2016 03:16
well if you don't mind, all of them, just like that :-)
Paul Martin
@Spazholio
May 31 2016 03:16
openaps use pump mmtune
openaps use cgm oref0_glucose --hours "25.0" --threshold "100"
openaps use ns-glucose shell []
openaps use pump model
openaps use pump read_clock
openaps use tz clock --timezone "EDT" --date None monitor/clock.json
openaps use pump read_temp_basal
openaps use pump reservoir
openaps use pump read_battery_status
openaps use pump status
openaps use pump iter_pump_hours 3
openaps use pump iter_pump_hours 28.0
openaps use tz rezone --timezone "EDT" monitor/pumphistory.json
openaps use tz rezone --timezone "EDT" settings/pumphistory-24h.json
openaps use iob shell monitor/pumphistory-zoned.json settings/profile.json monitor/clock-zoned.json []
openaps use tz rezone raw-cgm/glucose.json
openaps use meal shell settings/pumphistory-24h-zoned.json settings/profile.json monitor/clock-zoned.json monitor/auto-sens-glucose.json settings/basal_profile.json 
openaps use auto-sens shell monitor/auto-sens-glucose.json settings/pumphistory-24h-zoned.json settings/insulin_sensitivities.json settings/basal_profile.json settings/profile.json
openaps use pump read_bg_targets
openaps use pump read_insulin_sensitivities
openaps use pump read_carb_ratios
openaps use pump read_selected_basal_profile
openaps use pump read_settings
openaps use get-profile shell settings/settings.json settings/bg_targets.json settings/insulin_sensitivities.json settings/basal_profile.json max_iob.json settings/carb_ratios.json 
openaps use oref0-determine-basal shell monitor/iob.json monitor/temp_basal.json monitor/auto-sens-glucose.json settings/profile.json settings/auto-sens.json monitor/meal.json 
openaps use pump set_temp_basal enact/suggested.json
openaps use pebble shell monitor/glucose.json monitor/iob.json settings/basal_profile.json monitor/temp_basal.json enact/suggested.json enact/enacted.json monitor/meal.json []
openaps use ns-status shell monitor/clock-zoned.json monitor/iob.json enact/suggested.json enact/enacted.json monitor/battery.json monitor/reservoir.json monitor/status.json
openaps use tz rezone --timezone "EDT" --date timestamp --date dateString --date start_at --date end_at --date created_at --date display_time --date dateString raw-cgm/glucose.json
openaps use ns shell format-recent-history-treatments settings/pumphistory-24h.json settings/model.json
openaps use ns shell upload-non-empty-treatments nightscout/recent-treatments.json
openaps use cgm oref0_glucose --hours "25.0" --threshold "100"
openaps use status shell monitor/clock.json oref0-moitor/iob.json oref0-predict/oref0.json oref0-enacted/enacted-temp-basal.json monitor/battery.json monitor/reservoir.json monitor/status.json
openaps use tz rezone --adjust "replace" raw-ns/ns-glucose.json
openaps use tz select raw-ns/ns-glucose-zoned.json raw-cgm/glucose-zoned.json --date "dateString"
openaps use tz lsgaps raw-ns/ns-glucose-zoned.json raw-cgm/glucose-zoned.json --minutes "10" --date "dateString" --before "1 day ago" --after "now"
openaps use clean-nightscout-glucose shell monitor/glucose.json
Ben West
@bewest
May 31 2016 03:16
muahahahaha
nice
Paul Martin
@Spazholio
May 31 2016 03:16
I really should have run that before, just so I can see what you’re cackling about. =)
Ben West
@bewest
May 31 2016 03:17
ok, that's the honest version
you can scroll up earlier
you pasted it far earlier several times
see: openaps use tz rezone --timezone "EDT" --date timestamp --date dateString --date start_at --date end_at --date created_at --date display_time --date dateString raw-cgm/glucose.json o
that's a a great example right there
Paul Martin
@Spazholio
May 31 2016 03:17
You know what would be an AWESOME tool, that I suspect is well within your abilities? Something that goes through my setup, and identifies cruft.
Ben West
@bewest
May 31 2016 03:17
it used to say other stuff
nah, human eye is very useful
you can probably just run oref0 template stuff
and import all the templates and just keep it running mostly
Paul Martin
@Spazholio
May 31 2016 03:18
Yeah, but human eye glazes over. =)
Once AMA gets pulled into master, I may start over and try the oref0 template route.
I’m honestly terrified of completely torquing things. Yeah, I know I can just make a backup, but still.
Ben West
@bewest
May 31 2016 03:20
I mean most of the templates I don't think interfere with what you're doing
so I doubt they'd change anything other than giving more reports/aliases to choose from
Ben West
@bewest
May 31 2016 03:38
@Spazholio thanks for testing bunch of stuff, it's been useful
Paul Martin
@Spazholio
May 31 2016 03:39
I love that you’re thanking me. =) You’ve been helping me out with so much lately. So right back atcha.
Oh, and check your private messages when you get a moment.
Ben West
@bewest
May 31 2016 03:40
I noticed in your last report show --cli there are several reports with remainder = []
those seem likely to cause issues
Paul Martin
@Spazholio
May 31 2016 03:41
Yeah, I’m removing them as I go. What was that originally designed for?
Ben West
@bewest
May 31 2016 03:42
ah, it's just a bug/sloppy programmer from yours truly
Paul Martin
@Spazholio
May 31 2016 03:42
Gotcha.
Ben West
@bewest
May 31 2016 03:42
it's a coding thing where a list of items would turn into the code expression for a list of things like ['foo', 'bar']
or empty list []
either way it's not good, and been fixed
can just re-add the report with same use now
now you can actually copy/paste and it works
aha
openaps report show --json |json -g | json -c "this[this.name].remainder == '[]'"
this shows all the reports with remainder = []
Paul Martin
@Spazholio
May 31 2016 03:47
I have got to learn the json command a little better.
And that remainder bit is absolutely not needed anywhere? I haven’t wholesale removed it because…well, that’s how thing get broke. =)
Ben West
@bewest
May 31 2016 03:50
openaps report show --json |json -g | json -c "this[this.name].remainder == '[]'" |json -e "delete this[this.name].remainder"
lists a veresion with it deleted
hmm
on this system, I had to remove the report and re-add it or re-import it
Paul Martin
@Spazholio
May 31 2016 03:55
Really? I can’t just edit the openaps.ini and yank it outta there?
Hmm.
Ben West
@bewest
May 31 2016 03:55
you can, yes
but the script to find them looks something like that
Paul Martin
@Spazholio
May 31 2016 03:56
Ahhh, ok.
Oh, that line showed me three reports, but it missed the ones that had remainder = instead of remainder = [].
ryanchen26
@ryanchen26
May 31 2016 03:57
Is it a bad sign if /etc/group shows crontab:x:107: without my "pi" username after it? Is this the reason I can't schedule openaps loop to run in a loop?
Ben West
@bewest
May 31 2016 03:58
openaps report show --json |json -g | json -c "this[this.name].remainder == '[]'" | json -a name | while read name ; do openaps report show $name --json | json -e "delete this[this.name].remainder" | (openaps report remove $name && openaps import - ); done
that snippet may or may not fix the remainder = [] problem if any exists
remainder = is fine
it's the [] that's the problem
@ryanchen26 what does groups say?
and what about crontab -l
ryanchen26
@ryanchen26
May 31 2016 04:02
it shows me the crontab schedule : SHELL=/bin/bash PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin */2 * * * * cd /home/pi/aps && openaps retry-loop 2>&1 | logger -t openaps-loop it just never runs. I'm trying to find out why. I'm stuck.
typing groups...presents this list: pi adm dialout cdrom sudo audio video plugdev games users input netdev gpio i2c spi
Ben West
@bewest
May 31 2016 04:06
anything in tail /var/log/syslog?
ryanchen26
@ryanchen26
May 31 2016 04:15
interesting....it is definitely running , just not well. I'm seeing "bad zero CRC?" a lot.
ryanchen26
@ryanchen26
May 31 2016 04:21
rebooting, it looks like it's just tripping over itself.
Ben West
@bewest
May 31 2016 04:47
bad zero CRC might be just fine
JaysonEwer
@JaysonEwer
May 31 2016 04:52
Has anyone seen or heard from @danamlewis or @scottleibrand this evening?
ryanchen26
@ryanchen26
May 31 2016 04:56
@bewest thanks for the feedback.
May 30 21:55:47 raspberrypi openaps-loop: ns-status://JSON/shell/monitor/upload-status.json May 30 21:55:47 raspberrypi openaps-loop: reporting monitor/upload-status.json May 30 21:55:47 raspberrypi openaps-loop: Input file entries.json does not exist.
in syslog, this is where it fails.
works when executed manually...mysterious.
Scott Leibrand
@scottleibrand
May 31 2016 08:43
@elodaille01_twitter Normally you'll want to use your existing Nightscout site for OpenAPS. If you already have BG from Dexcom share bridge, that is great. You can either have openaps download that BG info from NS, or it can get BG data independently from the local receiver. Either way, you'll want to upload treatment and status info from OpenAPS to NS so you can see what OpenAPS has been doing.
Hichame Yessou
@hichameyessou
May 31 2016 08:47
What about uploading the BG from OpenAPS to nightscout? How I should do it?
Scott Leibrand
@scottleibrand
May 31 2016 08:48
@Spazholio would the "You CGM time is" error be clearer if it said "the last CGM glucose reading was"?
Scott Leibrand
@scottleibrand
May 31 2016 09:02
@hichameyessou uploading BG from OpenAPS to NS can be done, but isn't documented well yet, and isn't something I'm doing yet.
@JaysonEwer your evening was middle of the night here in Norway. :-)
@Spazholio I'm still reading all of Gitter, so feel free to drop me questions, and I'll answer at some point.
Didie
@elodaille01_twitter
May 31 2016 09:47
@scottleibrand @bewest hello !!please must we use the Web site we already have built for our daughter (to see nightscout from the motorola connected to the dexcom receiver) ? We are a little bit lost.I guess that the answer is no..indeed, the motorola we use today to see the dexcom graphs is connected to dexcom..but here there s no connection between dexcom and ..any phone..
Scott Leibrand
@scottleibrand
May 31 2016 11:23
@elodaille01_twitter Normally you'll want to use your existing Nightscout site for OpenAPS. If you already have BG from Dexcom share bridge, that is great. You can either have openaps download that BG info from NS, or it can get BG data independently from the local receiver. Either way, you'll want to upload treatment and status info from OpenAPS to NS so you can see what OpenAPS has been doing.
JaysonEwer
@JaysonEwer
May 31 2016 12:26
@scottleibrand No worries. I'm all set.
Dana Lewis
@danamlewis
May 31 2016 12:27
:+1:
Paul Martin
@Spazholio
May 31 2016 13:40
I'm a little concerned about something I just noticed. I've been hanging out around 60-70, and can't seem to get my numbers up. I checked NS, and I see the following:
blob
The pill is showing that my last BG reading was 95. I have NO IDEA where it's getting that. But it's setting a higher-than-normal temp basal based on that.
Tom Boudreau
@TomBoudr_twitter
May 31 2016 13:45
@ryanchen26 I'm stuck in the same spot on my build... let me know if you figure it out. I have it working manually, but cron is doing zippy
Paul Martin
@Spazholio
May 31 2016 14:00
I'm wondering if something's wrong with share. I'm looking at my monitor/glucose.json and seeing flat-out wrong sgv values in there. It says the device is "share2" but the values I'm seeing simply aren't reflected on my receiver.
Paul Martin
@Spazholio
May 31 2016 14:14
@bewest Just a heads up - it appears some of the work on my monitor/glucose.json we did yesterday is/was the culprit. Merging the CGM and NS data. Moving back to my older method seemed to fix the issues. No clue why.
ryanchen26
@ryanchen26
May 31 2016 16:14
@jaylagorio Yes, I'll mail you the defective pump if you really think you can do something with it. Out of curiosity I did crack open the pump to look at the internals so if you're going to access the guts anyway I saved you a step. PM me your address.
ryanchen26
@ryanchen26
May 31 2016 16:23
@TomBoudr_twitter Yes, I'm going to hit this every night until I get it. I wish I could sign a waiver and just get the cold hard fact regarding the setup without all the ambiguity. I'll let you know.
Chris Oattes
@cjo20
May 31 2016 16:40
@ryanchen26 what do you mean by "the cold hard fact regarding the setup without all the ambiguity"?
if you mean you want to be able to just have a magic setup that just works, it isn't like that for a very good reason. An APS consists of a bunch of parts and you need to know how all of them fit together to be able to use it effectively, to be able to make adjustments, to be able to troubleshoot and understand why it is doing what it is doing. If someone just told you how to fit it all t
ogether, you wouldn't understand all of that.
Dana Lewis
@danamlewis
May 31 2016 17:12
:+1:
Colin Lennon
@colinlennon
May 31 2016 17:14
hey folks. I think I might have missed this... has AMA been merged into the latest dev now?
dhermanns
@dhermanns
May 31 2016 18:10
Hi! I tried to install openaps on a Pi but get the Error:
SyntaxError: '<' operator not allowed in environment markers
error: Setup script exited with error in setup command: Error parsing /tmp/easy_install-j8XR62/mock- 2.0.0/setup.cfg: SyntaxError: '<' operator not allowed in environment markers
I dont know much about python, so any help would be appreciated ;)
mm22dl
@mm22dl
May 31 2016 18:43
Good evening everyone! :-) Question of the day: when I'm using openaps, does the latter automatically detect when I'm bolusing for a meal? Or do I have to unplug the whole system?
Ben West
@bewest
May 31 2016 18:49
do everything normally, including bolus for meal
mm22dl
@mm22dl
May 31 2016 18:57
Thanks!
Is openaps directly reading the bolus history on the pump for that, or rather relying on a suddenly rising BG, for example?
Ben West
@bewest
May 31 2016 19:04
it's reading history from the pump
JaysonEwer
@JaysonEwer
May 31 2016 19:13
@dhermanns are you working from the
mm22dl
@mm22dl
May 31 2016 19:18
(y)
JaysonEwer
@JaysonEwer
May 31 2016 19:20
@dhermanns if you can provide the command you used to install Openaps, that would be helpful. Also, if you are following the doc in the link above, what step are you on and do you feel there is a lack of clarity that could be improved? If so it is always helpful to submit a PR with added clarity.
ryanchen26
@ryanchen26
May 31 2016 19:37
@cjo20 Yes, you are right and I agree with you. My comment was more about wishing to see under the hood of a working example (having reached a particular threshold in my own rig by following the docs) so I can understand why my setup isn't working properly. The loop runs and updates NS fine when triggered manually but not when triggered with crontab (the loop is triggered but fails right after status-update). I just need to be more patient and keep troubleshooting.
Andrew Baugh
@baughaw
May 31 2016 19:42
@ryanchen26 Can you paste the contents of your cron so we can see what you have? Could be something as simple as the PATH hasnt been set
Ben West
@bewest
May 31 2016 20:41
@ryanchen26 it sounds like there are some environment variables being used that are defined when you log in (maybe from ~/.bashrc?) but not defined via cron
there are two ways to fix it: 1) (my preferred way) is to use nightscout autoconfigure-device-crud to create a new ns device and use that instead of ns-upload $NIGHTSCOUT_HOST $API_SECRET using openaps use ns shell upload
2.) add NIGHTSCOUT_HOST=aaaa.mysite.com and API_SECRET=mydemosecret in the crontab
dhermanns
@dhermanns
May 31 2016 21:07
@JaysonEwer I openend this Ticket with some more Information:
openaps/openaps#95
I followed the github
docu so far. So I tried:
sudo python setup.py develop
And
sudo pip install -U openaps
Nathan Morse
@morsen
May 31 2016 22:24
at this point i’m only curious: has anyone gotten a cellular radio to work with pi or edison to act as an uploader for an openaps implementation?