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

27th
Feb 2016
Jason Calabrese
@jasoncalabrese
Feb 27 2016 00:54
got my HooToo, even smaller than I expected
Dana Lewis
@danamlewis
Feb 27 2016 00:56
I love it. :+1:
Ben West
@bewest
Feb 27 2016 00:56
This message was deleted
fda was explicit with us "we just gave dexcom this de novo thing, you should do the same thing"
it's explicitly for secondary display real-time, which is different from non-secondary display ;-)
for share cloud, it's not clear that this is primary display, for apple health, apple health is hard to recognize as a primary display...
Howard Look
@HowardLook
Feb 27 2016 01:52
The de novo was “We are making a new classification for retrospective data review software, and we will consider it Class 1/Exempt.” It’s category 862.2120, product code PHV. Specifically calls out “retrospective” but not 3 hours. If I had to guess (but do not know), it was a conversation something like “Well 3 hours is clearly retrospective, becaust insulin lasts about that long, but less than that could be used for real-time therapy). Stayce was very clear with us, and agreed for us to document, that real-time secondary display was Class2/Exempt.
Kevin Wells
@wkcwells
Feb 27 2016 02:23
Looks like my problem with NS treatment display is that it displays treatments 8 hours before the y occurred. Must be a timezone thing :(
Ring any bells?
Scott Leibrand
@scottleibrand
Feb 27 2016 02:32
Do you have your files -zoned?
with tz?
I think that part is documented
and if you have them law reports, are you calling the -zoned versions for upload?
Jim Matheson
@jmatheson
Feb 27 2016 02:52
has anyone with an edison setup gotten bluetooth set up with a hotspot (instead of wifi)? Using bluetoothctl I was able to pair the Edison with the phone, but can’t seem to get it to connect. Getting an error "a2dp-source profile connect failed. Protocol not available”. It seems like some sort of audio profile, so not sure the complaint and most google links are taking me to an intel page behind a wall.
Scott Leibrand
@scottleibrand
Feb 27 2016 02:53
I was not able to get it working. Tried the exact same config that works on a Pi. Maybe a package or kernel version thing?
Jim Matheson
@jmatheson
Feb 27 2016 02:54
did you get the same error? I do remember you having issues...
Scott Leibrand
@scottleibrand
Feb 27 2016 02:55
No, I was doing it differently than it sounds like you are.
Jim Matheson
@jmatheson
Feb 27 2016 02:55
k. will keep you posted if i get it working :)
Scott Leibrand
@scottleibrand
Feb 27 2016 02:55
I'm not using a2dp (which is Bluetooth stereo IIRC)
what my phone uses to stream Bluetooth audio to my car.
Jim Matheson
@jmatheson
Feb 27 2016 03:09
damn now my /home/root is full
Jason Calabrese
@jasoncalabrese
Feb 27 2016 03:36
medtronic wants me to register a serial number to be able to get supplies via insurance, wondering if that's a bad idea. Talked to the local San Diego rep, she never heard of OpenAPS and was trying to sell me on their sensors, sent her to the .org site and twitter...
Rachel Sandlain
@audiefile
Feb 27 2016 03:43
I'm using a circa 2004 pump and they didn't say anything to me... I did say I was test driving going back to a tubed (implied I might upgrade) I think they want to verify you actually have a pump and it's not insurance fraud
Jim Matheson
@jmatheson
Feb 27 2016 03:55
cleaned out my logs and anything obvious i could find. ran apt-get cleanup. still maxed out.
/usr/lib has 420M in it, /usr/share 383
Jason Calabrese
@jasoncalabrese
Feb 27 2016 04:11
There's a command to find big packages
Jim Matheson
@jmatheson
Feb 27 2016 04:11
i’ve found all the biggest files but seem to be important
i probably have packages i have installed but dont need anymore. not sure how to remember what those are though :)
Scott Leibrand
@scottleibrand
Feb 27 2016 04:13
dpkg-query -Wf '${Installed-Size}\t${Package}\n' | sort -n
Jim Matheson
@jmatheson
Feb 27 2016 04:16
thanks
seems like maybe a volume of packages rather than large ones
that command has 669 results in it.
Matthias
@ceben80
Feb 27 2016 04:30
A wonderful good morning and greetings from Germany. I have again running a great loop night with a totally flatline.
It's so great :sparkles: :sparkles: :sparkles:
Scott Leibrand
@scottleibrand
Feb 27 2016 04:31
:-D
Dragan Maksimovic
@dm61
Feb 27 2016 04:38

hmm, tried that git gc to clean things up, and got this:

pi@raspberrypi ~/myaps2 $ git gc
Counting objects: 185954, done.
Delta compression using up to 4 threads.
Compressing objects: 100% (158964/158964), done.
fatal: unable to read fcc3004d0ce44e8e4ae8995c7e78eca84340da3c
error: failed to run repack

looks like I have something more seriously corrupted?

Kevin Wells
@wkcwells
Feb 27 2016 06:17
@scottleibrand Thanks for the suggestions. I had everything working. Something must have changed and I forgot something I did, or maybe updated a repo… something. Still digging.
Scott Leibrand
@scottleibrand
Feb 27 2016 11:11
@dm61 does git status fail?
dougtakeuti
@dougtakeuti
Feb 27 2016 11:12
Indirect consequences of closing the loop: we became more aware of the different parameters, such as carb ratios, basal rates and isfs, and could do little tweaks on them. We also decided to zero the carbs in the morning snack, since the annoying school schedule has it 45 min before lunch. I know that a couple of weeks isn't a ton of data yet, but it makes me happy when I see in the reports a decrease in average BG +- std dev from 150 mg/dL +- 50 mL to 130 mg/dL +- 30 mg/dL. Time in range hasn't changed from 75% yet. The 25% off range is mostly high BG during school. Now that I'm done with the edison rig I hope to see improvements during the day. That came to us at the right time, as adolescence is starting to kick in!
Scott Leibrand
@scottleibrand
Feb 27 2016 11:13
:-)
live4sw
@live4sw
Feb 27 2016 12:51
Hi guys - making progress slowly and finally up to the "enact" stage, just want to make sure I am thinking about this logically
First, it seems clear that enacted.json should be a JSON report that uses suggested.json to set_temp_basal
So I'm doing that as openaps report add enact/enacted.json JSON pump set_temp_basal enact/suggested.json
That seems simple enough. Then the issue is that I want the alias to make sure that (i) I am clearing out the enacted folder, then (ii) invoking suggested.json, (iii) then using grep to check to see if duration is true or false, (iv) then if it is true enacting the enacted.json report, and (v) if it is false, doing nothing or maybe echoing something like nothing to enact if I want to be verbose
live4sw
@live4sw
Feb 27 2016 12:57
does that sound logical?
Oskar Pearson
@oskarpearson
Feb 27 2016 13:35
@live4sw Seems about right to me :)
BTW: I know many people put lots of logic into aliases - but I generally try keep it in my loop script
so it's in bash
live4sw
@live4sw
Feb 27 2016 13:43
thanks! will keep that in mind as I keep working on this
Oskar Pearson
@oskarpearson
Feb 27 2016 14:09
@bewest Hey there. We've got a situation where decocare isn't picking up the latest glucose data from the medtronic sensor. If we look at the pump screen it has data that's a minute old, but decocare only has data from a few hours back
I've tried changing to the dev branch, but still no data
I'm busy organising to pull the page
Interestingly, though, it's the same page number as bewest/decoding-carelink#145
(From November last year)
dougtakeuti
@dougtakeuti
Feb 27 2016 14:16
@oskarpearson could this be the same issue that @johnmales and I have observed? Some days, due to the alignment of the planets, mood of the Greek gods and brightness of the Sirius star, the MDT pump sends info to compile all reports BUT glucose. It happened to me once and to him twice, I think. Therefore not frequently enough to see any patterns.
Oskar Pearson
@oskarpearson
Feb 27 2016 14:17
Yeah - pretty sure it's the same thing :)
Have you got a github issue for me to compare, perhaps?
dougtakeuti
@dougtakeuti
Feb 27 2016 14:17
I'm afraid not
we should start tracking these events
It lasted a few (4-6) hours and then got back to normal
I hadn't closed the loop by then
Jason Calabrese
@jasoncalabrese
Feb 27 2016 15:20
@juliatakeuti running the loop at school has helped us almost as much as running it at night, working so much better than I expected. With the predictable schedules of school (at least in lower grades) I'm hoping it's repeatable.
dougtakeuti
@dougtakeuti
Feb 27 2016 15:21
First day of meal-assist resulted in more time in range, ~90%, but 4 mild (~70 mg/dL) lows, when the average would be 0 or 1 low per day. We've been with Julia all the time to take action quickly. Apparently meal-assist is a bit too agressive and, since her basal rate is an average of 0.50 UI/h, there isn't much room to low-temp while BG is falling quickly. Any suggestions on how to make adjustments? I also have to be extra careful since I'm using Enlites.
@jasoncalabrese great to hear that Jason!
Dragan Maksimovic
@dm61
Feb 27 2016 15:24
@scottleibrand git status shows I have some untracked files (which is probably ok - I have some manually added files) but it does not fail. The loop seems to work fine, no unusual stuff in the log file. Here is what oref0-reset-git returns:
$ oref0-reset-git
Attempting to fix git corruption.  Please wait 15s.
Git repo does not appear to be corrupt.
oref0-fix-git-corruption: no process found
Jason Calabrese
@jasoncalabrese
Feb 27 2016 15:26
And git gc still fails? If it was me I'd create a new repo
Dragan Maksimovic
@dm61
Feb 27 2016 15:29
yup, git gc still fails; will start a new repo - thanks
Jason Calabrese
@jasoncalabrese
Feb 27 2016 15:36
Good seems like a time bomb, better now that when it's broke
It you rm -rf the .git dir, and then go up one level and do openaps init YOUR_DIR you should be set
Dragan Maksimovic
@dm61
Feb 27 2016 15:55
Great, that's super simple - thanks!
Scott Leibrand
@scottleibrand
Feb 27 2016 16:42
@juliatakeuti great feedback, thanks. Can you share some NS screenshots showing what happened prior to those lows?
also, is this the dev branch with meal assist, or the meal-assist branch?
dougtakeuti
@dougtakeuti
Feb 27 2016 17:09
blob
@scottleibrand this is the dev branch with meal-assist.
Oskar Pearson
@oskarpearson
Feb 27 2016 17:16
Created bewest/decoding-carelink#173 for it. But it seems to have started working again :/
Scott Leibrand
@scottleibrand
Feb 27 2016 17:16
Ok. You might want to upgrade to the meal-assist branch then. It has some smoother phase-in/out of meal assist, as well as auto sensitivity detection that you can configure. Looks like only the 4am one was maybe caused by the meal assist high temps, and in that case it had failed to execute any temps for an hour prior to the low, when it should've been temping to zero. Are you using carelink or one of the longer range sticks?
dougtakeuti
@dougtakeuti
Feb 27 2016 17:23
I'm using the TI stick
i'll beout for a few hours but will check gitter at night. Thanks for your help!
Jim Matheson
@jmatheson
Feb 27 2016 17:28
What parameters are people using as "in range"? I am hoping to hit 90% at some point but since looping i am around 80 most weeks, hoping to hit 85 this week wrapping up. I am using 75-150
Scott Leibrand
@scottleibrand
Feb 27 2016 17:29
That's a YDMV thing.
Jim Matheson
@jmatheson
Feb 27 2016 17:29
?
Scott Leibrand
@scottleibrand
Feb 27 2016 17:29
I would be quite happy with 85% 75-150, but everyone is different.
Your Diabetes May Vary
depends on age, hormones, diet, etc. etc.
Dana Lewis
@danamlewis
Feb 27 2016 17:30
Whereas @bewest uses 80-180. So make sure to ask before you compare yourself to someone else's numbers ;)
Jim Matheson
@jmatheson
Feb 27 2016 17:30
Of course. I am only asking what range people are using
Dana Lewis
@danamlewis
Feb 27 2016 17:31
My DIYPS calculates TIR for 80-160
Sometimes I'd like to do 75-160 because I bet % would go up a ton :)
Jim Matheson
@jmatheson
Feb 27 2016 17:31
Totally, thats where i want to cheat my range too :)
May be why i am using 75 actually. 70 seemed pushing it
Dana Lewis
@danamlewis
Feb 27 2016 17:33
Yea. I often touch 78-79 but don't really consider that a low for me anymore
Jim Matheson
@jmatheson
Feb 27 2016 17:34
I dont until 70. At least if i am just sitting at a desk
Plus it helps to drag my average down :)
it will be interesting to see how i do when i stress test the system more. Been eating fairly low carb and low exercise since starting.
Jason Calabrese
@jasoncalabrese
Feb 27 2016 17:38
maybe instead of cheating we track time in unsafe range
the red zones in NS
even showing that side by side would help with the mid 70's
Rachel Sandlain
@audiefile
Feb 27 2016 18:08
:+1: I've noticed that I'll stay "mildly high/ yellow zone" for longer periods of time as openaps low temps to prevent a predicted low which means more time out of range but also much much less "red zone" I think time in range and time not in dangerous range are both important
Scott Leibrand
@scottleibrand
Feb 27 2016 18:12
@juliatakeuti what I like to do in these situations where we got a less than ideal outcome is go back and figure it what loop decisions could've been made differently, given the info available at the time, to help improve the outcome.
Dana Lewis
@danamlewis
Feb 27 2016 18:12
:+1: @audiefile
Scott Leibrand
@scottleibrand
Feb 27 2016 18:18
@audiefile interesting. We usually don't see low temps start until <160.
Jim Matheson
@jmatheson
Feb 27 2016 18:23
@danamlewis how do you prebolus/correct to 80? How do you figure out how much and include any iob etc?
Dana Lewis
@danamlewis
Feb 27 2016 18:25
It's "eating soon", so if I know I plan to eat within next hour, do a correction bolus down to 80 rather than normal target (for me 100) to get insulin active in the body.
I use DIYPS, which shows my netIOB including loop activity, and auto does the reco for me after I tap a button
But you can do same thing manually, just use netIOB as part of your calc for targeting 80.
(Current bg - target (80)) x your ISF - net iob.
So if you have negative it may end up being bigger than you'd expect. But it's really important when lots negative, because your body will have a very dif rewacti
Jim Matheson
@jmatheson
Feb 27 2016 18:27
iob in ns does not include high temps from oref right?
Dana Lewis
@danamlewis
Feb 27 2016 18:27
*reaction to carbs than when you have some or normal insulin activity going
Correct. If you have openaps running, assume it's temped appropriately to target, bolus between target and 80.
Jim Matheson
@jmatheson
Feb 27 2016 18:28
Is there a true iob in my ns view anywhere?
Dana Lewis
@danamlewis
Feb 27 2016 18:28
By true, you mean "net" - not yet
Use server auditor to check logs via phone if you're on the go
To get net iob
Or do the above, if you think Openaps is treating you to target, just do bolus for target to 80
Jim Matheson
@jmatheson
Feb 27 2016 18:39
Oh that might work. Will try that
Rachel Sandlain
@audiefile
Feb 27 2016 18:39
content://media/external/file/34523
Jim Matheson
@jmatheson
Feb 27 2016 18:40
I dont monitor server on the go. Have not figured out how to do that yet
Or maybe you just mean the drill down on the openaps pill
Rachel Sandlain
@audiefile
Feb 27 2016 18:40
trying to add a screenshot for you @scottleibrand
Jason Calabrese
@jasoncalabrese
Feb 27 2016 18:42
a dropbox link works if you're on a phone
Rachel Sandlain
@audiefile
Feb 27 2016 18:43
Basically after a big meal I was still high but eventual bg was 54 so it low temper at about 190 and flat
Dana Lewis
@danamlewis
Feb 27 2016 18:43
No @jmatheson Not the pill
Get the free server auditor app
Then you can ssh to PI or Edison to whatever very easily just as you would from comp
Create a shortcut to see logs
I tap "l" and it runs :)
Rachel Sandlain
@audiefile
Feb 27 2016 18:46
Always forget drop box
Dana Lewis
@danamlewis
Feb 27 2016 18:47
:+1: @audiefile He's driving but will see in a bit.
What branch are you using?
Rachel Sandlain
@audiefile
Feb 27 2016 18:49
Should be dev
Jim Matheson
@jmatheson
Feb 27 2016 18:49
I could never get serverauditor to connect. May need to try it again. Would always just spin
Dana Lewis
@danamlewis
Feb 27 2016 18:49
@audiefile With meal assist turned on?
@jmatheson Yes, try again
Dragan Maksimovic
@dm61
Feb 27 2016 18:54
@audiefile looks like you have lots of IOB, openaps does not know you apparently have lots of COB, so it is projecting an eventual low, which is why it's low temping.
Rachel Sandlain
@audiefile
Feb 27 2016 18:54
Meal assist is on but because I use 712 no carb records so mostly just wtfassist really
I agree @f
Sorry I agree @dm61
I understand why it happened, not even upset about it, I just know that it was the wrong decision since I had all the info
Yes post a big winner so lots of carbs in addition to iob
Dana Lewis
@danamlewis
Feb 27 2016 18:58
@audiefile Still want to go back and see wh t it was doing to see if we can make it do better in those situations.
John Males
@johnmales
Feb 27 2016 18:58
@juliatakeuti @oskarpearson @bewest
Hi. Exactly the same issue as Iskar mentioned. Has happened three times that I can think of. All reports run, but the glucose data from the pump is hours out of date, even though the glucose report has just been done.
Dana Lewis
@danamlewis
Feb 27 2016 19:01
@audiefile Also wondering if meal assist branch would do better there, ramps wtf assist in more smoothly
John Males
@johnmales
Feb 27 2016 19:01
Aargh iPhone - that was meant to be Oskar :)
Rachel Sandlain
@audiefile
Feb 27 2016 19:04
Certainly worth a try, I was thinking I should update this weekend...
Jim Matheson
@jmatheson
Feb 27 2016 19:04
Serverauditor on iphone with rig connected via hotspot. Not sure how to get it to connect.
Dana Lewis
@danamlewis
Feb 27 2016 19:06
@audiefile Just PRed meal assist to dev - still open, so would be good for feedback and testing. @amazaheri Says working well for him so far.
Rachel Sandlain
@audiefile
Feb 27 2016 19:15
:+1: I'll give it a go
Andrew Warrington
@andrew-warrington
Feb 27 2016 19:17
@oskarpearson, just used this documentation, THANK GOODNESS you created it. Nice. Thank you.
thanks as well to others who contributed. I see @jmatheson did some as well.
Rachel Sandlain
@audiefile
Feb 27 2016 19:18
@danamlewis meal assist or advanced meal assist?
Jason Calabrese
@jasoncalabrese
Feb 27 2016 19:32
@jmatheson edison.local should work from your phone if it's on the hotspot
Jim Matheson
@jmatheson
Feb 27 2016 19:38
Home now and finally got it to connect here. Will try again once mobile.
eyim
@eyim
Feb 27 2016 20:02
For us, we like it that it begins to low temp early - even well above 200. If it waits to low temp in the mid 100s then it is too late for us. Here it low temps starting in mid / low 200s and it still isn't even enough
blob
eyim
@eyim
Feb 27 2016 20:08
For us, the normal pattern is a large spike after he eats - even with pre bolus and the other things we have tried. So if we see a drop start like this soon after he eats we know we are in trouble. On top of the low temp, there was probably 50+ rescue carbs he ate. Was having stomach problems at the time so maybe he didn't digest all the carbs in the food.
Rachel Sandlain
@audiefile
Feb 27 2016 20:08
And if it was me I'd have probably ended up staying at least a bit higher than in would like. Just shows how true YDMV is
Jim Matheson
@jmatheson
Feb 27 2016 20:08
was the entry of the 80/120 carbs timing wise when it ate or before or after?
Rachel Sandlain
@audiefile
Feb 27 2016 20:09
Sorry about the tummy trouble. Lows are the worst then. You have to eat and it's the last thing You want to do
eyim
@eyim
Feb 27 2016 20:10
normally bolus 15 min or more before. The rise before that was an unaccounted fruit snack that he mindlessly ate without dosing
Scott Leibrand
@scottleibrand
Feb 27 2016 20:10
@audiefile it shouldn't temp to zero if you're actually flat (+0 over 15m)
eyim
@eyim
Feb 27 2016 20:10
@audiefile Ya he was upset with me forcing him to eat while in the bathroom doing his business - lol
Jim Matheson
@jmatheson
Feb 27 2016 20:10
ok, that makes more sense.
Scott Leibrand
@scottleibrand
Feb 27 2016 20:11
it will ramp in the low temp proportionally to how fast you're falling relative to expectedDelta.
Jim Matheson
@jmatheson
Feb 27 2016 20:12
that’s what i would expect
or desire
Scott Leibrand
@scottleibrand
Feb 27 2016 20:13
@audiefile Looks like it did wait until you started to fall. Would be better if you could get carbs entered somehow. We get ours from NS (entered with DIYPS)
meal-assist branch would probably also be a bit better. Advanced meal assist definitely should be, once I finish it.
Jim Matheson
@jmatheson
Feb 27 2016 20:14
@eyim what are you using for length of insulin ? IA something. forget the acroynm
Scott Leibrand
@scottleibrand
Feb 27 2016 20:14
@eyim have you tried earlier preboluses, eating soon style (60m before meal)?
eyim
@eyim
Feb 27 2016 20:14
5 hours
Scott Leibrand
@scottleibrand
Feb 27 2016 20:14
DIA
duration of insulin action
@jmatheson what host/IP are you connecting to with ServerAuditor ?
Jim Matheson
@jmatheson
Feb 27 2016 20:16
interesting.
eyim
@eyim
Feb 27 2016 20:16
Yes we pre-bolus based on how high he is. Sometimes even an hour earlier. Even tried superbolus where we give him 2 hours of basal up front and then 0 temp for the 2 hours.
Jim Matheson
@jmatheson
Feb 27 2016 20:16
mursefree.local :)
eyim
@eyim
Feb 27 2016 20:18
Long DIA + fast carb absorption rate (40+ an hour) + teenage attitude = difficult spike to manage
Jim Matheson
@jmatheson
Feb 27 2016 20:18
it does not look like there was a spike from his meal
eyim
@eyim
Feb 27 2016 20:19
Nope - that was what was strange and we knew something was off
Jim Matheson
@jmatheson
Feb 27 2016 20:19
it almost looks like the meal bolus/correction was too much.
eyim
@eyim
Feb 27 2016 20:20
Yep definitely does
That is why I don't think he digested all the food correctly or maybe we just miscounted carbs
Jim Matheson
@jmatheson
Feb 27 2016 20:21
was there any insulin given just off the left of the graph?
eyim
@eyim
Feb 27 2016 20:21
No nothing for hours before
Dragan Maksimovic
@dm61
Feb 27 2016 20:22
wasn't that spike due to another meal some time earlier?
Jim Matheson
@jmatheson
Feb 27 2016 20:22
interesting. and it looks like the bolus was @ 11.1 ratio for all 4, so there was no correction bolus added.
seems like it temped to zero pretty early in the process
Dragan Maksimovic
@dm61
Feb 27 2016 20:23
I'd say it looks like it temped to zero correctly
Jim Matheson
@jmatheson
Feb 27 2016 20:24
one thing i do is tend to over esitmate carbs when i start out higher and underestimate when i am starting out lower. not sure if it is some sort of subliminal thing or what.
eyim
@eyim
Feb 27 2016 20:24
@jmatheson Yes - he didn't check his bg and didn't enter it
@dm61 Yes the spike up before he ate dinner was a fruit snack he ate unaccounted for
Basically I just wanted to show this as an example of how the early low temp works for us -we like it.
Jim Matheson
@jmatheson
Feb 27 2016 20:24
:)
eyim
@eyim
Feb 27 2016 20:25
But understand it doesn't work for everyone as @audiefile said YDMV
Jim Matheson
@jmatheson
Feb 27 2016 20:26
helpful to look at graphs from others.
Dragan Maksimovic
@dm61
Feb 27 2016 20:27
indeed
Jim Matheson
@jmatheson
Feb 27 2016 20:27
changes the way i think about it which is good
eyim
@eyim
Feb 27 2016 20:28
Yes I agree as well. Definitely a big difference for people who have very good control vs those that struggle a bit in comparison. Hard to develop an algorithm that works with both
That is why I am interested to see how the Bigfoot system works. They said they will only ask two questions at startup - what is your basal rate and how worried are you about lows and then they will handle everything else although you have to tell them you are about to eat and roughly how much
Jim Matheson
@jmatheson
Feb 27 2016 20:30
i am most interested right now in improving my pre-bolus. takes forever for the boat to stop rocking once there is a swell, but if i can avoid the swell it’s generally calm waters. always have struggled with that swell :)
Dragan Maksimovic
@dm61
Feb 27 2016 20:33
@jmatheson, have you tried meal assist?
Jim Matheson
@jmatheson
Feb 27 2016 20:33
i think bigfoot’s first question is actually hard to answer correctly on day 1. their second question i would think is just how low the target BG is.
i have.
i have been a bit easy on the system by eating low carbs for the most part, so boat has not been rocking much. just thinking back in life where i’ve struggled most.
i think if i can get better with the pre-bolus though i will be that much better. its fun to only have little things left to improve :D
Scott Leibrand
@scottleibrand
Feb 27 2016 20:40
:-)
Jason Calabrese
@jasoncalabrese
Feb 27 2016 21:18
the new careportal pebble app seems like it would be good for entering carbs
we normally use the bolus wizard so I never got around to trying it
Scott Leibrand
@scottleibrand
Feb 27 2016 21:22
Is that in the pebble App Store?
would be cool to reply with an initial bolus suggestion
Jim Matheson
@jmatheson
Feb 27 2016 21:24
@jasoncalabrese yea no dice on hotspot. Just spins never connects. Works fine at home when edison at iphone both on home network
Scott Leibrand
@scottleibrand
Feb 27 2016 21:26
You might try putting avahi-daemon restart into cron.
          • sudo /etc/init.d/avahi-daemon restart
That's * * * * * sudo /etc/init.d/avahi-daemon restart
Jason Calabrese
@jasoncalabrese
Feb 27 2016 21:29
installed the pebble app, will give it a try for next carb correction
Rachel Sandlain
@audiefile
Feb 27 2016 21:33
The only bad thing about the pebble care portal app is you have to remember to use it
@scottleibrand I agree on entering carbs. Didn't realize I could pull from nightscout. Is there a write up on that by any chance?
Scott Leibrand
@scottleibrand
Feb 27 2016 21:37
Nope. Pretty sure we're the only ones using it. Lemme see if I can find my config.
Jason Calabrese
@jasoncalabrese
Feb 27 2016 21:39
well pebble app didn't work on the first couple tries, maybe I'll try again later
Scott Leibrand
@scottleibrand
Feb 27 2016 21:40
openaps alias add ns-meal-carbs '! bash -c "egrep -q carbs.:0, monitor/meal.json && curl -m 30 -s \"$NIGHTSCOUT_HOST/api/v1/treatments.json?find[created_at][\$gte]=date -d \"3 hours ago\" -Iminutes&find[carbs][\$exists]=true\" > monitor/carbhistory.json && oref0-meal monitor/pumphistory-zoned.json settings/profile.json monitor/clock-zoned.json monitor/carbhistory.json > monitor/meal.json; exit 0"'
Rachel Sandlain
@audiefile
Feb 27 2016 21:43
3 hours because dia=3?
Scott Leibrand
@scottleibrand
Feb 27 2016 21:44
Ya
Rachel Sandlain
@audiefile
Feb 27 2016 21:44
:+1: perfect thank you
Jason Calabrese
@jasoncalabrese
Feb 27 2016 21:44
looks like some backticks in there, around the data command
Scott Leibrand
@scottleibrand
Feb 27 2016 21:44
Yep
Jason Calabrese
@jasoncalabrese
Feb 27 2016 21:45
just a formatting thing with gitter
Scott Leibrand
@scottleibrand
Feb 27 2016 21:45
copy paste will preserve them I think
Rachel Sandlain
@audiefile
Feb 27 2016 21:47
I've had good results with the pebble app. Did the config page load ok @jasoncalabrese ?
Jason Calabrese
@jasoncalabrese
Feb 27 2016 21:48
yeah, and preserves data when opening up again
seems to be stuck in mmol too
Rachel Sandlain
@audiefile
Feb 27 2016 22:22
Weird
Ben West
@bewest
Feb 27 2016 22:42
you need curl -g -g flag to turn globbing OFF
if you've got any [{]} characters in the url
@scottleibrand ^^
hmmm in that query if you're lucky it might not produce a glob I guess
Scott Leibrand
@scottleibrand
Feb 27 2016 22:45
It works, one way or the other
live4sw
@live4sw
Feb 27 2016 22:55
Hi guys - up to the loop part of the docs. I understand what the loop alias does, but not sure what the sample retry-loop code is doing
the `"until( ! mm-stick warmup || openaps loop); do sleep 5; done"'
"until( ! mm-stick warmup || openaps loop); do sleep 5; done"
is it essentially checking to make sure the carelink stick is working, and if it is, it runs loop, and if not it waits 5 seconds and then tries again?
Scott Leibrand
@scottleibrand
Feb 27 2016 23:05
In English that means "until warmup fails or loop succeeds, wait 5 seconds and try again"
LadyViktoria
@LadyViktoria
Feb 27 2016 23:06
i found a handy device to extend openaps ^^
http://www.spiegel.de/fotostrecke/spektrometer-von-scio-analysiert-feste-und-fluessige-stoffe-fotostrecke-134860.html
https://www.consumerphysics.com/myscio/
it can analyse food and show its kCal and carbs for example <3
live4sw
@live4sw
Feb 27 2016 23:07
aha, thanks scott.