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

4th
Jul 2017
Dirk Gastaldo
@dirkgastaldo
Jul 04 2017 00:24
To set up two WiFi routers, do I repeat the "network={" and "}" lines, or just those two lines once with the ssid and psk lines for each WiFi router, like this:

network={
ssid="first network"
psk="first wifi password"
ssid="second network"
psk="second wifi password"
}

or

network={
ssid="first network"
psk="first wifi password"
}
network={
ssid="second network"
psk="second wifi password"
}

cameronrenwick
@cameronrenwick
Jul 04 2017 00:24
hey so answering my own question from above, when you edit the oref0 set up instructions be careful for extra spaces when putting the microbolus in. I had done a sloppy cut/paste which added an extra space and lead to all my problems. I'm smbing now -- woot woot!
Jorge Aguilar
@jonahtaxt
Jul 04 2017 00:24
Hi. Does anybody know which alias or report sends checking system clock against pump clock? Im getting a null value when the loop tries checking deliverat:
I just want to know which command it is to run manually and try to figure out the error
Scott Leibrand
@scottleibrand
Jul 04 2017 00:37
@jonahtaxt that is almost always the symptom of another problem, such as stale BG data
@dirkgastaldo the latter is what you want. one network= per network
Dirk Gastaldo
@dirkgastaldo
Jul 04 2017 00:38
@scottleibrand thanks, Scott
Jorge Aguilar
@jonahtaxt
Jul 04 2017 00:58
@scottleibrand the thing is that mdt is getting and formatting data correctly
John Benjamin
@Bender1061
Jul 04 2017 01:02
@scottleibrand I don't think it's a good idea to comment out the openaps mmtune in the oref0-setup.sh
the loop will not run correctly untill I manually run it
Starting supermicrobolus pump-loop at Mon Jul 3 20:34:35 EDT 2017 with 7 second wait_for_silence:
MDT CGM configured; not waiting
Radio ok. Listening: .No pump comms detected from other rigs
mmeowlink.exceptions.CommsException: No acknowledgement from pump on wakeup. Is it out of range or is the battery too low?
Checking pump status (suspended/bolusing): mmeowlink.exceptions.CommsException: No acknowledgement from pump on wakeup. Is it out of range or is the battery too low?
grep: monitor/status.json: No such file or directory
Error, retrying
Sleeping 10;
just repeats that
wait maybe I spoke too soon, cause the problem is still happening
Jorge Aguilar
@jonahtaxt
Jul 04 2017 01:04
I sense a looong night
Which are the aliases that upload retrieved bg data to ns?
John Benjamin
@Bender1061
Jul 04 2017 01:07
one the first rig I reinstalled, I ended up running the ccprog. and it worked after that. Though I don't know why that would be needed.
cameronrenwick
@cameronrenwick
Jul 04 2017 01:22
having seen various approaches to SMB and how ppl are entering carbs, what in your opinion is the best way? I routinely IFTTT eating soon and then bolus/carb via bolus wizard. Should I be just entering carbs via care portal or maybe make up some more IFTTT buttons just for carbs and let smb do the bolus? Kinda wondering ppl's thoughts
Dana Lewis
@danamlewis
Jul 04 2017 01:23
YDMV/it depends on the type of meal. Some people have found eating soon + entering carbs + letting SMB bolus works well for them. Others do ES + enter carbs + an initial partial bolus, and let SMB pick up rest (I often do this). The exact split/best approach will require testing person to person depending on the type of meals you eat, how your body reacts, and if you’re throwing anything into the mix like walking to/from the meal or exercising shortly after.
also depends on whether you’re optimizing for less work (i.e. do one thing and let the system do the rest) or flatter post-meal curves <—that optimization will also drive which strategy you might choose
John Benjamin
@Bender1061
Jul 04 2017 01:24
:+1: on the above
cameronrenwick
@cameronrenwick
Jul 04 2017 01:27
as I personally like to take conservative approach to this, I think I'll continue to do the eating soon, BW with carbs and bolus (but perhaps roll the bolus back a bit) and just see how that works. I like the idea of IFTTT buttons although I'd need a lot of them as I tend to count carbs in 1's and 2's not 10's and 15's sort of thing. Thanks for the suggestions tho...I'll see how she goes
+1
tynbendad
@tynbendad
Jul 04 2017 01:30
given that my son just ate 80 carbs in five minutes, and doesn't pre-bolus, i think i'll keep having him bolus the entire amount on pump ;)
cameronrenwick
@cameronrenwick
Jul 04 2017 01:35
this question is particularly important to me as I'm on fiasp and am going to be super attentive to what's going on. My hunch is that things should be pretty good - maybe a few bumps in the road getting there - but I just want to know what's best practice so thanks everyone for sharing. I've altered my DIA and have been using fiasp on my looping pump for a few weeks so this is just another add on... I'll keep y'all posted
David Rimmer
@zjedr
Jul 04 2017 01:42
I wired a Qi charger coil onto my rig last weekend and was using it to charge my rig overnight. I noticed that I had several periods where I dropped back to default basal rates (no temp running). Last night I went back to usb charging and had no drop outs.
Anyone else had issues with Qi chargers interfering with comms?
jmcrawford
@jmcrawford
Jul 04 2017 01:46
The drops could be due to overheating, the Qi devices generate a fair bit of heat.
Dana Lewis
@danamlewis
Jul 04 2017 01:55
@cameronrenwick :+1:
@zjedr not sure about comms, but know @tim2000s & others had other problems (heat I think?) with Qi and most everyone gave up on them who had tried them out
John Benjamin
@Bender1061
Jul 04 2017 01:57
@zjedr are you using the explorer board for the Qi charger? if so where are you soldering it onto?
David Rimmer
@zjedr
Jul 04 2017 02:04
I'm using a 500mAh adafruit micro lipo charger and 4000mah battery (+ explorrer board). I've also noticed that it interferes with the digital radio signal on my clock radio so I'm not confident, I think Qi will be off my list as well.
David Rimmer
@zjedr
Jul 04 2017 02:10
Its mid winter here now, if I get heat problems now I hate to think what it will be like mid summer
Dana Lewis
@danamlewis
Jul 04 2017 02:18
we’ve got a potential fix for the decay-carbs-while-no-CGM-data issue - openaps/oref0#518 needs testing if anyone has experienced that and has a sensor restart upcoming.
tynbendad
@tynbendad
Jul 04 2017 02:45
@danamlewis it looks like that is in a branch openaps:interpolate-bg-for-cob, which is pretty even with dev
but, i'm still git illiterate - how do i switch my rig from dev to that?
we just started a cgm blackout a few minutes ago...
Dana Lewis
@danamlewis
Jul 04 2017 02:46
even except for those small tweaks, ya. just that fix on that branch, which forked from dev this am. same process you did for the other branch except use interpolate-bg-for-cob as the branch name instead of other-branch-name-whatever-it-was-called
tynbendad
@tynbendad
Jul 04 2017 02:48
cd ~/src/oref0 && git fetch && git checkout ignore-carb-corrections && git pull
found it
thx
i'm not sure what to look for - do we need to bolus for carbs during the cgm blackout period or is cob enough?
Dana Lewis
@danamlewis
Jul 04 2017 02:51
It's more relevant if insulin dosed because if not then minimum will kick in. But bolus or SMB doesn't matter
But - point is for cob during blackout and it to be decayed properly when CGM comes back
tynbendad
@tynbendad
Jul 04 2017 02:54
ok, i've got it on rig in use currently, will keep an eye and hope i can tell if it helped ;)
Dana Lewis
@danamlewis
Jul 04 2017 02:57
:+1:
Your son doesn't want to eat more right now? For science? ;)
tynbendad
@tynbendad
Jul 04 2017 02:59
lol, i'm scared to ask him - maybe he will anyway... he's already had 200+ carbs. here is last smb before blackout: "COB: 40, Dev: 79, BGI: -15.41, ISF: 66, Target: 89, minPredBG 131, IOBpredBG 53, COBpredBG 136, UAMpredBG 256; insulinReq 0.53; setting 90m zero temp. Microbolusing 0.2U. ","units":0.2,"IOB":2.643}
you don't have a software fix for dexcom g5 blackouts do you? :)
Dana Lewis
@danamlewis
Jul 04 2017 03:03
Get a G4 and run two receivers? ;)
tynbendad
@tynbendad
Jul 04 2017 03:04
lol
xdrip probably more likely
Dana Lewis
@danamlewis
Jul 04 2017 03:04
#nothelpfulIknow
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:05
lol
tynbendad
@tynbendad
Jul 04 2017 03:07
would watching mealCOB in meal.json help tell us anything?
Scott Leibrand
@scottleibrand
Jul 04 2017 03:09
If you feel like it you can switch back and forth between dev and the interpolate branch to see he difference in oref0-meal output in meal.json
That will tell you for sure if it's helping.
tynbendad
@tynbendad
Jul 04 2017 03:13
i have two rigs, one is on dev and one on interpolate... lemme check
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:14
don't know if I mentioned it, but love the ascii art.
       .-. .,-.  .-. .--. .-.  .,-. .--.
      (   )|   )(.-' |  |(   ) |   )`--.
       `-' |`-'  `--''  `-`-'`-|`-' `--'
           |                   |
           '                   '
:)
Dana Lewis
@danamlewis
Jul 04 2017 03:15
:smile:
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:16
tell Scott we recently dropped $company_name in ascii like this on every machines motd. Ascii art is making a comback!
Cameron Chunn
@applehat
Jul 04 2017 03:16
My rigs all have MOTD mods so I can remember which I'm on (live or dev)
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:17
awesome.
Cameron Chunn
@applehat
Jul 04 2017 03:17
I tend to put Ascii Art text in my code sometimes too, if its a single file with a few different logical pieces. Makes seeing it in the Minimap (a feature that VS Code / Atom / Sublime all support) easy!
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:18
I'm just dipping into atom from vi. I'll have to check that out.
Dana Lewis
@danamlewis
Jul 04 2017 03:19
I use those files to remember which rig is in which case
And to correct a certain someone for misnaming the cases ;)
Cameron Chunn
@applehat
Jul 04 2017 03:19
I liked Atom a lot, but moved to VS Code recently. Seems slightly faster on my dated machines.
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:20
interesting
@danamlewis "a certain someone" lol
Cameron Chunn
@applehat
Jul 04 2017 03:21
I read "misnaming" as "misspelling" and thought I was being talked about for half a second there.
I didn't learn how to spell good in my alabama schoolin
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:22
I think all our ears would have poped up if she had said misspelling!
@danamlewis are you running oref1 these days?
Dana Lewis
@danamlewis
Jul 04 2017 03:24
Yes
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:25
so very cool.
How involved is your day to day interaction with NS/rig?
or rigs
Cameron Chunn
@applehat
Jul 04 2017 03:26
Im also on oref1, if you want multiple sources of info =)
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:26
totally do, thanks Cameron. Same question as above
please
Cameron Chunn
@applehat
Jul 04 2017 03:26
Are you asking how often I have to interact with Nightscout or my rig?
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:26
yes
Cameron Chunn
@applehat
Jul 04 2017 03:27
So first, just so you're aware of my setup (this is sort of important): Medtronic Pump, Dexcom Sensors using xDrip (and xDripAPS on rig + NS upload), Wifi at home, Bluetooth Tether while out.
At home i hardly pay any attention at all
I have a tablet or two around the house that show Nightscout
When I eat at home, its generally low carb (or at least not HIGH carb) - I just use my "bolus wizard" on the pump
95% of the time, doing this, SMB will take up any slack that I miss, i'll stay in range, and everything is gravy.
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:29
wow
Cameron Chunn
@applehat
Jul 04 2017 03:29
Now, I pay closer attention while im not at home
But this is mainly because my phone is 3 years old and sometimes likes to be a jerk about BT tethering
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:30
I get the same on my 7+ with edison.
very interesting.
Cameron Chunn
@applehat
Jul 04 2017 03:30
And if BT tethering is down while im mobile, both my xDrip => xDripAPS AND my Upload to NS (this is the backup, since openaps will try pulling from there) are down - as well as the rigs internet connection.
So I do give periodic checks (especially if im eating out and consuming way too many carbs)
I actually found that I never loose my BT-Pan network tho - I only lose internet on the rig
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:31
that's still much less than my involvement now. Same bolusing for meals method, just the wizard on the pump?
thats interesting.
Cameron Chunn
@applehat
Jul 04 2017 03:32
If I know that Im going to be eating something with more then say 30 carbs, I'll click my "eating soon" button on my phone.
so I can low-temp to 80
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:32
android button?
Cameron Chunn
@applehat
Jul 04 2017 03:32
starting at 80 means the rise will be more likely to stay in range.
IFTTT button, but yes, Im using android.
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:33
right was just reading Dana's article on that.
Cameron Chunn
@applehat
Jul 04 2017 03:33
an alternative to that (and something that I need to try still) is setting up the hot-button app
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:33
that's a great idea
Cameron Chunn
@applehat
Jul 04 2017 03:33
IFTTT has the benefit of working over the internet, but thats also a a downfall if your connection is spotty
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:34
so within the hour before eating you hit the button and things are good.
Cameron Chunn
@applehat
Jul 04 2017 03:34
Yep. Sometimes I forget to do that too, gotta make a habbit
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:34
are you mixing oref0 & 1 at all?
Cameron Chunn
@applehat
Jul 04 2017 03:35
oref1 is just the stuff in "dev"
that deviates from the original oref0 spec
(because it issues bolus)
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:35
I have 2 rigs currently for home coverage and am making a 3rd for work / carry on.
Cameron Chunn
@applehat
Jul 04 2017 03:35
oref0 is still the core spec
Wanna see my "i may have eating 150g carbs at dinner" screen shot? lol
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:36
lol
sure!
Cameron Chunn
@applehat
Jul 04 2017 03:37
Screenshot from 2017-07-03 22-35-11.png
So i ate out, and ordered something that I couldn't figure out if it had 100g or 160g from the info I found, so I went "safe" and bolus'd for 100g.
Spiked to 191. Didn't even break 200.
I think thats pretty dang good for that much "bad diabetic"
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:38
yeah thats wild.
Scott Leibrand
@scottleibrand
Jul 04 2017 03:38
grumble grumble stigma grumble
I know, scare quotes and all, but still
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:38
I dunno, 160 gram breakfast sounds amazing.
:)
Cameron Chunn
@applehat
Jul 04 2017 03:39
Haha, maybe OpenAPS can change that stigma @scottleibrand ;)
Scott Leibrand
@scottleibrand
Jul 04 2017 03:39
yeah, that rocks that you can do that and have a reasonable outcome. :)
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:39
@applehat have you run into any issues being on the dev branch?
Cameron Chunn
@applehat
Jul 04 2017 03:40
I feel like its pretty reasonable, for sure. Not something I wanna do every meal, but considering i've tested some friends BG after lots of carbs and seen them at 180mg/dL I figure 191 for me is pretty :100:
Dana Lewis
@danamlewis
Jul 04 2017 03:41
Keep in mind dev is living breathing thing. Lots of improvements and changes ongoing, even last 24 hr (especially). So keep that in mind if you're trying to compare dev experiences
Cameron Chunn
@applehat
Jul 04 2017 03:41
@kimstdennis_twitter not any that I can recall, no. @scottleibrand is pretty good about not merging in bad code =)
But yeah, head @danamlewis's warning for sure
Its important to have a decent grasp of how to track down problems
Scott Leibrand
@scottleibrand
Jul 04 2017 03:42
well, there was that tokenauth regression a few days ago...
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:42
yeah, I've held back for that. I can only dive into this every few weeks, so I'd be behind on updates.
Prosulpump
@Prosulpump
Jul 04 2017 03:42
This message was deleted
Cameron Chunn
@applehat
Jul 04 2017 03:42
I actually take that back, I've run into 2 separate issues where I've caused race conditions by letting my phone die and loosing BG data, but updating dev and reinitialized have fixed them both times.
Prosulpump
@Prosulpump
Jul 04 2017 03:42
This message was deleted
Scott Leibrand
@scottleibrand
Jul 04 2017 03:43
@Prosulpump this channel is not appropriate for this discussion.
Cameron Chunn
@applehat
Jul 04 2017 03:43
I actually take that back, I've run into 2 separate issues where I've caused race conditions by letting my phone die and loosing BG data, but updating dev and reinitialized have fixed them both times.
Scott Leibrand
@scottleibrand
Jul 04 2017 03:43
@Prosulpump has now been banned from this channel.
Cameron Chunn
@applehat
Jul 04 2017 03:43
:hammer:
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:44
:thumbsup:
just pulling down dev updates and re init doesn't seem bad though.
Cameron Chunn
@applehat
Jul 04 2017 03:45
not at all
tynbendad
@tynbendad
Jul 04 2017 03:46
@scottleibrand :+1:
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:51
I just passed my year anniversary for looping, so this is a great next step. I want to run autotune for a while as recommended, but I think I'll do that off the dev branch.
Dana Lewis
@danamlewis
Jul 04 2017 03:51
:+1:
Happy loopaversary too :smile:
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:52
thanks! Last year I was hanging out with Ali (and our kids) at Diabetes camp and then we had a openaps meetup in orange county.
Cameron Chunn
@applehat
Jul 04 2017 03:53
I feel like such a newbie =P
I dont even know my (n=1)* number
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:55
pssh. I feel like each time I build a new rig, I learn looping all over again. so consider me a n00b. :smile:
Cameron Chunn
@applehat
Jul 04 2017 03:57
I still only have the 1 that I keep in my pocket. I eventually want to get whole-house coverage going. Especially if we can get on-rig G5 sensor readings. Being able to leave my phone on my charger in my bedroom while at home would be nice. =P
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:57
cool. Thanks for all the fantastic info on your experience with the oref revision @applehat .
Dana Lewis
@danamlewis
Jul 04 2017 03:57
@applehat you're part of the uncountable in exact order group because the community got so big ;)
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 03:59
I had considered that but figured if I ran multiple rigs with different calibrations that would be an issue eventually.
hmm, pick up some used g5 receivers and go cable -> rig?
you can run a bunch of those, then leave your phone on the changer.
tynbendad
@tynbendad
Jul 04 2017 04:02
can you really have >1 g5 receiver connecting at a time?
Cameron Chunn
@applehat
Jul 04 2017 04:02
I don't use the G5 receiver
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:02
the g5 receiver & iphone app sync at some point. I would think a few g5 receivers would do the same.
Cameron Chunn
@applehat
Jul 04 2017 04:02
So - actually, no.
tynbendad
@tynbendad
Jul 04 2017 04:02
for some reason i thought it was 1 receiver + 1 phone only
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:02
oh that might be the case.
Cameron Chunn
@applehat
Jul 04 2017 04:02
The receiever (or xDrip) physically "bond"
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:02
I can test.
Cameron Chunn
@applehat
Jul 04 2017 04:03
The iPhone listens in "passive" mode.
(unless the receiver is not around)
its a really weird, dumb implamentation
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:03
curious.
yeah
Cameron Chunn
@applehat
Jul 04 2017 04:04
If I have multiple rigs, all doing forced deauth + rebond, then whichever is closest should just "win" and get the signal tho
There was also talk of g5 sensors potentially still working on the old G4 signaling (tho I wont believe that till I see it) - if thats the case, some Wixel's would solve the issue.
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:05
I haven't gone the xdrip route yet. that sounds interesting.
wixels might. I haven't checked the compatibility in a while.
Cameron Chunn
@applehat
Jul 04 2017 04:05
xDrip saved me so much headache
Im all about making diabetes as easy as possible, an xDrip is soooooo nice.
No restarting sensors every 7 days.
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:06
I have a g4 with share and g5 receiver. I can test 2 g5s + g4 as soon as they have a charge.
Cameron Chunn
@applehat
Jul 04 2017 04:06
Auto calibrations based off Bolus Wizard (or bluetooth meter)
and xDrip lets you replace g5 batteries too!
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:07
seriously?
Cameron Chunn
@applehat
Jul 04 2017 04:07
and xDrip lets you replace g5 batteries too!
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:07
that's interesting.
Cameron Chunn
@applehat
Jul 04 2017 04:08
Yeah. It can bypass all the "safety" / "forced expire" stuff in dexcom transmitters
So even if a dexcom transmitter has reached its max allowed days, xdrip will keep on using it as long as the battery voltage is good.
I still have 1 fresh g5 transmitter, but as soon as that one is in active use, I plan to replace the batteries in the first one.
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:09
wild.
Cameron Chunn
@applehat
Jul 04 2017 04:10
Same story with the G4 transmitter fwiw
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:12
hmm first rx only allows 6/7 serial numbers.
Cameron Chunn
@applehat
Jul 04 2017 04:13
Someone else was going to test reading a G5 sensor with a G4 and had the same issue
We'd need someone with a virgin, unpaired G5 and a wixel to verify the claim.
tepidjuice
@tepidjuice
Jul 04 2017 04:14
Does anyone have any power saving tips for an Edison rig? It's permanentally on a bluetooth tether to my phone, using xdripaps, I set temp targets via SSH and only needs to upload to nightscout occasionally.
Cameron Chunn
@applehat
Jul 04 2017 04:14
Like I said, I don't believe it in the slightest, but I also don't want to potentially miss a huge easy-of-use fix for the community if it is the cae.
@tepidjuice how long is your battery lasting?
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:14
yeah I think the g4 receiver kills that option.
Cameron Chunn
@applehat
Jul 04 2017 04:15
Maybe I haven't put mine to the test hard enough yet, but I haven't had my rig die on me yet and I only have a 2000mAh battery
Maybe I haven't put mine to the test hard enough yet, but I haven't had my rig die on me yet and I only have a 2000mAh battery
tepidjuice
@tepidjuice
Jul 04 2017 04:16
Not long enough. I'd estimate about 8 hours on a 2000mah battery.
Cameron Chunn
@applehat
Jul 04 2017 04:16
You could always go the @rsilvers129 route and slap 2x 2500mAh batteries on your rig
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:17
tepidjuice
@tepidjuice
Jul 04 2017 04:17
Size is
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:18
if capacity is the goal. 6600mAh :)
tepidjuice
@tepidjuice
Jul 04 2017 04:18
a problem. I've read people get 14hours on. 2000mah
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:19
I think I got 8-9 today on 2000mAh on the new rig.
it was with an explorer + edison
tynbendad
@tynbendad
Jul 04 2017 04:19
we get ~16 hours with 2500mah
tepidjuice
@tepidjuice
Jul 04 2017 04:20
Ok. Are there some unnecessary Cron jobs I could get rid of?
tynbendad
@tynbendad
Jul 04 2017 04:20
image.png
does anyone else find the 4 prediction lines spooky looking?
Cameron Chunn
@applehat
Jul 04 2017 04:20
im of the mindset that as long as it doesn't get any bigger then my cell phone, i'm doing alright on size - Im also an adult tho and realize that for kids it may be a bigger concern.
tepidjuice
@tepidjuice
Jul 04 2017 04:21
I'm an adult who doesn't like things in their pockets :)
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:21
I guess it would be radio use / bt scanning that would get the biggest reduction in battery consumption.
Cameron Chunn
@applehat
Jul 04 2017 04:22
@tepidjuice snag one of @rsilvers129's belt loop cases =P
Dana Lewis
@danamlewis
Jul 04 2017 04:22
@tepidjuice there’s certainly some efficiencies that could be done w/ NS uploads - no one’s just wanted to do them
@tynbendad spooky? tell me more about what you mean?
tynbendad
@tynbendad
Jul 04 2017 04:22
just looks a bit scary, like tentacles ;)
Cameron Chunn
@applehat
Jul 04 2017 04:22
lol
tepidjuice
@tepidjuice
Jul 04 2017 04:22
So if I turn off the WiFi attenna as I only really use Bluetooth?
Dana Lewis
@danamlewis
Jul 04 2017 04:22
how can it be scary when they’re PURPLE tentacles? ;) :)
Cameron Chunn
@applehat
Jul 04 2017 04:22
@tepidjuice yeah, wifi eats more then bluetooth
tynbendad
@tynbendad
Jul 04 2017 04:23
lol
:worried:
tepidjuice
@tepidjuice
Jul 04 2017 04:23
And maybe switch to NS uploads every 30mins as I only really need NS except for data visualis.
ation
Cameron Chunn
@applehat
Jul 04 2017 04:24
@tepidjuice yeah, no harm in that if you're not relying on NS for your loop
tynbendad
@tynbendad
Jul 04 2017 04:26
that might only save you battery if you disconnect the bt/wifi for the time you don't need it.
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:26
@tynbendad that was the first thing I thought when I saw the NS lines.
tynbendad
@tynbendad
Jul 04 2017 04:27
i think the crontab has the loop running all the time, or every minute... you could probably make it run a bit less often.
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:27
you should be able to shutdown the wifi interface if on bt teather.
that might save you some battery life.
Cameron Chunn
@applehat
Jul 04 2017 04:28
iwconfig wlan0 down
or ifconfig wlan0 down
i forget which one
probably need to modify oref0-online to not bring it back up tho
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:28
ah
yeah
tynbendad
@tynbendad
Jul 04 2017 04:29
maybe just remove the wpa_supplicant line
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:29
I think you would want to test and make sure you're connected too.
Cameron Chunn
@applehat
Jul 04 2017 04:30
plug your rig to the phone and disable all the radios and USB tether. This of course will try and charge the phone, tho, so maybe dont do this.
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:30
:D
zombie-carbs :laughing:
tynbendad
@tynbendad
Jul 04 2017 04:42
Dana didn't realize they were sentinels
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:43
ha!
Cameron Chunn
@applehat
Jul 04 2017 04:43
You guys obviously haven't read the loop code.
Kim St. Dennis
@kimstdennis_twitter
Jul 04 2017 04:43
@applehat yeah you're right, the second g5 receiver just has the BT sign blinking. it's been 20min now or so.
tynbendad
@tynbendad
Jul 04 2017 04:43
lol
Cameron Chunn
@applehat
Jul 04 2017 04:52
Yeah. I know the loopKit G5 code has a "passive listen" mode thats designed to sort of just sniff for the receiver talking and grab the reading when it floats through the air.
Martin Haeberli
@mhaeberli
Jul 04 2017 05:02
ifconfig wlan0 down
@danamlewis btw, what is total openaps and Loop count like now?
^^ everybody can bookmark/remember to look there ;)
Martin Haeberli
@mhaeberli
Jul 04 2017 05:04
k thx
Cameron Chunn
@applehat
Jul 04 2017 05:06
I want my name on a list so i can feel special and feed my narsasism tho....
Cameron Chunn
@applehat
Jul 04 2017 05:07
(i was joking on myself btw - wasn't trying to take a shot at anyone else - realized it sounded sarcastic after I posted it)
Dana Lewis
@danamlewis
Jul 04 2017 05:08
I have links for (almost) everything ;)
Cameron Chunn
@applehat
Jul 04 2017 05:09
Do you remember those "It's in the box" commercials? You can be "It's in the docs"
Dana Lewis
@danamlewis
Jul 04 2017 05:10
::snorts::
gregtcm
@gregtcm
Jul 04 2017 05:10
@danamlewis Amazing how many have started looping since I did my first PR to that doc! :)
Dana Lewis
@danamlewis
Jul 04 2017 05:11
:+1: :clap:
Matthias
@ceben80
Jul 04 2017 05:12
My forecast by the end of the year, 400-500 :star2:
Dana Lewis
@danamlewis
Jul 04 2017 05:13
remember, that’s only who’s filled out the form. lots don’t fill out the form.
David Rimmer
@zjedr
Jul 04 2017 05:17
for full time BT tether I just have
@reboot iwconfig wlan0 txpower off
in cron
Can restart via hot button or juicessh if needed
tynbendad
@tynbendad
Jul 04 2017 05:18
hmm, at the end of the 2 hour blackout, cob was 0 on both dev & interpolate-bg-for-cob branch rigs
it was nearly DIA hours after the last carbs tho
sorry, probably not a great test
Dana Lewis
@danamlewis
Jul 04 2017 05:18
@tynbendad given the small amount left over, that seems about right for 2 hours
no worries
If anyone has a new rig, you can test the new “super duper easy install” (someone else’s name for what we call wifi-bootstrap) process. It helps you find your wifi and installs master right now, but you can switch to dev after that if you want dev on that rig. Dev docs have this now, but note it’s just the one page you’ll test from: http://openaps.readthedocs.io/en/dev/docs/Build%20Your%20Rig/OpenAPS-install.html#installing-openaps-on-your-rig-with-an-already-flashed-edison (everything else in dev docs is not done yet, so don’t be surprised if you click around and spot an empty page or two)
gregtcm
@gregtcm
Jul 04 2017 05:25
@danamlewis I fixed my issue by adding a sleep 10 before the line
echo -ne "\nWifi SSID: "; iwgetid -r
Dana Lewis
@danamlewis
Jul 04 2017 05:26
@gregtcm will you comment on https://gist.github.com/danamlewis/ea2dc03304bd516895cb5e4600243524 with the note about what you saw & that solution, so we can leverage it if anyone else runs into the same snag? Thanks!
gregtcm
@gregtcm
Jul 04 2017 05:26
:+1:
Dana Lewis
@danamlewis
Jul 04 2017 05:33
@gregtcm thanks!
@thanksegon per Q about NS - not sure what you’re expecting to see in NS? Sounds like you might have some regular setup issues, so you might want to post here re “My NS site doesn’t show these yet"
you may have just missed a few things to be enabled, or need to toggle on the OpenAPS pill, etc.
gregtcm
@gregtcm
Jul 04 2017 05:58
@danamlewis You're welcome
Dirk Gastaldo
@dirkgastaldo
Jul 04 2017 07:44
Good evening. I'm working on BT tethering. I drove away from my WiFi router and the rig connected to my iPhone's hotspot. I've been home over 30 minutes and I still see a blue bar on my phone with "1 connection". I'm not used to having my personal hotspot on, so maybe this is normal. According to my Papertrail log my rig is connected to WiFi (192.168.1.6). Is is normal to see a blue bar with "1 connection" on my phone when I leave hotspot on?
image.png
Dirk Gastaldo
@dirkgastaldo
Jul 04 2017 07:53
doh, I think it was my computer. I plugged my iPhone into my computer. Not sure what would have connected to my hotspot through that USB connection, but I guess my computer? When I unplugged the USB cable from my phone the blue bar went away.
Joey
@Cooljoe04_twitter
Jul 04 2017 08:03
Likely. Your iPhone will accept USB clients to its hotspot connection, so your PC probably detected it as a USB cell modem.
Dirk Gastaldo
@dirkgastaldo
Jul 04 2017 08:48
@Cooljoe04_twitter Thanks, Joey, I think you're exactly right.
Luuc
@luucvp_twitter
Jul 04 2017 09:37
MDT cgm data retrieveError, retrying this happened to me a few times the last couple of days, it usually cures itself after a while. Battery levels are good, reboot doesn't have any effect. Do you guys have an idea what this could be?
It almost looks like logging into the system via SSH and then checking out the logs cures it haha
peterfulvi
@peterfulvi
Jul 04 2017 10:39
Ive been trouble shooting this issue for 5 days ever since i started offline looping with the G5. @tim2000s said it was corrupt BG entries due to a doprout in BT sir wifi signal. I looked in the xdrip/glucose.json and found nothing unusual int the entries. The thing that gets me is this The last BG data was read at Tue Jul 04 2017 06:47:12 GMT-0400 (EDT) but your system time currently is Tue Jul 04 2017 06:31:13 GMT-0400 (EDT). I'll post the whole output again. Like i said before everything works, but sometimes the loop will stop due to this problem with the time
Starting supermicrobolus pump-loop at Tue Jul 4 06:28:59 EDT 2017 with 9 second wait_for_silence:
Waiting up to 4 minutes for new BG: glucose.json newer than pump_loop_completed
Radio ok. Listening: .No pump comms detected from other rigs
Preflight OK. Old pumphistory, waiting for 9 seconds of silence: Radio ok. Listening: .No pump comms detected from other rigs
Refreshed pumphistory
Profile less than 60m old. glucose.json newer than pump_loop_completed.
Temp refreshed
Autotune exists! Hoorah! You can use microbolus-related features.
{"carbs":0,"boluses":0,"mealCOB":0,"currentDeviation":-4.31,"maxDeviation":0,"minDeviationSlope":0}
BG data is too old (it's probably this), or clock set incorrectly.  The last BG data was read at Tue Jul 04 2017 06:47:12 GMT-0400 (EDT) but your system time currently is Tue Jul 04 2017 06:30:13 GMT-0400 (EDT)
  "reason": "BG data is too old (it's probably this), or clock set incorrectly.  The last BG data was read at Tue Jul 04 2017 06:47:12 GMT-0400 (EDT) but your system time currently is Tue Jul 04 2017 06:30:13 GMT-0400 (EDT)"
Checking system clock against pump clock:
sudo: unable to resolve host petersrig2
Checking deliverAt: null is within 1m of current time: Tue Jul  4 06:30:27 EDT 2017
date: invalid date ‘null’
Checking pump status (suspended/bolusing): {"status":"normal","bolusing":false,"suspended":false}
Error, retrying
Sleeping 7;
Another part im wondering about is the error at the end Checking pump status (suspended/bolusing): {"status":"normal","bolusing":false,"suspended":false} Error, retrying Sleeping 7;
cameronrenwick
@cameronrenwick
Jul 04 2017 10:50
@peterfulvi I get the same error at the end you report too... trying to get to the bottom of it now
if you figure it out please pm/post
peterfulvi
@peterfulvi
Jul 04 2017 11:43
@cameronrenwick same here Thanks
Iain
@iainct
Jul 04 2017 12:47
I'm not sure if this is the best place to ask but. My xDrip android app just stopped uploading bgs to entries, for no apparent reason. I had upgraded my xdrip app earlier to the latest nightly so i have downgraded it to what i was using before but still only able to upload locally to my rig and not via nightscout.... thanks
Tim Street
@tim2000s
Jul 04 2017 12:51
@peterfulvi what's your clock time on your xDrip+ device?
Iain
@iainct
Jul 04 2017 12:57
android phone? it is the correct local time
it is uploading to the local ip address with no worries but not directly to nightscout how usually use it....
Iain
@iainct
Jul 04 2017 13:05
I need to go to sleep but i appreciate any tips thanks.
John Benjamin
@Bender1061
Jul 04 2017 13:31
@peterfulvi I get that same error a lot with MDT.
John Benjamin
@Bender1061
Jul 04 2017 13:45
@luucvp_twitter That error is caused sometimes when the rig is attempting to get the CGM data from the pump openaps report invoke monitor/cgm-mm-glucosedirty.json sometimes errors out and gets a decocare.errors.DataTransferCorruptionError: Page size too short My guess would be an issue with decocare. we recently pushed a work around for that in dev, when when it happens it will retry, and that almost always works. well at least for me. I think this has been happening for a while, but since it is just being run in an Alias we did not see it happening other than what you are seeing. we got rid of all the alias and have it all running in pump-loop.sh now so the problem started showing itself.
peterfulvi
@peterfulvi
Jul 04 2017 14:02
@tim2000s @cameronrenwick @Bender1061 . My loop for now does not show that error. My phone was on venezuela time zone which was thye right time. Im Eastern Standard Time. So i switched it to that and it was still an hour off so i manually adjusted it. But xdrip seems to be 3 minures ahead in the apploicatgion
Scott Leibrand
@scottleibrand
Jul 04 2017 14:04
Luuc: are you on the latest dev? There were some updates to the MDT CGM logic this week.
@luucvp_twitter (what @Bender1061 said)
Tim Street
@tim2000s
Jul 04 2017 14:17
@peterfulvi I'd just kill the phone and restart it.
John Benjamin
@Bender1061
Jul 04 2017 14:33
Killing it seams a bit excessive, not like it's a decepticon
stefanakerblom
@stefanakerblom
Jul 04 2017 14:37
@danamlewis Some maths. If 400 apser help 3 T1Ds every 3 months to start looping, when will there be 10 million?
John Benjamin
@Bender1061
Jul 04 2017 14:38
probably not that many working Loopable pumps
stefanakerblom
@stefanakerblom
Jul 04 2017 14:43
@Bender1061 You are right, but if there were? This was a math quiz! Lucky me to have gotten hold on a 722 for 400 bucks. Will switch to AAPS and Dana as soon as my insurance will pay for the Dana (in a year or so) - until then there might there might even be something better around. Crossing my fingers that my 12 year old 722 will last until then.
Dana Lewis
@danamlewis
Jul 04 2017 14:44
Dana*R pump ;)
Insurance does not cover me ;)
LadyViktoria
@LadyViktoria
Jul 04 2017 14:45
:-D
stefanakerblom
@stefanakerblom
Jul 04 2017 14:52
@danamlewis I believe very well that this group would cover you :)
Dana Lewis
@danamlewis
Jul 04 2017 14:54
;) :smile:
stefanakerblom
@stefanakerblom
Jul 04 2017 14:59
grafik.png
Why does this one say 0.9.2 and
grafik.png
says its up to date. I thought 0.9.3 ist the most recent version?
Scott Leibrand
@scottleibrand
Jul 04 2017 15:08
The automatic update checker is probably checking master. You need dev for OpenAPS.
stefanakerblom
@stefanakerblom
Jul 04 2017 15:14
and wher do I get that one?
PieterGit
@PieterGit
Jul 04 2017 15:22
@danamlewis @scottleibrand : openaps/oref0#520 does not only fix color and spacing, but also disables token based authentication by default. I would say that from safety perspective secure by default , https://en.wikipedia.org/wiki/Secure_by_default should be a guiding principle. Is there a reason why token based authentication is not yes by default? BTW the docs in openaps/docs#862 can also be merged to dev since token based has been merged to oref0 dev.
Dana Lewis
@danamlewis
Jul 04 2017 15:24
Until we do more testing and can make sure it more widely works, I think it should be off by default
PieterGit
@PieterGit
Jul 04 2017 15:27
you'll get more testing if it's on by default :smile: .
stefanakerblom
@stefanakerblom
Jul 04 2017 15:41
grafik.png
Well. I am still a novice. I need to merge the update but someone needs to approve and that fails
Dana Lewis
@danamlewis
Jul 04 2017 15:42
Try swapping. You should be able to merge from them to you without perms. You might have it backward
stefanakerblom
@stefanakerblom
Jul 04 2017 15:43
dont understand.......
Dana Lewis
@danamlewis
Jul 04 2017 15:48
Base fork should be your fork. That's what you're merging to. Head should be where you're pulling from.
There's a button to easily swap those and see if that clears your ability to merge yourself
Geoff Aumaugher
@geoffro10
Jul 04 2017 16:29
@danamlewis I'm trying to test the new "super duper easy install" process. I reflashed my rig with Jubilinux but that's about as far as I got. The docs say to login with root and no password. Well, the Jubilinux flash automatically sets the pw to edison as default. Minor detail... I logged in, copied and pasted the commands, and it didn't do anything after typing in my WiFi info. Thoughts?
DaisyHK
@laurathiessen
Jul 04 2017 16:30
when you are setting up shirt cuts. in nano do you only input the entire line "alias autotune="tail -n 100 -F /var/log/openaps/autotune.log"" or just the red part "tail -n 100 -F /var/log/openaps/autotune.log" ?
Dana Lewis
@danamlewis
Jul 04 2017 16:30
Red part
Sorry
Full line
DaisyHK
@laurathiessen
Jul 04 2017 16:30
k thanks!
Dana Lewis
@danamlewis
Jul 04 2017 16:30
Just red part of you want to run that command manually
@geoffro10 say more about what it looks like when "it didn't do anything"?
Geoff Aumaugher
@geoffro10
Jul 04 2017 16:33
Updates the WiFi configuration and the goes to command prompt.
Dana Lewis
@danamlewis
Jul 04 2017 16:34
Doesn't continue to the script?
Geoff Aumaugher
@geoffro10
Jul 04 2017 16:36
Nope... actually, it appears I'm not connected to the WiFi.
Dana Lewis
@danamlewis
Jul 04 2017 16:37
Ok. Maybe try again?
And if it doesn't work
Can you pm a screenshot of what that looks like so we can see if any clues for fixing? On a plane so may have to look at it later. One other thing is to see if it looks like what Greg saw and try his solution if so: https://gist.github.com/danamlewis/ea2dc03304bd516895cb5e4600243524#gistcomment-2139111
DaisyHK
@laurathiessen
Jul 04 2017 16:38
image.png
@danamlewis i believed i follow the directions of copying and pasting red parts into nano but the commands did not stay?
Dana Lewis
@danamlewis
Jul 04 2017 16:40
@laurathiessen sorry on plane so can't view images well and answer Qs as fast. Remember to control x and say yes to save when you exit the file.
DaisyHK
@laurathiessen
Jul 04 2017 16:48
oh ok I will wait for someone else to check, as I believe I did control x and y. However i exit the file and try to edit the wifi with vi /etc/network/interfaces now this comes up??
image.png
so now I am really confused lol
jmcrawford
@jmcrawford
Jul 04 2017 17:02
Type e for edit. This message means you have edited this file before, but the editing session didn't close cleanly.
Scott Leibrand
@scottleibrand
Jul 04 2017 17:48
@PieterGit to be pedantic, setting NS to denied doesn't increase security, it just increases privacy, at the expense of ease of use and interoperability (with things like the data commons). I'm fine making that an option, but I think leaving read-only access open by default is better. That said, we can have OpenAPS rigs use token auth even without denying read-only access. So once we get all the docs reviewed and merged, I would be fine making token auth the default again for the next dev version after we release 0.5.0 to master. I just haven't tested any of the token auth functionality yet, so didn't want to risk breaking things for people who don't need/want it as a default, especially just when we're trying to get a release out (of lots of things many more people do want).
I do get that there are some security benefits of restricting permissions per token rather than using the API_SECRET for everything, and agree we should move toward that. I just don't want to move too fast and break things for anyone who hasn't opted in to test them, and don't want to disable read-only access by default.
cameronrenwick
@cameronrenwick
Jul 04 2017 18:12
well said @scottleibrand ... I agree with your thoughts and approach. I too think there are far more nefarious things that can happen...
Tom Petrillo
@tompetrillo
Jul 04 2017 19:07
Hi. I just started looping and experience from time to time long term communications difficulty. Over the course of the day I would say I get 10 hours of communications and 14 without communications, each coming in blocks of several hours. I've tried changing pump batteries and toggling the remote option to see if this would fix it. I've also rebooted many a time. Will the cut trick help with this issue? Any othersuggestions? I'm using a 512, so the age could certainly be a factor.
Dana Lewis
@danamlewis
Jul 04 2017 19:11
What do your logs say when it's not communicating?
Tom Petrillo
@tompetrillo
Jul 04 2017 19:12
"916.636", 0, -99 Starting pump-loop at Tue Jul 4 12:07:17 PDT 2017: Radio ok, Listening: .No pump comms detected from other rigs Old pumphistory: Error, retrying Starting pump-loop at Tue Jul 4 12:08:25 PDT 2017: Radio ok, Listening: .No pump comms detected from other rigs Old pumphistory: Error, retrying Starting pump-loop at Tue Jul 4 12:09:35 PDT 2017: Radio ok, Listening: .No pump comms detected from other rigs Old pumphistory: Error, retrying Starting pump-loop at Tue Jul 4 12:10:42 PDT 2017: Radio ok, Listening: .No pump comms detected from other rigs Old pumphistory: Error, retrying Listening: .
it looks like it does some type of "tuning" discovers the frequency as 916.636 and then fails to communicate.
Dana Lewis
@danamlewis
Jul 04 2017 19:14
Yea -99 means no comms. Where is pump in relation to rig at those times?
Tom Petrillo
@tompetrillo
Jul 04 2017 19:14
its very close right now
touching
the rig is in a plastic bag if that could somehow effect things
Dana Lewis
@danamlewis
Jul 04 2017 19:17
Shouldn't
Jake Punshon
@JakePunshon
Jul 04 2017 19:18
whenever i get a 916.636 0 -99 message, I reboot my APS and it starts right up.
Tom Petrillo
@tompetrillo
Jul 04 2017 19:24
@jkp1 That has at times fixed the problem, but a) its not fixing it right now and b) it happens to often for that solution
Jake Punshon
@JakePunshon
Jul 04 2017 19:25
do you use wifi or bt?
Tom Petrillo
@tompetrillo
Jul 04 2017 19:25
both
right now i'm in my house, so wifi in the house
Jake Punshon
@JakePunshon
Jul 04 2017 19:26
does it look like you lose comms when you are switchingbetween networks>
Tom Petrillo
@tompetrillo
Jul 04 2017 19:26
I dont think thats related
Jake Punshon
@JakePunshon
Jul 04 2017 19:27
ok. it seemed to be related for me and I now get fewer -99 errors just staying on BT permanently
Tom Petrillo
@tompetrillo
Jul 04 2017 19:27
well... let me shut off my phones blue tooth and restart
you think I should disable wifi entirely?
Jake Punshon
@JakePunshon
Jul 04 2017 19:28
its something to try if you have done all the easy stuff already
Tom Petrillo
@tompetrillo
Jul 04 2017 19:28
ok... can you comment on the antenna cutting suggestion?
is that at all reasonably likely to help with this type of issue?
Jake Punshon
@JakePunshon
Jul 04 2017 19:28
i haven't cut my antenna
Tom Petrillo
@tompetrillo
Jul 04 2017 19:28
I'm nervous to do something like that since it can't be undone.
Jake Punshon
@JakePunshon
Jul 04 2017 19:29
it can be undone with a bit of solder : )
the instructions on cutting it are pretty good, i think. it was on my list of things to do but my comm has been better since i went BT all the time, so i haven't cut
Tom Petrillo
@tompetrillo
Jul 04 2017 19:30
are you able to ssh in while its on BT?
Jake Punshon
@JakePunshon
Jul 04 2017 19:30
yep
Tom Petrillo
@tompetrillo
Jul 04 2017 19:31
how?
Jake Punshon
@JakePunshon
Jul 04 2017 19:31
i use juice SSH on my GL G6
Tom Petrillo
@tompetrillo
Jul 04 2017 19:31
@gregtcm thats a good idea and I can borrow one from work tomorrow!
gregtcm
@gregtcm
Jul 04 2017 19:32
It seems to help with my rig, although I haven't done any testing. Just real world experience.
Jake Punshon
@JakePunshon
Jul 04 2017 19:32
you need your bt IP address, there are instructions on how to obtain that in the BT tethering section of the setup docs
Tom Petrillo
@tompetrillo
Jul 04 2017 19:32
thanks.
gregtcm
@gregtcm
Jul 04 2017 19:33
@jkp1 Can JuiceSSH connect even w/o an external network connection?
Scott Leibrand
@scottleibrand
Jul 04 2017 19:33
@Redshift83 what frequency and rssi did mmtune get the last time it got a better rssi than -99?
Tom Petrillo
@tompetrillo
Jul 04 2017 19:34

Scott:

"916.516", 3, -80 Starting pump-loop at Tue Jul 4 03:01:33 PDT 2017:
"916.492", 1, -85 Starting pump-loop at Tue Jul 4 09:35:18 PDT 2017:
"916.492", 1, -89 Starting pump-loop at Tue Jul 4 10:32:43 PDT 2017:
"916.516", 2, -80 Starting pump-loop at Tue Jul 4 10:43:15 PDT 2017:
"916.492", 1, -87 Starting pump-loop at Tue Jul 4 11:26:33 PDT 2017:
"916.468", 1, -86 Starting pump-loop at Tue Jul 4 11:41:33 PDT 2017:
"916.492", 1, -88 Starting pump-loop at Tue Jul 4 12:19:35 PDT 2017:
"916.516", 1, -88 Starting pump-loop at Tue Jul 4 12:31:18 PDT 2017:

(grepped a few)
gregtcm
@gregtcm
Jul 04 2017 19:34
I was having trouble with Termius and others when I was offline a few weeks ago. Looping was still working (I'm using xDripAPS) but I couldn't access the rig via my hotspot.
Scott Leibrand
@scottleibrand
Jul 04 2017 19:35
Do you have another Explorer board or another pump (even non-loopable) you could try?
Jake Punshon
@JakePunshon
Jul 04 2017 19:35
@gregtcm yes, it connects via the BT
Tom Petrillo
@tompetrillo
Jul 04 2017 19:35
@scottleibrand is that at me?
Scott Leibrand
@scottleibrand
Jul 04 2017 19:35
Yeah
Tom Petrillo
@tompetrillo
Jul 04 2017 19:35
i do have another pump
but what would i do with it?
(its non loopable)
Scott Leibrand
@scottleibrand
Jul 04 2017 19:36
Configure it as if it were loopable, and see if it can read data off it. It just won't be able to set temps.
gregtcm
@gregtcm
Jul 04 2017 19:36
@jkp1 Ok, I'll try it ... I'm connecting via wifi hotspot ... and now I'll bow out and not interrupt anymore. :)
Scott Leibrand
@scottleibrand
Jul 04 2017 19:36
If that works, it may be a flaky loopable pump. If it doesn't, you might need a new EB
Tom Petrillo
@tompetrillo
Jul 04 2017 19:36
you mean turn on the remote feature in the utilities menu and enter the serial when I reinstall openaps script?
Dana Lewis
@danamlewis
Jul 04 2017 19:37
You might want to copy your current setup so you don't have to redo your hard x12 work
Scott Leibrand
@scottleibrand
Jul 04 2017 19:37
The remote feature isn't needed, but yeah. You could install it to a new directory instead of myopenaps to leave your existing setup alone.
Tom Petrillo
@tompetrillo
Jul 04 2017 19:38
ok
Scott Leibrand
@scottleibrand
Jul 04 2017 19:38
You could even put the non-loopable pump into PC Connect mode to let it look for a couple minutes before it times out.
*loop
Tim Street
@tim2000s
Jul 04 2017 19:46
Anyone seen this before when trying to flash an Edison? I keep getting similar errors:
Flashing rootfs, (it can take up to 10 minutes... Please be patient)
Download    [===========              ]  47%    760795136 bytesdfu-util: Error during download
I think it's an edison error
Tom Petrillo
@tompetrillo
Jul 04 2017 19:51
@scottleibrand I'm probably not completely doign everything right, but within a couple minutes of install ing the new system, I got a message: "916.732", 5, -43 Listening: .No pump comms detected from other rigs
I assume that means the issue I'm facing is likely with the 512's communications module
PieterGit
@PieterGit
Jul 04 2017 19:51
@scottleibrand about https://gitter.im/nightscout/intend-to-bolus?at=595bd501703e565c3373a817 : Token based authentication is going further than only increase privacy. Using roles to restrict permissions (Role-based access control) of devices is a quite common way to increase security. With internet connected devices such as oref0 that are able to influence insulin, we should take security more serious, and my opinion is that not having a secure by default policy in 0.5.0 is a wrong decission. If stuff is broken with token based it should be treated as a major bug (for example if curl is still used somewhere instead of using the ns device). Bugs must be fixed and sometimes require a patchrelease. There is no reason why we cannot fix such issues in a 0.5.1 release, just like we did release a hotfix for the MRAA issues. If people want to use readable and not use role based access, it's possible to answer no and use the old API_SECRET authentication. It's also possible to use readable in combination with token based authentication, so if people do not want to disable read-only access to their Nightscout, there is no problem.
Scott Leibrand
@scottleibrand
Jul 04 2017 19:52
@redshift83 yes, a -43 is a very strong signal.
Scott Leibrand
@scottleibrand
Jul 04 2017 19:57
@PieterGit I'm happy to change oref0-setup to use token auth by default once we've had some people run through it end to end and verify it is easy to set up as documented. And there's no reason we can't add support for token auth in 0.5.0, and then enable it by default in a 0.5.1 release. I understand the security concerns, but I'm also mindful of the support concerns that come with making users do something new and different.
Tom Petrillo
@tompetrillo
Jul 04 2017 20:00
@scottleibrand so I should source another pump, correct?
@scottleibrand or, try the cutting technique, or install an antenna?
RedCyclist1
@RedCyclist1
Jul 04 2017 20:00
Hey @scottleibrand or @PieterGit , did the master file ever get updated with SMB?
Scott Leibrand
@scottleibrand
Jul 04 2017 20:01
@RedCyclist1 that release (0.5.0) is what we're planning now.
The dev branch is ready for testing if you'd like to help.
@redshift83 yes, I would start looking for another pump. You can try the antenna stuff, but that will likely only improve signal up to 10 dB, not fix if it's completely non responsive.
RedCyclist1
@RedCyclist1
Jul 04 2017 20:03
We tried once, LOL, and were told we should wait until it releases.... We couldn't figure out the flag and nightscout pieces. @scottleibrand
Tom Petrillo
@tompetrillo
Jul 04 2017 20:03
ok.... this is a major bummer
Martin Haeberli
@mhaeberli
Jul 04 2017 20:04
@redshift83 as a thought (and a giant pain) have you tried noting (say, in a journal) or looking at your logs to note times AND places where it works and where it doesn’t? In case there may be some kind of crazy RF (radio frequency) interference
Tom Petrillo
@tompetrillo
Jul 04 2017 20:04
@mhaeberli the problem occurs in the bedroom, so even if true, its a major road block.
RedCyclist1
@RedCyclist1
Jul 04 2017 20:05
@scottleibrand Reading now
Martin Haeberli
@mhaeberli
Jul 04 2017 20:05
@redshift83 also, have you tried changing to a fresh pump battery, especially the energizer ultimate lithium batteries
Tom Petrillo
@tompetrillo
Jul 04 2017 20:06
yes
Martin Haeberli
@mhaeberli
Jul 04 2017 20:06
@redshift83 i can see why bedroom would be a problem… as a backup, have you tried adding an RPi / TI stick rig to your bedroom ? ‘cause sometimes the TI stick can have better range?
Tom Petrillo
@tompetrillo
Jul 04 2017 20:07
its an idea, but given the size of the problem, not going to sink money into that when I suspect things wont work out regardless.
Martin Haeberli
@mhaeberli
Jul 04 2017 20:07
we (on behalf of our almost-off-to-college T1 teen) have some problems w/edison in the bedroom that have been mitigated by hanging the Edison from its charging cable kind of above the bed a few feet
Dana Lewis
@danamlewis
Jul 04 2017 20:07
@RedCyclist1 you can try again on dev with a fresh pull using setup script. We changed the way you enable it.
Martin Haeberli
@mhaeberli
Jul 04 2017 20:07
understood - cheering you on
Tom Petrillo
@tompetrillo
Jul 04 2017 20:08
i might try dismantling the pump and seeing if some spring cleaning on the inside improves thigns
RedCyclist1
@RedCyclist1
Jul 04 2017 20:09
@danamlewis Will we have to change nightscout to the dev build?
Dana Lewis
@danamlewis
Jul 04 2017 20:11
If you want to see the oref1 related predictions, yes
Martin Haeberli
@mhaeberli
Jul 04 2017 20:11
@redshift83 let me know if you have good instructions on pump dis - assembly / assembly - those could come in handy here, too. as an EE / Computer Science guy (and long - time amateur radio guy) i’m a little skeptical that dis - and re- assembly of pump with internal cleaning can inherently help - but wishing you well
RedCyclist1
@RedCyclist1
Jul 04 2017 20:12
@danamlewis Is there a link with some clear instructions on how to do that? :)
Tim Street
@tim2000s
Jul 04 2017 20:13
Every edison I have flashed on an explorer board with jubilinux 0.1.1 has failed to run WPA_supplicant daemon correctly.
francesc0-cgm
@francesc0-cgm
Jul 04 2017 20:13
@scottleibrand will be ever available a GUI for openaps to bolus from a phone?
Tim Street
@tim2000s
Jul 04 2017 20:13
Does anyone have any idea how to fix that?
It's really starting to get on my nerves
Martin Haeberli
@mhaeberli
Jul 04 2017 20:14
@tim2000s what kind of failures do you see? we haven’t seen that here … (3 edisons, 3 explorers, more not yet flashed, all 0.1.1)
Dana Lewis
@danamlewis
Jul 04 2017 20:14
@francesc0-cgm yes. Part of the app being built for read/write to rig. No timeline though.
Eric
@ecc1
Jul 04 2017 20:15
I'd try a different USB cable, preferably connected directly to your computer (versus a hub)
@tim2000s ^^
Dana Lewis
@danamlewis
Jul 04 2017 20:15
@RedCyclist1 yes. NS has good docs for updating.
francesc0-cgm
@francesc0-cgm
Jul 04 2017 20:15
Thank you @danamlewis this is a beautiful news. As a parent one of the best feature i had from loop is the possibility to never bring on hands my son pump.
Eric
@ecc1
Jul 04 2017 20:16
@tim2000s re: your partial flashing problem, not wpa_supplicant ...
John Benjamin
@Bender1061
Jul 04 2017 20:16
@scottleibrand @danamlewis So there is something seriously an issue with the Dev release. it's extremly unreliable now. I've done this on 3 Explorer board rigs now. One of them works somewhat reliable, but the others I'm lucky if it completes a loop an hour. Reduced running of mmtune, so that may be part of the issue, but It's in the Where it should be running the Precheck. I get several different errors mmeowlink.exceptions.CommsException: No acknowledgement from pump on wakeup. Is it out of range or is the battery too low? so with that, I decided to do the ccprog. Once I do that, stderr: 'fatal: Not a git repository (or any of the parent directories): .git' happens about 80% of the time. Now on my Non Explorer board rigs, it works fine. So I'm thinking it has something to do with the explorer board. And this was not happening on the #514 so it's something that was added before we added this to Dev.
francesc0-cgm
@francesc0-cgm
Jul 04 2017 20:16
Using openaps i miss this feature a bit
Dana Lewis
@danamlewis
Jul 04 2017 20:17
@Bender1061 can you create an issue so we can better track? @Kdisimone and I have been discussing but need to pinpoint the error messages and frequency etc to better drill down
francesc0-cgm
@francesc0-cgm
Jul 04 2017 20:18
@Bender1061 me too se the related issue but only on reboot or on starting. After a mmtune it is quite stable
Tim Street
@tim2000s
Jul 04 2017 20:18
Okay - then I need some new USB-C to USB-Micro cables
Dana Lewis
@danamlewis
Jul 04 2017 20:19
May also be the adapter if you're using adapter - had someone w new Mac that had a lot of issues but old Mac w regular USB worked fine for same edison
John Benjamin
@Bender1061
Jul 04 2017 20:20
openaps/oref0#522 Created
Dana Lewis
@danamlewis
Jul 04 2017 20:20
Ty @Bender1061
PieterGit
@PieterGit
Jul 04 2017 20:32
i issued a comment in openaps/oref0#522 . My goal was to rewrite the init stuff (including the WW-init stuff) and better integrate in the openaps pump device. That's quite some work and it takes quite some time to understand how the openaps pump and mmeowlink stuff internals works.
Tim Street
@tim2000s
Jul 04 2017 20:58
May need to go back to the Linux box for flash.
RedCyclist1
@RedCyclist1
Jul 04 2017 21:03
Trying to run setup for dev and getting this error Please, commit your changes or stash them before you can merge.
Aborting
Couldn't pull latest oref0
Tom Petrillo
@tompetrillo
Jul 04 2017 21:07
@RedCyclist1 cd ~/src and type git status
and then likely type git stash
RedCyclist1
@RedCyclist1
Jul 04 2017 21:09
@redshift83 Got this message "fatal: Not a git repository (or any of the parent directories): .git"
After typing in what you suggested above
Tom Petrillo
@tompetrillo
Jul 04 2017 21:11
cd ~/src/oref0 and git status
look at the files that have been changed and carefully consider stashing them by git stash
PieterGit
@PieterGit
Jul 04 2017 21:27
@Bender1061 : what type of case do you use if you have radio problems described in openaps/oref0#522 ? I think some cases have worse rig-pump communication due to the metal of the battery and too little space between rig and lipo
John Benjamin
@Bender1061
Jul 04 2017 21:33
I actually have it just sitting on my desk. I use a plastic project box normally, with a part of a sock around the battery so that the battery is not directly against the rig, which also has heat shrink around it. But the issue right now is most def not the battery getting in the way
PieterGit
@PieterGit
Jul 04 2017 21:49
ok, just trying to figure out what the root cause is where fighting
Scott Leibrand
@scottleibrand
Jul 04 2017 21:49
@RedCyclist1 I assume you haven't made any changes to your oref0 source code, so cd ~/src/oref0 and then git stash is likely all you need, and then you can re-run the command that failed.
Geoff Aumaugher
@geoffro10
Jul 04 2017 22:30
@danamlewis Ok, I reflashed and reran the new script. Worked fine.
gregtcm
@gregtcm
Jul 04 2017 23:37
I'm trying to update my rig (on dev) and am running into an error with running nightscout autoconfigure-device-crud. Could it have anything to do with the new token based authentication code? Has anyone else come across this error?
nightscout autoconfigure-device-crud https://xxxxx.herokuapp.com xxxxxxxx
added process://ns/nightscout/ns NIGHTSCOUT_HOST API_SECRET
Traceback (most recent call last):
  File "/usr/local/bin/openaps-device", line 35, in <module>
    app( )
  File "/usr/local/lib/python2.7/dist-packages/openaps/cli/__init__.py", line 52, in __call__
    self.epilog( )
  File "/usr/local/lib/python2.7/dist-packages/openaps/cli/__init__.py", line 75, in epilog
    self.create_git_commit( )
  File "/usr/local/lib/python2.7/dist-packages/openaps/cli/__init__.py", line 98, in create_git_commit
    self.repo.index.commit(msg)
  File "/usr/local/lib/python2.7/dist-packages/git/index/base.py", line 954, in commit
    author_date=author_date, commit_date=commit_date)
  File "/usr/local/lib/python2.7/dist-packages/git/objects/commit.py", line 316, in create_from_tree
    parent_commits = [repo.head.commit]
  File "/usr/local/lib/python2.7/dist-packages/git/refs/symbolic.py", line 176, in _get_commit
    obj = self._get_object()
  File "/usr/local/lib/python2.7/dist-packages/git/refs/symbolic.py", line 169, in _get_object
    return Object.new_from_sha(self.repo, hex_to_bin(self.dereference_recursive(self.repo, self.path)))
  File "/usr/local/lib/python2.7/dist-packages/git/refs/symbolic.py", line 119, in dereference_recursive
    hexsha, ref_path = cls._get_ref_info(repo, ref_path)
  File "/usr/local/lib/python2.7/dist-packages/git/refs/symbolic.py", line 136, in _get_ref_info
    assert(len(tokens) != 0)
AssertionError
Traceback (most recent call last):
  File "/usr/local/bin/openaps-device", line 35, in <module>
    app( )
  File "/usr/local/lib/python2.7/dist-packages/openaps/cli/__init__.py", line 52, in __call__
    self.epilog( )
  File "/usr/local/lib/python2.7/dist-packages/openaps/cli/__init__.py", line 75, in epilog
    self.create_git_commit( )
  File "/usr/local/lib/python2.7/dist-packages/openaps/cli/__init__.py", line 98, in create_git_commit
    self.repo.index.commit(msg)
  File "/usr/local/lib/python2.7/dist-packages/git/index/base.py", line 954, in commit
    author_date=author_date, commit_date=commit_date)
  File "/usr/local/lib/python2.7/dist-packages/git/objects/commit.py", line 316, in create_from_tree
    parent_commits = [repo.head.commit]
  File "/usr/local/lib/python2.7/dist-packages/git/refs/symbolic.py", line 176, in _get_commit
    obj = self._get_object()
  File "/usr/local/lib/python2.7/dist-packages/git/refs/symbolic.py", line 169, in _get_object
    return Object.new_from_sha(self.repo, hex_to_bin(self.dereference_recursive(self.repo, self.path)))
  File "/usr/local/lib/python2.7/dist-packages/git/refs/symbolic.py", line 119, in dereference_recursive
    hexsha, ref_path = cls._get_ref_info(repo, ref_path)
  File "/usr/local/lib/python2.7/dist-packages/git/refs/symbolic.py", line 136, in _get_ref_info
    assert(len(tokens) != 0)
AssertionError
process://ns/nightscout/ns https://xxxxx.herokuapp.com xxxxxxxx
Traceback (most recent call last):
  File "/usr/local/bin/openaps-import", line 89, in <module>
    app( )
  File "/usr/local/lib/python2.7/dist-packages/openaps/cli/__init__.py", line 52, in __call__
    self.epilog( )
  File "/usr/local/bin/openaps-import", line 50, in epilog
    super(ImportToolApp, self).epilog( )
  File "/usr/local/lib/python2.7/dist-packages/openaps/cli/__init__.py", line 75, in epilog
    self.create_git_commit( )
  File "/usr/local/lib/python2.7/dist-packages/openaps/cli/__init__.py", line 98, in create_git_commit
    self.repo.index.commit(msg)
  File "/usr/local/lib/python2.7/dist-packages/git/index/base.py", line 954, in commit
    author_date=author_date, commit_date=commit_date)
  File "/usr/local/lib/python2.7/dist-packages/git/objects/commit.py", line 316, in create_from_tree
    parent_commits = [repo.head.commit]
  File "/usr/local/lib/python2.
Scott Leibrand
@scottleibrand
Jul 04 2017 23:41
Did you select token auth or API secret?
gregtcm
@gregtcm
Jul 04 2017 23:43
Did run-again ... so wasn't given the question ... which makes me think that's the problem and I need to run fresh. :) Thanks for the insight!
John Benjamin
@Bender1061
Jul 04 2017 23:43
does anyone know what refreshgit.exc.InvalidGitRepositoryError: means? obviously something is wrong with the git. but I have no idea what that really means