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

30th
Jun 2015
diabeticgonewild
@diabeticgonewild
Jun 30 2015 00:08
At JDRF research update. Not many people are here yet but it's going to be a nice event. Anyone have questions? http://m.imgur.com/ry4tnwO
diabeticgonewild
@diabeticgonewild
Jun 30 2015 00:18
Anyone? Or forever hold your peace.
Darrell Wright
@beached
Jun 30 2015 00:21
Quicker insulin :) will it happen?
As in shorter 1/2 life so that it is out of the system
not inhaled
diabeticgonewild
@diabeticgonewild
Jun 30 2015 00:38
Got it. He's sitting in the row in front of me
Darrell Wright
@beached
Jun 30 2015 01:07
Was there a site that had a bunch of medtronic rf data packet captures? I thought I saw it months ago
Paul Andrel
@stavlor
Jun 30 2015 01:47
I know @bewest had a few
Darrell Wright
@beached
Jun 30 2015 02:02
I hexdump to cat a mmcommander and a looking now and pushing them through the medtronic_rf tools
I hooked hexdump up to...
diabeticgonewild
@diabeticgonewild
Jun 30 2015 02:30
Yeah he was like WTF you can go to jail prison etc for redistributing code
😔
diabeticgonewild
@diabeticgonewild
Jun 30 2015 02:51
That was stressful to me.
diabeticgonewild
@diabeticgonewild
Jun 30 2015 02:59
Like I haven't been feeling well since I got my infusion on Thursday and I slept all day and now I am exhausted from the event and stressed out. Life is a drag. 👿 Nobody knows how I truly feel but it's something like this 💩
Jason Calabrese
@jasoncalabrese
Jun 30 2015 03:11
I don't get jdrf's response to all this, they actually blocked several of us from speaking at the T1 nation events, and that was after Ben spent a day in Aaron's office
diabeticgonewild
@diabeticgonewild
Jun 30 2015 03:13
Yeah he was like posting code is sooooooooooooo bad
I have the recoding on my Dropbox if you want to listen to the event. My question was in the last 5 minutes.
I asked about we are not waiting and openaps and why the JDRF doesn't take an active role in it, as it develops safety and efficacy and bugs that would not be fixed are fixed
Cause of the FDA being slow
I told him that my cousin who puts in applications for medical devices that it takes 3 years for a class 3 device in Europe while its 5 years for a class 3 device in the US
He said no doubt that it made the FDA do things faster but it seriously stressed me out about everything else pretty much.
Jason Calabrese
@jasoncalabrese
Jun 30 2015 03:20
it's so disappointing , sure you didn't need the extra stress
diabeticgonewild
@diabeticgonewild
Jun 30 2015 03:22
Nope, but i will get over it. At least I'm not like 😲🔫 (yes I have a history of having that problem and I don't really care what anyone thinks about it). But I am calming down. We are all fighting the same disease and he used scare tactics
There is a disconnect I guess
But thanks. I don't need that stress. I think being on two forms of immunosuppression has done me in.
He pretty much condemned anyone who has publicly posted code for dosing insulin in any form
diabeticgonewild
@diabeticgonewild
Jun 30 2015 04:00
Like it's weird. They befriend people who develop products or make it into fruition directly into industry but don't want anything to do with anyone outside of product development, in industry. Like the JDRF in the past 10 or so years has been more about product development than research (by far) and perhaps that's one of the reasons, of many, why the JDRF doesn't stand for anything as an acronym
Like if it's not a commercialized thing or is potentially going to be commercialized then they don't want anything to do with it. It's super weird and even kind of elitist.
diabeticgonewild
@diabeticgonewild
Jun 30 2015 04:08
They also have had people serve on the board of the JDRF whose companies have received direct funding for product development.
There was some controversy, like around 7-8 years ago when the JDRF stood for the Juvenile Diabetes Research Foundation, and they were funding mainly product development
Scott Leibrand
@scottleibrand
Jun 30 2015 04:38
@amazaheri you around?
Ali Mazaheri
@amazaheri
Jun 30 2015 04:44
10 4
Scott Leibrand
@scottleibrand
Jun 30 2015 04:44
so I checked your repo, and noticed you reverted the timestamp to date change you emailed me about?
Ali Mazaheri
@amazaheri
Jun 30 2015 04:44
right
Scott Leibrand
@scottleibrand
Jun 30 2015 04:44
wondering what the backstory is there. : )
Ali Mazaheri
@amazaheri
Jun 30 2015 04:45
data, at this point I am using date and looks like the iob is being calculated considering the bolus
Scott Leibrand
@scottleibrand
Jun 30 2015 04:46
so I've seen the same issue when sorting by timestamp, but I can't see how sorting by date would help
Ali Mazaheri
@amazaheri
Jun 30 2015 04:46
i will update the repo doing some testing
Scott Leibrand
@scottleibrand
Jun 30 2015 04:49
with regard to the pump commands failing, it might be related to not having enough history, or having something in the history that decocare can't handle...
@bewest doesn't think it should be a firmware version problem
Ali Mazaheri
@amazaheri
Jun 30 2015 04:49
I noticed that by timestamp bolus was not being included as sort was putting bolus as first item in array and temp basal with older date at end and check was failing
Scott Leibrand
@scottleibrand
Jun 30 2015 04:50
what I usually do when I suspect messed-up pump data confusing decocare is issue about 100 repeated temp basal commands. :)
Ali Mazaheri
@amazaheri
Jun 30 2015 04:50
with sorting based on date it seems to be working, coupld be data related (pump history)
:)
Ben West
@bewest
Jun 30 2015 04:51
can you dump some json where this happens?
like known input where the resulting output is not known
maybe just update your repo at that state
or if it's in tehre
just which commit
Scott Leibrand
@scottleibrand
Jun 30 2015 04:52
so we basically just want to check out a copy of your repo in the state where iob.js is causing the issue but sorting by date fixes it
we might be able to repro another way, but if you already have a repro we can use, all the better
Ali Mazaheri
@amazaheri
Jun 30 2015 04:52
ok
I do that
in a sec
Ben I sent u the list of commands not working with 512 1.4, as of now read_settings, read_bg_targets and Status are not working
Ben West
@bewest
Jun 30 2015 04:54
thanks
Ali Mazaheri
@amazaheri
Jun 30 2015 04:54
give me a sec and I send the json post manual basal
diabeticgonewild
@diabeticgonewild
Jun 30 2015 04:58
I'm soooo so tired. :(
Just ate some immunosuppressants....yummy. One time I was in the hospital and they wouldn't give them to me even though my neurologist approved me taking it (duh) cause it was a "dangerous drug" (it's an anti-rejection drug). I'm on two forms of immunosuppression. This goes for my T cells, the infusions I had a month ago goes after my B cells. Fun.
Ben West
@bewest
Jun 30 2015 05:01
great ,thanks, @amazaheri
new Date( )
Ali Mazaheri
@amazaheri
Jun 30 2015 05:16
done, you can take a look
Ben West
@bewest
Jun 30 2015 05:16
sudo easy_install -ZU openaps
Scott Leibrand
@scottleibrand
Jun 30 2015 05:20
sleibrand@diyps:~/openaps-dev$ nodejs iob.js pumphistory.ali.json
{"iob":-0.06597132541229761,"activity":-0.12325809523809521}
@amazaheri does that look right?
Ali Mazaheri
@amazaheri
Jun 30 2015 05:23
after .8 bolus, if I sort by date I get {"iob":0.8684284107355555,"activity":0.09338835978835976}, and sorting by timestamp I get
{"iob":0.07301440193555557,"activity":-0.07180105820105814}
Scott Leibrand
@scottleibrand
Jun 30 2015 05:23
I'm suspecting not
k
Ali Mazaheri
@amazaheri
Jun 30 2015 05:23
i also pass profile.json along with pumphistory.json
Ben West
@bewest
Jun 30 2015 05:24
the history .json is fine
it should be higher or around 0.8
Scott Leibrand
@scottleibrand
Jun 30 2015 05:24
oh yeah, you updated iob.js to take profile as well
Ali Mazaheri
@amazaheri
Jun 30 2015 05:24
yeap
Scott Leibrand
@scottleibrand
Jun 30 2015 05:24
did that change get pushed/pulled?
Ali Mazaheri
@amazaheri
Jun 30 2015 05:24
it should
Ben West
@bewest
Jun 30 2015 05:24
not 0.07 or 0.06
Ali Mazaheri
@amazaheri
Jun 30 2015 05:25
@bewest that's when u sort by .date
Ben West
@bewest
Jun 30 2015 05:25
ok
got it
have noahter
Ali Mazaheri
@amazaheri
Jun 30 2015 05:26
I haven't updated the iob.js but if we are good with change I push it
Ben West
@bewest
Jun 30 2015 05:27
amazaheri/openaps-dev@bb3d601 ?
if we have any patients left, I'd like to try treatments.sort(function (a, b) { return Date.parse(a.timestamp) < Date.parse(b.timestamp); }); if we can
patience even
haa
Ali Mazaheri
@amazaheri
Jun 30 2015 05:28
i think i did that
let me test again
in a sec
Ben West
@bewest
Jun 30 2015 05:29
oh!? and you still found .date was better?
Ali Mazaheri
@amazaheri
Jun 30 2015 05:31
{"iob":-0.099955552,"activity":-0.0017777777777777774}
Ben West
@bewest
Jun 30 2015 05:31
oh
well
Scott Leibrand
@scottleibrand
Jun 30 2015 05:33
ok, so I just tested ali's fix, and another of mine (to use now instead of lastTimestamp), and it looks good to me.
ali, you want to re-push your fix, or should I just re-do it on my branch?
Ali Mazaheri
@amazaheri
Jun 30 2015 05:33
let me push
done
Scott Leibrand
@scottleibrand
Jun 30 2015 05:46
ok, resolved the merge conflict and pushed the result to https://github.com/scottleibrand/openaps-dev
we really need to move this code out of the data repo
Ali Mazaheri
@amazaheri
Jun 30 2015 05:48
:+1:
now we need to merge your other iob.js updates
Ali Mazaheri
@amazaheri
Jun 30 2015 06:03
along with getprofile.js and determine-basal.js
Scott Leibrand
@scottleibrand
Jun 30 2015 06:03
just added you as a member of the plugin-authors team
which gives you perms to push to this repo
go ahead and clone it, copy your changes over, and push them
Ali Mazaheri
@amazaheri
Jun 30 2015 06:03
ok
Ben West
@bewest
Jun 30 2015 06:08
{ "bin" : { "myapp" : "./cli.js" } }
Ali Mazaheri
@amazaheri
Jun 30 2015 06:14
:clap:
Scott Leibrand
@scottleibrand
Jun 30 2015 07:03
created an npm package that should let us install it with npm install, but can't get it working for some reason
but, time for bed
Jason Calabrese
@jasoncalabrese
Jun 30 2015 07:17
some updated NS docs, if anyone over here is interested https://github.com/nightscout/cgm-remote-monitor/blob/wip/maker/README.md
Ali Mazaheri
@amazaheri
Jun 30 2015 08:01
Awesome stuff @jasoncalabrese thank you!
Ali Mazaheri
@amazaheri
Jun 30 2015 15:10
@jasoncalabrese for cob and other add ons do I need to deploy from dev? I have iob but others not showing up
Geramy
@Geramy
Jun 30 2015 15:37
@bewest did you end up sorting the data?
Paul Andrel
@stavlor
Jun 30 2015 16:11
@amazaheri Yes, I believe master is missing most of the plugin stuff for pushover/iob/cob/bwp/basal and stuff thats in dev ifttt is still in wip/maker while @jasoncalabrese is working on it
Jason Calabrese
@jasoncalabrese
Jun 30 2015 17:23
yeah dev or wip/maker, I'll merge maker after someone else gets it setup
Paul Andrel
@stavlor
Jun 30 2015 17:25
was just deploying it now to check for issues will let you know
Jason Calabrese
@jasoncalabrese
Jun 30 2015 17:34
:+1:
should discuss it in the main room if you find anything
Ali Mazaheri
@amazaheri
Jun 30 2015 18:28
ok I go with Dev, have some ideas about integrating NS with OpenAPS :smile:
Jason Calabrese
@jasoncalabrese
Jun 30 2015 18:59
I see NS being a dashboard for OpenAPS, show what has happened, what might happen next, but not control an pumps
Dana Lewis
@danamlewis
Jun 30 2015 19:00
:)
Matthias Granberry
@mgranberry
Jun 30 2015 19:04
@jasoncalabrese I think that would be useful even after commercial APS systems come out, assuming they don't lock them up and refuse to give the user the keys.
coughmedtroniccough
Jason Calabrese
@jasoncalabrese
Jun 30 2015 19:14
I also want to make NS even more hackable, so if you wanted to hook up something custom you can do it in a way that won't make it impossible to stay current. Think we're mostly there on the server, but the client side still needs a good amount of work
Ali Mazaheri
@amazaheri
Jun 30 2015 20:29
Exactly, I am using powerbi now but we can do that with NS
Paul Andrel
@stavlor
Jun 30 2015 20:33
@amazaheri interesting
Paul Andrel
@stavlor
Jun 30 2015 21:02
@amazaheri how are you getting data from NS in to power bi?
Dana Lewis
@danamlewis
Jun 30 2015 21:06
@mgranberry Exactly my thoughts, too
Ali Mazaheri
@amazaheri
Jun 30 2015 21:45
right now I am pushing the data directly to an Azure SQL
I can also have a simp0le job porting the data from Mongo to SQL but I was thinking about ditching PowerBI and visualize the data on NS directly
Ali Mazaheri
@amazaheri
Jun 30 2015 21:52
@scottleibrand I am going to delete my openaps-dev repo and push to openaps-js moving forward
Ben West
@bewest
Jun 30 2015 21:53
@amazaheri the openaps-js repo is for code only
Ali Mazaheri
@amazaheri
Jun 30 2015 21:53
correct
Ben West
@bewest
Jun 30 2015 21:53
your openaps-dev repo is for your data and config only
or you can have others
Ali Mazaheri
@amazaheri
Jun 30 2015 21:54
10 4
Ben West
@bewest
Jun 30 2015 21:54
:-)
Ali Mazaheri
@amazaheri
Jun 30 2015 21:54
any help on 512 :)
Ben West
@bewest
Jun 30 2015 21:54
might have something next week
Ali Mazaheri
@amazaheri
Jun 30 2015 21:54
I hate to hard code stuff , like back to the future
Ben West
@bewest
Jun 30 2015 21:54
my 512 broke
Ali Mazaheri
@amazaheri
Jun 30 2015 21:54
:)
I love our tiny 512 @danamlewis :)
great
Ben West
@bewest
Jun 30 2015 21:55
the error you got made it seem like there was something up with decocare
but it works with other pumps?
Ali Mazaheri
@amazaheri
Jun 30 2015 21:55
works with 722
2.2 A
Ben West
@bewest
Jun 30 2015 21:55
and same commands are only weird with same setup on 512?
Ali Mazaheri
@amazaheri
Jun 30 2015 21:55
yes
in 1.4 we don't have basal profile a, b,c but I can still get std
also no way to add DIA
Ben West
@bewest
Jun 30 2015 21:56
yeah, some commands changed, vs some are absent in 512
Ali Mazaheri
@amazaheri
Jun 30 2015 21:56
but rest are ok
Ben West
@bewest
Jun 30 2015 21:56
so if that's the problem, just need to guard against the bad ones better somehow
there's a list of commands per model in models.py
would be a matter of moving the bad ones from the top in the base to 515+
then they ... will generate a different error
lol
will think about another way to generate a hint
so it doesn't show up in the menu
Ali Mazaheri
@amazaheri
Jun 30 2015 21:57
not urgent but I am sure we will have folks with 512
diabeticgonewild
@diabeticgonewild
Jun 30 2015 22:52
Yeah I feel like my batteries are running empty all the time. I'm completely worn out and out of the loop with OpenAPS. It's almost as bad as getting plasma exchange/plasmapheresis except it lasts longer which makes it worse. I'm pretty sure it's the immunosuppression.
diabeticgonewild
@diabeticgonewild
Jun 30 2015 22:58
I just woke up an hour ago and I have no energy even to watch TV