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

14th
Mar 2016
Peter Miller
@ochenmiller
Mar 14 2016 00:10
Matt: "These artificial pancreases are soooo cool! Like the new iPhone. It's no longer Diabetes. It's iAbetes!"
Scott Leibrand
@scottleibrand
Mar 14 2016 00:20
;-)
Brian Bonner
@itsn0tm3
Mar 14 2016 00:22
hmm my NS treatments seem to be 1 hour off
trying to figure out if it's NS that is messed up or if it's OpenAPS zoning that is messed up
Scott Leibrand
@scottleibrand
Mar 14 2016 00:23
did you change your pump time last night?
Brian Bonner
@itsn0tm3
Mar 14 2016 00:24
yup we were talking about it when I changed it at 2/3am this morning heh
Scott Leibrand
@scottleibrand
Mar 14 2016 00:24
hard keeping everyone straight when I only know them by their github handle. :)
Brian Bonner
@itsn0tm3
Mar 14 2016 00:24
(double checks again) yep 8:24pm
it's weird because the temp basals are also 1 hour of
I can see a 2U temp basal coming up in about 30 minutes
Eric Luhrs
@eluhrs
Mar 14 2016 00:29
I've got the same 1-hour-off problem. Kept the AP off overnight, then changed my pump and cgm first thing this morning, and verified that my phone was correct before powering up the RPi this morning (which also had the correct time). Not sure what else to check. Could it be NS-specific?
Brian Bonner
@itsn0tm3
Mar 14 2016 00:31
Hmm I should be able to check the json file and see if it's the zoning that is causing it, right?
Jason Calabrese
@jasoncalabrese
Mar 14 2016 00:31
Check the timezones, on the pi, phones, desktops, etc
Brian Bonner
@itsn0tm3
Mar 14 2016 00:32
already did, all are correct.
Jason Calabrese
@jasoncalabrese
Mar 14 2016 00:32
What time zone are you in?
Brian Bonner
@itsn0tm3
Mar 14 2016 00:32
EST
Scott Leibrand
@scottleibrand
Mar 14 2016 00:32
EDT now
Brian Bonner
@itsn0tm3
Mar 14 2016 00:32
hrmm
Jason Calabrese
@jasoncalabrese
Mar 14 2016 00:32
The city named timezones should auto adjust
Brian Bonner
@itsn0tm3
Mar 14 2016 00:33
let me check the timezone thru raspi-config and see what it's in
Mine is set to America > New_York
Jason Calabrese
@jasoncalabrese
Mar 14 2016 00:34
That should be good, what about the treatment profile in NS
Brian Bonner
@itsn0tm3
Mar 14 2016 00:34
US/Eastern
Oooh just scrolled through the list and found America > New_York in there too
just changed it to that
still showing 1 hour off tho
do I need to somehow clear the treatment history in NS and then re-upload through OpenAPS?
Jason Calabrese
@jasoncalabrese
Mar 14 2016 00:39
Maybe, my guess is it's something on the pi, maybe time zone set in multiple ways. Could check the dates on the zoned history file
Brian Bonner
@itsn0tm3
Mar 14 2016 00:41
looks correct to me..
"timestamp": "2016-03-13T19:49:51-05:00",
I had dinner at 7:49pm EDT
NS is showing it at 8:49pm tho
Jason Calabrese
@jasoncalabrese
Mar 14 2016 00:45
Thought it would be -4 instead of -5
Brian Bonner
@itsn0tm3
Mar 14 2016 00:46
me too but everything I check still refers to the timezone as -5
at least MS does..
I'll keep an eye on it tmw and see if it goes away
let me do a 1 carb entry and see what it comes up as now that I re-checked the config thru raspi-config
Jason Calabrese
@jasoncalabrese
Mar 14 2016 00:47
Says -4
Brian Bonner
@itsn0tm3
Mar 14 2016 00:48
hmm
maybe I inadvertently changed something to fix it
Jason Calabrese
@jasoncalabrese
Mar 14 2016 00:49
Time zones are such a pain, I'd be OK staying on UTC time, we'd adjust
Brian Bonner
@itsn0tm3
Mar 14 2016 00:49
heh
Rachel Sandlain
@audiefile
Mar 14 2016 00:55
I had to change the zones pump history report to get treatments and everything to work. EST to EDT. That was the key. Took a while to track done
*down where the hang up was
Funnily enough everything worked until I changed the pump time
Brian Bonner
@itsn0tm3
Mar 14 2016 00:56
in the report itself? I thought it followed the Pi TZ?
oooh I see what you mean
I just looked up my clock-zoned.json and it still says -5
thats odd
I just ran the clock-zoned report and it still shows up as -5
how did you fix it?
live4sw
@live4sw
Mar 14 2016 01:04
Yep, I'm having this same issue in NS, treatment showing as 1 hour in the future
Jim Matheson
@jmatheson
Mar 14 2016 01:05
are your Basal blue lines in NS working? mine stopped working, just shows default basal
live4sw
@live4sw
Mar 14 2016 01:11
my basal lines are working fine
just the treatment is in the future. openaps seems to be handling it ok, so i think it's just the NS data
Scott Leibrand
@scottleibrand
Mar 14 2016 01:16
OpenAPS doesn't use the zoning for much. So if it's zoning wrong due to the config still saying EST vs. EDT, that might do it.
Brian Bonner
@itsn0tm3
Mar 14 2016 01:17
but where is the config that says EDT vs EST?
I'm trying to figure out why my clock-zoned.json is saying -5 still when it should be saying -4
pi date says I'm in EDT so the timezone on the pi is correct
will re-adding the tz device fix it?
Scott Leibrand
@scottleibrand
Mar 14 2016 01:22
In openaps.ini
[report "monitor/pumphistory-zoned.json"]
use = rezone
reporter = JSON
astimezone = False
date = timestamp dateString start_at end_at created_at
adjust = missing
input = monitor/pumphistory.json
device = tz
timezone = PST
I bet if you change that from EST to EDT it fixes the offset
Dragan Maksimovic
@dm61
Mar 14 2016 01:23
I edited openaps.ini, and replaced MST with MDT in clock-zoned and pumphistory-zoned reports. That fixed the problem.
Scott Leibrand
@scottleibrand
Mar 14 2016 01:23
And if so, we need to update the TZ tool to properly accept either and switch when DST switches.
Brian Bonner
@itsn0tm3
Mar 14 2016 01:24
d'oh, how did I overlook that
Dragan Maksimovic
@dm61
Mar 14 2016 01:24
yes, that would be good
Brian Bonner
@itsn0tm3
Mar 14 2016 01:24
just fixed both clock-zoned and pumphistory-zoned to EDT
Scott Leibrand
@scottleibrand
Mar 14 2016 01:24
Someone should open an issue in the openaps repo on that.
Jason Calabrese
@jasoncalabrese
Mar 14 2016 01:29
in mine I have this

[report "monitor/pump-history-zoned.json"]
use = rezone
reporter = JSON
date = timestamp dateString start_at end_at created_at
adjust = missing
input = monitor/pump-history.json
device = tz
timezone = None
astimezone = True
not sure where I got it from
maybe it uses the system timezone?
eyim
@eyim
Mar 14 2016 01:31

The only place i have a TZ is here:

[report "monitor/clock-zoned.json"]
use = clock
reporter = JSON
astimezone = False
date = None
adjust = missing
input = monitor/clock.json
device = tz
timezone = PDT

I just changed it to PDT it was PST all this morning with no problems

live4sw
@live4sw
Mar 14 2016 01:32
wow, i just changed my EST's to EDT's in clock-zoned and pump-history-zoned, and it is no longer showing basal history or treatment history other than a test one i just did after changing the timezones
Dragan Maksimovic
@dm61
Mar 14 2016 01:33
reload ns
I mean just reload the ns page
Jason Calabrese
@jasoncalabrese
Mar 14 2016 01:36
my clock zoned is like that too
but pump clock time in NS is correct
Rachel Sandlain
@audiefile
Mar 14 2016 01:48
Yeah it was tricky. It took forever to figure it out. What @scottleibrand and @eyim said. Only that one spot needed to be changed
I just opened an issue @scottleibrand should I add anything or was that clear?
live4sw
@live4sw
Mar 14 2016 02:04
Yep, seems to be working now. I changed EST to EDT in both clock-zoned and pump-history-zoned, appears to be working ok.
Rachel Sandlain
@audiefile
Mar 14 2016 02:07
Oh didn't notice clock-zones. Everything worked with out that...
Brian Bonner
@itsn0tm3
Mar 14 2016 03:13
doublecheck my cron if statement, it should be if (this > that); then dothis; fi; && other openaps stuff?
will that work?
I was trying to test it in my bash shell but I kept getting a > prompt so I figured i was missing something somewhere
nevermind, got it to work in bash so I guess it should work. Glad to answer my own question ha
I think I missed a ; somewhere
Scott Leibrand
@scottleibrand
Mar 14 2016 03:57
you don't want to mix if; then; fi with &&
Brian Bonner
@itsn0tm3
Mar 14 2016 04:01
I put it in a ( ) so I figured that was enough seperation?
(if (enact has duration); then do enact; fi;) && the rest of the loop
or is that still iffy?
or in theory I could take it out of cron itself and just put it in enact
Matt Pressnall
@logichammer
Mar 14 2016 04:06
Just wanted to say we are up on the dev NS branch and DAMN, I love the updated information on NS...looks awesome! Also, almost have a new up and Pi running with the TI stick so I should have meal assisted (using the dev source now) and increased range! Thanks so much for all the work being put into active development and for the tools available! Clearly, very stoked. :smile:
After that: Edison and TI...want to go mobile easier...
Scott Leibrand
@scottleibrand
Mar 14 2016 04:13
:+1:
Dana Lewis
@danamlewis
Mar 14 2016 04:26
:) @logichammer
Jim Matheson
@jmatheson
Mar 14 2016 06:44
@logichammer what’s your hardware strategy for the TI plus Edison?
Stephan
@MosiGitHub
Mar 14 2016 06:53
Got the TZ Problems yesterday as well - with timeszones set to Germany on the Edison ...glad that it was not only belonging to my setup. You guys switched to summertime yesterday, is that correct? so maybe because some of the servers related to my Nightscout setup are located in the states...
Nathan Morse
@morsen
Mar 14 2016 06:59
@scottleibrand : I'm just getting things set up, and do you have any time to chat about quick-packages.sh? I'm on Mac/OSX, and apt-get is not available, so I am modifying this script to work on OSX... (P.S. - I'm also pretty new to Gitter. What's the best way to have a direct 1-to-1 conversation? In case we don't want to noise up the main channel.) Thanks!
Scott Leibrand
@scottleibrand
Mar 14 2016 07:01
Ew... I know it's possible, but not sure how to get everything you need. Have you tried brew or macports? Probably best to keep it here, as others may be more helpful than me.
Jim Matheson
@jmatheson
Mar 14 2016 07:02
i can be helpful in the way of moral support :clap:
Scott Leibrand
@scottleibrand
Mar 14 2016 07:03
:-p
Nathan Morse
@morsen
Mar 14 2016 07:04
k. Yeah, I’m just using brew. It has “new official” status over MacPorts. The problem is, I seem to already have had all the packages I need. In my OS X-specific block in the script, I added a “brew update && brew upgrade”, but no installs, and the entire script seems to work… I’m just right there now though; I’ll continue with the setup and see if I encounter any problems, but I wanted to engage you, since you wrote the script! :)
thx
Scott Leibrand
@scottleibrand
Mar 14 2016 07:05
I would do each line manually rather than trying to run the whole thing.
Nathan Morse
@morsen
Mar 14 2016 07:08
ok
Scott Leibrand
@scottleibrand
Mar 14 2016 07:10
Or check the docs further down for the full manual install directions (optional alternative to running the quick scripts)
Nathan Morse
@morsen
Mar 14 2016 07:15
openaps-install-udev-rules is not happy:
Unknown:scripts morsen$ sudo openaps-install-udev-rules
/usr/local/bin/openaps-install-udev-rules: line 43: /etc/udev/rules.d/80-medtronic-carelink.rules: No such file or directory
failed /etc/udev/rules.d/80-medtronic-carelink.rules
/usr/local/bin/openaps-install-udev-rules: line 43: /etc/udev/rules.d/80-dexcom.rules: No such file or directory
failed /etc/udev/rules.d/80-dexcom.rules
/usr/local/bin/openaps-install-udev-rules: line 49: udevadm: command not found
Scott Leibrand
@scottleibrand
Mar 14 2016 07:25
Not sure if OS X does udev
might try skipping that.
Oh, wait, are you trying to use Carelink with the Mac?
I don't think they've gotten that working yet...
bewest/decoding-carelink#179
Nathan Morse
@morsen
Mar 14 2016 07:32
Yikes, okay. Yes, that’s what I’m going to try. Well that’s no fun. :) Thanks.
Does the basic OpenAPS system work on Linux? Windows? Both?
Scott Leibrand
@scottleibrand
Mar 14 2016 07:33
Linux. Most people use Raspberry Pi to start with.
Not much use in running it on a desktop OS given the lack of range of the carelink.
Nathan Morse
@morsen
Mar 14 2016 07:37
Ok, well that is my misunderstanding. I started this a few days ago before my Pi arrived, and I’m just getting back to it, and it was not apparent that the "Setting Up openaps and Dependencies” steps needed to be run on the Pi vs. on my desktop computer.
Scott Leibrand
@scottleibrand
Mar 14 2016 08:09
Ah. Yes, everything happens on the Pi. You only use the computer for ssh.
diabeticgonewild
@diabeticgonewild
Mar 14 2016 13:03
Microcenter has the Raspberry Pi 2s on sale for Pi Day!
$29.99!
I'm going to buy 5 so I can have a "receiverless" home!
Also Intel came up with a Raspberry Pi type board that is almost identical in design:
diabeticgonewild
@diabeticgonewild
Mar 14 2016 13:09
But the good news is I think it can run MATLAB!!!!!!!!!!!!!!!!!!!!!!!
Samuel
@mayaalinder
Mar 14 2016 13:22
@diabeticgonewild What matlab toolbox where you looking at for this? Control System?
JaysonEwer
@JaysonEwer
Mar 14 2016 13:31
Regarding the time change, did I do something wrong? 😜
The only thing I needed to do to fix stuff was change the time on my pump and my cgm, oh and restart my NS Azure service.
I did check to make sure my Pi automagically updated its time and I noticed it is using EDT, so I think I'm good. @jmatheson my blue line is working fine, but my ND site was not getting updates properly yesterday which prompted me to restart my azure site.
*NS site
JaysonEwer
@JaysonEwer
Mar 14 2016 13:36
@diabeticgonewild Microcenter has been selling Pi 2's at that price point for a while. I know I picked up one for that price a month ago. A price of 3.14 would be nice. 😉
diabeticgonewild
@diabeticgonewild
Mar 14 2016 15:15
Awesome!
Matt Pressnall
@logichammer
Mar 14 2016 16:37

@jmatheson , my hardware strategy is to hit up @eyim and ask him to tell me what to buy...but I am going to shoot for the lego style clip on hardware and most likely find a decent lipo battery to solder on. Once I have that squared away, I'll look at soldering with the Edison (and TI again). Sound decent? Anything I am missing?

I can't believe how fast development is moving. 4 months ago I was saying "Man, this is incredible! I'd love to have increased range and size (in that order) at some point" and then like a month later, both had been tackled!

Dana Lewis
@danamlewis
Mar 14 2016 16:38
:) ^^ +100
Jim Matheson
@jmatheson
Mar 14 2016 16:40
by lego style do you mean sparkfun?
Dana Lewis
@danamlewis
Mar 14 2016 16:48
The latest: #OpenAPS is (n=1)*40!
JaysonEwer
@JaysonEwer
Mar 14 2016 16:51
👍
Colin Lennon
@colinlennon
Mar 14 2016 17:08
woohoo :clap:
Rachel Sandlain
@audiefile
Mar 14 2016 18:10
@JaysonEwer check the set up on your zoned reports. My pi switched timezones but the pump history and clock reports did not. That meant everything was off by an hour
Microcenter is doing a pi zero give away on their social media
Ben West
@bewest
Mar 14 2016 20:06
the carelink stick won't work on osx, it has to do with osx usb
the TI stick and other variants will work on osx
openaps itself will work on mac and windows, yes
linux trivially supports all the hardware involved, depending on the hardware, you can do everything on your laptop, but openaps itself is fairly agnostic
eyim
@eyim
Mar 14 2016 20:44
@logichammer
Hi Matt - if you want to go the sparkfun route - here is what you need:
This one to connect the edison: https://www.sparkfun.com/products/13045
This one for an already wired battery: https://www.sparkfun.com/products/13037 but it is a small battery so can use this one and solder a larger battery if you want: https://www.sparkfun.com/products/13727
And you need a board to act as a spacer between the base board and the battery board. Otherwise you can get a short. Here are two options:
https://www.sparkfun.com/products/13033
https://www.sparkfun.com/products/13034
and pick up one or two of these to connect it all up:
https://www.sparkfun.com/products/13187
This will allow you to do the TI stick over USB in a much smaller package than rPi
Eric Luhrs
@eluhrs
Mar 14 2016 21:01
@eyim Thanks very much for providing this list of components. I'm eager build a smaller edison-based openaps device. The only thing I'm not clear on is your note about adding a spacer board. It sounds like this is done simply to create some space for the battery, so why not use taller standoffs? Or does the spacer board provide additional functionality as well?
Scott Leibrand
@scottleibrand
Mar 14 2016 21:01
you need a spacer board to connect the IO bus between the boards above/below
extra spacer nuts would prevent the boards from connecting
the problem is the battery wires sticking out the bottom of the battery board will contact and short out on the microUSB thing immediately below if you plug the battery board directly into the main board
so you have to use a board with no metal right below those wires
Jim Matheson
@jmatheson
Mar 14 2016 21:05
i am not using a spacer board. i ended up avoiding the short with non conductive tape
Scott Leibrand
@scottleibrand
Mar 14 2016 21:05
yeah, you can do that, but you have to keep an eye out for the sharp wires eventually wearing through the tape and shorting
Eric Luhrs
@eluhrs
Mar 14 2016 21:06
Okay, I get it. Thanks.
scottleibrand @scottleibrand likes to be comprehensive on safety issues for the benefit of others reading.
Jim Matheson
@jmatheson
Mar 14 2016 21:08
@eluhrs you already have a TI stick?
I ended up using this connector from the OTG to the TI stick. http://www.amazon.com/gp/product/B00K4RZQWQ
Eric Luhrs
@eluhrs
Mar 14 2016 21:10
No, I only have a standard RPi rig so far, but this stuff is addicting....
Jim Matheson
@jmatheson
Mar 14 2016 21:11
well you will need one of those if you are going the sparkfun route (someone correct me if i am wrong about that)
pretty sure you can’t use a carelink
Eric Luhrs
@eluhrs
Mar 14 2016 21:12
Yes, I realize that. I'm eager to ditch carelink. In fact, I have been for several years now....
Jim Matheson
@jmatheson
Mar 14 2016 21:13
haha. decades?
pretty sur it’s been over 10 years since i first got one
Eric Luhrs
@eluhrs
Mar 14 2016 21:15
Yea, something like that. I will be VERy happy to see it go. Shoot me a photo of your rig, if you get the chance.
Stephan
@MosiGitHub
Mar 14 2016 21:16
@scottleibrand by the way: I tried to use your timezone fix (EDT instead of EDT) with my timezone (CET was set, I changed it to CEDT) ... now I tried CEST, but still got the problem and still working on it. As backgurnd: the mongo server I have selected as well as Azure are located in the States. Germany will switch to summer time in two weeks. Just for info ... do you want a PR for this topic?
Scott Leibrand
@scottleibrand
Mar 14 2016 21:18
Probably need to figure out which project the problem is in before we can open an issue
JaysonEwer
@JaysonEwer
Mar 14 2016 21:22
@audiefile thank you, did you just change the Openaps.ini to have the correct time zone? I noticed I have two instances of EST rather than EDT in the ini file.
Scott Leibrand
@scottleibrand
Mar 14 2016 21:27

Some input from an email I just got:

"The new Pi3 has of course come out to great fanfare. It's a hot item… quite literally. It runs REALLY hot! Like "singe you skin hot" in some places. This is not the case with all boards but since you are making devices that people are going to have on their person or tuck away in tight quarters you might want to take note of this.

When it comes to a DIY development board, we are big fans of the Beagle Bone. Although it is less well know we think it is far more powerful, better made and easier to incorporate into designs because of it's layout and tons of I/O capabilities. Just something to think about."

Stephan
@MosiGitHub
Mar 14 2016 21:29
@scottleibrand : okay, I will try to find it out tomorrow :+1:
Rachel Sandlain
@audiefile
Mar 14 2016 21:38
Yes @JaysonEwer I did just edit the .ini directly
Brian Bonner
@itsn0tm3
Mar 14 2016 21:40
@scottleibrand That makes me wonder about the recommended case with the Pi3 then, not sure if it would have enough airflow to deal with those heat issues if it does have that.
Beaglebone doesn't need the SD card, so that would be one less thing to deal with too.
Scott Leibrand
@scottleibrand
Mar 14 2016 21:41
no idea how beaglebone compares to Edison...
Brian Bonner
@itsn0tm3
Mar 14 2016 21:42
I haven't looked up beaglebone in a long time, I don't think it has wifi or BT onboard, so the edison probably is a bit better in some respects
Nope, doesn't have wifi/bt
eyim
@eyim
Mar 14 2016 21:45
@eluhrs Eric - here is an explanation of why the spacer board is needed.
https://learn.sparkfun.com/tutorials/sparkfun-blocks-for-intel-edison---battery-block?_ga=1.254428296.1007630407.1451956042#using-the-battery-block
As @jmatheson said, I did the electric tape fix for a while with no problems.
Here is another cable that can be used to hook up the sparkfun to the TI via USB that is a straight on connector:
http://www.amazon.com/gp/product/B00B4GGW5Q?psc=1&redirect=true&ref_=oh_aui_detailpage_o03_s00
Jim Matheson
@jmatheson
Mar 14 2016 21:49
@eyim what does your setup look like?
i set mine up side by side since it was the same size as mt 2000mAh battery, but curious what others look like
The one on the left is with no space board. The middle one is the soldered TI stick to Intel Breakout board and the one on the right is spacer with a USB battery added
Jim Matheson
@jmatheson
Mar 14 2016 21:52
what battery is used for #1 on the left?
IMG_4535.jpg
is there no battery or is it one of those small 850s under that is just not visible?
eyim
@eyim
Mar 14 2016 21:54
It is just the small battery that you can buy at sparkfun. It just isn't visible. You can see it here:
https://www.dropbox.com/sc/rx4kqqqcbgv92u8/AACOgHNAAN78l6Pbb9yHzuKqa
Jim Matheson
@jmatheson
Mar 14 2016 21:55
how long does it last?
eyim
@eyim
Mar 14 2016 21:55
With the extra USB battery it all fits into a soap box:
https://www.dropbox.com/sc/p59510mk6frj4hs/AABSXFDOiG91BDnT3mJDYRm9a
Set up #1 - I don't know. I never fully ran it down but I don't think a full school day. Setup #2 (the far right one) and the picture above with the USB battery lasted about 10 -12 hours
Jim Matheson
@jmatheson
Mar 14 2016 21:58
mine is ~12hrs. would like it to be longer though. wondering if i combined an 850 with a 2000. i just tape over most of mine and shove it in my pocket.
eyim
@eyim
Mar 14 2016 22:00
I like yours. Much better than my usb external battery attachment. However, I now use the soldered TI stick to Intel breakout board
Jim Matheson
@jmatheson
Mar 14 2016 22:00
i can almost fit it in an icebreakers plastic mint box. my battery is a tiny bit too big in the corners.
yeah, i was trying for that route too but failed :D
eyim
@eyim
Mar 14 2016 22:01
Ya - took me a few tries as well. Lost a TI stick from smoking it twice. And twice soldered the wrong connections on the board itself but have it working now
Jim Matheson
@jmatheson
Mar 14 2016 22:08
actually just got it to fit in the icebreakers box. super tight. wonder if i can cut a hole for chargingand to get to the switches
JaysonEwer
@JaysonEwer
Mar 14 2016 22:09
Do you have a picture of the icebreaker a box?
Jim Matheson
@jmatheson
Mar 14 2016 22:10
crap now i can’t get it to fit back in :)
although i got the wintercool flavor :+1:
which i promptly tossed
Brian Bonner
@itsn0tm3
Mar 14 2016 22:14
the edison doesn't get warm?
surprised it doesn't need ventilation. I have the Pi in my pocket and it's a bit warm
but then it's always cold here so I'm happy it's keeping me warm
Jim Matheson
@jmatheson
Mar 14 2016 22:15
way less than the pi
let me be clear, i just fit it into this box 5 minutes ago. zero testing. but the Edison itself is not hot in my pocket. but it is a little warm
Brian Bonner
@itsn0tm3
Mar 14 2016 22:18
cool. The edison will probably be my next project after I get the 2 TI's I have on the way here working
JaysonEwer
@JaysonEwer
Mar 14 2016 22:22
@jmatheson, let me know how that icebreaker a box works for you. I was going to put mine in either a clear plastic playing cards case or a plastic electrical tape case, the only problem with both of those options is they don't have a real good clasp so I'll end up possibly taping them closed.
Jim Matheson
@jmatheson
Mar 14 2016 22:22
i did look at the playing cards too
JaysonEwer
@JaysonEwer
Mar 14 2016 22:26
@audiefile Just curious, how did you determine your pump history was an hour off? Just running the report?
Brian Bonner
@itsn0tm3
Mar 14 2016 22:30
I think most of us noticed on NS but you could also notice in the clock-zoned or pumphistory-zoned files by looking at the timezone (-5 or -4 etc)
JaysonEwer
@JaysonEwer
Mar 14 2016 22:31
Thanks @itsn0tm3
JaysonEwer
@JaysonEwer
Mar 14 2016 22:39
Interesting, this is why I'm slightly confused I think. My output seems to indicate it's -5, but it appears to be the correct time?
'''
'''
pi@Ewer-Open-APS:~/Documents/openaps $ cat monitor/clock-zoned.json
"2016-03-14T18:32:06-05:00"pi@Ewer-Open-APS:~/Documents/openaps $
Brian Bonner
@itsn0tm3
Mar 14 2016 22:41
Right, the time will show up correctly, the zoning is there (correct me if I'm wrong) to tell other programs what time zone it is using
so in this case since it's -5, it wouldn't take in account daylight savings
EST is -5, EDT is -4
so NS would incorrectly format it, unless NS is also using EST
the easy fix is to change EST to EDT in the openaps.ini
Jim Matheson
@jmatheson
Mar 14 2016 22:46
i’d say the ice breakers is just a little too small. i think the fit is too tight and will cause issues
Ben West
@bewest
Mar 14 2016 22:49
added some annotations for oref0 recomendations in the help messages
additional help filling in these types of things would be welcome

bewest@bewest-MacBookPro:~/Documents/openaps$ openaps use pump  -h | grep oref0
    model               Get model number [#oref0] [#recommended] [#safe]
                        Check battery status. [#oref0]
    read_bg_targets     Read bg targets. [#oref0]
    read_carb_ratios    Read carb_ratios. [#oref0]
    read_clock          Read date/time of pump [#oref0]
                        Read insulin sensitivities. [#oref0]
                        Fetch the currently selected basal profile. [#oref0]
    read_settings       Read settings. [#oref0]
    read_temp_basal     Read temporary basal rates. [#oref0]
    set_temp_basal      Set temporary basal rates. [#oref0]
bewest@bewest-MacBookPro:~/Documents/openaps$
also...
these all do reasonable things now:
 2164  openaps use pump set_clock --to now
 2165  openaps use pump set_clock --to 
 2166  openaps use pump set_clock 
 2167  openaps use pump set_clock --to bad
 2168  openaps use pump set_clock --to $(date -Iseconds)
 2169  openaps use pump set_clock --to now
 2170  openaps use pump set_clock  monitor/clock.json 
 2171  openaps use pump read_clock
 2172  openaps use pump set_clock --to now
 2173  openaps use pump set_clock  monitor/clock.json 
 2174  openaps use pump set_clock --to $(date -Iseconds)
 2175  openaps use pump  iter_pump 100
 2176  openaps use pump  -h
Tim Howard
@tghoward
Mar 14 2016 22:51
:+1:
Ben West
@bewest
Mar 14 2016 22:54
s$ openaps use pump bolus -h
usage: openaps-use pump bolus [-h] input

Send bolus command. [#warning!!!]

positional arguments:
  input

optional arguments:
  -h, --help  show this help message and exit

  Beware! This is a powerful command because it can give a lot of
  insulin.  Please be careful!
  Not a part of oref0.
  -----------------------
  Requires json input with the following keys defined:
    * `units` - Number of units to bolus.
  Zero point one units:
  { "units": 0.1 }
  Two units:
  { "units": 2 }

bewest@bewest-MacBookPro:~/Documents/openaps
Kevin Wells
@wkcwells
Mar 14 2016 22:55
:+1:
JaysonEwer
@JaysonEwer
Mar 14 2016 22:56
@bewest Thank you! I'm going to be building a new rig soon and my plan is to provide A LOT MORE feedback. I was a bit (insert ALOT) selfish the first time around, it's been running great, and I plan to go through the whole process again. I'll do what I can to assist with things like you just mentioned.
Dana Lewis
@danamlewis
Mar 14 2016 22:56
:+1:
Ben West
@bewest
Mar 14 2016 22:57
bit late for this time change, but should be useful for the next one
Dana Lewis
@danamlewis
Mar 14 2016 22:57
(For both @JaysonEwer Comment and @bewest New warning!)
Ben West
@bewest
Mar 14 2016 22:57
now, assuming ntp will move you through a time transition, you can tell openaps to set the time shortly after these transitions occur
Jason Calabrese
@jasoncalabrese
Mar 14 2016 22:57
would need to do the same for the dex too
Brian Bonner
@itsn0tm3
Mar 14 2016 23:02
Oooh I'm excited. it would be great to be able to sync the pump time via NTP to keep it accurate. That would help synchronize times with different things.
Ben West
@bewest
Mar 14 2016 23:02
openaps schedule add time-change-event "every $(dst-generate-rrule)" or something
Brian Bonner
@itsn0tm3
Mar 14 2016 23:05
(thumbsup)
Jason Calabrese
@jasoncalabrese
Mar 14 2016 23:09
I wouldn't trust something only run 2 times a year
would need to find a way to have it get used more often
Brian Bonner
@itsn0tm3
Mar 14 2016 23:12
in theory you could synchronize pump time with NTP every week or every other week or something like that so it gets used. and have an additional sync when DST rolls around
(depends on how picky you are about the time being right)
Jason Calabrese
@jasoncalabrese
Mar 14 2016 23:13
or anytime it's off by more than a minute or some other threshold
Ben West
@bewest
Mar 14 2016 23:13
or by next few cycles we'll have UI to send messages back and forth: "looks like time change is happening in a few weeks, confirm schedule time change update automatically? {OK, Cancel}"
my pi's sense of time is incredibly flaky
Jason Calabrese
@jasoncalabrese
Mar 14 2016 23:15
yeah, with the pi the time set was going in the other direction
but all my pi's are doing nothing now
Ben West
@bewest
Mar 14 2016 23:15
need hardware clock and known transaction points with ntp
the TI stick is perfectly usable on mac/osx, so maybe doing from laptop is good enough
Jason Calabrese
@jasoncalabrese
Mar 14 2016 23:16
or Edison
Ben West
@bewest
Mar 14 2016 23:17
not sure edison has real clock either
Jason Calabrese
@jasoncalabrese
Mar 14 2016 23:17
thought it did, but not sure why I thought that
Ben West
@bewest
Mar 14 2016 23:19
oh, great
needs some support to make it effective, looks like
Jason Calabrese
@jasoncalabrese
Mar 14 2016 23:19
something about an extra battery
Ben West
@bewest
Mar 14 2016 23:20
yeah, most devices have a tiny battery just for this clock
Brian Bonner
@itsn0tm3
Mar 14 2016 23:20
yeah all RTC's need a battery to keep them as a RTC
Ben West
@bewest
Mar 14 2016 23:20
they recommend a capacitor in the post, but I think that just helps, really need a battery
it helps if you guarantee certain minimums in uptime/downtime I guess
Brian Bonner
@itsn0tm3
Mar 14 2016 23:22
J18 - pin 5 V_VBAT_BKUP RTC backup battery input.
Ben West
@bewest
Mar 14 2016 23:23
the pump itself has both a capacitor and a little battery
Eric Luhrs
@eluhrs
Mar 14 2016 23:33
@jmatheson @eyim thanks the pics. My rig is fairly compact since i found a 9000mah battery that matches the Pi's profile, but these are really inspiring. Time to break out the soldeting iron! Two questions: 1) Have you tried to find a plastic version of an Altoids box? Or perhaps just line a metal one with something non-conductive? Could add some holes if a little ventilation was required. 2) Is it possible to wire two batteries in series? Would be great to have two of those 2000mAh LIPO batteries to draw from.
Eric Luhrs
@eluhrs
Mar 14 2016 23:42
Well after a moment of thought, I concluded that the TI probably wouldnt work very well in a metal box....😜
JaysonEwer
@JaysonEwer
Mar 14 2016 23:44
👍😊
it was a tight fit but i have everything in. the main issue is i dont want to try to shove everything in over and over
i ended up cutting a hole for a usb cord but that made it so i had to tape the whole thing up. if i need to hard reboot it’s going to suck.
but something around that size maybe a smidge bigger would be good. someone said they are using a large tic tac case. that may work but is a bit bigger
Jim Matheson
@jmatheson
Mar 14 2016 23:51
the biggest issue i find with (no fancy 3d) cases is these same things. ability to flip power on battery, press reset button on sparkfun board and to plug into the micro USB charger
Eric Luhrs
@eluhrs
Mar 14 2016 23:55
The icebreakers box is what made me think of looking for a plastic altoids box. The squared corners would make battery placement easier. A rubber band or one wrap of electrical tape would keep things secure yet accessable.