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

20th
Feb 2015
Matthias Granberry
@mgranberry
Feb 20 2015 00:00
@diabeticgonewild the spec that I have access to has a pretty large command set and only a very small number of commands documented, but I have done this kind of thing before and the tidepool uploader gives me a pretty powerful set of reverse engineering tools. I'm pretty sure that they issue commands through USB in their testbed
diabeticgonewild
@diabeticgonewild
Feb 20 2015 00:00
Yeah, unless we get help from Tandem, I don't think it can be done without reprogramming the whole pump, as they are the only ones that have access to the firmware upgrade that activates the Bluetooth chips in current t:slims
very nice!
I mean, there has to be a way to copy the RTOS (real time operating system) used on the t:slim, and just use an IDE to simulate it and try to program a driver for the Bluetooth chip
but IDK, not an area of expertise for me
Matthias Granberry
@mgranberry
Feb 20 2015 00:05
it gets tricky when you want to deal with compiled code. It is pretty slow going, even if you have the tools to do the analysis, and those get expensive pretty quickly. Optimizing compilers obfuscate code a lot and the fixed 32-bit instructions and barrel shifter that ARM uses make even reading what constants are used in the raw assembly a little tricky
and you lose all the function names, etc.
diabeticgonewild
@diabeticgonewild
Feb 20 2015 00:06
Oh, so that's not going to work
Dominik Stefan
@DominikStefan
Feb 20 2015 00:07
@scottleibrand Thanks for the nudge yesterday! :smile: I definitely needed it. Not there yet, but definitely made some good progress since. My next step is figuring out how to give decoding-carelink access to my mongo/nightscout sensor glucose values data.
still have yet to wrap my head around that one
Matthias Granberry
@mgranberry
Feb 20 2015 00:10
@diabeticgonewild it's still possible that it is just a command that needs to be issued over USB to enable bluetooth discovery, but even if I had that, I have no idea what discovery or the protocol would look like. Most of the pump commands would work well over BLE, but not all of them. I wish I knew an engineer out there that I could take out drinking.
diabeticgonewild
@diabeticgonewild
Feb 20 2015 00:14
LOL, I know a dude who is an EE who worked in controls who has a daughter with T1D. He is working on sniffing the Omnipod
Matthias Granberry
@mgranberry
Feb 20 2015 00:15
I have all the hardware here to do that. I just don't know anyone with one and I have a teflon allergy, so no patch pumps for me.
diabeticgonewild
@diabeticgonewild
Feb 20 2015 00:16
Nice! But the teflon allergy has to suck! I am allergic to all tape ftw!
Matthias Granberry
@mgranberry
Feb 20 2015 00:16
it just leaves me with steel infusion sets as the only option. They seem to work well, though, and kinked cannulas are a non-issue.
diabeticgonewild
@diabeticgonewild
Feb 20 2015 00:17
Cool!
Toby Canning
@TC2013
Feb 20 2015 01:43
@mgranberry email me with your address and I will send you the omnipod stuff (toby at canning dot us)
Scott Leibrand
@scottleibrand
Feb 20 2015 04:19
Dominik: we've seen that the pump will implement something like a 0.675U/hr basal when you ask it to do something like 0.7. Seems the pump's basal rate units are based on non-decimal fractions.
It doesn't hurt anything, as long as you design the system to be ok with the basal rate being "close enough" (within .1U/hr)
you can ask it to set it to whatever you want, and it will pick the closest it's capable of.
Dominik Stefan
@DominikStefan
Feb 20 2015 06:37
As long as I know it's there, I can work with/around it no problem.
Scott Leibrand
@scottleibrand
Feb 20 2015 06:38
Ya
Dominik Stefan
@DominikStefan
Feb 20 2015 06:44
Plus, once the loop is "closed" I won't be mentally giving it nearly as much thought anyways (as majority of basal decisions won't be made by me anymore! :smile:
Dana Lewis
@danamlewis
Feb 20 2015 06:45
Exactly :)
Dominik Stefan
@DominikStefan
Feb 20 2015 06:45
so yeah, it's definitely "close enough" for me :smile:
Dominik Stefan
@DominikStefan
Feb 20 2015 07:13
with all the variables that influence bg's, I wouldn't see the effects of a difference of .02u/h anyways, but I'm glad to hear that it wouldn't hurt the system in any way
diabeticgonewild
@diabeticgonewild
Feb 20 2015 10:58
Woke up to a 48. This is the second time tonight with a BG less than 50, plus I have had nights recently low. Time to change my insulin dosages overall to avoid a crisis
diabeticgonewild
@diabeticgonewild
Feb 20 2015 11:07
I am changing them overall because for nights it's like landing an airplane. You have to have a stable landing to actually land stably.
Like you have to have a stable landing "at night", to wake up with a decent BG "in the morning"
  • at night when you go to bed. Anyways, I need to chill out. Still low.
diabeticgonewild
@diabeticgonewild
Feb 20 2015 12:17
I meant that you have to have a stable descent (stable blood sugar when you go to bed) in order to have a stable landing (decent blood sugar when you wake up in the morning). That's why I am changing my insulin dosages overall. I was just confused when I wrote the above, and my low blood sugar stupor finally just went away.
Dominik Stefan
@DominikStefan
Feb 20 2015 13:12
48's suck! :-1: We're fairly well acquainted though, 48's and me. :smile:
too well if you ask me :smile:
diabeticgonewild
@diabeticgonewild
Feb 20 2015 13:13
LOL!
Matthias Granberry
@mgranberry
Feb 20 2015 13:48
I've been losing about .75 lbs a week for the past year and basal rate adjustments are a constant source of pain because my liver also appears to overproduce glucose. I can't go more than 2 weeks without decreasing them, but the peaks and valleys also move a little around weight plateaus.
Dominik Stefan
@DominikStefan
Feb 20 2015 13:49
They're trouble makers, those 48's (lows in general)!
Chris Oattes
@cjo20
Feb 20 2015 14:11
mgranberry: what have you been doing to lose weight?
I could do with being about 40lb lighter
Dominik Stefan
@DominikStefan
Feb 20 2015 14:17
@mgranberry I couldn’t agree more. The need for constant basal rate adjustments is incredibly frustrating. I'm lucky to make it a day with the same basal rate working for me for the entire duration of the day. The day (food, exercise, stress, etc) dictates basal needs. No fixed/predetermined amount ever will. Something I think many of us have painstakingly figured out over the years. This is akin to trying to predict the future – and what PWD are currently having to resort to when determining their basal rates. Stabilizing my overnight bg is probably the biggest reason I'm as excited as I am about OpenAPS and DIYPS! :smile:
Chris Oattes
@cjo20
Feb 20 2015 14:21
I'm quite lucky overnight, my basal requirements don't seem to change that much
Matthias Granberry
@mgranberry
Feb 20 2015 15:22
@cjo20 I exercise daily and fill up on fiber. I've essentially set up a weightlifting gym at home and have put on a fair bit of muscle, along with exercise almost exclusively as my corrections to minimize insulin use. It seems to be working, but things go to hell control wise if I stop doing it for any reason
Chris Oattes
@cjo20
Feb 20 2015 15:23
I think i probably need to put on muscle as well as lose fat. I can do about 4 pushups at the moment :P
Matthias Granberry
@mgranberry
Feb 20 2015 15:27
That's where I was. The body builds up quite quickly if you give muscles a day to recover between workouts. Just aim for lifting enough that <10 reps tires you out and it goes quickly and you get better results as well as encouraging bigger glycogen stores, which helps with post-meal control too. Just watch for 7-12h delayed hypos if you overdo it.
Chris Oattes
@cjo20
Feb 20 2015 15:50
heh, < 10 proper pressups wear me out
Toby Canning
@TC2013
Feb 20 2015 16:01
Is a pressup a pushup?
Chris Oattes
@cjo20
Feb 20 2015 16:08
TC2013: yes
@TC2013 they can be used interchangably
Toby Canning
@TC2013
Feb 20 2015 16:37
Then, I am not very good at either pressups or pushups either!
However, I am a master of the pressdown, pushdown, pulldown, etc. I have an exceptional mastery of gravity.
I think it is the extra mass working in my favor.
My apologies for all the bragging.
Scott Leibrand
@scottleibrand
Feb 20 2015 17:04
Lol
Dana Lewis
@danamlewis
Feb 20 2015 17:18
:)
Chris Oattes
@cjo20
Feb 20 2015 17:37
@TC2013 I understand exactly where you're coming from
Toby Canning
@TC2013
Feb 20 2015 21:53
@scottleibrand @danamlewis One of the things that may help with openaps is to have a post or central document with our definitions, common variables, and abbreviations. This way we can avoid confusion and all use a common language, rather that ending up with things like Bolus on Board, Insulin on Board, Carbs on Board, Active Insulin, Active Carbs, etc left to be interpreted and defined by each developer.
I'm hoping you'll take the first crack at it? fingers crossed
Dana Lewis
@danamlewis
Feb 20 2015 21:55
Yes :)
Anything else we should define?
Toby Canning
@TC2013
Feb 20 2015 21:57
Probably, but I'm not sure what terms to give them
Insulin curve - the rate at which insulin becomes active in the BG after delivery
Predicted BG should probably just be defined by a simple equation
Dana Lewis
@danamlewis
Feb 20 2015 21:58
Yea. Worst case I can start it and can add to it, just like everything else we are doing
Toby Canning
@TC2013
Feb 20 2015 22:01
Especially some safety concepts should be defined. Like the "Max IOB" the autopilot can give following the insulin curve. Ahhh DIA, Duration of Insulin Activity
Min, Max and Target BG
Simple definition, but best to all use same terms and abbreviations
Dana Lewis
@danamlewis
Feb 20 2015 22:02
Any chance you want to dump this all in an email to me? :D if not I can copy paste later
Toby Canning
@TC2013
Feb 20 2015 22:03
Will do!
Dana Lewis
@danamlewis
Feb 20 2015 22:03
Thanks!
Toby Canning
@TC2013
Feb 20 2015 22:03
You're getting stream of consciousness!
Dana Lewis
@danamlewis
Feb 20 2015 22:23
You rock, thanks for dropping that in email (my to do list) :)