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

24th
Sep 2017
br
@brmccollum
Sep 24 2017 00:28
No improvement. All results 0-99
Dana Lewis
@danamlewis
Sep 24 2017 00:39
Do you have: another pump (even if it's not loopable, that you could sub in SN and help validate to make sure the board is fine vs pump radio being the problem?)
Assuming this is first rig and you don't have another explorer board to test
br
@brmccollum
Sep 24 2017 00:42
I do have a 723 with 2.5b software.
What is the easiest way to switch out SN number?
br
@brmccollum
Sep 24 2017 00:50
Well, even without changing the SN, I think I have a result. To put in the battery into the 723 and the numbers changed to "916.684", 1, -89 "
Dana Lewis
@danamlewis
Sep 24 2017 00:54
Nah
You'd need to change the sn
br
@brmccollum
Sep 24 2017 00:56
Okay, Ill try that, I wonder why it gave me the new number after seeing nothing but this prior :"916.636", 0, -99
br
@brmccollum
Sep 24 2017 01:10
Thanks @danamlewis I switched the SN and now receive this "mmtune: "916.660", 5, -44 "
Based on that is seems like my 712 pump is the issue.
br
@brmccollum
Sep 24 2017 01:17
I want to try one more time on the 712 before I try to buy a used one. Is there anything else I should test while on the 723's SN?
Eric
@ecc1
Sep 24 2017 01:19
@brmccollum late to the party, but are you sure your 712 is a US pump, not a WW one?
br
@brmccollum
Sep 24 2017 01:20
Correct. 712LNAS
Eric
@ecc1
Sep 24 2017 01:22
it's worth trying mmtune quite far above and below the normal range. that pump is old, and its tuning might be off. someone else will have to tell you if there's a way to do that other than by editing the code
also, did it come with an RF remote, and does that work OK with it?
br
@brmccollum
Sep 24 2017 01:24
No RF remote. I've had this one for a long time and 712 was new when I got it.
Eric
@ecc1
Sep 24 2017 01:25
Looks like pumps that old might have a menu item Utilities > RF Options > Off/On that needs to be set On
br
@brmccollum
Sep 24 2017 01:31
Thanks for the suggestion, I just looked through utilities... the RF option did ring a bell, however this unit does not have it in there. Remote and Meter are set currently set to "on".
Scott Leibrand
@scottleibrand
Sep 24 2017 01:43
You might also try openaps use pump model even if mmtune gives -99
br
@brmccollum
Sep 24 2017 01:47
I think I have the syntax incorrect, how would I type in that command?
br
@brmccollum
Sep 24 2017 02:05
@scottleibrand thanks for the help, however I can't figure out how to set it to 712. Using -h with openaps-use it says "Once a device is registered in openaps.ini, it can be used." yet, the file for openaps.ini appears blank.
Scott Leibrand
@scottleibrand
Sep 24 2017 02:20
You just have to cd ~/myopenaps first.
The use pump commands don't set anything, they just try to talk to the pump and display the results.
Katie Aldridge
@kcrcgm
Sep 24 2017 02:34
@scottleibrand I'd like to understand more about eSMB and how it works and what makes it safe. Is there a place I can read about that? Thank you. 😊 So far I'm loving the new insulin and CA curves!
Dana Lewis
@danamlewis
Sep 24 2017 02:35
@kcrcgm :+1: have you already read https://github.com/openaps/oref0/pull/650#issue-256564183 ?
br
@brmccollum
Sep 24 2017 03:11

Hi @scottleibrand , here is what I received one I typed in the correct spot:

File "/usr/local/lib/python2.7/dist-packages/mmeowlink/handlers/stick.py", line 222, in power_control
raise CommsException("No acknowledgement from pump on wakeup. Is it out of range or is the battery too low?")
mmeowlink.exceptions.CommsException: No acknowledgement from pump on wakeup. Is it out of range or is the battery too low?

Scott Leibrand
@scottleibrand
Sep 24 2017 03:22
@brmccollum sounds like your pump radio might have a problem then. :-(
Katie Aldridge
@kcrcgm
Sep 24 2017 03:42
@danamlewis Yes
@danamlewis Yes but would be interested to know more. Maybe I'm missing something on that page... will check again 😊
Dana Lewis
@danamlewis
Sep 24 2017 03:48
@kcrcgm :+1: for reading again, but that's all the documentation so far, so please do ask questions around where you feel it's unclear, and that will help us more quickly figure out what needs documentation.
br
@brmccollum
Sep 24 2017 05:35

Well, I dug out another old 712 from the attic (seriously) that had a broken ACT button, I got the button to work and tried to get it to loop. Its signal isn't as strong as the 723 I used but it is better than -99:

mmeowlink.exceptions.CommsException: Could not get subg_rfspy state or version. Have you got the right port/device and radio_type?
Listening for 30s silence before mmtuning: .No pump comms detected from other rigs
mmtune: "916.468", 1, -90 waiting for 60 second silence before continuing
Radio ok. Listening: .No pump comms detected from other rigs
Listening: .No pump comms detected from other rigs
Old pumphistory: IndexError: bytearray index out of range

Error, retrying
pump_loop_completed more than 15m old; waiting for 30s silence before mmtuning
Radio ok. Listening: .No pump comms detected from other rigs
Listening for 30s silence before mmtuning: .No pump comms detected from other rigs
mmtune: "916.636", 0, -99 waiting for 78 second silence before continuing
Radio ok. Listening: .No pump comms detected from other rigs

As you can see from the log, it goes back and forth on if it gets a signal and has continued to do so. This line makes me think there is user error:

mmeowlink.exceptions.CommsException: Could not get subg_rfspy state or version. Have you got the right port/device and radio_type?

Scott Leibrand
@scottleibrand
Sep 24 2017 05:55
or it could just need a reboot
what do the two LEDs next to the microUSB ports on the Explorer board show?
br
@brmccollum
Sep 24 2017 06:07
Only the LED between the two USB ports is on (Solid Red)
Scott Leibrand
@scottleibrand
Sep 24 2017 06:14
ok, then you probably don't need to run ccprog
Martin Haeberli
@mhaeberli
Sep 24 2017 06:29
Br - I always defer to Scott’s wisdom, but have had luck in similar circumstances running ccprog.
Scott Leibrand
@scottleibrand
Sep 24 2017 06:34
I'd try a hard power-cycle first, and then check to make sure your oref0-runagain has the proper spidev5.1 port it in. if neither of those is the problem, ccprog is worth a shot
you can search the docs for the page on how to run it
Niels Hartvig
@hartvig
Sep 24 2017 07:12
Does anyone know what's needed in order to not get a lot of nightscout items under "Remove future items from mongo database". I've tried to check dates on pump, edison and nightscout site and they all seem in sync. This mostly means that I can't see the bolus dots in nightscout which I'd like to get working before switching to oref1
br
@brmccollum
Sep 24 2017 07:13
Scott, I am calling it a night. Thank you and everyone for trying to fix my issue. I had the correct port and tried ccprog. No luck.
Mike
@libxmike
Sep 24 2017 07:16
anyone else has a problem with latest NS 0.10 and openAPS basal rate not displaying (just the default NS profile one) after update from 0.92 ?
Niels Hartvig
@hartvig
Sep 24 2017 07:21
@libxmike I'm running 0.10 with basal rate working just fine
Mike
@libxmike
Sep 24 2017 07:24
for me must be the 0.10 as my rigs did not change, something is wrong
all 3 rigs, same thing
Niels Hartvig
@hartvig
Sep 24 2017 07:25
What about the timezone or basal settings in nightscout? Perhaps it "lost" info in the upgrade process (just a wild guess tbh)
ie. in the profile editor
Mike
@libxmike
Sep 24 2017 07:26
time zone looks ok
Niels Hartvig
@hartvig
Sep 24 2017 07:34
@libxmike another long shot - is render basal set to none (from the hamburger menu). https://www.dropbox.com/s/2b7079nllxvrxcn/Screenshot%202017-09-24%2009.33.26.png?dl=0
Mike
@libxmike
Sep 24 2017 07:35
already checked that, is default
Matthias
@ceben80
Sep 24 2017 11:40
@scottleibrand @danamlewis @tim2000s does it makes sense to use maxSMBBasalMinutes with Humalog or it's more effectively when using Fiasp :question:
Mike
@libxmike
Sep 24 2017 11:56
@hartvig found the problem, it was the pump clock, I changed my battery in the pump at the same time as I updated to 0.10 and set the date to 2018, every time I change the battery I have to set the clock and prime
live4sw
@live4sw
Sep 24 2017 12:12
Random question - I've been looping for about 1.5 years now with good results and am finally considering moving to the oref1 features with my daytime rig when I upgrade to 0.5.3 today (leaving the nighttime/backup rigs on 0.5.3 with oref0), is there any info/educated guesses as to what % of loopers are on oref1 now?
Niels Hartvig
@hartvig
Sep 24 2017 14:25
I just switched my sons rig to oref1 last night. It's surreal so good it is. I can't believe what we're seeing. We've really tried to put maximum pressure on the system (because I can monitor him all day). He's 14 years old, super active yet his been between 5.2 and 11.4 all day (the 11 was only in 15 minutes and after consuming 110g of porridge oat (~60g of carb) + eating ~40g of Pringles chips). We've also tested that he ate two pears without doing ANYTHING (ie. UAM) and it peaked at 8.2(!!!) before lowering back to 5.6. He's also been jumping on trampoline where he often goes low and needs juice and he stayed within range.
I haven't been able to tell my son "just go eat a pear" for seven years (since he got d1t). To all of those behind openaps and oref1 - you're heroes. This is so incredible. Thank you so much!
live4sw
@live4sw
Sep 24 2017 14:39
Very exciting, I'm trying it out now
Do you guys just use one rig? Did you upgrade them all to oref1?
jaylagorio
@jaylagorio
Sep 24 2017 14:44
I use one, and have used oref1 for maybe a month or so
It definitely took some adjustment and I think I'm still getting used to it
But I'm pretty happy with it
live4sw
@live4sw
Sep 24 2017 14:46
Do you run autotune nightly with your loop?
jaylagorio
@jaylagorio
Sep 24 2017 14:47
I do, it's been iterating over data since February
Niels Hartvig
@hartvig
Sep 24 2017 14:48
Just one, autotuning since May
live4sw
@live4sw
Sep 24 2017 14:48
ah, interesting. i just run it periodically on a virtual machine, perhaps i should start running it as part of the loop
what has been holding me back is that one of my I:C ratios (dinner) is significantly off from my others
jaylagorio
@jaylagorio
Sep 24 2017 14:49
Out of curiosity, what makes dinner so different from your other meals?
live4sw
@live4sw
Sep 24 2017 14:51
I have a long walk home and then eat soon afterward, so i think i am more insulin sensitive at dinnertime. but it's also just possible that the lower dinner I:C ratio is just offsetting something else
jaylagorio
@jaylagorio
Sep 24 2017 14:51
Oh, that makes sense
live4sw
@live4sw
Sep 24 2017 14:52
it may be worth testing going to one I:C ratio, staying lower carb at dinner for a little while, and seeing what happens with SMB/UAM
Tim Street
@tim2000s
Sep 24 2017 14:52
@ceben80 It makes sense to use a greater value with Humalog. I am doing that with Novorapid right now. It means that it will give you more insulin earlier in the SMB cycle, which is hugely helpful in reducing the post prandial highs. It’s not as effective as Fiasp but it still makes a difference.
jaylagorio
@jaylagorio
Sep 24 2017 14:52
The thing I like about oref1/SMB is the long period of zero temping so if you start to rise out of control you can adjust, but it seems to take care of that itself automatically
live4sw
@live4sw
Sep 24 2017 14:53
i find when i am very active during the day that the zero temping in oref0 often does not work on its own and i have to manually treat
jaylagorio
@jaylagorio
Sep 24 2017 14:53
Everyone has the things they care about most vs. what other people care about, my personal focuses are reducing the number of or severity of hypos
live4sw
@live4sw
Sep 24 2017 14:54
i sort of wonder if the microboluses with the longer zero temp could even help, since a lot of the time i'll go to my max temp for a while, then be outside, then go low and it takes too long for the zero temp to get my basal IOB negative
jaylagorio
@jaylagorio
Sep 24 2017 14:54
Preferably without going over 200
Matthias
@ceben80
Sep 24 2017 14:54
@tim2000s thx, 45m for the 1st approach?
jaylagorio
@jaylagorio
Sep 24 2017 14:55
You're probably more active than I am because a rag doll is probably more active than I am, so I can't speculate too much on how that plays into things, but it does sound like the microboluses could help
live4sw
@live4sw
Sep 24 2017 14:55
i only get strenuous exercise 1-2x/wk, but i walk a ton (re: work and childcare) so that is the main cause of lows
i almost never have severe lows, but am in the 60's multiple times per week
it's tough, i was able to get a1c's down from low 7's to mid 6's with openaps without having more lows, but i still have hypos more often than i'd like and kind of wonder if maybe my targets (110-120) should be slightly higher
jaylagorio
@jaylagorio
Sep 24 2017 15:00
60s is in my realm of unacceptable. OpenAPS brought my A1c from 7.3 to 6.9 (6 months ago) to 6.3 (last month) with minimal actual work on my part - I suspect it was primarily nighttime use. 6.3 is the best result I've ever had. What I noticed in that time is that I had to raise my nighttime targets from 110 - 120 to 120 - 130, the result of which has me waking up right around 100.
Before that I would wake up around 90, which was OK but still a little lower than I like
Tim Street
@tim2000s
Sep 24 2017 15:00
@ceben80 yes, move forward in 15min steps.
Matthias
@ceben80
Sep 24 2017 15:03
:+1:
jaylagorio
@jaylagorio
Sep 24 2017 15:09
@live4sw To be clear I don't count the difference in waking up vs. what the target ranges are against OpenAPS - I count it against the sensor values which, while close to reality, vary up and down from actual numbers throughout the course of the night
live4sw
@live4sw
Sep 24 2017 15:09
Sure, makes sense
Random question for anyone with multiple rigs - does it cause any problems if only one rig is set up to do autotune automatically? To test out oref1, I'm just updating my daytime rig to 0.5.3 with oref1 and will leave the others either on 0.5.2 or 0.5.3 with oref0
Tim Street
@tim2000s
Sep 24 2017 15:14
The rig with Autotune will use a different set of basal values as it populates the base data from Autotune. If you aren’t varying too widely from your starting point pump settings then it is unlikely to make a huge difference.
The bigger aside is whether you are planning to use oref1 with SMB?
live4sw
@live4sw
Sep 24 2017 15:15
Yes, I am planning to enable SMB and UAM on the oref1 rig, which is my daytime rig and will do an automated nightly autotune
and then leave my 3 backup rigs on oref0 without autotune
Just trying to think through whether that would cause any problems
The 3 backup rigs are in my house and are just for running during the night and when i'm otherwise at home. My daytime rig I carry with me at all times (basically)
live4sw
@live4sw
Sep 24 2017 15:28
alternatively, I suppose I can start by just doing a manual autotune, making sure that my current settings are close to the autotune, and then test out oref1 w/SMB on the daytime rig while leaving the home rigs on oref0. maybe that's a better first step
jaylagorio
@jaylagorio
Sep 24 2017 15:36
That's what I did back in May for a couple weeks since I have to be manual during the work day anyway. Being able to add up all the data from February through April and using that in May was super helpful as a first step.
Being manual during the work day sucks but at least I have more insight than I'd otherwise have just by way of being able to run analysis over a long period
live4sw
@live4sw
Sep 24 2017 15:39
Thanks! That's super helpful. My sympathies for not being able to loop during the workday
jaylagorio
@jaylagorio
Sep 24 2017 15:39
It's cool, that just gives me another thing to look forward at the end of the day and on the weekend!
live4sw
@live4sw
Sep 24 2017 15:40
there was a period where work wifi would not let my rig on, but i just tethered to my phone instead
jaylagorio
@jaylagorio
Sep 24 2017 15:41
My rig is fully self-contained, the G5 receiver is directly connected so I'm pretty happy about that
live4sw
@live4sw
Sep 24 2017 15:42
Ah, makes sense
jaylagorio
@jaylagorio
Sep 24 2017 15:43

Git question! I have the following script with the goal of giving me the comment on the most recent local commit and to check whether it's the same as the most recent remote commit of oref0 dev:

pushd .                                                                                              
cd /root/src/oref0                                                                                   
git show --oneline -s                                                                                
git status                                                                                           
popd

The output right now is this:

1ce5d70 autotune/autosens on by default; don't ask about ~/myopenaps (#664)                          
On branch dev                                                                                        
Your branch is up-to-date with 'origin/dev'.                                                         
nothing to commit, working directory clean

But this is wrong because I'm at least two commits behind what's current on GitHub. What am I doing wrong and what might I be able to change to have the script tell me I'm not current?

Dana Lewis
@danamlewis
Sep 24 2017 17:20
@live4sw are you using high temp targets (ie activity mode) in advance of your walks and other activities?
Eric
@ecc1
Sep 24 2017 17:37
@jaylagorio git doesn't "know" you're N commits behind the remote repository until you do a "git pull" or equivalent
(for the purposes of your script, you probably just want to do "git fetch", not attempt to merge)
jaylagorio
@jaylagorio
Sep 24 2017 17:40
Wouldn't fetch grab all the changes? I just want a quick "hey, you're out of date" message instead of running setup to see if I'm out of date. I don't want it to pull new stuff down, I have setup to do that.
live4sw
@live4sw
Sep 24 2017 17:46
@danamlewis I'm not, maybe I should be doing that. I know you are moving a lot with all your travel, do you manually set your target range up to, say, 140 pretty often and then change it back?
Dana Lewis
@danamlewis
Sep 24 2017 17:49
Yes for using temp targets - makes a huge difference. I have a variety of "high" targets: 110 target for 45 minutes (mostly to turn off SMB if needed), 120 for an hour for a bit of buffer, 140 for 1-2 hours (what I usually set a while before exercise, to give it time to cut down on iob), and recently added a 200 for 20 minutes to essentially force a zero temp to start.
Also have a cancel temp target button in case I change my mind/plans. Very easy to use. I do most from my pebble, but you can use the ifttt widget on phone just as easy if you don't have pebble.
(Ditto Alexa etc)
And you can also use google calendar if your walks home are regular and you want to preschedule the week's worth of targets
live4sw
@live4sw
Sep 24 2017 17:56
Makes sense. And am I understanding it correctly that there are several ways to do this, including (1) change the target in bolus wizard and then manually change it back, (2) set it in care portal, (3) set it using the ifttt docs you just linked?
Dana Lewis
@danamlewis
Sep 24 2017 17:58
Correct. 1 is too annoying (imo) to do for temporary scenarios, but is great for offline purposes of changing target. And 3 means you can then use phone, watch, google calendar, etc however you want to use.
live4sw
@live4sw
Sep 24 2017 17:59
Awesome. Yep, #1 is super annoying which is why I did it a few times and then gave up. Will spend more time on #2 and #3.
Dana Lewis
@danamlewis
Sep 24 2017 18:01
:+1: whenever I have an idea for a new ifttt button (like a low treatment - 4g carbs and a 110 target combo for 45 min) I drag my feet...then finally build the button and kick myself because it really is a lot easier and faster then my brain makes it out to be.
Scott Leibrand
@scottleibrand
Sep 24 2017 18:16
@jaylagorio git pull updates your branch with the remote changes. git fetch just updates the backend "remote" branches so your local git knows their state and can do useful comparisons. you can and should run git fetch any time you are going to do a remote comparison. (git pull runs git fetch under the hood)
@hartvig thanks so much for sharing how well it's working for you. fingers crossed we should have some more eSMB stuff ready for you to test here shortly. I'm really excited about giving people the option to not interact with OpenAPS at all if they don't feel like it: with the no-bolus option we already have a "real APS" IMO, but if we can get all the way to no-meal-announcement at least for normal-sized meals, that would be huge. BTW, are you on Fiasp, or just Novolog/Humalog/Apidra?
sadafakahmad
@sadafakahmad
Sep 24 2017 18:20
How do I go in and change my basal setting and target BG values on my rig that I first entered when installing orefO?
Scott Leibrand
@scottleibrand
Sep 24 2017 18:21
you enter those in your pump, not in oref0
sadafakahmad
@sadafakahmad
Sep 24 2017 18:22
Didn't oref0 also ask for those values when asking the series of questions that it does when you first install it in the rig?
Tim Street
@tim2000s
Sep 24 2017 18:59
No, it doesn’t.
Niels Hartvig
@hartvig
Sep 24 2017 19:06
@scottleibrand we're on NovoRapid, but are likely to change to Fiasp by the end of the year. Even though it's not "allowed" for children by default in DK, the hospital is open for a dispensation.
sadafakahmad
@sadafakahmad
Sep 24 2017 19:13
Thank you for your responses. My current basal is 0.15 and based on what I've read I can't use SMB unless my basal is at least 0.3. I guess I'm stuck in oref0.
Scott Leibrand
@scottleibrand
Sep 24 2017 19:32
That's because the smallest bolus most MDT pumps can give is 0.1U, which would be 40m of basal all at once. Loopable 523s can do 0.05 though...
@hartvig that is great that you're getting such good results with eSMB even with just NovoRapid.
jaylagorio
@jaylagorio
Sep 24 2017 19:49
@scottleibrand Doing a fetch then a status is the right answer. Thank you!
BTW, this was breakfast on NovoLog without a previous lower temp target:
image.png
This morning was awesome on 0.6.0
sadafakahmad
@sadafakahmad
Sep 24 2017 19:51
I did the flakey wifi fix per recommendations here on the forum and I'm still getting the following error every few seconds.
wl_host_event: mismatched OUI, bailing
Niels Hartvig
@hartvig
Sep 24 2017 19:51
Does anyone know if nightscout (or openaps) have any kind of webhooks? I'd love to ping a phone (pushover / txt message) if the rigs battery is running low.
Scott Leibrand
@scottleibrand
Sep 24 2017 19:52
I think NS already has low-battery pushovers
@jaylagorio nice
sadafakahmad
@sadafakahmad
Sep 24 2017 19:52
It hard to do anything on the Terminal because it keeps filling up with three of those in a row every half minute. I feel its draining my battery because I only get 4 hours on a 2000 LiPO
Scott Leibrand
@scottleibrand
Sep 24 2017 19:55
@/all: I'm about ready to start testing some recent improvements to autotune included in openaps/oref0#668 and could use some assistance from folks who've had problems with autotune making less than perfect recommendations. In particular, if you would be willing to PM me your NS URL, along with the output of cat ~/myopenaps/autotune/autotune_recommendations.log | egrep -v "\| *\| *$", which version of oref0 you're running, and any notes about where autotune's recommendations have been off for you, I'd appreciate it. I'll run autotune on my own VM against your NS with both the old and new versions, and see if it converges on a more accurate set of recommendations.
Niels Hartvig
@hartvig
Sep 24 2017 20:01
@scottleibrand brilliant - that was fast to setup. Thank you for the pointer!
@scottleibrand is this already in the openaps setup docs - else I'll write it up and send a PR!
Scott Leibrand
@scottleibrand
Sep 24 2017 20:01
dunno, but search for pushover should find it if so
Dana Lewis
@danamlewis
Sep 24 2017 20:03
I think we don't specify that - but might be nice to note on NS setup (that page could use some love) . I think our only pushover references are explaining carbs req and insulinreq for SMB
Scott Leibrand
@scottleibrand
Sep 24 2017 20:31
also happy to test autotune against anyone's data who thinks it is working well, just to make sure these changes don't make anything worse
katie disimone
@Kdisimone
Sep 24 2017 21:21

re low battery text messages via NS....iphones won’t let apps send SMS. so you have to setup a convoluted way of doing that. i managed to do it, using papertrail integrated with another service. it’s kind of a pain because setting up the filter to alert for with papertrail. but, that’s the only way i was able to get sms messages to iphone working.

pushover service is much easier to get going than the sms option.

sadafakahmad
@sadafakahmad
Sep 24 2017 21:38
how do I change the timezone that was originally selected when installing oref0?
Scott Leibrand
@scottleibrand
Sep 24 2017 21:48
you can also search for tzdata in the openaps docs
ericmtx
@ericmtx
Sep 24 2017 21:48
upgraded to .6, enabled various SMB options (as I had previously) and set up "rapid-acting" curve to match daughter's Humalog. All three rigs running well. Is minguardbg a new variable being sent to the openaps pill? It's showing minguardBG 999
Scott Leibrand
@scottleibrand
Sep 24 2017 21:48
yes, that is new. 999 means it didn't get set. can you paste the context?
ericmtx
@ericmtx
Sep 24 2017 21:49
I reran previous setup script options - is it set during a full setup script run from scratch? Let me grab some data to paste from the log
image.png
seeing that on multiple rigs
Scott Leibrand
@scottleibrand
Sep 24 2017 21:50
minGuardBG isn't a preference: it's the lowest predicted BG
ericmtx
@ericmtx
Sep 24 2017 21:51
ah
we haven't bolused yet - been running about 2 hours
Scott Leibrand
@scottleibrand
Sep 24 2017 21:51
in your case it should be 71. it may be only getting set if SMB is enabled: I'll have to check.
ericmtx
@ericmtx
Sep 24 2017 21:52
hmm - SMB should be enabled on all of these (was before) and set prefs to true. microbolus flag is on the runagain
And where it's predicting a low it's not zero-temping as I'd expect. Because of the more aggressive curve?
Scott Leibrand
@scottleibrand
Sep 24 2017 21:54
make your putty window bigger and you can get some more context
ericmtx
@ericmtx
Sep 24 2017 21:56
image.png
Scott Leibrand
@scottleibrand
Sep 24 2017 21:56
the lack of zero temping is from bolus snooze: did you do any manual boluses, or has it been SMBing?
ericmtx
@ericmtx
Sep 24 2017 21:56
manual bolus before rig upgrade, about 5 hours ago
then smbing
image.png
Scott Leibrand
@scottleibrand
Sep 24 2017 21:57
ok, I think bolus snooze isn't working quite right with SMBs, and may need to be disabled and replaced with minGuardBG logic. but let me check why minGuardBG isn't getting set first
ericmtx
@ericmtx
Sep 24 2017 21:58
yeah, looks like bolus snooze keeps increasing while temp basal is being delivered. keeps extending it
Scott Leibrand
@scottleibrand
Sep 24 2017 22:00
found the bug in minGuardBG and pushed a fix: can you test it please?
cd ~/src/oref0 && git fetch && git checkout minGuardBG && git pull
ericmtx
@ericmtx
Sep 24 2017 22:01
sure - doing it now
updated - global install required?
Scott Leibrand
@scottleibrand
Sep 24 2017 22:04
nope, not if you already had dev
ericmtx
@ericmtx
Sep 24 2017 22:04
ok - watching the logs now. Will post next pump-loop
Scott Leibrand
@scottleibrand
Sep 24 2017 22:04
thx
opened openaps/oref0#673 to merge the fix to dev if it looks better
hmm, no, I missed one, one sec
can you git pull again please?
ericmtx
@ericmtx
Sep 24 2017 22:09
sure
done - watching loop now.
sadafakahmad
@sadafakahmad
Sep 24 2017 22:16
Thank you Scott, Jay and Katie. The error message is gone now.
ericmtx
@ericmtx
Sep 24 2017 22:29
better - zero-temping now as expected. still showing minguardbg 999 if that's important:
image.png
Scott Leibrand
@scottleibrand
Sep 24 2017 22:45
something is still not right with minGuardBG.
oh, I see it: stupid case-sensitive typo
fixed. pull one more time please?
ericmtx
@ericmtx
Sep 24 2017 22:51
sure
the rPi rig shows minGuardBG 12 now with that change. watching main rigs logs now also. so far so good
looks good
image.png
ericmtx
@ericmtx
Sep 24 2017 22:56
she's enjoying her low candy. been a while since we saw a deep low hehe
Scott Leibrand
@scottleibrand
Sep 24 2017 22:57
do you think the SMBs caused the low, or something else?
jaylagorio
@jaylagorio
Sep 24 2017 23:14
@sadafakahmad Great, glad that worked!
ericmtx
@ericmtx
Sep 24 2017 23:28
It was the lack of Zero temping seen since basal snooze was active and increasing prior to your code fix
Scott Leibrand
@scottleibrand
Sep 24 2017 23:31
ok. I'll prioritize eliminating the basal vs. bolus IOB distinction then as a more permanent fix for that, but in the mean time the minGuardBG should catch any time the predBGs dip too low
sadafakahmad
@sadafakahmad
Sep 24 2017 23:36
ok, new issue: when the rig is tethered to iPhone through bluetooth, Dexcom app on iPhone says no internet connection and won't connect to Dexcom Share, so rig doesn't know what blood sugar values are. Running G5. Ideas?