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

2nd
Oct 2016
Jason Calabrese
@jasoncalabrese
Oct 02 2016 00:11
@scottleibrand main reason to use copy-fresher is to be able to use a report invoke instead a more complicated alias, also get more of an audit trail
garykidd
@garykidd
Oct 02 2016 00:12
@scottleibrand at one point recently, I thought you may have indicated that ther ewas an oref0 pill in NS....might you have meant the OpenAPS pill...or is there a separate oref0 pill?
Jason Calabrese
@jasoncalabrese
Oct 02 2016 00:12
I think jq merge works good when it's fully rebuilt each time merging back to the same file seems most problematic
openaps report add monitor/glucose-merge.json JSON merge shell monitor/glucose-local.json nightscout/glucose-recent.json
So in that case there's no copy, if, etc, just an invoke
Scott Leibrand
@scottleibrand
Oct 02 2016 00:20
does it do that based on the file contents?
Jason Calabrese
@jasoncalabrese
Oct 02 2016 00:25
Merge is just a wrapper for jq, combined the 2 files and sorts
openaps device add merge process --require 'short long' 'jq -s ".[0] + .[1]|unique|sort_by(.dateString)|reverse"'
garykidd
@garykidd
Oct 02 2016 00:26
@jasoncalabrese would your monitor/glucose-local.json be cgm derived and the latter..obviously NS derived?
Jason Calabrese
@jasoncalabrese
Oct 02 2016 00:28
Yeah, I use copy-fresher for that
openaps report add monitor/glucose-local.json text oref0 shell copy-fresher /home/indy/cgm-loop/monitor/glucose-raw.json
garykidd
@garykidd
Oct 02 2016 00:31
and monitor/glucose-raw.json?
Jason Calabrese
@jasoncalabrese
Oct 02 2016 01:24
The actual reports there don't matter, just an example of the pattern
garykidd
@garykidd
Oct 02 2016 01:40
@scottleibrand o my 401 loop is working....I had not realized that my iphone battery was fully depleted...but my g5 receiver was plugged into rig and that kept me looping successfully while iphone was out.....
but was actually surprised to see that my BG data didnt get uploaded to NS (but my pump data did). I thought the rig would've acted as an uploader...I guess that's not right...or is there an issue?
Brandon Faloona
@bfaloona
Oct 02 2016 03:14
Anyone planning to attend the Nightscout Foundation Hackathon Oct 26 in SF? http://www.nightscoutfoundation.org/hackathon16
I just learned about it and hope to attend.
Dana Lewis
@danamlewis
Oct 02 2016 03:19
I'll be there at the tail end of the afternoon for the hackathon, also will be at ddata on Thursday and the dmine innovation summit Friday if anyone else is going to either of those.
Looking forward to meeting, @bfaloona!
Mitchell Slep
@moomoobloo
Oct 02 2016 03:20
Didn't know about that either but i work right down the street.
Just registered
Dana Lewis
@danamlewis
Oct 02 2016 03:20
:+1: @moomoobloo
Scott Leibrand
@scottleibrand
Oct 02 2016 04:40
@garykidd that's part what I was working on today. might have to give up and just go ahead and implement @jasoncalabrese's method from the wiki. openxshareble is just too unreliable to do any other way afaict
Jason Calabrese
@jasoncalabrese
Oct 02 2016 04:46
@bfaloona I'll be there, but don't think it will be a competitive hackathon, hoping to get a group together to design and build some new features
Mitchell Slep
@moomoobloo
Oct 02 2016 04:48
@bfaloona @jasoncalabrese @danamlewis let me know if you want to visit Google SF office down the street
Lynne Butts
@flutelynne_twitter
Oct 02 2016 06:05
Posting now, but will check back in the morning: I rebuilt my ns website early evening because BG numbers were slow keeping up. I kept my website name and API Secret, so everything in cron, etc. is the same. Problem: I have no ns info: TBRs are not showing, neither are the pills populating information. The OpenAPS is working with the pump, but no visualization. I did regenerate another API key, and put that into the heroku settings, but do I need to do something with the Pi2 configuration? Thank you.
Dana Lewis
@danamlewis
Oct 02 2016 06:06
Did you re-enable everything on the NS side that was enabled before?
Lynne Butts
@flutelynne_twitter
Oct 02 2016 06:07
Yes. I triple-checked with the Loop website (which is working fine) to compare.
I also restarted dynos.
I also shut down the pi and restarted it. This was over an hour ago.
Thanks. I will check back in the morning.
Jim is CrushingT1D
@CrushingT1D
Oct 02 2016 07:54
anyone have experience with Novolog vs Novorapid? I recently moved from the US to Australia and i have noticed that Novorapid does not work as well for me. my resistance is much higher. In the US i found Novolog worked better than Humalog and now i am wondering what to do or try next.
i assumed they were the same but now i am skeptical
C-Ville
@C-Ville
Oct 02 2016 08:06
@CrushingT1D I also always thought it was just a difference in Trademark.
Jim is CrushingT1D
@CrushingT1D
Oct 02 2016 08:17
Me too
i can’t figure anything else that has changed in the same timeframe except the insulin. possible it’s a bad batch but kind of doubt it.
C-Ville
@C-Ville
Oct 02 2016 08:29
@CrushingT1D I am living in Europe so I only have access to NovoRapid, so I can not send you a Pen cartridge for direct comparison
Jim is CrushingT1D
@CrushingT1D
Oct 02 2016 08:30
i’ve used Novolog for the last 10 years and just switched to Novorapid 3 weeks ago. Thanks for thinking of me though :)
Matt Walker
@walker0
Oct 02 2016 10:35
@CrushingT1D I can't see anything on the information sheets that looks different, a number of the graphs are shared across the two info sheets
Tim Street
@tim2000s
Oct 02 2016 16:07
It's the same insulin @CrushingT1D , just a different name. No reason Whitley the result should be different!
garykidd
@garykidd
Oct 02 2016 17:10
So how do you change max_iob immediately with oref0-setup?
Tim Mecklem
@tmecklem
Oct 02 2016 17:21
@scottleibrand I tried out oref0-setup. Went smoothly except that it didn't move back into the openaps directory after compiling/installing shareble so the next step failed.
Scott Leibrand
@scottleibrand
Oct 02 2016 17:26
@tmecklem thx, I'll fix that. A second run should work fine, as it won't have to install again.
@garykidd there is a command-line option. Look at the help line provided when you first run it.
Tim Mecklem
@tmecklem
Oct 02 2016 17:37
Second run did work. Thanks for making the setup script. Looks like a significant amount of effort.
garykidd
@garykidd
Oct 02 2016 17:56
@scottleibrand sorry do you mean when you run the script again? I'm not seeing it
garykidd
@garykidd
Oct 02 2016 18:25
@scottleibrand got it sorry
Scott Leibrand
@scottleibrand
Oct 02 2016 18:29
Not sure why that help line is so invisible to most people. ;-)
Tim Street
@tim2000s
Oct 02 2016 18:59
Hi @scottleibrand would you mind reposting that link with the testing scenarios? I've misplaced it.
Scott Leibrand
@scottleibrand
Oct 02 2016 19:01
Which testing scenarios?
Tim Street
@tim2000s
Oct 02 2016 19:27
The missed meal stuff.
Sorry, unexpected meals.
Unannounced meal testing. :-)
Tim Street
@tim2000s
Oct 02 2016 19:49
Thanks.
Tim Street
@tim2000s
Oct 02 2016 20:24
I think I need to build on Heroku. Keep hitting Azure data limits.
Matthias
@ceben80
Oct 02 2016 20:25
Fuxxing azure limits....occured again today..web app is stopped :worried:
Rob Kresha
@rkresha
Oct 02 2016 20:31
What I'm finding with oref0-setup is that the way it is doing the compare is that it is always trying to use the ns-glucose.json, regardless of whether I'm tethered or not to the G4, which causes me a problem when I'm not connected to an AP... leading to the following in the pump-loop.log BG data is too old, or clock set incorrectly. Your CGM time is Sun Oct 02 2016 10:11:47 GMT-0500 (CDT) but your system time is Sun Oct 02 2016 10:42:00 GMT-0500 (CDT)
Rob Kresha
@rkresha
Oct 02 2016 20:37
the way my manual loop is working is if the report to pull from a tethered cgm fails (secondary rig that relies on NS), then and only then does it go to NS for glucose data, because the tethered cgm will always be newer data IMHO. Working on adjusting oref0-setup to setup something similar
that'll save some data usage on the azure side of things too
C-Ville
@C-Ville
Oct 02 2016 20:43
does anyone know the diameter of the screws that attach the edison to the Intel Breakout board? I need to attach my Edison tightly to the sparkfun board, but the screws that come with the Sparkfun Blocks are way to big
John Benjamin
@Bender1061
Oct 02 2016 20:45
@rkresha that's a good idea, I'm using the enlite on my pump and there is almost no reason to go to ns to get the glucose valves.
C-Ville
@C-Ville
Oct 02 2016 20:47
I wish I could get my BG data easily into my edison
Rob Kresha
@rkresha
Oct 02 2016 20:52
@garykidd did you ever get the raw-cgm/raw-entries.json working... I'm having similar problems... except my failure on invoke is Traceback (most recent call last): File "/usr/local/bin/openaps-report", line 4, in <module> __import__('pkg_resources').run_script('openaps==0.1.0', 'openaps-report') 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-0.1.0-py2.7.egg/EGG-INFO/scripts/openaps-report", line 82, in <module> app( ) File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.0-py2.7.egg/openaps/cli/__init__.py", line 51, in __call__ self.run(self.args) File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.0-py2.7.egg/EGG-INFO/scripts/openaps-report", line 75, in run output = app(args, self) File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.0-py2.7.egg/openaps/cli/subcommand.py", line 52, in __call__ return self.method.main(args, app) File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.0-py2.7.egg/openaps/reports/invoke.py", line 24, in main task = app.actions.commands['add'].usages.commands[device.name].method.commands[report.fields['use']] KeyError: 'oref0_glucose'
Scott Leibrand
@scottleibrand
Oct 02 2016 20:54
@rkresha your openaps is super old should be 0.1.5 not 0.1.0
Rob Kresha
@rkresha
Oct 02 2016 20:54
:o:
Scott Leibrand
@scottleibrand
Oct 02 2016 20:54
You should be able to just run step 0
Rob Kresha
@rkresha
Oct 02 2016 20:55
step 0?
oh
I've been using npm to install new versions
Scott Leibrand
@scottleibrand
Oct 02 2016 20:56
npm only does oref0. That's step 1.
Rob Kresha
@rkresha
Oct 02 2016 20:56
failure
which part of step 0
Rob Kresha
@rkresha
Oct 02 2016 20:57
quick-packages?
Scott Leibrand
@scottleibrand
Oct 02 2016 20:58
Ya, unless you have local mods in ~/src/
Rob Kresha
@rkresha
Oct 02 2016 21:00
What's the purpose of ~/src/
since $PATH refers to /usr/local/ where npm install everything
Rob Kresha
@rkresha
Oct 02 2016 21:11
so it pulls oref0 into src and then does the npm install locally, why not just do the npm install -g git://......
oh because of the log rotation piece I suppose

@scottleibrand

@rkresha your openaps is super old should be 0.1.5 not 0.1.0
Alright works now

thanks! :thumbsup:
Scott Leibrand
@scottleibrand
Oct 02 2016 21:17
Yep, and because I don't want to do npm releases every time I change something. :smile:
C-Ville
@C-Ville
Oct 02 2016 21:21
I am running at 90% on /usr on my edison, I know that it is recomended to move /usr into /home however, when I do this /usr is so crowded that I can not install cgm-remote and mongodb on my Edison at the same time. Any suggestions what else I can do to get rid of some data in /usr?
Rob Kresha
@rkresha
Oct 02 2016 21:24
hmm, I pull my mod to oref0 via npm and do nothing with releases, just point it to my branch on github
:dunno
Rob Kresha
@rkresha
Oct 02 2016 21:30
now that @scottleibrand helped me fix the monitor-cgm and raw-cgm/raw-entries.json which will actually pulling glucose data data over the tether, we'll see if it works offline and gets me past my earlier situation of it not working when not connected to NS
Scott Leibrand
@scottleibrand
Oct 02 2016 21:41
@C-Ville I don't move all of /usr over. I do /var/log first.
C-Ville
@C-Ville
Oct 02 2016 21:46
thanks, I will try that, 90% is getting to close for my taste now
John Benjamin
@Bender1061
Oct 02 2016 21:48
since the Edison needs to connect to a time server to get the time and date. and in order for OpenAPS to run (correctly) we need the time, has anyone done anything for getting the pump time and use that to set the time on the edison? I'm looking at what I would need to do to run my rig while on a cruise in a few months. I'm not needing to pull BG from nightscout since i'm using enlite, but i'm thinking if i ever have to shut down for whatever reason during the trip my loop would stop working.
C-Ville
@C-Ville
Oct 02 2016 22:03
you could use openaps use pump read_clock
Eric
@ecc1
Oct 02 2016 22:05
^^ combined with the "date" command's --set and +FORMAT options
C-Ville
@C-Ville
Oct 02 2016 22:06
@ecc1 I was just looking for the date command options
@Bender1061 I like your idea, unfortunately I do not really know how to get a screen output in a textfile and then read that textfile as an argument for the date command
John Benjamin
@Bender1061
Oct 02 2016 22:14
might have to be a python command, take the json file that we make with a report invoke getting the pump time, then tearing apart the string, and use that to set the date and time. probably need to check to see if the ntp server is running, and do this afterwords, and make this a script that runs only once when the Edison first starts.
oh ideas now
ok just found oref0/bin/clockset.sh it was written for the pi, but that's a good place to start
Jason Calabrese
@jasoncalabrese
Oct 02 2016 23:49
That should work for the Edison