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

15th
Mar 2016
eyim
@eyim
Mar 15 2016 00:01
Some one tried it with an altoids metal box and it didn't work well. Plastic seems to be okay.
Jim Matheson
@jmatheson
Mar 15 2016 00:02
it looked bulky (thick) to me. i am a sticker for minimum. if this mint box i have here was square it’d have tons of room (borderline too much but would be good.
i like the plastic deck of cards pack idea if it’s not too small
i have it in my pocket now and it is for sure bulkier than what i had before which was just the hardware itself with athletic tape over it.
Scott Leibrand
@scottleibrand
Mar 15 2016 00:16
@bewest if we can figure out a way to reset the display time on the dexcom, that would be awesome. then we could even do things like have the Pi autodetect (the same way your laptop does) that it has changed timezones, and switch the pump and dexcom clocks to the new timezone automatically :)
Ben West
@bewest
Mar 15 2016 00:21
there's a analogous command for reading/setting time on dexcom
Eric Luhrs
@eluhrs
Mar 15 2016 00:26
Might be worth perusing a Muji store to see if they have any small plastic boxes that might work. I might be near one later this week. If so, I'll take a look and report back.
Scott Leibrand
@scottleibrand
Mar 15 2016 00:28
is time setting already in dexcom_reader? or just something we know could be done?
Rachel Sandlain
@audiefile
Mar 15 2016 00:29
@JaysonEwer my pump history was wrong on ns, the temp displayed did not match the pump, and the urchin watchface had negative minutes left
JaysonEwer
@JaysonEwer
Mar 15 2016 00:29
Ah, ok, thanks
Ben West
@bewest
Mar 15 2016 00:44
@audiefile yikes!?
Jason Calabrese
@jasoncalabrese
Mar 15 2016 00:54
really like my headphone case, tempted to try finding another
Ben West
@bewest
Mar 15 2016 00:54
it's pretty good :-)
or find a 3d model like it?
Jason Calabrese
@jasoncalabrese
Mar 15 2016 00:55
have the 2000mAh battery in there now
a clear case with a hinge and latch would be great
heh, lasso got pegged as a project worthy of copy/pasting by code-cognizant bots
I've used programcreek and similar sites to find specific code examples
can be much better when code exists but discussion does not
Jim Matheson
@jmatheson
Mar 15 2016 01:05
those are big pockets @jasoncalabrese
Ben West
@bewest
Mar 15 2016 01:05
literally typed in "dexcom commands"
Jason Calabrese
@jasoncalabrese
Mar 15 2016 01:06
Lasso could have really been good
Was able to ack server alarms
lasso was really nice
if it was under development, I'd be buying lots of android watches, lol
Jim Matheson
@jmatheson
Mar 15 2016 01:23
what are people doing in their loop to deal with multiple rigs going at the same time?
JaysonEwer
@JaysonEwer
Mar 15 2016 01:23
@jasoncalabrese I like that headphone case
Scott Leibrand
@scottleibrand
Mar 15 2016 01:26
@jmatheson I use a modified ruby bin/mmtune that bails immediately if it hears another rig communicating with the pump
that combined with a 120s sleep on preflight fail works well even with 3-4 rigs in range
Jim Matheson
@jmatheson
Mar 15 2016 01:27
will that catch a carelink? does it only work with ruby version?
Scott Leibrand
@scottleibrand
Mar 15 2016 01:28
with carelink range is so short that just the 120s preflight fail sleep is good enough
Jim Matheson
@jmatheson
Mar 15 2016 01:28
on both sides?
for example if i have my carelink at my side by my desk and i am carrying my edison with TI say in my pocket?
Scott Leibrand
@scottleibrand
Mar 15 2016 01:28
whenever either rig fails it sleeps for 2 minutes, to give the other one time to finish up
Jim Matheson
@jmatheson
Mar 15 2016 01:29
when i first got my TI going i just bailed out on the pi, but it really left me in a lurch when i had not looked at it in weeks
Scott Leibrand
@scottleibrand
Mar 15 2016 01:29
do the mmtune thing as well on the edison, and should be fine
best to space them out a bit so their ranges don't overlap too much, but even if they do it will work fine.
eyim
@eyim
Mar 15 2016 01:39
I've got the 120s sleep on preflight but not the mmtune that bails if it hears another rig. Was that difficult to implement?
Scott Leibrand
@scottleibrand
Mar 15 2016 01:55
requires using the ruby mmtune. I have a branch in my copy of the repo with everything you'd need
and if it's useful we could port it to the python version without too much trouble I'm sure
Brian Bonner
@itsn0tm3
Mar 15 2016 02:06
I'd be interested.. I'm planning to put a few rigs around my apt so I don't have to carry anything. Probably do the same at work
I've accumulated quite a few Pi's over time so I'd only really need the TI's and then get the code going
Brian Bonner
@itsn0tm3
Mar 15 2016 02:12
has anyone ever tried running openaps on a Pi1?
JaysonEwer
@JaysonEwer
Mar 15 2016 02:14
So I changed my EST to DST in my OpenAPS.ini file and now NS is not updating properly. Does that mean I need to modify something in NS?
Brian Bonner
@itsn0tm3
Mar 15 2016 02:14
I don't think anyone had to, just needed to refresh the page
I also refreshed the app too just to be safe but I don't think it needed it
and you mean EDT right, not DSDT?
err DST?
JaysonEwer
@JaysonEwer
Mar 15 2016 02:17
Doh, my bad I totally dorked that up. I modified it incorrectly.
Scott Leibrand
@scottleibrand
Mar 15 2016 02:19
@itsn0tm3 we have two Pi1 rigs. works fine, just a bit slower.
Brian Bonner
@itsn0tm3
Mar 15 2016 02:20
that's great. Will be good to get some use out of them again
Peter Miller
@ochenmiller
Mar 15 2016 02:25
Battery fires are real: https://goo.gl/photos/kHgL52YoDGBVQ5d36
Brian Bonner
@itsn0tm3
Mar 15 2016 02:26
?!?
Peter Miller
@ochenmiller
Mar 15 2016 02:27
Internal short
Brian Bonner
@itsn0tm3
Mar 15 2016 02:27
wow.
hope nothing else got hurt? looks like that melted through quite a bit
Peter Miller
@ochenmiller
Mar 15 2016 02:28
It was like a blow torch coming through the enclosure. Everyone's fine.
Definitely rethinking this design though.
Brian Bonner
@itsn0tm3
Mar 15 2016 02:30
yea wow
(eyeballs his power supply)
Peter Miller
@ochenmiller
Mar 15 2016 02:30
I'm pretty sure it was because I put point pressure on the outside of the battery.
Ben West
@bewest
Mar 15 2016 02:31
@scottleibrand mmtune is in mmeowlink directly now, https://github.com/oskarpearson/mmeowlink/blob/master/mmeowlink/mmtune.py
added 10+ dexcom commands
Scott Leibrand
@scottleibrand
Mar 15 2016 02:32
@bewest does that have the feature to abort immediately on detecting an abnormal packet? or does it proceed to try to tune anyway?
Jason Calabrese
@jasoncalabrese
Mar 15 2016 02:33
wow @ochenmiller very scary
what did you have to do to put it out
you think something punctured the battery?
Peter Miller
@ochenmiller
Mar 15 2016 02:36
Yeah, I think point pressure on the outside of the battery caused an internal short.
Ben West
@bewest
Mar 15 2016 02:36
@ochenmiller wow, bet that smelled lovely...
Peter Miller
@ochenmiller
Mar 15 2016 02:36
burnt plastic, yum
Jason Calabrese
@jasoncalabrese
Mar 15 2016 02:37
need to get a better case or safer battery
like the ones Sulka is using
Peter Miller
@ochenmiller
Mar 15 2016 02:37
@jasoncalabrese, I took it apart and disconnected the battery, but it had pretty much finished discharging by then I think
yeah, will be rethinking this entirely.
Jason Calabrese
@jasoncalabrese
Mar 15 2016 02:38
there was a flame till you disconnected it?
was looking at a dexcom receiver that I cut open and they use similar batteries, but have a metal shield to protect it
Peter Miller
@ochenmiller
Mar 15 2016 02:39
hard to say, I think it was starting to die down as I disconnected. First thought was overcurrent caused be a short in the rig, but I think it was battery internal
Ben West
@bewest
Mar 15 2016 02:40
you're probably lucky it didn't explode
not sure getting hands on is right tactic if it's on fire?
dunno how accessible connector is
Peter Miller
@ochenmiller
Mar 15 2016 02:40
feeling very lucky, yes
Scott Leibrand
@scottleibrand
Mar 15 2016 02:41
I read up on lithium battery safety, and their advice is to use water (or even soda, whatever is nearby) on it
Ben West
@bewest
Mar 15 2016 02:41
nice, was just wondering that
Scott Leibrand
@scottleibrand
Mar 15 2016 02:41
mostly just cools down the surrounding materials to keep it from spreading
the lithium fire will burn itself out pretty quickly for small batteries
how long did yours last @ochenmiller ?
Peter Miller
@ochenmiller
Mar 15 2016 02:42
fire was less than a minute
Jason Calabrese
@jasoncalabrese
Mar 15 2016 02:42
that would be more than enough time for a really really bad burn
Ben West
@bewest
Mar 15 2016 02:42
if that happens in your backpack, things could get ugly
Peter Miller
@ochenmiller
Mar 15 2016 02:42
yeah, I'm pretty shaken up about that possibility
Jason Calabrese
@jasoncalabrese
Mar 15 2016 02:42
yeah
Scott Leibrand
@scottleibrand
Mar 15 2016 02:42
they also mention that if you start to see the battery swell up, try to get it outside or at least onto a non-flammable surface
Ben West
@bewest
Mar 15 2016 02:42
yeah, really bad in a pocket
if it's swelling, it's already corrupt, I think
Scott Leibrand
@scottleibrand
Mar 15 2016 02:43
yeah, swelling = about to blow (maybe)
Ben West
@bewest
Mar 15 2016 02:43
hmm, there's probably a few I shouldn't be using anymore
Scott Leibrand
@scottleibrand
Mar 15 2016 02:44
we should probably recommend pre-packaged battery packs whenever possible in the docs
Ben West
@bewest
Mar 15 2016 02:44
I havnen't been super careful, just toss them in bag with rest of electronics
Scott Leibrand
@scottleibrand
Mar 15 2016 02:44
or figure out a good way to shield them
Jason Calabrese
@jasoncalabrese
Mar 15 2016 02:44
this is a good reason to try coming up with a case that will work for everyone too
Ben West
@bewest
Mar 15 2016 02:44
yeah
Peter Miller
@ochenmiller
Mar 15 2016 02:44
Fire started when we dropped the unit on the floor, so I think there was already internal damage
Jason Calabrese
@jasoncalabrese
Mar 15 2016 02:45
maybe cases for the batteries
Scott Leibrand
@scottleibrand
Mar 15 2016 02:45
I'll probably figure out a way to wrap (or tape) up the batteries @danamlewis is using when she gets home.
Jason Calabrese
@jasoncalabrese
Mar 15 2016 02:45
something for the 2000 and 850?
Scott Leibrand
@scottleibrand
Mar 15 2016 02:46
wasn't a puncture though, right? just internal damage?
Ben West
@bewest
Mar 15 2016 02:46
there's a lot of RC car/toy kits that use hard-pack batteries
Peter Miller
@ochenmiller
Mar 15 2016 02:47
There were pressure points from components pushing against the battery case. I didn't leave enough room, and "pressed" when I assembled.
There are commercial hard pack batteries, thinking about a design with separable batteries now.
Pretty sure it wasn't a short in the rig, because the powerboost 1000C has a current limiting circuit.
Jason Calabrese
@jasoncalabrese
Mar 15 2016 02:55
Pete Schwamb
@ps2
Mar 15 2016 03:14
@ochenmiller Scary! I was careful in designing the RL case to have no sharp parts where the battery goes. Also, the bottom of the RL board is smooth. The one issue I have is that the closure system isn’t strong enough; I usually wrap tape around the case to make sure it stays closed. If it popped open, and the lipo bounced around freely that would be another safety issue.
Jason Calabrese
@jasoncalabrese
Mar 15 2016 03:16
@bewest I put the latest openaps/dev on a couple edisons, seeing lots of git cpu each report
think doing a git gc each time is too much
eyim
@eyim
Mar 15 2016 03:24
I ordered a few of these because I was worried about the fire potential. Thinking this was safer. They are sold out now. http://www.amazon.com/gp/product/B00XOFU2Q0?psc=1&redirect=true&ref_=oh_aui_detailpage_o05_s00
I also had bought this http://www.amazon.com/gp/product/B00RHD2CL8?psc=1&redirect=true&ref_=oh_aui_detailpage_o06_s00 but the reviews aren't so good and I think Scott had issues with it.
Ben West
@bewest
Mar 15 2016 03:29
hmmm
spike in cpu on gc?
not sure connector is right
Scott Leibrand
@scottleibrand
Mar 15 2016 03:31
Yeah, no way that PS4 battery is 2000mAh.
Plug types are all wrong though. Do you have a plug that would adapt that to Edison?
Jason Calabrese
@jasoncalabrese
Mar 15 2016 03:32
Yeah, watching it on top, doing tons of stuff
Several git commands using 99% CPU
Peter Miller
@ochenmiller
Mar 15 2016 03:33
Rig boots up fine, so it was definitely an internal short in the battery.
Ben West
@bewest
Mar 15 2016 03:33
in that case, maybe in preflight
should I take out the gc line?
make it optional and "off" by default?
I think in preflight would be my recommendation
interactively, it'll take longer, but it actually comes with sanity check that git is ok
Ben West
@bewest
Mar 15 2016 03:35
and ensures that the subsequent git operations will not run you out of inodes ;-)
eyim
@eyim
Mar 15 2016 03:35
@scottleibrand I agree that battery doesn't last anywhere near as long as my other 2000 battery. In the reviews, someone said they cracked it open and it was something much smaller. I'll be testing it for the next few days to see how long it lasts me
Ben West
@bewest
Mar 15 2016 03:36
yeah, reflog :-D
Jason Calabrese
@jasoncalabrese
Mar 15 2016 03:36
Too much for each report
Ben West
@bewest
Mar 15 2016 03:36
it's every report "group"
so every invoke
Jason Calabrese
@jasoncalabrese
Mar 15 2016 03:36
Still too much
Ben West
@bewest
Mar 15 2016 03:36
so I can make it optional switch --with-gc
and it's default false
but what about taking it out and putting git gc in preflight?
Jason Calabrese
@jasoncalabrese
Mar 15 2016 03:38
Look at git gc --auto, think it's what we want
Ben West
@bewest
Mar 15 2016 03:39
nice......
nice........... it even highlights this as use case
Jason Calabrese
@jasoncalabrese
Mar 15 2016 03:40
Yeah, think that will be perfect
Ben West
@bewest
Mar 15 2016 03:40
"some commmands run git gc --auto after performing operations that could create many loose objects"
wow
yeah, might be perfect
Jason Calabrese
@jasoncalabrese
Mar 15 2016 03:41
I have a good test setup now if you want to put a branch up
Ben West
@bewest
Mar 15 2016 03:42
working on it
found the terminal
Jason Calabrese
@jasoncalabrese
Mar 15 2016 03:42
We might want to warn people to do a manual git gc before upgrading
Ben West
@bewest
Mar 15 2016 03:43
ok, hmmm repo.git.gc( )....
Jason Calabrese
@jasoncalabrese
Mar 15 2016 03:43
Was wondering if it was in the python library
Ben West
@bewest
Mar 15 2016 03:45
openaps/openaps#80
nice, looks like git gc --auto is pretty smart
Jason Calabrese
@jasoncalabrese
Mar 15 2016 03:50
looking good, not seeing git in top
Ben West
@bewest
Mar 15 2016 03:50
if that works, I'll likely cherry-pick that commit onto dev
Jason Calabrese
@jasoncalabrese
Mar 15 2016 03:50
way better than what's on dev
Ben West
@bewest
Mar 15 2016 03:51
on dev now
Jason Calabrese
@jasoncalabrese
Mar 15 2016 03:51
:+1:
updating other rig
Ben West
@bewest
Mar 15 2016 03:52
nice, much simpler than both my suggestions
and the manual indicates that there are lots of scenarios where tooling can create extra stuff
so that's "normal"
there's enough tooling to even provide fine grained exception list in .keep or even override the global behavior
very nice
Jason Calabrese
@jasoncalabrese
Mar 15 2016 03:54
bet this helps with the git corruption issues
Ben West
@bewest
Mar 15 2016 03:55
I hope so
I made three changes at same time...
Jason Calabrese
@jasoncalabrese
Mar 15 2016 03:55
:fire:
Ben West
@bewest
Mar 15 2016 03:55
I wanted any possible magic in there, we can isolate which one makes the difference later
bunch new uses for dexcom, think I filled out most of the read uses
there are few I didn't expose to openaps per se, because they seemed to hang the ble driver somehow
Jason Calabrese
@jasoncalabrese
Mar 15 2016 04:00
I need to try getting it setup again
Screenshot 2016-03-14 20.59.38.png
busy rigs
JaysonEwer
@JaysonEwer
Mar 15 2016 04:02
@bewest that battery you posted above is Quite large. 😳 5.3 x 1.7 x 1.6 inches
Item Weight 1.1 pounds
Ben West
@bewest
Mar 15 2016 04:03
would love to get the writes down, there is a write command for: system_time, display_time_offset, and charger_current_setting I would love to get a good understanding with
@JaysonEwer how heavy is an iphone 6?
hehehe
ok, quarter pound
This message was deleted
but smaller than iphone 6 with battery still and would last several days
Jason Calabrese
@jasoncalabrese
Mar 15 2016 04:05
charger_current_setting can detect if the receiver is plugged in?
Ben West
@bewest
Mar 15 2016 04:06
depending on what you mean, yes
Jason Calabrese
@jasoncalabrese
Mar 15 2016 04:06
never got receiver battery into NS, would be nice to add that
Ben West
@bewest
Mar 15 2016 04:06
my understanding is that if you are plugged into usb power and this is turned "off", it'll stop charging
ah, there's a special use for combining the status and level into single json
Jason Calabrese
@jasoncalabrese
Mar 15 2016 04:07
I was thinking it was to read
Ben West
@bewest
Mar 15 2016 04:08
openaps use share battery ...
{
  "status": "NOT_CHARGING", 
  "level": 50
}
but this could be used to arm/disarm charging from phone/usb
could help extend life of NS/openaps rig
Jim Matheson
@jmatheson
Mar 15 2016 04:33
@JaysonEwer, i’ve found that @bewest likes to carry a large gym bag around :) a murse would be way too small for Ben
Ben West
@bewest
Mar 15 2016 04:39
at this point, it's relatively small, if it's roughly size of a phone it's ok if everything is inside the box
and it needs to last more like pebble/dexcom
needs to really last 3 days at least
and get in the box
every time I try to put it in a box, bits and pieces change, hehehe
keeps hopping out
Scott Leibrand
@scottleibrand
Mar 15 2016 04:43
@bewest hates charging things more than he hates carrying them. :)
Jim Matheson
@jmatheson
Mar 15 2016 04:46
nothing wrong with a gym bag!
this thread makes me remember walking the .75mi home with a car battery from Costco in my backpack. was less than ideal.
Ben West
@bewest
Mar 15 2016 06:08
lol
Alin Artiom Kenibasov
@ktomy
Mar 15 2016 06:56
blob
Entered here just to say thank you :) Now our nights look like this:
eyim
@eyim
Mar 15 2016 06:59
:+1:
Scott Leibrand
@scottleibrand
Mar 15 2016 07:00
:-D
Brian Bonner
@itsn0tm3
Mar 15 2016 10:27
blah got git corruption during the night
I guess that will be 1 night looping and 1 night gitting ha
Brian Bonner
@itsn0tm3
Mar 15 2016 10:37
oh well, at least git status confirmed the corrutption and oref0-reset-git fixed it
Brian Bonner
@itsn0tm3
Mar 15 2016 10:43
doesn't look like it was able to fix itself tho so I guess I need a git check in the loop?
dougtakeuti
@dougtakeuti
Mar 15 2016 11:40
@itsn0tm3 I have * * * * * cd /home/edison/mypenaps && oref0-reset-git in my cron. It think Jason or Scott suggested that.
Brian Bonner
@itsn0tm3
Mar 15 2016 12:26
hmm that doesn't seem quite sane to me
that would reset git every minute?
or does it check it, then if needed reset it?
JaysonEwer
@JaysonEwer
Mar 15 2016 12:28
@jmatheson "Car battery in a backpack" BRILLIANT, why didn't I think of that. 😎 LOL
We could run the jumper cables directly to the Apple IIe in @bewest gym bag, GOLDEN!
Brian Bonner
@itsn0tm3
Mar 15 2016 12:33
(chuckling)
JaysonEwer
@JaysonEwer
Mar 15 2016 13:09
All kidding aside @ochenmiller , I'm very glad no one was injured during your scary incident. We are all better off now thinking of ways to improve our own designs to ensure something horrible doesn't happen. Thank you for sharing your story and images!
Jason Calabrese
@jasoncalabrese
Mar 15 2016 13:42
@ktomy :+1:
JaysonEwer
@JaysonEwer
Mar 15 2016 13:53
@ktomy Nice Graph!
fallon225
@fallon225
Mar 15 2016 14:54
So, right now we're not actually looping because we had to go back to the pods. Waiting on some silhouettes to get here - the mios just don't work. Moving on. Is there a way to make the pi the uploader, and not the cell phone? Take the apps off the phone that pull the info from the dex, and put them on the pi. I'm just relaying the question, so hopefully that makes sense.
Dana Lewis
@danamlewis
Mar 15 2016 16:23
@ktomy :)
@fallon225 yes. You just plug the cgm into the pi or use a cable and you can use that.
fallon225
@fallon225
Mar 15 2016 16:32
our dex is plugged into the pi. So have we been uploading everything twice then? having the dex go to the xdrip on his phone?
Dana Lewis
@danamlewis
Mar 15 2016 16:35
it doesn’t hurt to send data to multiple places, in some cases that’s a good thing. but if you’re trying to cut out devices, there’s different options.
I’m not sure if anyone’s combined an xdrip with Pi, so you’d need to have the receiver running and plugged to the Pi for what I was describing
fallon225
@fallon225
Mar 15 2016 16:36
Yeah, we're trying to get his personal cell doing less things.
when he gets his silhouettes in a few days we'll try it out.
Dana Lewis
@danamlewis
Mar 15 2016 16:40
:+1:
Scott Leibrand
@scottleibrand
Mar 15 2016 16:57
@itsn0tm3 it will do a git status first and only reset if that fails.
If you get the dev version it will try oref0-fix-git-corruption before doing a reset.
Matt Pressnall
@logichammer
Mar 15 2016 18:11

@eyim, awesome! Thanks so much for passing along!!! Also, what is the biggest battery you are using? This:
https://www.sparkfun.com/products/13037

And I'm assuming I can solder the TI to any of these boards (for better power consumption rather than using the USB port) when I get around to it. Is that correct?

eyim
@eyim
Mar 15 2016 18:19
I purchased these batteries
https://www.sparkfun.com/products/8483
but never got around to using them with the sparkfun stuff as I switched over to Intel breakout board + Solder TI
After the discussion last night about fire risk I am looking at other options.
I haven't heard of anyone soldering TI stick to the sparkfun board. Not sure that is possible. People are using the Intel Breakout Board if they want to solder the TI stick
Matt Pressnall
@logichammer
Mar 15 2016 18:23
@eyim, Gotcha and thanks for the clarification around which boards I can solder the TI to...last question, do you have that script that will restart your Wi-Fi if it is dead? Want to get that implemented as well...thanks a ton!
William Fallon
@inform880
Mar 15 2016 18:26
My website keeps saying "future" for the pump status. I've fixed all the times for DST, I don't understand why it still says this.
Stephan
@MosiGitHub
Mar 15 2016 18:35
I am playing with the timezones as well but until now no change ... BG is correct and the other values are in the future ...
eyim
@eyim
Mar 15 2016 18:38
@logichammer https://github.com/TC2013/edison_wifi I have been told you have to run it in sudo crontab -e
example here:
/2 * /home/edison/edison_wifi/wifi.sh google.com
Scott Leibrand
@scottleibrand
Mar 15 2016 18:49
if you end up getting the https://www.sparkfun.com/products/13276, the GPIO block (https://www.sparkfun.com/products/13038) has solder points you could use to connect a TI
fallon225
@fallon225
Mar 15 2016 19:04
@MosiGitHub are your basals on the nightscout site off by an hour then too? The basal line is ahead an hour, but the bg dots are correct.
Milos Kozak
@MilosKozak
Mar 15 2016 19:04
@scottleibrand in mealassist was bolusiob renamed to bolussnooze?
Stephan
@MosiGitHub
Mar 15 2016 19:07
@fallon225 same for me and the treatments are off.... I played a little bit with the timezones in the openaps.inibut so far no success...
Colin Lennon
@colinlennon
Mar 15 2016 19:12
this had probably come up before, but I've just been looking at c.h.i.p
sounds like it might be a good fit for openAPS... built-in wifi and BLE. 4GB might not be enough though?
Stephan
@MosiGitHub
Mar 15 2016 19:24
@ochenmiller : i was just wondering why my Edison was nat able to communicate with my sons pump anymore. The solution was that my son removed the edison out of his room after I have shown him the pictures of your burned Lipo ... Now he asks me to buy some safer batteries... clever kid, but I really like the small Lipos I have :-(
Scott Leibrand
@scottleibrand
Mar 15 2016 19:27
@MilosKozak yes. because it has never represented "normal" bolus IOB: it decays at double speed
there is also a new basaliob, so if you want true bolus IOB you can subtract that from net IOB
Ben West
@bewest
Mar 15 2016 19:33
@inform880 are you using the timezone tool to add zone info?
Milos Kozak
@MilosKozak
Mar 15 2016 19:37
thx. updating code to support mealassist but need to keep backward compatibility with master until i switch
Rachel Sandlain
@audiefile
Mar 15 2016 19:46
@inform880 check what time zone is being used for your zoned reports
For me everything was set properly except the time zone for zoned pump history and clock. That was the root of all my problems
Matt Pressnall
@logichammer
Mar 15 2016 19:55
Thanks @eyim and @scottleibrand!
William Fallon
@inform880
Mar 15 2016 20:03
@bewest I don't know what timezone tool you are talking about.
Stephan
@MosiGitHub
Mar 15 2016 20:08
@audiefile got the same problem, presently I am switching the timezone used for the zoned reports between GMT, CET and EET to see if it helps
Ben West
@bewest
Mar 15 2016 20:10
@audiefile yeah, that's nice in some ways, but it's a gotchain other ways
Stephan
@MosiGitHub
Mar 15 2016 20:12
interesting is that equal what I use for the zoned reports the temp basla the future by 50 minutes (for EET, GMT or CET) ...
Maybe I should change Azure to an Europ. server as well as the mlab database to Europe
Milos Kozak
@MilosKozak
Mar 15 2016 20:17
@scottleibrand one more question. what's the difference between treatment.date and treatment.started_at? https://github.com/openaps/oref0/blob/meal-assist/lib/iob/total.js
Ben West
@bewest
Mar 15 2016 20:19
should the timezone be a property of the "device" rather than a "report?"
right now zone is considered property of reports... could decrease maintenance if some of that logic moves into device itself, so all data from a device is influenced same way
Scott Leibrand
@scottleibrand
Mar 15 2016 20:22
@MilosKozak bolus treatments have date, temp basals have started_at
Stephan
@MosiGitHub
Mar 15 2016 20:24
@bewest that sounds easier, but I still don't see a difference, regardless what TZ I used ... maybe i try a US TZ next, just for finding any effect of it
Rachel Sandlain
@audiefile
Mar 15 2016 21:06
Property of device might make more sense. Kind of like think of it as a unit of measure? Like mgdl versus mail?
I also switched the time zone on nightscout to be the gmt offset instead of a city name. That might help too if your site is still showing treatments at the wrong time
JaysonEwer
@JaysonEwer
Mar 15 2016 21:36
I'm finally getting around to backing up my
Stephan
@MosiGitHub
Mar 15 2016 21:50
by the way: even with TZ set to EST the difference is still 50 minutes. So it does't make any difference :-(
JaysonEwer
@JaysonEwer
Mar 15 2016 21:51
Pi. I know, I know, it's crazy that I haven't done a full disk image yet, but I just did one. What method do you all use to backup your rigs? I just used Disk Utilty on my Mac, are you all using other tools, like dd? Thanks in advance for any feedback.
Stephan
@MosiGitHub
Mar 15 2016 22:00
has anybody an idea what I can do with the adjust = missing in the reports I am generating as "-zoned"? can I manipulate it with +1 or something like that? I really apprecciate any advise.
tazitoo
@tazitoo
Mar 15 2016 22:00
@JaysonEwer - I'm mostly sitting at linux box, so I use dd. Whatever you use, try to write your backup out to a spare card and see if that will boot your rig. If not, troubleshooting lies ahead..
@scottleibrand - is there a gist for setting up the sensitivity detection?
Scott Leibrand
@scottleibrand
Mar 15 2016 22:11
not unless @amazaheri or @jasoncalabrese or someone made one when they set it up
tazitoo
@tazitoo
Mar 15 2016 22:16
hmmm, and they both look inactive at the moment
Jason Calabrese
@jasoncalabrese
Mar 15 2016 22:20
there was a word doc floating around
tazitoo
@tazitoo
Mar 15 2016 22:21
happen to have a link?
Jason Calabrese
@jasoncalabrese
Mar 15 2016 22:22
I haven't tried it yet, was created by @amazaheri and @eyim https://www.dropbox.com/s/5u4e8ybbbof6s04/auto-sensitivity%20feature.docx?dl=0
I got a couple rigs updated to meal-assist last night, was hoping to try getting this setup tonight
Scott Leibrand
@scottleibrand
Mar 15 2016 22:22
we should get that pasted into a document in the dev branch of openaps/docs so we can share it around and start improving it
tazitoo
@tazitoo
Mar 15 2016 22:23
thx
Jason Calabrese
@jasoncalabrese
Mar 15 2016 22:23
feel that I don't know where docs go now, can't find anything directly in the docs repo
I'd be happy with a wiki
Stephan
@MosiGitHub
Mar 15 2016 22:33
@fallon225 : I recognized that the reports of pumphistory-zoned.json are missing the +01:00 after the timestamp. Changed the report. Now it is generating reports with the timestamp. Trying tomorrow wether this was sufficient. Maybe you check your reports as well.
dougtakeuti
@dougtakeuti
Mar 15 2016 22:49
Following up the battery conversation from yesterday, the 850 mAh lipo on the edison rig just heated up very quickly 30 min ago. Temperature was certainly above 60 Celsius because it was hard to touch it when I disconnected. It really felt like it was going to burn. My wife freaked out big time for a good reason. Also, I tested some power banks in the physics lab and found out that they vary a lot in terms of nominal and real capacity. I tested three: a cheap (who-knows-what-brand) 36,000 mAh, a 10,000 mAh (http://www.gpbatteries.com/int_en/304c), and a "local" (obviously made in China) good quality 5200 mAh (http://www.iwill.com.br/web/produtos.php?id=29|carregador-portatil-magico). I discharged them at a current of 100 mA and monitored the whole process until they were dead. The results were 1700 mAh, 7100 mAh, and 2700 mAh respectively.
Ben West
@bewest
Mar 15 2016 22:50
ok, openaps now support changing dexcom receiver's clock :-D
{ 
  "requested": {
    "clock": "2016-03-15T15:47:01.368530",
    "offset": -21429.63147
  },
  "enacted_at": "2016-03-15T15:47:02.481538",
  "result": {
    "ACK": true
  },
  "offset": -21429.0,
  "clock": "2016-03-15T15:47:01.368530"
}
that's response for: openaps use share UpdateTime --to now
never pressing the buttons again
Scott Leibrand
@scottleibrand
Mar 15 2016 23:01
:+1:
Tim Howard
@tghoward
Mar 15 2016 23:01
@bewest :+1:
Scott Leibrand
@scottleibrand
Mar 15 2016 23:01
@dougtakeuti did the 850mAh have any trauma to trigger that?
dougtakeuti
@dougtakeuti
Mar 15 2016 23:11
@scottleibrand It has a few scratches from 2 weeks ago when I was trying to make it fit in a TicTac box. No recent traumas. Today it was in a separate pouch just next to the pump sitting on Julia's stomach while she was on the couch watching tv.
Dana Lewis
@danamlewis
Mar 15 2016 23:11
@jasoncalabrese https://github.com/openaps/docs/blob/master/SUMMARY.md but doesn't include any of the bleeding edge or dev related stuff.
dougtakeuti
@dougtakeuti
Mar 15 2016 23:25
After the battery scare I have no rig to send to school as the power banks shut down due to the low consumption of the edison. I'm considering sending the Pi in the backpack. Any idea of its power consumption?
Scott Leibrand
@scottleibrand
Mar 15 2016 23:25
we can run a Pi all day on a large (15000mAh) battery. could probably do a school day on a smaller one.