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

27th
Feb 2017
chookchook1
@chookchook1
Feb 27 2017 01:39
@dlittle2 it means you don't have a trend arrow for your sensor at the moment.
peterfulvi
@peterfulvi
Feb 27 2017 02:40
Feb 26 21:38:17 localhost root: Uploaded upload/ns-status.json to https://petersopenaps.herokuapp.com
I saw this meesage in the cron jobs but i see no output on my ns
image.png
Michael Maniscalco
@gtscalco_twitter
Feb 27 2017 03:03
@peterfulvi I'm new to this but is your DB connected properly to your webapp?
dmdfreak
@dmdfreak
Feb 27 2017 05:00
Does anyone happen to know how to get back data in xdrip+ into my newly setup nightscout site? I've had the site a week and I still only have it from the start of the site. No old data.
chrisharris751
@chrisharris751
Feb 27 2017 08:56

I've drawn up a case for the EB utilising a 18650 battery. Check it out.

https://github.com/chrisharris751/OAPSEB18650/blob/master/OpenAPS%20Explorer%20Case%200.4.stl

I don't have a 3d printer, so its untested.

Anyone want to print it out for me as a test??

Paul Dickens
@thebookins
Feb 27 2017 10:13
Not sure if this is the right place to ask, please redirect as needed! When setting a temp target in care portal via nightscout using my phone, the time show as 'now' is correct. However, when the pop up box displays the details for confirmation the time is out by exactly 1 hour (its behind). This also happens for the start time if I set a temp target in advance. Any suggestions on what I need to change? It works perfectly if I do it through the laptop.
chrisharris751
@chrisharris751
Feb 27 2017 10:56
How do I check up on what AMA is doing?
Chris Oattes
@cjo20
Feb 27 2017 11:05
What do you mean by "check up on what AMA is doing?"
chrisharris751
@chrisharris751
Feb 27 2017 11:10
How soon it high temps after eating.
Chris Oattes
@cjo20
Feb 27 2017 11:28
As I understand it, it high temps when the 3 purple lines don't think you'll go low between now and DIA/2
tepidjuice
@tepidjuice
Feb 27 2017 14:01
Hi. Does anyone know how to get the entries over a specific date range from nightscout? I'm trying https://my-url/api/v1/entries/sgv?find[dateString][$gte]=2017-02-25&find[dateString][$lte]=2017-02-26 but I only get a few entries not all entries which are in that range.
tepidjuice
@tepidjuice
Feb 27 2017 14:09
nvm I worked it out missing &count[200].
dmdfreak
@dmdfreak
Feb 27 2017 14:28
@tepidjuice do you have any idea how to get all the info from xdrip+ to import into nightscout? I can only get data from the date I started the site. Can't get the rest to import over.
ejoe132
@ejoe132
Feb 27 2017 14:46
What or how do most find optimal setting for max iob? Originally set at max bolus I had, which was 10, found the system was driving me low, so now at 5, but still seeing some of this low. Seems like five would be ok so it could try to regulate highs but seems with basal since its going in slower it calculate iob lower than Id imagine, any tips?
Scorponok
@Scorponok
Feb 27 2017 14:47
@ejoe132 AFAIK max iob is really just a "how comfortable are you with openaps controlling your basals" setting - we kept it at 0 for our first couple of weeks, so it would only low-temp us, not high-temp us
Then gradually increased it after that - if it's doing the wrong thing that's going to be caused by something else, not the max iob
If your basal is being pushed up and driving you low, click the openaps pill and see why it's doing what it's doing
Chris Oattes
@cjo20
Feb 27 2017 14:54
If it's driving you low, it probably means some of your settings (DIA, isf or i:c ratio) are wrong
Tim Street
@tim2000s
Feb 27 2017 14:55
Agreed.
Chris Oattes
@cjo20
Feb 27 2017 14:55
max_iob is just a limit on how low it can drive you if your settings are wrong or if you get bad CGM data
ejoe132
@ejoe132
Feb 27 2017 15:02
Ok, Im constantly making changes to pump settings(based off 7 days of cgm data) but just getting low each morning after being high overnight. Thanks for clearing this up for me
jcorbett80
@jcorbett80
Feb 27 2017 15:09
I flashed Edison the other day. have been working on it a little at a time. Today i noticed the "Failed to start hostname service" I reflashed and still have same fail.
dan lightner
@danimaniac
Feb 27 2017 15:10
@chrisharris751 I like the case with the top and bottom secured with screws. Pretty cool...be sure to PR it into the docs.
Scott Leibrand
@scottleibrand
Feb 27 2017 15:13
@ejoe132 you might want to look at autotune to see what it thinks needs adjusting.
live4sw
@live4sw
Feb 27 2017 15:14
Has anyone polled the OpenAPS community on which meters they use? I've used the Freestyle Lite for at least 8 years now, but after reading this great article (https://www.ncbi.nlm.nih.gov/pubmed/27697848) and doing about 15 head-to-head tests where the Lite was systematically anywhere from 0-16 points lower than a Contour Next EZ, it makes me wonder whether I have been unknowingly targeting higher BGs than I thought for years. It certainly puts my high 6/low 7 A1c's into perspective given that my Freestyle Lite and Dexcom (calibrated from Freestyle) averages would suggest 6.0
Scott Leibrand
@scottleibrand
Feb 27 2017 15:17
@chrisharris751 @cjo20 when the "COB" predBG line (usually the middle one) is predicted to be above target, from 30m from now until DIA, then AMA can start high-temping as required to bring eventualBG down to target. If you want to get into more details than that, you can either look at the determine-basal code, or watch the pump-loop.log and NS after a meal.
@live4sw you should read Chris Hannemann's article on "a craftsman blames his tools" and related #samedrop stuff.
live4sw
@live4sw
Feb 27 2017 15:22
Yep, when I read his article last year, I was skeptical that it wasn't rigorous enough - plus his benchmark was defined by the meters themselves rather than by a lab test - but this recent journal article definitely supports his implied finding of the Contour being the most accurate.
It really raises some major issues for the community, since AP is fantastic for lowering variability and targeting a user-set definition of euglycemia. But if large parts of the community are targeting the same number, and those numbers actually mean very different things, it means that we're often not fully aware of the risk/control tradeoff we're making. I mean, it's good that the Freestyle seems to err on the side of conservatism, but I certainly didn't know that until recently.
Chris Oattes
@cjo20
Feb 27 2017 15:27
live4sw: take in to account that BG->a1c conversions are inherently inaccurate
There's a bunch of stuff on an individual level that can skew the conversion
live4sw
@live4sw
Feb 27 2017 15:31
make sense
Chris Oattes
@cjo20
Feb 27 2017 15:33
on average, they're about right, but you might just be one of the people that have higher a1cs than the formulas predict
(to go from 6.0 to 7.0, your meter would have to be reading 126mg/dl or 7.0mmol/l when you are really 154mg/dl or 8.6mmol/l
Chris' data suggests that there's only a 3 mg/dl difference between contour and freestyle at 150mg/dl
live4sw
@live4sw
Feb 27 2017 15:37
Yep, one doctor suggested I might be a "slow glycosolator". But if the Lite was systematically low by 10 points (and i didn't have any major highs during that time - as #'s went up, the deviation was greater), that's about 0.4 A1c points - a material portion of the perceived difference. Chris's results along with the lower tested accuracy from the Freestyle in the journal article suggest there could be a real issue there.
Chris Oattes
@cjo20
Feb 27 2017 15:39
The article text isn't freely available
Eric
@ecc1
Feb 27 2017 15:45

I just found a bug in decocare's decoding of the carb ratio field in Bolus Wizard pump history records. Actually I found it in my Go code, then saw that I'd "inherited" it from decocare, as did @ps2 's minimedkit code. On newer pumps, this code:

     carb_ratio = (((self.body[2] & 0x07) << 8) + self.body[3]) / 10.0

misses a bit of the value. It should be:

    carb_ratio = (((self.body[2] & 0x0F) << 8) + self.body[3]) / 10.0

The important question is whether anything in openaps or Loop depends on this?

Pete Schwamb
@ps2
Feb 27 2017 15:52
Loop doesn't use bolus wizard entries
Eric
@ecc1
Feb 27 2017 15:53
Also, this may only occur when the carb units are Exchanges (I saw it in a value that should have decoded to 2500 ...)
Chris Oattes
@cjo20
Feb 27 2017 15:58
it'll only make a difference in values over 204.8 (vs 409.6)
@ecc1 You've got a carb ratio of 1u:250g?
(I'm assuming that's how that value is used)
Eric
@ecc1
Feb 27 2017 16:10
When carb units are set to Exchanges, the carb ratio value is represented as 100x units/exchange. So if your ratio is 2.5 units per exchange, the value stored is 2500.
You're right that when units are Grams, the value (represented as 10x grams/unit) would have to be very high.
Chris Oattes
@cjo20
Feb 27 2017 16:16
Might be worth testing setting a carb ratio of 1u:206g and seeing if that decodes properly (eith / without the change)
Eric
@ecc1
Feb 27 2017 16:22
My 522 won't let me set a ratio greater than 150g/U. Anyone have a x23 handy to check there too?
Chris Oattes
@cjo20
Feb 27 2017 16:30
554 wraps to 200g/u
I don't know the bolus wizard code off the top of my head, but it may need a switch based on the type of units, to make sure it doesn't decode badly
chrisharris751
@chrisharris751
Feb 27 2017 17:02
@danimaniac I certainly will. I just need to get it printed and test it first.
peterfulvi
@peterfulvi
Feb 27 2017 17:15
https://myopenaps.herokuapp.com does this look like the right url to direct my rig to this is were my app aka nightscout is. Somehow its still blank but it says its uploading to it. I also configured the framework to be node.js which i do not know if that has anything to do with it This is what my crontab says:
Feb 27 11:59:07 localhost root: Unable to upload to https://petersopenaps.herokuapp.com
Feb 27 11:59:10 localhost root: Uploaded upload/ns-status.json to https://petersopenaps.herokuapp.com
this is the app logson my heroku site


2017-02-27T17:11:23.327003+00:00 app[web.1]: GRANTED:  74.92.27.60 api-secret api:treatments:read

2017-02-27T17:11:31.584538+00:00 app[web.1]: tick Mon Feb 27 2017 17:11:31 GMT+0000 (UTC)

2017-02-27T17:11:32.588827+00:00 app[web.1]: searching tretaments q { find: { created_at: { '$gte': '2017-02-19T17:11:32.587Z' } },

2017-02-27T17:11:32.588838+00:00 app[web.1]:   sort: { created_at: 1 } }

2017-02-27T17:11:32.600173+00:00 app[web.1]: Load Complete:

2017-02-27T17:11:32.600177+00:00 app[web.1]:      profiles:1, devicestatus:1

2017-02-27T17:11:32.601357+00:00 app[web.1]: all buckets are empty

2017-02-27T17:11:32.602598+00:00 app[web.1]: delta calculation indicates no new data is present

2017-02-27T17:11:32.602300+00:00 app[web.1]: WS: running websocket.update

2017-02-27T17:12:06.779813+00:00 heroku[router]: at=info method=GET path="/api/v1/entries/sgv.json?find[date][$gte]=1488129126029&count=1000" host=petersopenaps.herokuapp.com request_id=3f2eb667-2a7b-4174-956e-38d4fbf3bd71 fwd="74.92.27.60" dyno=web.1 connect=0ms service=10ms status=200 bytes=233

2017-02-27T17:12:06.771887+00:00 app[web.1]: GRANTED:  74.92.27.60 no-token api:entries:read default

2017-02-27T17:12:11.544887+00:00 heroku[router]: at=info method=GET path="/api/v1/treatments.json?find[created_at][$gte]=2017-02-27T06:12-0500&find[eventType][$regex]=Target" host=petersopenaps.herokuapp.com request_id=25d1b90e-a535-458b-921e-b96afaf110f4 fwd="74.92.27.60" dyno=web.1 connect=1ms service=16ms status=200 bytes=225

2017-02-27T17:12:11.529469+00:00 app[web.1]: GRANTED:  74.92.27.60 no-token api:treatments:read default

2017-02-27T17:12:18.335747+00:00 heroku[router]: at=info method=GET path="/api/v1/treatments.json?find[carbs][$exists]=true" host=petersopenaps.herokuapp.com request_id=dc98e154-8cdd-45dd-8417-f401d7622365 fwd="74.92.27.60" dyno=web.1 connect=0ms service=8ms status=200 bytes=225

2017-02-27T17:12:18.328489+00:00 app[web.1]: GRANTED:  74.92.27.60 no-token api:treatments:read default

2017-02-27T17:12:30.764409+00:00 app[web.1]: GRANTED:  74.92.27.60 no-token api:entries:read default

2017-02-27T17:12:30.785077+00:00 heroku[router]: at=info method=GET path="/api/v1/entries.json?count=300" host=petersopenaps.herokuapp.com request_id=786490d3-c4df-4f57-8497-91cce552877a fwd="74.92.27.60" dyno=web.1 connect=1ms service=10ms status=200 bytes=233

2017-02-27T17:12:31.588556+00:00 app[web.1]: tick Mon Feb 27 2017 17:12:31 GMT+0000 (UTC)

2017-02-27T17:12:32.604227+00:00 app[web.1]: searching tretaments q { find: { created_at: { '$gte': '2017-02-19T17:12:32.599Z' } },

2017-02-27T17:12:32.604236+00:00 app[web.1]:   sort: { created_at: 1 } }

2017-02-27T17:12:32.610176+00:00 app[web.1]: Load Complete:

2017-02-27T17:12:32.610179+00:00 app[web.1]:      profiles:1, devicestatus:1

2017-02-27T17:12:32.610866+00:00 app[web.1]: all buckets are empty

2017-02-27T17:12:32.611684+00:00 app[web.1]: WS: running websocket.update

2017-02-27T17:12:32.611833+00:00 app[web.1]: delta calculation indicates no new data is present

2017-02-27T17:12:37.981781+00:00 app[web.1]: GRANTED:  74.92.27.60 api-secret api:entries:read

2017-02-27T17:12:37.981796+00:00 app[web.1]: GRANTED:  74.92.27.60 api-secret api:entries:create
Trying to connet the dots here
katie disimone
@Kdisimone
Feb 27 2017 17:38
@peterfulvi It doesn’t look like you have all your variables configured correctly
@peterfulvi how are you thinking you are getting your BGs into NS (because they aren’t)…bridge from dexcom? x-drip? You are missing openaps and pump in your ENABLE line. You don’t have your BASE_URL filled in. You don’t have DEVICESTATUS_ADVANCED set to true. You should have SHOW_FORECAST with openaps (not just ar2).
katie disimone
@Kdisimone
Feb 27 2017 17:44
@peterfulvi basically you are missing a lot of the variables in your heroku setup to make openaps work. YOu need to address those first
katie disimone
@Kdisimone
Feb 27 2017 18:00
@peterfulvi also, remove commas from the values in your ENABLE line. they should only be separated by spaces.
Sacha M
@coolestkidsever
Feb 27 2017 18:25
Hi all, I have the Pump low battery alarm going on NS, and I snooze it for as long as I can (120min), but essentially I'm awake every two hours to snooze the pump low battery alarm. I assume I can just delete the PUMP_URGENT_BATT_U Field from Heroku to stop this from happening? Or is there a better way? What do others do?
katie disimone
@Kdisimone
Feb 27 2017 18:31
Change the value to a lower value that you’d actually want to NOT ignore, or yes…you can turn that field off/delete it. I lowered mine to 1.3
I have warn set to 1.3 and urgent is deleted. Urgent alarms have shorter snoozes, so i dislike those overall.
Sacha M
@coolestkidsever
Feb 27 2017 18:32
What does the 1.3 represent? (not time?)
I think I've only got the Urgent field actually, whats the diff between PUMP_URGENT_BATT_U and PUMP_URGENT_BATT_V
peterfulvi
@peterfulvi
Feb 27 2017 18:37
@Kdisimone So I basically forked over cgm remote from github and deployed from there setup a free heroku account. So basically i should start over. What step am i missing im on an enlite sensor and was told it will read the cgm data from my rig do i need another app on my phone?
katie disimone
@Kdisimone
Feb 27 2017 18:39
@coolestkidsever U is for battery percentage, V is for voltage. I use PUMP_WARN_BATT_V at 1.3 for OpenAPS (since it reads pump battery in voltage). So when pump battery gets to 1.3 volts...
Sacha M
@coolestkidsever
Feb 27 2017 18:39
ok, thanks :-)
katie disimone
@Kdisimone
Feb 27 2017 18:40
@peterfulvi go into your heroku settings and configure vars…you need to get them properly updated. I’ve given you some things to change. Do you know how to get into your config vars?
sdneufer
@sdneufer
Feb 27 2017 18:42
ns-w-ed.jpg
I switched from a pi3 to the edison on Saturday. It is running, my basal pill works, however the openaps and pump pills are not correct. What do I look at to figure this out?
I set eating soon to show the temp basal working.
katie disimone
@Kdisimone
Feb 27 2017 18:44
@sdneufer do you have DEVICESTATUS_ADVANCED set to true? And check for typos carefully. Usually when it toggles between showing pump and openaps info…there’s a typo in that line or it just plain isn’t there.
sdneufer
@sdneufer
Feb 27 2017 18:49
Yes
All the pills were working when I was using the pi.
peterfulvi
@peterfulvi
Feb 27 2017 18:55
@Kdisimone yes i saw that in the setup
I'll do it later im working now. Thank You
@Kdisimone I dont see Device statu and show forcast
peterfulvi
@peterfulvi
Feb 27 2017 19:04
I found it on mlab and its set to false' @Kdisimone
im trying to turn it to true
peterfulvi
@peterfulvi
Feb 27 2017 19:11
@Kdisimone can i add them amanually?
katie disimone
@Kdisimone
Feb 27 2017 19:13

@peterfulvi they are not in mlab. You need to go to your Heroku settings, config vars.

Yes, you can add them manually if they aren’t already there. All the ENABLE vars need to be on a single line, separated by spaces and not commas

gwynethrocks_·_Settings___Heroku.jpg
Screenshot_2_27_17__11_16_AM.jpg
Here’s an idea of how you may want to setup your config vars in heroku for openaps use specifically…there are obviously more lines than just these, but these are the ones that are specifically relevant to openaps use
peterfulvi
@peterfulvi
Feb 27 2017 19:18
all these have to go in the ENABLE ver first then i can add the right?
katie disimone
@Kdisimone
Feb 27 2017 19:22
@peterfulvi sorry, but that question didn’t make sense to me. Not sure how to answer it
gregtcm
@gregtcm
Feb 27 2017 19:33
@live4sw @scottleibrand @cjo20 I've recently been made aware of the meter issue, too, and use the OneTouch Verio. I ran across someone who did their own test with mutiple Freestyle Lites and OneTouch meters and compared with lab blood tests (not Chris Hannemann's article). His results were the Freestyle lites measured 10-20 points lower than the lab tests and the OneTouch meters measured 10-20 points higher than the lab tests. I wish I could find that web page again! At my last endo appointment, my meter tested at exactly 20 points higher and amazingly enough, my Dexcom estimated a1c was also exactly 20 points higher than my actual a1c. This was essential info as my measured lows of 70 are probably more like 50! I've since adjusted my target, high, and low up 20 points and things are much, much better! I know everyone's mileage might differ, but I think people really need to be made aware of this.
peterfulvi
@peterfulvi
Feb 27 2017 19:49
@Kdisimone All those fields you posted above do they have to go on the enable var first,or do i just put them in? Sorry this is all a learning curb even the terminology
katie disimone
@Kdisimone
Feb 27 2017 19:57
@peterfulvi let’s do this on chat maybe so that you can share screenshots without accidentally showing any private info you don’t want out there...
tepidjuice
@tepidjuice
Feb 27 2017 21:15
@dmdfreak No sorry I don't. My guess is that you'd have to upload it to mlab somehow... maybe someone on the xdrip+ or cgm-uploader chat knows?
dmdfreak
@dmdfreak
Feb 27 2017 21:17
@tepidjuice thank you. OK I'll check there
jmcrawford
@jmcrawford
Feb 27 2017 21:18
@dmdfreak what do the xdrip+ records look like? Might be possible to write a conversion to make it look the CGM data that gets uploaded via ns-loop, and upload it that way.
dmdfreak
@dmdfreak
Feb 27 2017 21:21
I'm not sure. I'm still very new. I'll try to look at it and check into ns-loop.
jmcrawford
@jmcrawford
Feb 27 2017 21:22
Have a look at cgm/glucose.json
dmdfreak
@dmdfreak
Feb 27 2017 21:23
OK thank you @jmcrawford
PopcornPam
@PopcornPam_twitter
Feb 27 2017 22:43
Just getting started here, and have a pump question. I have 2 minimed paradigm pumps, and am trying to figure out which one would be better to use (if there is a difference). Planning to donate the other one to a clinic in Haiti that needs a way to read enlite sensor data, so want to figure out which one I should keep
jmcrawford
@jmcrawford
Feb 27 2017 22:45
What models do you have?
feraridurango
@feraridurango
Feb 27 2017 22:53
where do I find my mmeowlink full port address??
Tim Street
@tim2000s
Feb 27 2017 23:10
@sdneufer Are your Nightscout settings correct in NS.ini ? Your issues look like either you key or URL are incorrect for OpenAPS. You can check if your data is being uploaded by using "tail -f /var/log/syslog | grep upload"
If the data can't be uploaded you'll see a message saying so and you should go back and check.
Scott Leibrand
@scottleibrand
Feb 27 2017 23:23
@popcornpam the docs also detail how to check for PC Connect to see if they're loopable. If you end up with two loopable pumps, lots of people have newer non-loopable ones they'd love to trade you, so you can donate a newer one in better condition (with full Enlite support), assuming the clinic only needs read-only access.
PopcornPam
@PopcornPam_twitter
Feb 27 2017 23:28
Ah, that would be great! One is 722 and the other is 723
gregtcm
@gregtcm
Feb 27 2017 23:31
@PopcornPam_twitter I just started a chat with you about your pumps.
PopcornPam
@PopcornPam_twitter
Feb 27 2017 23:31
I have been using tandem for the past three, years, so don't have a ton of supplies for medtronic. Trying to figure out if I can access the menu's without opening and using a resevoir (it keeps getting stuck on a screen demanding I load the resevoir
Scott Leibrand
@scottleibrand
Feb 27 2017 23:32
You can stick your pinky or an eraser down the reservoir hole to stop the plunger and make it "detect" a reservoir.
PopcornPam
@PopcornPam_twitter
Feb 27 2017 23:40
So the 723 has firmware 2.5, but it also has on option for pc connect
katie disimone
@Kdisimone
Feb 27 2017 23:52
Not going to work for looping @PopcornPam_twitter