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

25th
Oct 2016
Sgoldberg89
@Sgoldberg89
Oct 25 2016 00:21
Working through the setup... When I saw the nightscout app on someone's phone, it had two lines. I have merged my files on Github, disconnected and reconnected on Azure, but still looks identical. How do I know if I've done it properly?
Dana Lewis
@danamlewis
Oct 25 2016 00:30
“two lines” - do you mean two purple lines in prediction, or something else like two BG lines?
Sgoldberg89
@Sgoldberg89
Oct 25 2016 00:42
yes, there were two sets of trends lines. Though to be honest I don't know exactly what the difference was
Dana Lewis
@danamlewis
Oct 25 2016 00:45
The AR2 forecast would’ve been light teal. "AR2 generates alarms based on forecasted values. AR2 is enabled by default if no alarm thresholds are set or if the ALARM_TYPES variable includes predict. You can adjust AR2 behavior with the following extended settings variables:
AR2_USE_RAW (to forecast using rawbg values when standard values don’t trigger an alarm; default is false)
AR2_CONE_FACTOR (to adjust size of cone, use 0 for a single line; default is 2)"
Sgoldberg89
@Sgoldberg89
Oct 25 2016 00:49
Thanks Dana. So confused but really enjoying this process! (I don't really even know where I should enter the information you provided above) :(
Scott Leibrand
@scottleibrand
Oct 25 2016 00:50
you probably don't need to
what are you missing? the cyan ar2 lines, or the purple prediction line(s), or something else?
Dana Lewis
@danamlewis
Oct 25 2016 00:50
@Sgoldberg89 I wouldn’t worry about the forecasting in NS right now. I would focus on make sure you’re getting BG data, etc. at first, then the requisite OpenAPS data after you get a loop running
Sgoldberg89
@Sgoldberg89
Oct 25 2016 00:51
@danamlewis I was able to get nightscout setup actually. Have BG data on my phone and online. I'm also fairly sure I was able to update to the newest "dev" version.
Dana Lewis
@danamlewis
Oct 25 2016 00:52
cool
Sgoldberg89
@Sgoldberg89
Oct 25 2016 00:52
Now I'm trying to figure out how to make the stick and the pump communicate
and incorporate Nightscout
Dana Lewis
@danamlewis
Oct 25 2016 00:52
http://nightscout.info/ is a great reference site for NS if you want to poke around there and learn about all the options and goodies it has
Sgoldberg89
@Sgoldberg89
Oct 25 2016 00:53
is it bad to run this multiple times: curl -s https://raw.githubusercontent.com/openaps/docs/master/scripts/quick-packages.sh | bash -
Scott Leibrand
@scottleibrand
Oct 25 2016 00:53
nope, that's fine
Dana Lewis
@danamlewis
Oct 25 2016 00:53
no, it won’t hurt anything
ditto for re-running the setup script multiple times
Sgoldberg89
@Sgoldberg89
Oct 25 2016 00:54
Ok, I did it before, but the pump didn't have a battery in it and the Carelink stick wasn't in the device, so it would be good to start over I'm assmuming
assuming*
Now they're next to eachother
Scott Leibrand
@scottleibrand
Oct 25 2016 00:55
neither of those is necessary to run oref0-setup, so it might Just Work now that you've fixed them :)
Dana Lewis
@danamlewis
Oct 25 2016 00:55
I do recommend going ahead and putting a battery in the pump, and getting out of suspend or prime, before it trips you up later
Sgoldberg89
@Sgoldberg89
Oct 25 2016 00:55
@scottleibrand Awesome... But how do I know if it is? (I'm laughing at myself, don't worry)
Dana Lewis
@danamlewis
Oct 25 2016 00:55
checking that the pump has a basal profile, your settings in it, and has been set to absolute rather than % temp
Sgoldberg89
@Sgoldberg89
Oct 25 2016 00:56
Yes, I hooked it up to a reservoir of water for now so it's primed and ready to go
Scott Leibrand
@scottleibrand
Oct 25 2016 00:56
in the docs it mentions how to do a tail -F of the pump-loop.log
that will show you what the rig is doing trying to talk to the pump
Sgoldberg89
@Sgoldberg89
Oct 25 2016 00:56
sweet, i'll look for that
Scott Leibrand
@scottleibrand
Oct 25 2016 00:56
it's on the oref0-setup page
Sgoldberg89
@Sgoldberg89
Oct 25 2016 00:57
pi@raspberrypi:~ $ tail -F /var/log/openaps/pump-loop.log
tail: cannot open ‘/var/log/openaps/pump-loop.log’ for reading: No such file or directory
Dana Lewis
@danamlewis
Oct 25 2016 00:57
and if it doesn’t work, run everything again ;)
Sgoldberg89
@Sgoldberg89
Oct 25 2016 00:57
Bingo
I'll do that
Also, when I entered the SN for the pump, it doesn't include the letters that preceded the numbers, is that correct?
Scott Leibrand
@scottleibrand
Oct 25 2016 01:01
correct, numbers only
Sgoldberg89
@Sgoldberg89
Oct 25 2016 01:03
"Do you need advanced features?" <- What does this include?
Dana Lewis
@danamlewis
Oct 25 2016 01:04
If you’re just starting, you probably want the basic loop before you add any additional features
Sgoldberg89
@Sgoldberg89
Oct 25 2016 01:04
I suppose... ;)
"Schedule openaps in cron?"
Dana Lewis
@danamlewis
Oct 25 2016 01:08
do you want to automate your closed loop? or run it in manual mode?
Sgoldberg89
@Sgoldberg89
Oct 25 2016 01:08
I want to be able to control it from the pebble watch, eventually.
Dana Lewis
@danamlewis
Oct 25 2016 01:09
do you want an automated closed loop, or an open loop? that’s what it’s asking, whether you want it to be automated and scheduled to run constantly, or for you to force it to run manually
Sgoldberg89
@Sgoldberg89
Oct 25 2016 01:09
Does that answer your question?
automatically is fine
Dana Lewis
@danamlewis
Oct 25 2016 01:10
don’t tell me, tell the setup script ;)
Sgoldberg89
@Sgoldberg89
Oct 25 2016 01:10
So should the answer be yes?
Sorry!
Dana Lewis
@danamlewis
Oct 25 2016 01:11
Does this note I just added to the docs help? :) openaps/docs@523325b
(and if not, would love for you and anyone else working through to PR in additional notes)
Sgoldberg89
@Sgoldberg89
Oct 25 2016 01:16
That IS helpful. One thing I would love to know (if you're honestly asking for feedback), is if I would need to rerun the entire thing from the top of the page to modify one of the options, or if there is a way to modify one specific option in this setup process without changing everything else.
Dana Lewis
@danamlewis
Oct 25 2016 01:16
You’d re-run the setup script, which is step 2
Sgoldberg89
@Sgoldberg89
Oct 25 2016 01:17
ok.
Thank you so much.
Starting pump-loop at Mon 24 Oct 18:12:29 MST 2016: <- frozen on this line, is that normal? I looked on the website for any instructions for settings that should be entered into the pump, but don't see anything.
Paul Martin
@Spazholio
Oct 25 2016 01:21
Just curious - is there any movement on merging the whole “can’t set temps for higher than 6.3u/hr” issue into master?
Dana Lewis
@danamlewis
Oct 25 2016 01:23
@Sgoldberg89 do you have basal in the pump, BG targets, ISF, etc. set?
There was someone who was going to start a troubleshooting page, but can’t remember who :) but if this rings a bell, you probably have stuff to add to this one!
diabeticgonewild
@diabeticgonewild
Oct 25 2016 01:29
I'm asking some "craft groups specifically for the Brother ScanNCut2" (a die cutting machine that cuts fabric to the shape of ANY pattern you program into the machine) if those expensive GriffGrips can basically be created by DIY methods on this machine.
Like designs like this: https://imgur.com/a/MQQ6U
The tape that GriffGrips and others use is a stretchy cotton/elastic fabric with an acrylic based adhesive, more commonly known as kinesiology tape.
Sgoldberg89
@Sgoldberg89
Oct 25 2016 01:30
@danamlewis I think the issue (or at least one of them) is with Nightscout. OpenAPS = Unkown and Pump = Nothing
diabeticgonewild
@diabeticgonewild
Oct 25 2016 01:30
RockTape H2O 4" with a hole cut in it for the transmitter is the best.
Dana Lewis
@danamlewis
Oct 25 2016 01:31
@Sgoldberg89 NS wouldn’t be causing issues for your loop, but if you’ve got a frozen loop and it’s not showing in NS, you’ve got a double sign of it not working
@Sgoldberg89 is it still stuck on that line? tail again? and if it’s still not working, try setup script again… (step 2)
diabeticgonewild
@diabeticgonewild
Oct 25 2016 01:31
But you could get one of those ScanNCut2 machines and even copy the GriffGrips designs, and pay for itself over time, using RockTape Kinesiology Tape...
So anxiously awaiting an answer...
Sgoldberg89
@Sgoldberg89
Oct 25 2016 01:32
NS show's BG info. Line is the same. Will start from #2 @danamlewis
diabeticgonewild
@diabeticgonewild
Oct 25 2016 01:36
But I feel good considering I had a very minor surgical procedure this morning, which is good because my body sucks at regulating itself. And general anesthesia is definitely harder on my body than most. So yay. But my mind is not in the right place right now...
Going to bed.
Good night.
Sgoldberg89
@Sgoldberg89
Oct 25 2016 01:39
@danamlewis It keeps trying different "ports" and retrying... is this normal?
Dana Lewis
@danamlewis
Oct 25 2016 01:40
@sgoldberg89 what line are you expecting to be different?
Sgoldberg89
@Sgoldberg89
Oct 25 2016 01:41
This is an example of what it's doing over and over: return self.query(ProductInfo)
File "/usr/local/lib/python2.7/dist-packages/decocare/stick.py", line 509, in query
return self.process( )
File "/usr/local/lib/python2.7/dist-packages/decocare/stick.py", line 490, in process
raw = self.send_force_read( )
File "/usr/local/lib/python2.7/dist-packages/decocare/stick.py", line 606, in send_force_read
self.link.write(reader.format( ))
File "/usr/local/lib/python2.7/dist-packages/decocare/link.py", line 47, in write
r = self.serial.write( string )
File "/usr/local/lib/python2.7/dist-packages/serial/serialposix.py", line 517, in write
raise portNotOpenError
serial.serialutil.SerialException: Attempting to use a port that is not open
Error, retrying
Dana Lewis
@danamlewis
Oct 25 2016 01:41
Are you using carelink or mmeowlink?
Sgoldberg89
@Sgoldberg89
Oct 25 2016 01:42
carelink rn. Going to use the TI stick eventually
Dana Lewis
@danamlewis
Oct 25 2016 01:42
Is carelink plugged in? If so unplug replug and see if that fixes
Sgoldberg89
@Sgoldberg89
Oct 25 2016 01:50
:( nope
maybe I
I'll restart it
diabeticgonewild
@diabeticgonewild
Oct 25 2016 02:03
Can't go to sleep yet...
Anyway, does anyone know the dude in the mmeowlink chat that was trying to get NS locally on the Edison via the explorer board and an SD card?!
Please reply if you know who it is. He said he was going to report back if he was successful.
Sgoldberg89
@Sgoldberg89
Oct 25 2016 02:09
@danamlewis Exciting news! The pump and the device are communicating!
Martin Haeberli
@mhaeberli
Oct 25 2016 02:09
:+1: @Sgoldberg89
Sgoldberg89
@Sgoldberg89
Oct 25 2016 02:09
But, NS still does not show any info
Dana Lewis
@danamlewis
Oct 25 2016 02:09
Progress!
Sgoldberg89
@Sgoldberg89
Oct 25 2016 02:09
Wait
Dana Lewis
@danamlewis
Oct 25 2016 02:09
Did it set a temp?
Sgoldberg89
@Sgoldberg89
Oct 25 2016 02:10
by any info, I mean specifically about the pump
yes!
Dana Lewis
@danamlewis
Oct 25 2016 02:10
Woohoo! :confetti_ball:
Sgoldberg89
@Sgoldberg89
Oct 25 2016 02:10
WOOOO
Dana Lewis
@danamlewis
Oct 25 2016 02:10
You're looping :)
Sgoldberg89
@Sgoldberg89
Oct 25 2016 02:10
With water... but I'm... PUMPED! (pun intended)
Dana Lewis
@danamlewis
Oct 25 2016 02:11
Ok. Next step is NS troubleshooting. Did you enable stuff in NS?
Sgoldberg89
@Sgoldberg89
Oct 25 2016 02:11
Yes... I think so
How would I know?
Dana Lewis
@danamlewis
Oct 25 2016 02:12
You have OpenAPS pill showing?
Sgoldberg89
@Sgoldberg89
Oct 25 2016 02:12
Yes
diabeticgonewild
@diabeticgonewild
Oct 25 2016 02:12
I got an answer on a Facebook group about being able to cut "GriffGrips" designs (to keep the glucose sensor retained on the arm, and to extend sensor life) via the Brother ScanNCut2 machine (although I don't fully understand the answer!!!)
Sgoldberg89
@Sgoldberg89
Oct 25 2016 02:12
it says "unknown"
Dana Lewis
@danamlewis
Oct 25 2016 02:13
Where are you getting BG from, NS or receiver?
diabeticgonewild
@diabeticgonewild
Oct 25 2016 02:14
The machine is expensive, but within a year, it certainly would pay for itself, in lieu of buying "GriffGrips"
Brother CM350 ScanNCut 2 Home & Hobby Cutting Machine https://www.amazon.com/dp/B014T1CI1E/ref=cm_sw_r_cp_api_e.RdybMEV0RXN
Dana Lewis
@danamlewis
Oct 25 2016 02:14
Cat your NS.ini to see what's there and confirm your url is right, no trailing slash, etc
Sgoldberg89
@Sgoldberg89
Oct 25 2016 02:16
I'm getting BG from G4 Share
bluetooth to phone
What is NS.ini (sorry)
Paul Martin
@Spazholio
Oct 25 2016 02:24
@diabeticgonewild Thinking of undercutting GrifGrips? =)
Dana Lewis
@danamlewis
Oct 25 2016 02:26
@sgoldberg89 at dinner so slow to respond, someone else may jump in and help you keep troubleshooting this part: and/or search back for NS.ini and other troubleshooting convos
Sgoldberg89
@Sgoldberg89
Oct 25 2016 02:26
@danamlewis Thank you! I'll call it a day for now. Need to jump on the bike for a while.
Much appreciated. Will tinker tonight and check in tomorrow.
Darrell Wright
@beached
Oct 25 2016 02:41
Is smartguard just rebranded low suspend?
diabeticgonewild
@diabeticgonewild
Oct 25 2016 02:41
@Spazholio Why not?! I think it's a waste of money to buy, and if you want "nice, clean cuts", you might as well find the type of machines that companies like GriffGrips uses. Plus, you can "steal the designs" from GriffGrips VERY easily with the ScanNCut2 machine! It's a win-win situation!!!
Paul Martin
@Spazholio
Oct 25 2016 02:42
Hey, I’m with you. I just don’t use the GrifGrips enough to justify that kind of cost. I can only get ~2 weeks out of a sensor, and SkinTac will last 2 weeks.
diabeticgonewild
@diabeticgonewild
Oct 25 2016 02:43
But in terms of spending $$$ OpenAPS, upgrade to t:slim x2 pump for $800 (ouch!!!), and a new sled/sledge for sled hockey are higher up on my priority list.
Dana Lewis
@danamlewis
Oct 25 2016 02:43
@trevettallen that might be one of the best first PRs ever ;)
Paul Martin
@Spazholio
Oct 25 2016 02:43
HAH!
@diabeticgonewild Thinking of actually setting up a site/shop to sell these? Let people customize for a couple bucks more, that sort of thing?
diabeticgonewild
@diabeticgonewild
Oct 25 2016 02:44
I don't use GrifGrips. I cut a rectangular piece of RockTape H2O 4" roll tape, and then in the middle of the rectangular piece, I cut a small rectangular hole for the transmitter to go through with an Xacto knife.
The problem is that it NEVER looks clean, and I can't cut straight so the alignment of the tape is never right.
Matt Walker
@walker0
Oct 25 2016 02:46
@beached yes I believe it is just rebranded predictive low suspend.
diabeticgonewild
@diabeticgonewild
Oct 25 2016 02:46
So that affects the life of the sensor adversely, even though the tape definitely extends it. And kinesiology tape is literally the best for transmitters and pods, that's for sure, because it wicks out moisture, and it stretches with your skin.
Scott Leibrand
@scottleibrand
Oct 25 2016 02:46
@beached for 530g, smartguard = low glucose (threshold) suspend. For 640G and 670G, it's PLGM (predictive low glucose management).
Paul Martin
@Spazholio
Oct 25 2016 02:46
I need the tape to stick to the plastic of the sensor, not the tape itself. The hole in the GrifGrip is just a TAD too large for that.
Darrell Wright
@beached
Oct 25 2016 02:46
They are using it on the new Canadian 630g
Scott Leibrand
@scottleibrand
Oct 25 2016 02:47
Yeah, that's the 530g algorithm
Darrell Wright
@beached
Oct 25 2016 02:47
that's not out until 2017, but loop/ns/openaps has set my expectations too high
Scott Leibrand
@scottleibrand
Oct 25 2016 02:47
(Don't bother unless you're doing it as a way to preorder a 670G)
Darrell Wright
@beached
Oct 25 2016 02:47
why would I though, the US 670g is a step back.
locked into their reporting(sucks) and a higher than needed (maybe im ignorant) target value
Scott Leibrand
@scottleibrand
Oct 25 2016 02:48
Maybe. It will be a bit easier to use, but if you're a power user, meh
Darrell Wright
@beached
Oct 25 2016 02:48
plus no urchin
diabeticgonewild
@diabeticgonewild
Oct 25 2016 02:49
I can't really "set up a shop". I'm on disability, but it's not that bad for me. I'm receive SSDI off of both mine (because I earned enough work credits working part time while at university, before my declared disability date at age 21) and my (deceased) dad's work record (because I was able to prove that I was unable to work before the age of 22). So I receive 75% of what my dad would have received for SSDI.
Darrell Wright
@beached
Oct 25 2016 02:49
At long as the FW on my new one is the same or works im good for another 5years. My 754 is good
Scott Leibrand
@scottleibrand
Oct 25 2016 02:49
670G will be better for a lot of people who haven't built OpenAPS, or who did but don't want to maintain it.
Dana Lewis
@danamlewis
Oct 25 2016 02:49
So sorry about those expectations, @beached ;)
Darrell Wright
@beached
Oct 25 2016 02:50
Will the 670G fill the need of all those others who are watching their kids at school. Seems to be a huge use case
diabeticgonewild
@diabeticgonewild
Oct 25 2016 02:50
My work record pays first into the benefit each month, then my dad's record pays the remaining amount, up to the 75% of my dad's benefit, that I am entitled to receive.
Scott Leibrand
@scottleibrand
Oct 25 2016 02:50
LOL. Yeah, for people who like to hang out in this channel, I doubt many will go 670G
diabeticgonewild
@diabeticgonewild
Oct 25 2016 02:50
So it's not really "double dipping"...government policies are weird though.
Darrell Wright
@beached
Oct 25 2016 02:50
Now that my carbs are in NS, im golden now :)
Sgoldberg89
@Sgoldberg89
Oct 25 2016 02:50
could someone remind me where I set target bg?
Scott Leibrand
@scottleibrand
Oct 25 2016 02:50
@beached not at first. Remote monitoring is supposed to be a fast follow-on
Sgoldberg89
@Sgoldberg89
Oct 25 2016 02:50
Sorry, my head is swimming in this
Darrell Wright
@beached
Oct 25 2016 02:51
We have yet to get remote monitoring in canada
Dana Lewis
@danamlewis
Oct 25 2016 02:51
@sgoldberg89 it will come from pump
Sgoldberg89
@Sgoldberg89
Oct 25 2016 02:51
mind = blown
what setting on the pump?
Dana Lewis
@danamlewis
Oct 25 2016 02:51
Target BG for bolus wizard
diabeticgonewild
@diabeticgonewild
Oct 25 2016 02:51
And I live with my family so it's not bad at all, being on disability, but it still sucks. I would rather be working any day!!!! Those years I was working were much happier days for me!
It's not healthy to be of working age—and not be working, whatever the reason may be. I can tell you that. And I was definitely "made to work". I'm wired in a way that I feel in need to accomplish and live up to certain standards through work.
But yeah...
Dana Lewis
@danamlewis
Oct 25 2016 02:53
@sgoldberg89 you should spend time going through all the menus on pump and seeing what it can do. You don't want to be the person/people who go 10 years without using the easy bolus button ;) (sorry @Spazholio ;))
Paul Martin
@Spazholio
Oct 25 2016 02:54
No no no, I earned that one...
Darrell Wright
@beached
Oct 25 2016 02:54
@danamlewis it's amazing what isn't used. You find out when the history record hasn't been seen before
Dana Lewis
@danamlewis
Oct 25 2016 02:54
Truth
Sgoldberg89
@Sgoldberg89
Oct 25 2016 02:55
@danamlewis I'm a serial manual boluser...
Had the medtronic pumps for over a decade... mostly just hit the up arrow haha
Paul Martin
@Spazholio
Oct 25 2016 02:56
Sure, rub it in.
diabeticgonewild
@diabeticgonewild
Oct 25 2016 02:56
I pretty much ONLY manual bolts.
Bolus
Dana Lewis
@danamlewis
Oct 25 2016 02:56
@sgoldberg89 :+1: same.
Sgoldberg89
@Sgoldberg89
Oct 25 2016 02:56
Can I ask ONE more question for the day?
Darrell Wright
@beached
Oct 25 2016 02:56
My pump thinks I only manual bolus. it's wrong
diabeticgonewild
@diabeticgonewild
Oct 25 2016 02:56
I don't trust bolus wizard and I only want to take what I feel like taking.
Paul Martin
@Spazholio
Oct 25 2016 02:56
@Sgoldberg89 Yes, but that was it. Sorry, man.
Dana Lewis
@danamlewis
Oct 25 2016 02:56
I bet you'll have more ;)
Paul Martin
@Spazholio
Oct 25 2016 02:56
(meaning that was your one question) =)
Sgoldberg89
@Sgoldberg89
Oct 25 2016 02:57
Then I'll reeally go jump on a bike because that san pelli is kicking in and i'm going up finally
@Spazholio :)
Paul Martin
@Spazholio
Oct 25 2016 02:57
@diabeticgonewild I don’t know you or your particular disabilities, but it seems like this might be easy enough to set up an online shop for something like this, no? Sounds like the machine does the bulk of the work and then you just mail out the product and rake in the bucks. =)
Scott Leibrand
@scottleibrand
Oct 25 2016 02:57
Anyone want to try https://github.com/Perceptus/iob/blob/master/insulin_on_board_pct.js with oref0? Seems pretty straightforward to integrate.
Sgoldberg89
@Sgoldberg89
Oct 25 2016 02:57
@danamlewis Oh I have a ton
@danamlewis So is it possible that the pump and OpenAPS pills aren't working because I haven't entered the appropriate information in Azure (like specifying what info should show with the pump?) or something along those lines?
Dana Lewis
@danamlewis
Oct 25 2016 03:00
Could be that you put wrong info into setup script (url is wrong or Api is wrong).
Sounds like you did enable stuff in azure already.
So it's probably the upload mechanism
Sgoldberg89
@Sgoldberg89
Oct 25 2016 03:01
Blurg
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:01
@Spazholio I am a "disabled dependent" on my parents'/dad's retiree benefit through his former employer via Aetna. I'm over age 26 (I'm actually 27) and so Aetna will send s*** in the mail and tell me that I need to prove that I am "incapable of self-support" and that my "mom provides more than 50% of my support" and that I am "declared as a dependent" on my mom's tax return.
Dana Lewis
@danamlewis
Oct 25 2016 03:01
cat NS.ini and see what's there
Might be an easy fix like fixing url or api
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:02
The issue is, Aetna, even when I got certified as permanently and totally disabled by them, like literally LESS THAN 3 months ago, doing the same exact same process.
Darrell Wright
@beached
Oct 25 2016 03:02
@danamlewis so the expectations are set such that Medtronic touts only 59min per day of <70mg/dl. I am sure that is very good and market leading. That is my weekish, 50 readings during the last week or 2%. But that was also a median of 3.8mmol/L(68) +- 0.4mmol(+-7mg/dl). So yeah, I feel spoiled and thankful
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:02
They will send a letter that is "30 days notice" and it's actually like "20 days notice"
Paul Martin
@Spazholio
Oct 25 2016 03:02
@diabeticgonewild That sucks. I don’t envy you having to deal with insurance companies on top of everything else.
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:03
And they tell us that they are TERMINATING my insurance at the end of the month (with no recourse) if I don't go through the process again, even though I did it 3 months ago.
Assholes.
Dana Lewis
@danamlewis
Oct 25 2016 03:03
@beached that's some icky fine print
Darrell Wright
@beached
Oct 25 2016 03:03
@danamlewis sorry, fixed my target from 4.4 to 4mmol/ 48 readings in last 31days of low
Paul Martin
@Spazholio
Oct 25 2016 03:04
@scottleibrand Unsure how someone would try that with oref0. Where would it be called, and where would the output go? Or would it just be an oref0 command and called via command-line?
Darrell Wright
@beached
Oct 25 2016 03:04
@danamlewis I am a bilingual diabetic. i speak mmol/L and mg/dl
Dana Lewis
@danamlewis
Oct 25 2016 03:04
@spazholio more work than that to test
Martin Haeberli
@mhaeberli
Oct 25 2016 03:04
@Sgoldberg89 can you send us an obscured copy of the line you used to add the nightscout url? (obscured ‘cause we don’t want to know the actual url or secret)
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:04
And it's like, I have to write down how much school I have completed, fill out a personal statement, write a letter, sign papers, get a statement filled out from my doctor, have my doctor certify that I have a disability that meets "blue book criteria" for social security, AND submit the past year's worth of medical records.
With less than 30 days notice, or they are terminating me!!!!
And I literally got certified less than 3 months ago!
Such BS!
Like seriously, once you meet the criteria by the US government as "permanently and totally disabled", nothing dramatic is going to change in 3 months!!!!
Paul Martin
@Spazholio
Oct 25 2016 03:07
@danamlewis I’m sorry, I don’t follow?
Martin Haeberli
@mhaeberli
Oct 25 2016 03:07
@Sgoldberg89
nightscout autoconfigure-device-crud https://xyzsite.azurewebsites.net/ api-secret
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:08
Aetna just hopes that I don't follow through and MAKE my doctors write down the SPECIFiC listings which I meet "blue book criteria" (they left it blank—and in FINE print Aetna said that this particular omission would be AUTOMATIC grounds for denial!)
Sgoldberg89
@Sgoldberg89
Oct 25 2016 03:08
@mhaeberli I'm working on that but get some error every time
Paul Martin
@Spazholio
Oct 25 2016 03:08
@diabeticgonewild Well, the only “good” part about this is that since you just did it 3 months ago, you likely have all the stuff you need already, and with a few exceptions, just send it in again?
Sgoldberg89
@Sgoldberg89
Oct 25 2016 03:08
json: error: input is not JSON: Unexpected 'N' at line 1, column 1:
Not an openaps environment, run: openaps init
^
json: error: input is not JSON: Unexpected 'N' at line 1, column 1:
Not an openaps environment, run: openaps init
^
json: error: input is not JSON: Unexpected 'N' at line 1, column 1:
Not an openaps environment, run: openaps init
^
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:08
Aetna just hopes you're being lazy and don't follow through or something or make a mistake in a circumstance like mine.
Martin Haeberli
@mhaeberli
Oct 25 2016 03:08
@Sgoldberg89 to get mine, I just did
history | grep autoconfigure
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:09
Well no, because I completed 1 class (which is scrutinized as being able to "partially work") plus other stuff
Dana Lewis
@danamlewis
Oct 25 2016 03:09
@Spazholio it will be more complicated to test than what you asked. @scottleibrand is driving and my fingers are tired so I'm summarizing ;) as a placeholder til one of us gets to a keyboard
Martin Haeberli
@mhaeberli
Oct 25 2016 03:09
possibly failing because you need to
cd myopenapsdirectory
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:09
It's a really complex matter.
Martin Haeberli
@mhaeberli
Oct 25 2016 03:09
or whatever it’s called ...
Sgoldberg89
@Sgoldberg89
Oct 25 2016 03:10
so enter "myopenapsdirectory" first
?
Paul Martin
@Spazholio
Oct 25 2016 03:10
@danamlewis Roger that. =)
@diabeticgonewild I’ve had to jump through hoops with the VA before. I understand the frustration. =/
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:10
And the thing is, if you think that taking classes online on your own time, part time, and not having a job or anything, with significant disability accommodations equates to working a job of any type, you're crazy.
Dana Lewis
@danamlewis
Oct 25 2016 03:10
@Spazholio ask stay tuned for follow up commentary ;)
Martin Haeberli
@mhaeberli
Oct 25 2016 03:10
let’s say my openapsdirectory is at /Documents/martinopenaps
Paul Martin
@Spazholio
Oct 25 2016 03:11
@danamlewis I’m likely heading out in a bit, but I’ll check back later. I’m definitely curious.
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:11
Like the taking of online courses is on MY TIME, not on my employer's time.
Martin Haeberli
@mhaeberli
Oct 25 2016 03:11
then type “cd /Documents/martinopenaps” first
then you have a chance that “nightscout autoconfigure-device-crud …” will work
we can go private if you like
Paul Martin
@Spazholio
Oct 25 2016 03:12
@diabeticgonewild God forbid you try to improve yourself, right?
Dana Lewis
@danamlewis
Oct 25 2016 03:12
@mhaeberli not sure that works for uploads yet, just downloads. That'd be a good PR to put in
(In setup scripts, I mean)
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:12
This is different. One time they INiTIALLY denied the "disabled dependent" claim, and they NEVER sent us the denial. Aetna actually NEVER sends me denials for anything, despite the fact that they DO deny significant stuff. Ironically, other stuff from Aetna in my name makes it to my address.
Martin Haeberli
@mhaeberli
Oct 25 2016 03:13
@danamlewis I see your point about uploads vs downloads
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:13
And HR told us not to worry about Aetna "taking their sweet time" (and them subsequently denying the claim...), and never sending us the denial.
So, we never knew the status, even though we HOUNDED them with line 10+ hours of phone calls.
And it was denied and we were never told it.
And we lost the chance to appeal it.
And we really had no recourse.
Except that, when they were taking their sweet time making the decision, after we submitted the paperwork, I happened to develop another disabiling impairment—a neurological disease that causes progressive muscle weakness among other things.
So, before we realized that in the meantime that I developed another disability.
There was no chance to appeal and we had lost the insurance.
Paul Martin
@Spazholio
Oct 25 2016 03:17
A friend of mine recently got diagnosed with something similar - CMT. =/
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:17
But because I had developed another disabling impairment, they "couldn't say no to a new and updated application"
Yeah, but this disease can also affect autonomic functioning too.
So technically I have 3 diseases that cause autonomic neuropathy.
But yeah, the insurance for retroactively reinstated after my 26th birthday.
I lost the secondary insurance for like a month. But I already had Medicare at that time.
I didn't apply for disability until age 24 (cause I knew that I needed to apply for SSDI—as Aetna asks if you're getting government benefits—and if you're not—it just gives them an easy reason to deny you and kick you off the vent by default) cause I knew that Medicare takes 24-29 months (in my particular case 24 months) to get AFTER your first SSDI payment—and you can only claim retroactively back for 1 year.
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:22
So, basically had to apply so that Aetna could not play games with us—and they still played a wildcard on us.
I applied for SSDI basically so I could stay on my parents' insurance past the age of 26z
Scott Leibrand
@scottleibrand
Oct 25 2016 03:25
@Spazholio I think we'd also need Perceptus/iob#1
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:29
So my intentions for applying for SSDI were ONLY for staying on my parents' insurance past the age of 26.
But, it was a good financial decision too.
Scott Leibrand
@scottleibrand
Oct 25 2016 03:30
If we get insulin activity added, we could pretty easily have https://github.com/openaps/oref0/blob/master/lib/iob/calculate.js call that function.
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:30
And if any of you think it's ridiculous, for what it's worth, I worked extremely hard when I was able to. And I even got an internship offer for 6 months with Intel for electrical engineering in Folsom, CA, at the beginning of the new year in 2012, at age 22.
But I had to decline it because I literally had been in and out of the hospital for the past 3 months and I was about to be diagnosed with a rare autoimmune neurological disease (the test had already been submitted to the Mayo Medical Laboratories—and we were just waiting on a result) and I wasn't in a position to take it.
I even got straight Fs in my classes because I had asked for Incomplete grades, but the undergraduate engineering dean at the University of Houston, who literally is the biggest jerk in the world, not only refused to give me incompletes—he even retaliated against me.
And I was already facing the enrollment cap for the state of Texas due to medical withdrawals solely—not due to academic reasons—and so is I had to retake ANY more classes, even for medical withdrawal, I would be subjected to out of state tuition charges for the rest of my academic career, unless the university provost gave an exception dornmezz
For me.
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:37
And willingly knowing this, and despite getting As on my first set of midterms and having record of that, he still refused to give me incompletes. So I ended up getting Fs, being on engineering probation, AND BEING FORCED to take medical withdrawal.
And having to repay for all of those courses—let alone working my ass off and showing up to school pretty much no matter how sick I was.
And this jerk also retaliated against me too.
So between, health and grades, working a highly coveted internship was out of the question.
Dana Lewis
@danamlewis
Oct 25 2016 03:39
This is from slack but wondering usefulness for Gitter potentially - using :rabbit: (since no :raccoon:) or :dolphin: as described here: http://randsinrepose.com/archives/shaping-your-slack/
Darrell Wright
@beached
Oct 25 2016 03:39
Has anyone seen this before? They;re ledgible :) https://sites.google.com/site/diabetesmapp/insulin-action-curves
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:40
And just thinking about it here and there, just having it cross my mind...the possibility of working a highly covered once in a lifetime internship as an electrical engineering student with Intel—made me sick to my stomach cause I had to decline itS
And, just in case you're wondering, I did end up having the rare disease...
Scott Leibrand
@scottleibrand
Oct 25 2016 03:41
@beached the nice thing about @bustavo's is it's already a js function.
Darrell Wright
@beached
Oct 25 2016 03:41
really! I didn't see that
Scott Leibrand
@scottleibrand
Oct 25 2016 03:42
Just need activity as well as IOB I think.
Darrell Wright
@beached
Oct 25 2016 03:43
If you have a function, doing a derivative "should" be easy. please note the quotes
@scottleibrand ok. looks like he generated a curve and stretched over different DIA
Scott Leibrand
@scottleibrand
Oct 25 2016 03:45
Yep. Same thing oref0 does, just with a more complex curve than my two straight lines. ;-)
Darrell Wright
@beached
Oct 25 2016 03:47
Looks like to get active insulin you need the rate of change, so yeah 1st derivative and flip the sign as the IOB is going down
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:49
@beached I know how to "convert" the insulin action curves into insulin on board calculations.
It's kind of an esoteric kind of thing, too.
Darrell Wright
@beached
Oct 25 2016 03:51
really curious, why is it esoteric.
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:56
I was going to show you a camera shot of the first few steps for doing this process, but it's deep in my twitter feed. And I had a long day today. Had a minor surgery today and had surgery drugs.
Darrell Wright
@beached
Oct 25 2016 03:57
ah ok. I assumed it was like 1st year University calculus
so I know what I forget but know how to know it again when needed
diabeticgonewild
@diabeticgonewild
Oct 25 2016 03:57
Anyway, there are only a few srudies (1-2 at best for each type of rapid acting insulin) using the appropriate protocol for euglycemic insulin clamp
And
And usually the originate from Germany, because of the specific tool that they use for the clamp (I forget what it's called—it's basically an "artificial pancreas from the 80s"...)
Scott Leibrand
@scottleibrand
Oct 25 2016 03:59
@diabeticgonewild I think you're making it unnecessarily complicated for the problem at hand. We already have the Walsh curves that come from doing all that hard work. We are just converting them into js and taking the derivative.
diabeticgonewild
@diabeticgonewild
Oct 25 2016 04:00
The "protocol" for deriving the insulin on board formula depends on the euglycemic clamp being performed to the right protocol.
And there are only a few published papers on how to correctly calculate the IOB formula. All from German authors
Scott Leibrand
@scottleibrand
Oct 25 2016 04:01
We already have hundreds of people looping successfully using "incorrect" IOB curves. This is an opportunity to perhaps start using something a bit closer to the best published ones, and see if we get better outcomes overall.
diabeticgonewild
@diabeticgonewild
Oct 25 2016 04:02
No I'm not @scottleibrand . John Walsh, who works with the Germans, including Heinemann, said that there are NOT enough valid clamp studies (and he specifically mentioned which ones can be used and cities them)
In a paper.
Scott Leibrand
@scottleibrand
Oct 25 2016 04:02
And we're discussing using his results.
diabeticgonewild
@diabeticgonewild
Oct 25 2016 04:02
And he said that the formula has to be derived in a specific way, accounting for confidence intervals, etc.
Scott Leibrand
@scottleibrand
Oct 25 2016 04:03
We're not trying to reproduce his work, just implement it in a hybrid closed loop.
diabeticgonewild
@diabeticgonewild
Oct 25 2016 04:03
seriously?
Scott Leibrand
@scottleibrand
Oct 25 2016 04:03
(Which Loop already does, FWIW)
Darrell Wright
@beached
Oct 25 2016 04:03
@scottleibrand Im lost in tabs, got to find the link to that js again, but you are entering time though. so just hack it and pop in t=0,1,2,3 and the difference between t=n, t=n+1 should give the devative no?
diabeticgonewild
@diabeticgonewild
Oct 25 2016 04:03
Like all I was saying
Scott Leibrand
@scottleibrand
Oct 25 2016 04:03
@beached yeah, you could do that
diabeticgonewild
@diabeticgonewild
Oct 25 2016 04:04
Was that you can calculate the SPECIFIC IOB for the pharmacokinetics of A SPECIFIC chart that @beached was referencing
You are taking this WAY too seriously.
Scott Leibrand
@scottleibrand
Oct 25 2016 04:04
I tried something similar, but the point in the code I was doing it was on the sum total, so it didn't work if there were boluses or temps going on.
diabeticgonewild
@diabeticgonewild
Oct 25 2016 04:04
And what a joke Scott, I don't need to be put down by you.
WTF
Scott Leibrand
@scottleibrand
Oct 25 2016 04:05
@beached But if you did that on each individual (micro)bolus you could get activity that way.
Darrell Wright
@beached
Oct 25 2016 04:05
@scottleibrand you need the function eh. it has to be calculated for each bolus/basal dose
diabeticgonewild
@diabeticgonewild
Oct 25 2016 04:06
This conversation didn't even involve you and you took a personal spin on it, which is completely irrelevant.
I mean seriously, what a joke.
I don't have to put up with your crap Scott, including your personal attacks.
Scott Leibrand
@scottleibrand
Oct 25 2016 04:08
@beached yeah. We calculate the IOB contribution and the activity contribution of each bolus and microbolus, and then sum them.
Darrell Wright
@beached
Oct 25 2016 04:09
I think the function is perfect as is no?
IOB = 0
 foreach( dose in doses )  {  if( dose_age < dose_dia ) { IOB += f( dose_age, dose_dia )*dose_amount ); } else { remove_dose; }
Scott Leibrand
@scottleibrand
Oct 25 2016 04:10
That's one half of what we need yes.
The other half is we need insulin activity to calculate BGI, deviations, etc.
diabeticgonewild
@diabeticgonewild
Oct 25 2016 04:11
@scottleibrand I didn't read your posts right. I'm sorry. I'm just not feeling well.
Darrell Wright
@beached
Oct 25 2016 04:11
Keep the last sum and use the difference
Scott Leibrand
@scottleibrand
Oct 25 2016 04:11
@diabeticgonewild no worries. Hope you feel better.
Darrell Wright
@beached
Oct 25 2016 04:11
you are running it every x minutes, so the difference is the derivative(sorta)
Scott Leibrand
@scottleibrand
Oct 25 2016 04:12
Except when a temp is running.
Then the difference is affected by that.
Darrell Wright
@beached
Oct 25 2016 04:12
IOBt1 - IOBt0
diabeticgonewild
@diabeticgonewild
Oct 25 2016 04:12
I thought that you were telling me that it was a waste of time to do these specific IOB curves that apply to a specific cjaetS
If you want me to work on this, I can do it.
Darrell Wright
@beached
Oct 25 2016 04:12
no, calculate with the basal rates as a dose every 5min
functions are cheap to run so it should be quick
diabeticgonewild
@diabeticgonewild
Oct 25 2016 04:12
I have all of the files and stuff needed to do it, and I can share them. Like tomorrow or something
@scottleibrand thanks. Too many drugs today and not enough sleep over the past couple of days due to not feeling well and stress.
Scott Leibrand
@scottleibrand
Oct 25 2016 04:13
@Beached Yeah, but we have to calculate the activity impact by doing the diff of IOBs from each dose, rather than doing the diff on the sum.
diabeticgonewild
@diabeticgonewild
Oct 25 2016 04:14
At least you know why "I got angry at you"...for no reason at all.
I can upload the folder now and share it.
Darrell Wright
@beached
Oct 25 2016 04:15
@scottleibrand let me code this up. It may not start in javascript but that won't matter as it will port easy
diabeticgonewild
@diabeticgonewild
Oct 25 2016 04:15
I'm going to try to go to bed soon.
Scott Leibrand
@scottleibrand
Oct 25 2016 04:15
@beached you might want to start with the derivative calculation.
As if we can do that symbolically it'll be the most efficient method.
(And easiest to debug)
@diabeticgonewild have a good night.
@beached if I have the time what I'd do is try to get wolframalpha to do the derivative for me. :-)
Darrell Wright
@beached
Oct 25 2016 04:17
@scottleibrand ok, im using paper here. because friends don't let friends math on the computer
Dana Lewis
@danamlewis
Oct 25 2016 04:17
Lol
Scott Leibrand
@scottleibrand
Oct 25 2016 04:17
Even better. ;-)
diabeticgonewild
@diabeticgonewild
Oct 25 2016 04:17
@scottleibdand thanks
diabeticgonewild
@diabeticgonewild
Oct 25 2016 04:28
@beached @scottleibrand Here is the "collection" of files needed for calculating Apidra/Novolog/Humalog IOB curves, among other things. It's literally EVERYTHING you will possibly need. The Heinemann_buch.PDF file is a treasure that I found, too! https://drive.google.com/open?id=0B8pPP-odNkv9SlozRmZlSXFxaTA
If you give me a month (as it will take around 50 hours total to make the calculations and verify them, in my experience), I can make various IOB and activity curves for Apidra/Novolog/Humalog, and dose dependent IOB activity such as when surface area to volume ratio differ significantly
I'm pretty sure that I can get it done in a month. I have gotten close to being at least halfway done before in the past, and this work isn't too complicated, like other stuff.
But I got to go to bed. Sad, but I'm going to add drugs upon the drugs I already took, cause I don't feel well and I have to "shut it off". I also have had trouble sleeping over the past couple of days and I really need a decent night's sleep.
@scottleibrand I'm sorry for what I said and for being mean in general. I really shouldn't be on the internet post-surgery as you're likely to say "dumb stuff" anyway, but it also keeps my mind kind of busy...so it's hard to resist. I hope you understand. I'm really not myself tonight and I have had a stressful few days.
Scott Leibrand
@scottleibrand
Oct 25 2016 04:33
Yeah. Take care.
Neal
@tnharvey
Oct 25 2016 04:59
Thanks @diabeticgonewild , the instructions worked well. I ended up having to flash my edison back to stock to get it to recognize, but after that, things went smoothly!
Darrell Wright
@beached
Oct 25 2016 04:59
I keep getting nerd sniped today
Tim Street
@tim2000s
Oct 25 2016 05:15
I managed to be stupid enough to not have ordered sensors in a timely enough fashion, so when my replacement failed I was left loopless. This is how it feels: http://www.diabettech.com/artificial-pancreas/the-loss-of-a-loop-recognising-the-life-benefits-of-hybrid-closed-loop-systems/
Hristina Dimova
@xpucuto
Oct 25 2016 09:12
@tim2000s I am sure there are a lot of people using dexcom in London . Can't u borrow a sensor until ur order arrive ? I am Loop addicted 😻😻😻 !
Tim Street
@tim2000s
Oct 25 2016 09:14
I'll wait now. It's two days or so, but it's amazing how much you get to stop thinking about things while running the loop!
Hristina Dimova
@xpucuto
Oct 25 2016 09:19
I admire ur patience . In less than a hour I can get a new sensor from a friend ... can't wait that long :) can I subscribe for ur website ? Stupid question I know
Tim Street
@tim2000s
Oct 25 2016 09:59
Yes. You can do it from the website @xpucuto
Tim Street
@tim2000s
Oct 25 2016 10:19
I've added in a new subscribe to blog via email option. Have another go.
Hristina Dimova
@xpucuto
Oct 25 2016 10:20
Great . I just left home so I will give it a try bit later :+1:
Chris Oattes
@cjo20
Oct 25 2016 10:28
@scottleibrand I know you can use wolfram alpha to help, but derivatives of simple polynomials is just going through each term and replacing a*(x^n) with a*n*(x^(n-1)) (e.g. 2x^3 -> 6x^2)
Chris Oattes
@cjo20
Oct 25 2016 10:46
(and removing the constant terms). But as was suggested last night, just working on the difference between two IOB figures is probably actually a better solution. The derivative gives the amount of insulin active at a particular instant, but we don't work with points in time, we work with 5 minute periods.
diabeticgonewild
@diabeticgonewild
Oct 25 2016 10:57
The surgical procedure I had yesterday is making me an emotional mess. I think it is from the general anesthesia. This has happened before in the past. I really don't know how I am going to get through the next couple of weeks, let alone today.
C-Ville
@C-Ville
Oct 25 2016 11:01
@tim2000s I liked your blog article. It is so true and shows nicely the impact @scottleibrand @danamlewis @bewest had on my/our lifes and all the others that contributed to that amazing project
diabeticgonewild
@diabeticgonewild
Oct 25 2016 11:06
I have no quality of life whatsoever.
Chris Oattes
@cjo20
Oct 25 2016 11:20
@scottleibrand you were saying about working on totals rather than doses, I think changing it to work on 'actual' IOB curve is just a case of replacing that line of code I keep complaining about with what beached suggested of just doing IOB(x+5) - IOB(x)
diabeticgonewild
@diabeticgonewild
Oct 25 2016 13:35
Seriously, the meds that I was given are making me into an emotional mess. My best option at this point of time is to "sleep it off". Sucks.
emsamom
@emsamom
Oct 25 2016 13:59
Screen Shot 2016-10-25 at 9.29.02 AM.png
Tested last night not connected to the pump. If i needed ANY more proof that I need to do this live I had it last night. I think I would have avoided this low entirely had I been looping!
elenavilla
@elenavilla
Oct 25 2016 14:31
Hi all, good morning! I got my temporary basals to display in NS already. I'm so happy. @danamlewis the update to grilled cheese worked, although it took a bit to start working. I woke up in the morning to a screen with all my basals there. I'll celebrate by eating a grilled cheese sandwich today for lunch.
Lexsus
@Lexsus
Oct 25 2016 14:46
Hi! All models of pumps(512/712, 515/715, 522/722) will work with OPenAps or need firmware 2.4 or less?
bschrag620
@bschrag620
Oct 25 2016 15:34
I think the main issue with pumps is just to make sure it doesn't have a pc connect option. I believe the 523/723 are the only ones that received a firmware update that added pc connect. So yes, I believe all those earlier pumps will work for sure. I'm new here though but that's my understanding.
emsamom
@emsamom
Oct 25 2016 15:36
Any suggestions why my Dexcom and raspberry pi clocks don't agree? They were in sync yesterday but now I have {"reason":"BG data is too old, or clock set incorrectly. Your CGM time is Tue Oct 25 2016 09:20:25 GMT-0400 (EDT) but your system time is Tue Oct 25 2016 09:36:03 GMT-0400 (EDT)"}
Notice the times are off by 16 minutes
Chris Oattes
@cjo20
Oct 25 2016 15:40
emsamom: it usually means your APS hasn't had fresh data from the CGM
bschrag620
@bschrag620
Oct 25 2016 16:05
Does the raspberry pi have an internet connection?
emsamom
@emsamom
Oct 25 2016 17:23
It looks like one or the other was not connected - but it resolved itself while work took my attention this morning :)
Dana Lewis
@danamlewis
Oct 25 2016 17:23
@Lexsus yes, all of those will work per @bschrag620’s comment.
albertotrg
@albertotrg
Oct 25 2016 17:24
Still looking for an old Medtronic pump for openAPS. I found one 722 in USA and I'm in Spain...does anybody know if I will have problems with Customs during the shipment from USA to Europe? I'm afraid since it's medical stuff...
Dana Lewis
@danamlewis
Oct 25 2016 17:24
@bschrag620 not sure what you’re asking re: pi and internet connection. do you mean built-in capabilities, or if it’s a requirement to have internet connection?
@elenavilla glad your NS basal render is working! :)
@emsamom :+1:
@albertotrg not sure re: shipping to spain, but perhaps @jberian knows?
Tim Street
@tim2000s
Oct 25 2016 17:35
@albertotrg I'm fairly sure there is a pump on medwow in Spain.
If you log in and then contact them directly rather than using medwow, you should get a response.
albertotrg
@albertotrg
Oct 25 2016 17:40
Thanks so much @tim2000s but unfortunately that one is out of my budget
Tim Street
@tim2000s
Oct 25 2016 17:41
The best approach is to get in contact and make an offer. The price they are showing for a 522 is way over market.
Dana Lewis
@danamlewis
Oct 25 2016 17:41
:+1:
Tim Street
@tim2000s
Oct 25 2016 17:42
(I even managed to find a 715 for £350 last week!)
albertotrg
@albertotrg
Oct 25 2016 17:45
350-500 pounds or euros would be ok...but this guy is asking 1.800$...out of my budget
Tom Boudreau
@TomBoudr_twitter
Oct 25 2016 17:45
I found asking folks on diabetes groups
Dana Lewis
@danamlewis
Oct 25 2016 17:45
No one would pay that
So definitely try making him a reasonable offer and pointing out he is 3x over market prices
Tom Boudreau
@TomBoudr_twitter
Oct 25 2016 17:46
I found 2 on https://myglu.org/ and I offered $250-300 US Dollars depending on the condition and had 2 takers
one was too new but I got a nearly mint 715 for $300
emsamom
@emsamom
Oct 25 2016 17:49
Monitoring my log it looks like the loop will enact a temp basal even when the current basal (or current temp basal) is the same as the requested change. I dug through the code a bit but wasn't able to confirm this or figure out why.... can anyone help me understand? Is it to extend the end time of the temp basal?
albertotrg
@albertotrg
Oct 25 2016 17:50
I will check that web, thanks @TomBoudr_twitter ! Is that community located in USA or Europe?
Dana Lewis
@danamlewis
Oct 25 2016 17:51
@emsamom that's a feature (you can turn off) to help you know if the loop is working if you don't have connectivity, just by checking presence of temp on pump
Tim Street
@tim2000s
Oct 25 2016 17:52
It's a global community, but I'm not sure where it's mostly based.
But contact the guy in Valencia. You never know what you might find.
My experience is that people will come down.
Matthias
@ceben80
Oct 25 2016 17:54
1800$ for a 522...Absolutely crushed in the head :astonished: ....I paid 150€ for my 522
Tom Boudreau
@TomBoudr_twitter
Oct 25 2016 17:55
yeah...
Dana Lewis
@danamlewis
Oct 25 2016 17:55
Ppl who price that high are thinking about how they paid $5000 out of pocket if they didn't have insurance coverage
So 1800 is sort of reasonable if you don't have any idea that a used pump is really worth 250ish
albertotrg
@albertotrg
Oct 25 2016 18:12
I understand
I'll make an offer and keep searching. Please if anybody knows about shipments between USA -> EU and Custom please let me know
Tim Street
@tim2000s
Oct 25 2016 18:17
Iirc, used goods into the EU aren't subject to duty.
So you shouldn't be paying any customs.
JBerian
@jberian
Oct 25 2016 18:31
@albertotrg I've never had issues sending things from the US to Spain. (Just in case you want to try)
Mike
@libxmike
Oct 25 2016 18:32
I got the minimed 754 2.6A with remote, 18 Mio Katheter, 20 Reservoir, new battery cover, clip, and some other accessories few days ago for 450€
Tim Street
@tim2000s
Oct 25 2016 18:32
Where was that from @libxmike
Mike
@libxmike
Oct 25 2016 18:33
Germany, where I'am
Tim Street
@tim2000s
Oct 25 2016 18:33
I ended up paying £600 for a 722 with all the bits and a remote.
Mike
@libxmike
Oct 25 2016 18:34
Thats bit more, just on the edge of too much
Darrell Wright
@beached
Oct 25 2016 18:35
Neat, I just modelled not my carb absorption rate. I wish
Tim Street
@tim2000s
Oct 25 2016 18:35
That's as much as I was willing to pay. The issue is still the lack of availability so it's rather a seller's market.
bschrag620
@bschrag620
Oct 25 2016 18:36
@danamlewis sorry I wasn't clear in my question. It was directed toward @emsamom in regards to the clock times being different. My thinking was if the pi had lost power and had not yet regained internet the clock time isn't going to be correct.
Dana Lewis
@danamlewis
Oct 25 2016 18:36
Gotcha :)
Mike
@libxmike
Oct 25 2016 18:37
But the seller don't know about the market for loop pump , so time to time you can get good deal if hunting for one
@danamlewis any news about adding additional Explorer boards to existing orders?
garykidd
@garykidd
Oct 25 2016 18:45
is there a way (easy) to get your hashed API secret without having command line access to your rig?
Dana Lewis
@danamlewis
Oct 25 2016 18:46
@libxmike I think Morgan will be at the hackathon tomorrow, back from his trip, and I can ask then
Mike
@libxmike
Oct 25 2016 18:49
OK
PieterGit
@PieterGit
Oct 25 2016 18:55
why is nodejs installed with dpkg instead of apt-get, on https://github.com/openaps/docs/blob/dev/docs/docs/walkthrough/phase-0/setup-edison.md
i would have expected apt-get install nodejs nodejs-dev instead of dpkg -P nodejs nodejs-dev
Scott Leibrand
@scottleibrand
Oct 25 2016 19:00
-P is purge
same as apt-get purge I believe
PieterGit
@PieterGit
Oct 25 2016 19:01
there was no nodejs on my jubilinux install
Scott Leibrand
@scottleibrand
Oct 25 2016 19:01
then that line should be a no-op
PieterGit
@PieterGit
Oct 25 2016 19:04
are people running nightscout on edison? or is that not recommended
Scott Leibrand
@scottleibrand
Oct 25 2016 19:12
If you want to run mongodb on edison, you need an SD card block. I think @jasoncalabrese just got NS running successfully with a json file backend, which should work on Edison...
Darrell Wright
@beached
Oct 25 2016 19:22
Has anyone tried using the glycemic index to adjust carb absorption, or a wag if mixed. So if one absorbs dextrose at 0.8g/min multiply that by a factor based on GCI. or is that not realistic?
Dana Lewis
@danamlewis
Oct 25 2016 19:23
Does it need to be more complicated?
Darrell Wright
@beached
Oct 25 2016 19:23
yes
until I get a biopancreas
Dana Lewis
@danamlewis
Oct 25 2016 19:24
What are you trying to solve for?
Colin Lennon
@colinlennon
Oct 25 2016 19:25
@scottleibrand You can run mongodb on an Edison without an SD card block, or any other additional storage. That's what I'm doing, on Yocto. I'm not running NS on it though, don't know if that'd fit.
Darrell Wright
@beached
Oct 25 2016 19:26

so dextrose is easy to test. OK. GCI is easy to lookup, store and make a nice interface for to get a rough guage of how much absoption is changed.

Not sure exactly, but I think I am either naive(always) or I might be able to model it with that if GCI is actually valid

Scott Leibrand
@scottleibrand
Oct 25 2016 19:30
the UI is the main problem I think. Loop "solved" that by creating three icons for foods of three different typical absorption times
we "solved" it by not trying to predict future increases in carb absorption at all
so the user just boluses according to their own experience of what works with that food type, and then AMA takes over based on actual observed carb absorption.
Darrell Wright
@beached
Oct 25 2016 19:31
@scottleibrand There is a third way, auto calculate absoption times based on food type then if wrong take over. But closer maybe
Scott Leibrand
@scottleibrand
Oct 25 2016 19:32
that requires additional user entry of food type, as with Loop. with oref0 we don't have a good place to enter anything except the basic carb count.
Dana Lewis
@danamlewis
Oct 25 2016 19:33
And the observation has been, it's really hard for people to do that well
Darrell Wright
@beached
Oct 25 2016 19:34
So even loops interface which allows for direct entry, but enter food type fast, med, slow and have a stored dex value for 100%
Even storing that and looking later might give insight
Scott Leibrand
@scottleibrand
Oct 25 2016 19:34
I think the number of people who would prefer to do all that extra work would be small
Darrell Wright
@beached
Oct 25 2016 19:35
I thik the UI could be a one time test to get dex abs rate. Then after that either the direct entry of time or the fast/med/slow buttons like on Loop
1time extra work. not sure
N=1 me wouldn't mind
Scott Leibrand
@scottleibrand
Oct 25 2016 19:35
that said, you could certainly work on it for you if it's worth it. :)
Darrell Wright
@beached
Oct 25 2016 19:36
hopefully my new pump tomorrow allows for this. otherwise I use old one for a while longer
Im just unaware of the Canadian FW versions
other than mine
Neal
@tnharvey
Oct 25 2016 19:43
quick question, I'm running through the interactive setup for oref0-setup.sh, if I'm using NS for CGM right now, what do I enter for "What kind of CGM are you using?"
technically I'm using xDrip (xBridge2) to NS, but don't have a dexcom receiver in use right now
Matthias
@ceben80
Oct 25 2016 19:46
With G4 I assume?
Neal
@tnharvey
Oct 25 2016 19:46
yes, sorry, with G4
Matthias
@ceben80
Oct 25 2016 19:47
That's the answer :smile:
Regarding the script
Neal
@tnharvey
Oct 25 2016 19:48
oh, so answering G4 won't create a device for that? I just wanted to make sure, since I'm using NS for CGM data
Scott Leibrand
@scottleibrand
Oct 25 2016 19:50
it will, but it won't do anything unless you plug in the G4 over USB
it downloads data from any attached CGM and from NS, and uses whatever is most current
Neal
@tnharvey
Oct 25 2016 19:50
Ooohh, awesome. thanks!
Eric
@ecc1
Oct 25 2016 19:51
@garykidd echo -n $API_SECRET | sha1sum assuming you have the unhashed $API_SECRET
garykidd
@garykidd
Oct 25 2016 20:01
I do have unhashed API secret..dont have access to rig via ssh or otherwise
Neal
@tnharvey
Oct 25 2016 20:05
Okay, just finished the full install on an edison/SRF with NS. I started fresh, switched to jubilinux and the whole thing was maybe 2 hours tops. That was brilliantly fast and easy. Thanks! @danamlewis
And the documentation was fantastic by the way
Dana Lewis
@danamlewis
Oct 25 2016 20:06
:) always room for improvement, as evidenced by questions asked along the way! PRs welcome!
Neal
@tnharvey
Oct 25 2016 20:09
One comment. Jubilinux wouldn't flash to my existing Ubilinux install. I ended up using Monica's video to flash on Windows, but what ultimately fixed it was reflashing back to stock with the Intel software, and THEN flashing Jubilinux. It may have worked fine from Linux at that point, not sure, I was already on Windows, so I just did it there. I'd love to turn that into a PR and update to the docs, but not sure if any of that is really helpful.
Tim Street
@tim2000s
Oct 25 2016 20:09
@beached Don't bother with putting estimated carb absorption in. It makes Loop a nightmare. OpenAPS manages it a lot better without trying to predict. The issue is that two meals that are in theory identical don't necessarily absorb at the same rate on different occasions so the auto calc mechanism works far better than the guesstimate you do otherwise.
The amount of fiddling you end up doing with loop to make the absorption rates work is a real pain.
Darrell Wright
@beached
Oct 25 2016 20:12
@tim2000s thanks Yeah loop I only adjust when using dextabs, otherwise default normallly. But I am thinking of how to approach this in code. Like if I need x units of active insulin at t=100 how to approach that and later on not go low, so maybe it cannot be done but only y units are available by lowering future insulin.
Jason Calabrese
@jasoncalabrese
Oct 25 2016 20:14
@PieterGit yeah, NS runs on the edision in a read-only mode using only the openaps json report files, you just need to set STORAGE_URI=openaps:///home/edison/path/to/config.jsonusing a structure like this: https://github.com/nightscout/cgm-remote-monitor/blob/dev/tests/fixtures/openaps-storage/config.json
that lets you map 1 or more report files to a NS collection
Scott Leibrand
@scottleibrand
Oct 25 2016 20:15
@tim2000s good feedback, thanks. that's the most direct feedback I've heard from an end user comparing the two.
looking forward to someone getting oref0 with AMA into Loop
Jason Calabrese
@jasoncalabrese
Oct 25 2016 20:15
it won't replace a NS server in many cases, but works great for offline mode
Scott Leibrand
@scottleibrand
Oct 25 2016 20:17
can you enter temp targets in that config?
Tim Street
@tim2000s
Oct 25 2016 20:20
I've started comparing the code bases @scottleibrand It's going to take me a while as there are some notable gaps, but I think I can get there eventually. Proper hack though. And not a lot of time at the mo.
Darrell Wright
@beached
Oct 25 2016 20:23
I am working through the oref0 code base too.
Jason Calabrese
@jasoncalabrese
Oct 25 2016 20:25
haven't added support for posts/inserts yet, but I do want to be able to enter carbs and set temp targets
I'm going to spend some time away from networks around Thanksgiving, so hoping to have something by then
was thinking I'd have an output file per collection that it would append too, and than that could be combined with NS treatments
had some problems on Saturday with intermittent network that created empty files and my merge didn't handle that well
Rick Warren
@riwarren
Oct 25 2016 20:30
I have a question. I am about to venture into openaps, but one thing I am not understanding. I have medtronic pump, and all the photos of openaps rigs show a pump, raspberry pi board, carelink usb, and then another device plugged into the raspberry pi that looks like an MP3 player.....what is that other device?
Scott Leibrand
@scottleibrand
Oct 25 2016 20:30
that's the dexcom receiver
Rick Warren
@riwarren
Oct 25 2016 20:31
Why would you use dexcom with a medtronic pump?
Are the dexcom cgm sensors cheaper or something?
Scott Leibrand
@scottleibrand
Oct 25 2016 20:32
because most people find it gives more accurate results than the old Enlite sensors
Dana Lewis
@danamlewis
Oct 25 2016 20:32
Because of sensor accuracy; ability to remote view data, etc
Scott Leibrand
@scottleibrand
Oct 25 2016 20:32
and the Dexcom sensors tend to last longer as well
Mike
@libxmike
Oct 25 2016 20:33
@scottleibrand @danamlewis someone should consider to take the carelink off the list as its not working sustainable with Rpi 2 or Zero.
Dana Lewis
@danamlewis
Oct 25 2016 20:33
Until explorer boards are shipping regularly, carelink is still the main best documented setup.
Rick Warren
@riwarren
Oct 25 2016 20:33
Are they any cheaper? Is this required? Or will enlight sensors still work ?
Dana Lewis
@danamlewis
Oct 25 2016 20:33
@riwarren look at the hardware page to see required hardware.
Rick Warren
@riwarren
Oct 25 2016 20:34
I have. It doesnt tell me that I need dexcom cgm for use with a paradygm pump
Mike
@libxmike
Oct 25 2016 20:34
@danamlewis but is working most for 2-4h until death and power cycle is needed
Scott Leibrand
@scottleibrand
Oct 25 2016 20:34
@libxmike did you get oref0-reset-usb working in cron?
Mike
@libxmike
Oct 25 2016 20:35
I tried even powering USB off/on and did not help
Scott Leibrand
@scottleibrand
Oct 25 2016 20:36
back when we were using Carelink, this worked reliably: https://github.com/scottleibrand/openaps-sh/blob/dev/setup.sh#L133
Mike
@libxmike
Oct 25 2016 20:36
@scottleibrand oref0-reset-usb never worked, even reboot is not enough, hub-ctrl works for 6h
Scott Leibrand
@scottleibrand
Oct 25 2016 20:36
with Pi 2?
Mike
@libxmike
Oct 25 2016 20:37
I thinking about doing some circuit and hard reset with GPIO
Scott Leibrand
@scottleibrand
Oct 25 2016 20:37
if oref0-reset-usb isn't sufficient with Pi 2, you might have a dud carelink
Mike
@libxmike
Oct 25 2016 20:37
my care link is working but limited to maybe 6-8h
Rick Warren
@riwarren
Oct 25 2016 20:38
So its not required. Perfect. Thanks.
Mike
@libxmike
Oct 25 2016 20:38
will try with Pi 2 and hub-ctrl (usb off/on)
but Pi 2 ist not portable enough
Zero ist the way to go if you have Pi
to work with
Dana Lewis
@danamlewis
Oct 25 2016 20:40
@libxmike speaking of portable, they are putting a contact email on the site tonight supposedly, so I encourage you to check tomorrow & email them to cancel your existing order & put a new one in with two, if they can’t easily add a second and charge you no shipping.
(for explorer boards, I mean)
Rick Warren
@riwarren
Oct 25 2016 20:40
Have you guys considered using arduino to communicate with medtronic pump directly instead of using carelink usb? The data is not encrypted.....
garykidd
@garykidd
Oct 25 2016 20:40
anybody have any pointers for getting wifi working reliably on edison. I got as far as the initial setup and that was fine but when I migrated to adding another wifi network in supplicant.conf...well it all failed. May have to reflash even.
Dana Lewis
@danamlewis
Oct 25 2016 20:41
@riwarren you can use a variety of radio sticks in your setup. example list here on the Edison-based setup pages: http://openaps.readthedocs.io/en/latest/docs/walkthrough/phase-0/hardware/edison.html (you can use TI for example with a PI rig very effectively)
the sky is basically the limits :) to build rigs however you like. some rigs have documentation support; others haven’t been designed yet so if you design one, you can add docs to support & help others set up that way, too
Mike
@libxmike
Oct 25 2016 20:42
@danamlewis CC for my 1st order was already charged, will check tomorrow what you said
@riwarren medtronic pump does not have usb
to talk to it directly
just radio
Rick Warren
@riwarren
Oct 25 2016 20:48
Yes I am aware. Two more questions for you: 1. When is the edison explorer board available? Can you use edison without that? 2. Has anyone considered simply engineering a superior pump that has this functionality built into it? I have my paradygm 722 in pieces currently as I am reverse engineering it. Woud make a lot more sense to combine everything into 1 device...
Dana Lewis
@danamlewis
Oct 25 2016 20:48
Explorer board can be ordered now, ships November, and yes uses an Edison.
Rick Warren
@riwarren
Oct 25 2016 20:49
When in november?
Dana Lewis
@danamlewis
Oct 25 2016 20:49
Lots of people have considered, several are working on an open source pump…will be excited if someone succeeds but until then, want to keep making it possible for people to loop today.
I don’t have a definitive date
Mike
@libxmike
Oct 25 2016 20:49
@riwarren I see, you want to make interface to talk to, best option will be control with GPIO from Pi Zero
Rick Warren
@riwarren
Oct 25 2016 20:50
Naturally. Okay. Thanks.
Dana Lewis
@danamlewis
Oct 25 2016 20:50
Naturally to which?
Rick Warren
@riwarren
Oct 25 2016 20:50
@libxmike thanks i will check that out
@danamlewis naturally want to be able to closed loop now in the mean time
Dana Lewis
@danamlewis
Oct 25 2016 20:51
:+1:
Scott Leibrand
@scottleibrand
Oct 25 2016 20:51
the other concern with an open source pump is how to prove it's reliable enough. using MDT pumps, we get to take advantage of all the work they did to prove it's reliable (and the FDA's certification of that).
so that means we only have to make sure that we're sending the right commands to the pump, not worry about whether the pump will do as it's told
Rick Warren
@riwarren
Oct 25 2016 20:52
My perspective is that I cannot believe I am forced to use a life support device whos patents were made in 1999. Something needs to be done about that.
Its not reliable though
Scott Leibrand
@scottleibrand
Oct 25 2016 20:52
how so?
Rick Warren
@riwarren
Oct 25 2016 20:53
MDT pump is the most horribly designed electronic device I have ever had to use.
Dana Lewis
@danamlewis
Oct 25 2016 20:53
looping actually reduces the number of interactions you have to do with the pump, though, which is nice
Rick Warren
@riwarren
Oct 25 2016 20:53
It doesnt properly sense no deliveries? Only tells me 8-12 hours after the fact
Scott Leibrand
@scottleibrand
Oct 25 2016 20:53
we're not talking about the UI or the RF. we're talking about reliability of actually delivering the amount of insulin instructed
Paul Martin
@Spazholio
Oct 25 2016 20:53
@riwarren If that’s true, your pump is quite defective.
Scott Leibrand
@scottleibrand
Oct 25 2016 20:54
are there newer pumps with more accurate no delivery sensors? I've heard that many of them simply don't bother because they can't do it well.
Rick Warren
@riwarren
Oct 25 2016 20:54
I have had several pumps all of them were the same in this respect
Scott Leibrand
@scottleibrand
Oct 25 2016 20:54
and :+1: to @Spazholio: at ~1U/hr basal rates, ours detects no delivery within an hour of a reservoir going empty
Rick Warren
@riwarren
Oct 25 2016 20:54
I have the latest veo 754
Paul Martin
@Spazholio
Oct 25 2016 20:54
They told you 8-12 hours after the fact that there was no delivery of insulin? I’m notified within an hour or so, if not sooner. It happens extremely rarely.
Rick Warren
@riwarren
Oct 25 2016 20:55
Im not talking about reservoir being emlty
Dana Lewis
@danamlewis
Oct 25 2016 20:55
I think there’s two things: 1) ineffective sites (the beginning) 2) the actual no delivery by the time there’s a full occlusion. that’s the no delivery alarm.
Rick Warren
@riwarren
Oct 25 2016 20:55
Im talking about a bent canula or bleeding
Paul Martin
@Spazholio
Oct 25 2016 20:55
@riwarren I wasn’t either. =)
Dana Lewis
@danamlewis
Oct 25 2016 20:55
Yea. That’s not always an occlusion to the criteria of what the pump will sense; often there’s SOME delivery at that point
Scott Leibrand
@scottleibrand
Oct 25 2016 20:56
@riwarren those don't result in increased pressure within the reservoir though, because the insulin is going somewhere, just not into the right part of your body
garykidd
@garykidd
Oct 25 2016 20:56
YOu may want to try medtronics new infusion sets with multiple outlets.
Rick Warren
@riwarren
Oct 25 2016 20:56
Only available in the US
In canada we dont have prosets yet
I wish
I have that problem weekly
Scott Leibrand
@scottleibrand
Oct 25 2016 20:56
so it's not actually possible to detect bent cannula with a pressure sensor. the no delivery alarms will only detect complete occlusion that stops all delivery and results in backpresssure
Paul Martin
@Spazholio
Oct 25 2016 20:57
@garykidd Do you have a link to those? I was unaware they existed...
Thanks. =)
Dana Lewis
@danamlewis
Oct 25 2016 20:57
What pump sites are you using? many people who have frequent cannula issues like that try the sure-t and those problems go away
garykidd
@garykidd
Oct 25 2016 20:57
just barely beat you to the punch
Rick Warren
@riwarren
Oct 25 2016 20:57
I put them on my ass cheeks. Only place I have fatty tissue
Scott Leibrand
@scottleibrand
Oct 25 2016 20:57
I really wish FDA would establish bidirectional approval agreements with CA, EU, and AU so that anything approved in one place could be made available everywhere
garykidd
@garykidd
Oct 25 2016 20:58
and there's always steel, however short lived they can be per use...but they dont have bent cannula issues
Dana Lewis
@danamlewis
Oct 25 2016 20:58
yes. if you’re experiencing weekly cannula issues, try Sure-T.
Rick Warren
@riwarren
Oct 25 2016 20:58
Sure t uses a metal canula no? Ill pass....
I wind up having over 33.3 mmol/L blood sugar at least once a week due to canula issues
Dana Lewis
@danamlewis
Oct 25 2016 20:59
@garykidd you had a wifi Q, right?
You may want to re-post, it may have gotten buried too quickly
garykidd
@garykidd
Oct 25 2016 21:00
I think you may be far better off with a different infusion set...if you're having that many problems...you should try other sets....medtronic will even replace infusion sets with others for things like this in my experience.
Rick Warren
@riwarren
Oct 25 2016 21:00
And I just wasted $300 trying to get an enlight sensor in because the inserter wouldnt release the sensor from it afyer insersion. Took 5 sensors to get one in properly
Sandra Keßler
@SandraK82
Oct 25 2016 21:01
I am amazed how wel sure-t works for me. Used the last years the Orbit with luer and good lots of infected insertion sites within 48h. Current sure-t is 52h in place and still no issues
Rick Warren
@riwarren
Oct 25 2016 21:01
Sandra: how does sure-t differ from others? Doesnt the metal canula hurt?
garykidd
@garykidd
Oct 25 2016 21:02
@danamlewis I do, may need to work on specifics of questions a bit. Anyway, was happy to get my edison rig working the other day. Followed directions on how to set up wifi...but when I tried to add a wifi network through nano'ing supplicant.conf...I couldn't connect to either...was wondering if there is something fairly obvious to check by chance.
Rick Warren
@riwarren
Oct 25 2016 21:03
Reset wifi router so it resets its routing tsble and will accept the new device?
Sandra Keßler
@SandraK82
Oct 25 2016 21:04
I used rapid-d a fees years ago and got lots of infected sites and had a lot problems with it. Switched to teflon and got very bad issues with the glue. Switched to Orbit and it worked somehow. But got worse the last six months. Now I'm one week with sure-t and they are great. I don't feel the insertion site
garykidd
@garykidd
Oct 25 2016 21:04
the medtronic sensors are tougher to insert than dex in my experience...also a bit more painful.....I had significant bleeding issues from medtronic enlites.
Rick Warren
@riwarren
Oct 25 2016 21:05
Are dex sensors any more expensive/cheaper?
Sandra Keßler
@SandraK82
Oct 25 2016 21:05
After I finish my blueReader project I will start working on a true open source community based reference design for a pump
Rick Warren
@riwarren
Oct 25 2016 21:05
I wish sensors were covered by insurance
Sandra: I would love to be apart of that. That is my current plan.
garykidd
@garykidd
Oct 25 2016 21:06
The other thing I like about dex sensors is the life...ppl relaly stretch them sometimes...but I seem to always get 2 weeks at least
Sandra Keßler
@SandraK82
Oct 25 2016 21:06
I have some ideas on how to check amounts delivered Insulin as I do Software engineering for a client I the field of very advanced sensors for the last three years
Rick Warren
@riwarren
Oct 25 2016 21:07
Really? 2 weeks?
garykidd
@garykidd
Oct 25 2016 21:08
sure, sometimes longer. You usually can tell because sensor results start to bounce around fairly erratically. Off label of course.
Sandra Keßler
@SandraK82
Oct 25 2016 21:08
From all the data I currently have it should be possible to detect delivery (or not delivery ) within the range of 0.005u to 10u per minute
Rick Warren
@riwarren
Oct 25 2016 21:08
Sandra: you know insulin delivered by the counts of revolution performed bybthe motor. I know the exact custom made count motor to deliver 0.025 units of insulin
Not delivery is determined by resistive force against the motor
I think you are over complicating something simple
What I would like to see is a pump that uses 3ml pen vials instead of having to purchase and fill 3rd party reservoirs
Sandra Keßler
@SandraK82
Oct 25 2016 21:10
That's one way and the currently used way to calculate it. It does not include speed, pressure and other stuff that will help to determine bad sites, bent infusion sets or loose connections
Rick Warren
@riwarren
Oct 25 2016 21:10
Thats just a stupid idea to increase revenues imho
garykidd
@garykidd
Oct 25 2016 21:11
@riwarren you used to be able to get prefilled cartridges
Mike
@libxmike
Oct 25 2016 21:11
@riwarren pen vials wont work, the resistance tho push is to big
Sandra Keßler
@SandraK82
Oct 25 2016 21:11
There was the d-Ton @riwarren it used the 3.x ml humalog penfills. It just did not work reliable
Rick Warren
@riwarren
Oct 25 2016 21:11
Speed? Speed is determibed by the pulse sent to the motor. Pressure is already being determined by magnetic sensor
garykidd
@garykidd
Oct 25 2016 21:12
I was referring to disetronic pump/reservoir--you used to be bale to get pre-filled cartridges...that part was nice
David
@david-t1
Oct 25 2016 21:12
I think Bigfoot's reincarnation of the Snap pump will use prefilled cartridges.
Rick Warren
@riwarren
Oct 25 2016 21:12
Thats too bad. Humalog? Why would a pump use something other than novorapid or apidra?
Sandra Keßler
@SandraK82
Oct 25 2016 21:12
If you want prefilled ones get an insight pump
St the time there was no novorapid as far as I remember
garykidd
@garykidd
Oct 25 2016 21:14
I stll use humalog...you know...insurance. Had liked apidra very much...just not worth the incremental cost to me...which is huge.
Rick Warren
@riwarren
Oct 25 2016 21:15
Ah yeah see I went from taking over 400 units of humulin N and R a day to straight novorapid
Paul Martin
@Spazholio
Oct 25 2016 21:15
I’m using Novolog, but I’ve heard good things about Apidra. Unsure what the diff is though….
Rick Warren
@riwarren
Oct 25 2016 21:15
Apidra has less blockage at infusoon sets
I used it until they had a shortage and it was unavailabke for like 6 month
David
@david-t1
Oct 25 2016 21:15
makes no sense to me either to remove insulin from one container and put it into another...
Mike
@libxmike
Oct 25 2016 21:16
I use Novorapid, never had issue with blockage but I use OmniPod
Paul Martin
@Spazholio
Oct 25 2016 21:16
Does Apidra work faster or leave the system quicker?
Rick Warren
@riwarren
Oct 25 2016 21:17
No
garykidd
@garykidd
Oct 25 2016 21:17
@Spazholio I found it quicker to kick in....
Rick Warren
@riwarren
Oct 25 2016 21:17
Drop in replacement for novorapid
Sandra Keßler
@SandraK82
Oct 25 2016 21:17
I had humalog a long time and can't use it any more. Humalog is to acid and my. Ody developed a resistance resulting inninternal burns on infusion sites and water bubbles. Had about 20 surgical sessions....
garykidd
@garykidd
Oct 25 2016 21:17
and I had less issues with infusion sets
Rick Warren
@riwarren
Oct 25 2016 21:17
Really? I wasnt aware it kicked in faster
@SandraK82 omg thats horrible!
Paul Martin
@Spazholio
Oct 25 2016 21:18
Yeah, I’m fairly certain it’s NOT a drop-in replacement.
Agh. Hello, BG death-spiral. My old friend...
garykidd
@garykidd
Oct 25 2016 21:18
may be more anecdotal than scientifically quanitifiable...but yes..was defintely a little faste rin my experience
Rick Warren
@riwarren
Oct 25 2016 21:18
I used it as a dropin replacement. Didnt change basal rates at all
Mike
@libxmike
Oct 25 2016 21:19
I read and also my Doc told me works same @Spazholio
Sandra Keßler
@SandraK82
Oct 25 2016 21:20
My doc tells me that speed seems to be very dependent on the person in question and a lot of other variables like hormone levels...
Paul Martin
@Spazholio
Oct 25 2016 21:21
Too many friggin’ variables with this disease, I swear...
Rick Warren
@riwarren
Oct 25 2016 21:21
Youre right
Sandra Keßler
@SandraK82
Oct 25 2016 21:22
:+1: @Spazholio
Rick Warren
@riwarren
Oct 25 2016 21:22
It does kick in faster apidra kicks in in minutes and tails off quicker
Paul Martin
@Spazholio
Oct 25 2016 21:22
I wonder if the VA would cover Apidra instead of Novolog….?
R. Dennis Steed
@rdsteed
Oct 25 2016 21:22
Eliminate the variables you can through good engineering and use closed loop to minimize the effects of the rest.
Rick Warren
@riwarren
Oct 25 2016 21:23
Not so great for high fat meals
VA?
Paul Martin
@Spazholio
Oct 25 2016 21:23
@rdsteed Hard to counteract individual biological variables though.
garykidd
@garykidd
Oct 25 2016 21:23
square wave
Sandra Keßler
@SandraK82
Oct 25 2016 21:23
That's now my way @rdsteed and so far it works like a dream
Paul Martin
@Spazholio
Oct 25 2016 21:23
@riwarren Veterans Administration. I’m a vet, and my T1D is considered a service-connected disability so they cover my stuff.
Rick Warren
@riwarren
Oct 25 2016 21:24
I may be switching back to apidra...
Ah
Apidra isnt that much more expensive is it?
Paul Martin
@Spazholio
Oct 25 2016 21:25
No clue.
Scott Leibrand
@scottleibrand
Oct 25 2016 21:25
@Spazholio interesting: wonder how they figure T1 is service-connected
Rick Warren
@riwarren
Oct 25 2016 21:25
I didnt even notice the difference in price. But ive got 80% coverage and pharmacare picks up the other 20%
Scott Leibrand
@scottleibrand
Oct 25 2016 21:25
is it just because you got it while active duty?
Rick Warren
@riwarren
Oct 25 2016 21:25
@scottleibrand my thoughts exactly
Paul Martin
@Spazholio
Oct 25 2016 21:25
Bingo.
Scott Leibrand
@scottleibrand
Oct 25 2016 21:26
I suppose that's as good a rule as any
Sandra Keßler
@SandraK82
Oct 25 2016 21:26
Today we decided to have diner at a nice restaurant here with three-course meal... ditched shortly 170md/dl and that was it
Paul Martin
@Spazholio
Oct 25 2016 21:26
There’s an anecdotal correlation between some servicemembers who got the pertussis vaccine and developing T1D but nothing directly causative.
Rick Warren
@riwarren
Oct 25 2016 21:27
Anyone heard anything about the TB vaccine curing T1D?
Hardwick1973
@Hardwick1973
Oct 25 2016 21:27
I'm struggling with my set-up for my 7 year old daughter - she often goes high quite quickly just after she falls asleep around 8pm. We typically treat this in manual mode with a higher basal rate for the evening time plus a one-off manual bolus as a hit. The method seems to work and we have a good feel for what is needed based on experience. Having moved to OpenAPS the algorithm is struggling to keep up and she's going really high just after she falls asleep. Is there anything I can do?
Paul Martin
@Spazholio
Oct 25 2016 21:27
Oh hell, I just jumped up ~50 points in 20 minutes while riding a low. This is gonna SUCK. =)
Rick Warren
@riwarren
Oct 25 2016 21:27
Ive been tempted to pretend to be a social worker jusy to get it and find out....
Sandra Keßler
@SandraK82
Oct 25 2016 21:27
AMA did not even used above 6.35u/h 😹
Dana Lewis
@danamlewis
Oct 25 2016 21:28
@Hardwick1973 does the evening rise seem to be because of meal? or hormonal needs?
Darrell Wright
@beached
Oct 25 2016 21:28
@riwarren did you say you are using a 754, which one ? Canadian or EU?
Rick Warren
@riwarren
Oct 25 2016 21:28
Is it rebounding from a low?
Paul Martin
@Spazholio
Oct 25 2016 21:28
Oh, how’s that going, btw? The >6.3u/h issue?
Dana Lewis
@danamlewis
Oct 25 2016 21:28
one thing to look at is if you are hitting the basal caps in that time period. and also look at if the basals 2 hours before hand need to be adjusted higher
Rick Warren
@riwarren
Oct 25 2016 21:29
@beached canadian
Hardwick1973
@Hardwick1973
Oct 25 2016 21:29
@danamlewis typically she eats around 5pm so it is more hormonal
Paul Martin
@Spazholio
Oct 25 2016 21:29
@riwarren Yes. I’ve entered all the carbs into NS so the loop is “prepared” but that can only do so much.
Dana Lewis
@danamlewis
Oct 25 2016 21:29
@Hardwick1973 ok. should be helped somewhat with increased basals ~2 or so hours prior, then can be ratched back down after that peak time.
if you were doing a bolus on top of a higher basal, it sounds like basal may need to be increased even more, esp. if loop is then struggling to keep up?
Darrell Wright
@beached
Oct 25 2016 21:30
@riwarren you said latest, is it still firmware 2.7A
Dana Lewis
@danamlewis
Oct 25 2016 21:30
but also is it hitting your max basal (set on pump); or hitting the maxsafebasal (calculated as 3x max daily and 4x max the current basal)?
Scott Leibrand
@scottleibrand
Oct 25 2016 21:30
@Spazholio just waiting on you: openaps/decocare#12
Rick Warren
@riwarren
Oct 25 2016 21:30
@beached yes 2.7A
Sandra Keßler
@SandraK82
Oct 25 2016 21:30
6.35 is possible without the patch @ecc1 supplied @Spazholio I applied the patch and tested it, so far OpenAPS never used it. I needed higher at one time two weeks ago when I was sick and discovered it just was not working
Rick Warren
@riwarren
Oct 25 2016 21:31
I meant the latest model. The pumo is 2 years old now
I checked its 2.7A
Paul Martin
@Spazholio
Oct 25 2016 21:31
I see myself hitting that 6.3u/hr limit….I don’t want to say “often” but often enough so that I’l notice it.
Darrell Wright
@beached
Oct 25 2016 21:31
@riwarren oh good, mine is that version and I get a new one tomorrow
Scott Leibrand
@scottleibrand
Oct 25 2016 21:31
@SandraK82 I think the testing you did might be sufficient to justify merging. Do you still have the results, or do you remember where you posted it? need to get that added into the PR for historical purposes.
Darrell Wright
@beached
Oct 25 2016 21:32
@riwarren oh ok.
Paul Martin
@Spazholio
Oct 25 2016 21:32
@scottleibrand I don’t know how to test the “decoding” bit. The “setting” seems fairly easy to verify. =)
Rick Warren
@riwarren
Oct 25 2016 21:32
@beached why new one?
Scott Leibrand
@scottleibrand
Oct 25 2016 21:32
@Spazholio if it shows up in NS's blue skyscraper view after it enacts, then it was decoded properly.
Darrell Wright
@beached
Oct 25 2016 21:32
@riwarren the 630g isn't loopable
@riwarren and Im over 5yrs
Paul Martin
@Spazholio
Oct 25 2016 21:33
@scottleibrand Sorry - skyscraper view? The blue dots? I’m coming off a low, be kind. =)
Eric
@ecc1
Oct 25 2016 21:33
Or just do openaps use pump temp_basal ...; openaps use pump read_temp_basal
Sandra Keßler
@SandraK82
Oct 25 2016 21:33
@scottleibrand i tested up to 10u/h as I have only one loopable pump and that one is on my body 😝 tested setting basal on the pump and temp basal setting
Rick Warren
@riwarren
Oct 25 2016 21:33
Ah. Too bad I have a 722 that is not being used here coild have saved you some $$
Scott Leibrand
@scottleibrand
Oct 25 2016 21:33
no just the temp basal display
@SandraK82 yeah I remember you posted the results into gitter, but it never made its way into the PR
Sandra Keßler
@SandraK82
Oct 25 2016 21:34
And tested like @ecc1 just suggested
Dana Lewis
@danamlewis
Oct 25 2016 21:34
@riwarren someone would love that if not being used & you’re willing to pass it on
Scott Leibrand
@scottleibrand
Oct 25 2016 21:34
@ecc1 we also want to make sure it shows up in pumphistory.json
which @SandraK82 did test by pasting her NS screenshot
Sandra Keßler
@SandraK82
Oct 25 2016 21:34
Ah sorry i just striked my comments in the or as soon as the second patch was added by @ecc1
Eric
@ecc1
Oct 25 2016 21:35
Yes, I did openaps use pump readhistory or whatever to verify that too. It shows the right values in the JSON output
Scott Leibrand
@scottleibrand
Oct 25 2016 21:35
anyway, I think we're good to merge, as soon as one of you puts your test results into the PR. :)
Rick Warren
@riwarren
Oct 25 2016 21:35
@danamlewis it is in pieces at the moment but I will see if I can get it all put back together and let you know
Eric
@ecc1
Oct 25 2016 21:36
@riwarren Do you have to crack and reglue the case, or is there a less destructive method?
Rick Warren
@riwarren
Oct 25 2016 21:37
It was very tedius but I got it out without damaging the case
You rip the rubber spacers out first
Then pull the boards out
Paul Martin
@Spazholio
Oct 25 2016 21:38
If you’ve disassembled it, I’d venture to say it’s likely not really safe to use anymore?
I mean…I wouldn’t trust it. But that’s just me.
Hardwick1973
@Hardwick1973
Oct 25 2016 21:38
@danamlewis - thanks...I don't believe insulin delivery is being limited by any of the safety settings. However, to check my understanding, why would changing the underlying pump basals impact the AP algorithm - if maxIOB hasn't been reached, doesn't the algorithm calculate how much insulin is needed based on ISF and deviation regardless of whether or not the underlying pump basal is say 0.1 or 1u/h?
Rick Warren
@riwarren
Oct 25 2016 21:39
Thats your own decsion. I would have no issue using it. Im no different from one of medtronics technicians
Dana Lewis
@danamlewis
Oct 25 2016 21:40
@Hardwick1973 yes, but it also has hardware limits of max basal (what you set on pump) and software limits (3x, 4x as described) that it checks against, too. So it may think you haven’t reached maxiob, and need buckets of insulin, but will set max between those 3 factors
Rick Warren
@riwarren
Oct 25 2016 21:40
You would use a refurbished unit they send you would you not? Thats what the self tests on power on are for.
Paul Martin
@Spazholio
Oct 25 2016 21:44
Respectfully, they’re technicians, trained and certified to work on these pumps specifically.
Rick Warren
@riwarren
Oct 25 2016 21:45
If someone else can do something. Theres no reason I cannot as well.
My two cents anyways. your only limited in what your capable of by your own mental limitations.
Paul Martin
@Spazholio
Oct 25 2016 21:46
Oh, I don’t disagree. But the difference (for me) is in training and accountability.
Rick Warren
@riwarren
Oct 25 2016 21:47
Well im not selling my backup pump to anyone. If I want to take it apart and put it bavk together im only accountable to myself.
Paul Martin
@Spazholio
Oct 25 2016 21:48
Fair enough. =)
fiberfan
@fiberfan
Oct 25 2016 22:00
I havent done an official test but applied the 6.4 patch manually and set max basal on the pump and software limits high enough they would make sone people nervous. I watched my loop like a hawk for the first couple of days and syill pay close at It has worked with no problems. Next time openaps does a basal above 6.4 I will add screen print to the pr.
diabeticgonewild
@diabeticgonewild
Oct 25 2016 22:41
@tnharvey You're welcome for the awesome video on how to flash an Edison with Ubilinux on Windows (to find the videos search for "diabeticgonewild" on YouTube.com). My goal was to create a reliable and easy-to-follow tutorial on this, which I had never seen online. Please share the video, as the Edison is superior to the Raspberry Pi setups and others, and is needed in order to get more users wearing OpenAPS devices 24/7.
Dana Lewis
@danamlewis
Oct 25 2016 22:42
(just note we’re recommending jubilinux moving forward)
diabeticgonewild
@diabeticgonewild
Oct 25 2016 22:58
@danamlewis Thanks. Also, there's only 2 minor differences in steps between flashing Ubi/Jubi. So it works as a tutorial for both.
Scott Leibrand
@scottleibrand
Oct 25 2016 23:08
yeah, any method that successfully flashes ubilinux should be usable for jubilinux as well.
fiberfan
@fiberfan
Oct 25 2016 23:46
@scottleibrand I just commented on openaps/decocare#12 with a clip of a ns screen shot showing a basal of 7.15.
Dana Lewis
@danamlewis
Oct 25 2016 23:47
:+1: