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

21st
Feb 2017
PieterGit
@PieterGit
Feb 21 2017 00:04
More good news. I 've been able to get ti chip on explorer board with 4.10rc4 kernel working! Succesfull mmtune and got my pump model back. Thanks @ecc1 for helping me with the last part. This makes newer kernel possible for people like me who can't get USB OTG Dexcom G4 working on stock 3.10 kernel with Explorer board with trace cut.
will work on patches to remove mraa for future kernels, perhaps make spi_serial work with both mraa and without it (because not using mraa probably requires 4.8 or 4.10+ kernel)
Scott Leibrand
@scottleibrand
Feb 21 2017 00:45
@lukas-ondriga we have an open issue about interpolating missing glucose values for the COB calculations. Even just assuming BG was flat when data is missing would be an improvement. Then the min_5m_carbimpact would kick in, or BGI, if insulin activity peaks while CGM data is missing. Yes, a PR on that subject would be most welcome.
Also, as Tim mentioned, I would make sure you have continuous CGM data whenever possible. AMA and autosens assume that.
peterfulvi
@peterfulvi
Feb 21 2017 00:51
I got my rig and phoned tethered and established a network connection. On my walk home I tried to tunnel in with the IP address if config had for it which I was able to access before when I shut the wifi of on my phone. It was saying it couldn't connect. One thing I do notice it's still port 22. Is there another port i need to configure my connection to
feraridurango
@feraridurango
Feb 21 2017 00:53

''
Attempting to communicate with pump:
openaps: no process found
pump://JSON/read_temp_basal/monitor/temp_basal.json
monitor/temp_basal.json raised ACK is 0 bytes:

Traceback (most recent call last):
File "/usr/local/bin/openaps-report", line 82, in <module>
app( )
File "/usr/local/lib/python2.7/dist-packages/openaps/cli/init.py", line 51, in call
self.run(self.args)
File "/usr/local/bin/openaps-report", line 75, in run
output = app(args, self)
File "/usr/local/lib/python2.7/dist-packages/openaps/cli/subcommand.py", line 52, in call
return self.method.main(args, app)
File "/usr/local/lib/python2.7/dist-packages/openaps/reports/invoke.py", line 40, in main
output = task.method(args, app)
File "/usr/local/lib/python2.7/dist-packages/openaps/uses/use.py", line 44, in call
self.before_main(args, app)
File "/usr/local/lib/python2.7/dist-packages/openaps/vendors/medtronic.py", line 62, in before_main
self.setup_medtronic( )
File "/usr/local/lib/python2.7/dist-packages/openaps/vendors/medtronic.py", line 161, in setup_medtronic
self.uart.open( )
File "/usr/local/lib/python2.7/dist-packages/decocare/stick.py", line 874, in open
log.info('%s' % self.product_info( ))
File "/usr/local/lib/python2.7/dist-packages/decocare/stick.py", line 515, in product_info
return self.query(ProductInfo)
File "/usr/local/lib/python2.7/dist-packages/decocare/stick.py", line 509, in query
return self.process( )
File "/usr/local/lib/python2.7/dist-packages/decocare/stick.py", line 496, in process
ack, response = self.command.respond(raw)
File "/usr/local/lib/python2.7/dist-packages/decocare/stick.py", line 71, in respond
raise AckError("ACK is 0 bytes:\n%s" % lib.hexdump(raw))
decocare.errors.AckError: ACK is 0 bytes:

''

has anyone seen this before?
Error, retrying
monitor/temp_basal.json  raised  ACK is 0 bytes:

pump://JSON/read_temp_basal/monitor/temp_basal.json
Traceback (most recent call last):
  File "/usr/local/bin/openaps-report", line 82, in <module>
    app( )
  File "/usr/local/lib/python2.7/dist-packages/openaps/cli/__init__.py", line 51, in __call__
    self.run(self.args)
  File "/usr/local/bin/openaps-report", line 75, in run
    output = app(args, self)
  File "/usr/local/lib/python2.7/dist-packages/openaps/cli/subcommand.py", line 52, in __call__
    return self.method.main(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps/reports/invoke.py", line 40, in main
    output = task.method(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps/uses/use.py", line 44, in __call__
    self.before_main(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps/vendors/medtronic.py", line 62, in before_main
    self.setup_medtronic( )
  File "/usr/local/lib/python2.7/dist-packages/openaps/vendors/medtronic.py", line 161, in setup_medtronic
    self.uart.open( )
  File "/usr/local/lib/python2.7/dist-packages/decocare/stick.py", line 874, in open
    log.info('%s' % self.product_info( ))
  File "/usr/local/lib/python2.7/dist-packages/decocare/stick.py", line 515, in product_info
    return self.query(ProductInfo)
  File "/usr/local/lib/python2.7/dist-packages/decocare/stick.py", line 509, in query
    return self.process( )
  File "/usr/local/lib/python2.7/dist-packages/decocare/stick.py", line 496, in process
    ack, response = self.command.respond(raw)
  File "/usr/local/lib/python2.7/dist-packages/decocare/stick.py", line 71, in respond
    raise AckError("ACK is 0 bytes:\n%s" % lib.hexdump(raw))
decocare.errors.AckError: ACK is 0 bytes:
Dana Lewis
@danamlewis
Feb 21 2017 01:09
It's a carelink error
Unplug replug
hilarykoch
@hilarykoch
Feb 21 2017 01:23
I think I got one of the Explorer boards in the first batch that went out (in November I think?) Can I hook our Dex receiver directly to it?
feraridurango
@feraridurango
Feb 21 2017 01:31
@danamlewis thank you!!!!!!!! :)
Scorponok
@Scorponok
Feb 21 2017 01:34
You should also be able to fix that by running oref0-reset-usb if you can ssh in
kenstack
@kenstack
Feb 21 2017 01:39
@ecc1 rc.local was the key for me - but all credit goes to @scottleibrand @danamlewis for pushing me to do a clean install and look at the differences - I haven't tried running it without screen but I would assume it works - on my list
Eric
@ecc1
Feb 21 2017 02:09
Yes, I extracted just the brcm_patchram_plus command and put it into my /etc/rc.local (with the BT address from the /dev/mmcblk0p5 partition hard-coded in). Seems to be working fine.
zaheria985
@zaheria985
Feb 21 2017 02:15
I joined here ages ago planning to close the loop and life got in the way.. but now I'm going to finally do it. Just have a silly question - if you buy a pump with the correct firmware on it, is it possible to accidentally upgrade the firmware with Medtronic somehow? You know - when you have a device hooked up to the computer or internet sometimes it just updates itself. Or your phone for example...
Dana Lewis
@danamlewis
Feb 21 2017 02:37
Welcome @zaheria985. Nope, you can't change the firmware on the pump yourself
zaheria985
@zaheria985
Feb 21 2017 02:38
Good, because if I could find a way to do it by accident, it would happen.
glenrob1940
@glenrob1940
Feb 21 2017 03:13
In the loop discussions they talk about an issue with 523/723 pumps and boluses > 10.0 where these pumps increase flow rate above normal boluses. The ref is https://github.com/loudnate/Loop/releases/tag/v0.11.0
for the fix. Does open APS handle this
Eric
@ecc1
Feb 21 2017 03:25
openaps never sends bolus commands, just temporary basal rates. I don't think it's possible to set an hourly temp basal rate that comes close to a 10U bolus, so I can't see how this could come up.
Darrell Wright
@beached
Feb 21 2017 03:39
10U/hr would be 0.6U/min. A bolus will push, what, 1U in a minute or more
Terri Lyman
@tatigger
Feb 21 2017 03:41
@hilarykoch If I read the doc correctly, we cannot plug the dexcom G4 directly into the older Explorer Board.
Terri Lyman
@tatigger
Feb 21 2017 03:47
How do you get your Dexcom data to your OpenAPS.....because, my most recent XDrip+ update seemed to have not been mirroring the dexcom data very well. And that is making me nervous. I reverted back to an older version that seems better for now. So, how do Y'ALL handle that?
David
@david-t1
Feb 21 2017 04:15
@ecc1 Eric, just wanted to let you know, I tried a different phone and bang, bt-pan client works. I also had to uncomment the patchram line which I had previously commented out to use openxshareble.
katie disimone
@Kdisimone
Feb 21 2017 04:37

Can someone help explain how autosens and autotune are affecting this situation right now?

My daughter normally has a basal rate around 0.6 to 0.8. This last day or so she’s been hit by hormones and a bad cold. Despite an increase in basal overnight, she stayed about 30 points above target bg and carried a fairly big chunk of positive iob the whole time.

So through today, we’ve drastically upped her basal rate to 1.3 and tried to bring her back closer to target. Lots of work today and we still haven’t gained ground like we normally would.

I pulled up the pump log to look specifically at what’s going on.

Starting pump-loop at Mon Feb 20 20:31:03 PST 2017:
Radio ok, Listening: .No pump comms detected from other rigs
Profile less than 60m old
glucose.json newer than temp_basal.json
Temp refreshed
{"carbs":0,"boluses":3.3,"mealCOB":0}
{"ratio":1.2}
{"iob":0.29,"activity":0.0039,"bolussnooze":0,"basaliob":0.29,"netbasalinsulin":0.3,"hightempinsulin":1.3,"time":"2017-02-21T04:31:45.000Z"}
{"delta":-3,"glucose":93,"short_avgdelta":-3.5,"long_avgdelta":-3.37}
{"duration":25,"rate":0,"temp":"absolute"}
Adjusting basal from 0.577 to 0.7
Adjusting sens from 42.857 to 35.7
Carb Impact: -2.8 mg/dL per 5m; CI Duration: 0 hours
Accel. Carb Impact: 10 mg/dL per 5m; ACI Duration: 0 hours
determine-basal://text/shell/enact/suggested.json
reporting enact/suggested.json
{"temp":"absolute","bg":93,"tick":-3,"eventualBG":66,"snoozeBG":66,"predBGs":{"IOB":[93,90,87,84,81,79,76,74,73,71,70,70,69,69,68,68,68,68,68,67,67,68,68,68,68,68,68,69,69,69,70,70,70,71,71,71,72,72,72,72,73,73,73,73,73,73,73,73,74]},"COB":0,"IOB":0.29,"reason":"COB: 0, Dev: -17, BGI: -0.7, ISF: 36, Target: 90; Eventual BG 66 < 90, setting -1.975U/hr, but 25m left and 0 ~ req 0U/hr: no action required"}
{"bg":96,"temp":"absolute","snoozeBG":61,"recieved":true,"predBGs":{"IOB":[96,92,88,85,81,78,76,73,71,70,68,67,67,66,66,65,65,65,64,64,64,64,64,64,64,64,64,64,64,64,64,65,65,65,65,65,66,66,66,66,67]},"rate":0,"reason":"COB: 0, Dev: -22, BGI: -0.61, ISF: 36, Target: 90; Eventual BG 61 < 90, setting -2.55U/hr","COB":0,"eventualBG":61,"timestamp":"2017-02-20T20:25:48.226011","duration":30,"tick":-4,"IOB":0.366}

The parts that stuck out to me:

Adjusting basal from 0.577 to 0.7
Adjusting sens from 42.857 to 35.7

It seems as though her 1.3 basal rate isn’t being considered at all. And her ISF in the pump is 30.

Why does it look like autosens is not using her current pump pattern at all?

I checked her settings’ basal_profile.json and the basal rate of 1.3 has been properly pulled from pump. Just can’t see where it is being used.

katie disimone
@Kdisimone
Feb 21 2017 04:48
Basically…i guess it looks like I may just need to reframe my understanding of autotune and autosens. Like in situations where I manually feel the need to significantly adjust basal rates in the pump…sounds like I need to understand that autotune and autosens will be running my basals instead. I need to turn those off if I want to control them with the pump settings instead during these big shifts?
Scott Leibrand
@scottleibrand
Feb 21 2017 05:07
One thing you might want to consider is allowing autosens to adjust more than 20% from the autotuned baseline. Or you could manually remove the autotune.json and re-run it from the new pumpprofile.
(Not seeing my other message) The updated pumpprofile won't be used directly if you have autotune enabled. It'll only be used to set the upper and lower bounds for autotune.
katie disimone
@Kdisimone
Feb 21 2017 05:14

Ok…thanks @scottleibrand That helps me reframe how I treat these instances. I think I understand how it’s working better now. I was unaware that the autotuned baseline locks in regardless of pump profile changes later that day. Makes sense when I stop and think about things logistically…but I wasn’t getting that earlier in the day.

So, it would seem like (for example) if you had a sudden increase in basal needs from something like a steroid pill in middle of day…you may absolutely need to delete the old autotune.json and start fresh with a new pump profile (assuming needs changed significantly and suddenly)?

Scott Leibrand
@scottleibrand
Feb 21 2017 05:18
We haven't worked out best practices for that yet, but yeah, that's a good start.
Dana Lewis
@danamlewis
Feb 21 2017 05:18
Yes
Sacha M
@coolestkidsever
Feb 21 2017 06:18
@jmcrawford thanks for the info on the failed pack ref, but I wasn't able to run that command-it didn't work. Any other ideas? Thanks

Am getting this error when looking at log:

ump://JSON/read_temp_basal/monitor/temp_basal.json
Traceback (most recent call last):
File "/usr/local/bin/openaps-report", line 82, in <module>
app( )
File "/usr/local/lib/python2.7/dist-packages/openaps/cli/init.py", line 51, in call
self.run(self.args)
File "/usr/local/bin/openaps-report", line 75, in run
output = app(args, self)
File "/usr/local/lib/python2.7/dist-packages/openaps/cli/subcommand.py", line 52, in call
return self.method.main(args, app)
File "/usr/local/lib/python2.7/dist-packages/openaps/reports/invoke.py", line 44, in main
app.epilog( )
File "/usr/local/bin/openaps-report", line 69, in epilog
super(ReportToolApp, 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 99, in create_git_commit
self.repo.git.gc(auto=True)
File "/usr/local/lib/python2.7/dist-packages/git/cmd.py", line 424, in <lambda>
return lambda args, **kwargs: self._call_process(name, args, kwargs)
File "/usr/local/lib/python2.7/dist-packages/git/cmd.py", line 873, in _call_process
return self.execute(call,
_kwargs)
File "/usr/local/lib/python2.7/dist-packages/git/cmd.py", line 687, in execute
raise GitCommandError(command, status, stderr_value, stdout_value)
git.exc.GitCommandError: Cmd('git') failed due to: exit code(255)
cmdline: git gc --auto
stderr: 'Auto packing the repository in background for optimum performance.
See "git help gc" for manual housekeeping.
fatal: Unable to create '/home/pi/myopenaps2017/.git/packed-refs.lock': File exists.

If no other git process is currently running, this probably means a
git process crashed in this repository earlier. Make sure no other git
process is running and remove the file manually to continue.
error: failed to run pack-refs

But don't know why it went wrong nor how to fix
lukas-ondriga
@lukas-ondriga
Feb 21 2017 06:40
@tim2000s the reason why I turned it off is that my rig is still not wearable and I play with the offline solution for now without downloading from Nightscout. I understand that there are ways how not to loose the data, but it would be nice somehow prepare the software for situations like this. For example you can start the sensor 1 hour after the meal and you can have problems ....
Tim Street
@tim2000s
Feb 21 2017 06:46
@lukas-ondriga Don't know if you saw @scottleibrand's comment. Why did you not have full CGM data available though? That's the bit I'm not clear on?
This message was deleted
lukas-ondriga
@lukas-ondriga
Feb 21 2017 06:57
I use xDripAPS, so my xDrip is set to upload data to the rig. When the rig is off, I do not have data. But there are ways how to get them. I just did not do that for this time.
Tim Street
@tim2000s
Feb 21 2017 07:09
Why not set it to upload data to Nightscout as well so that you never have a point where you lose data for the rig to use?
(Or rather that's much less likely)
I'm running xdripaps using the restAPI and uploading to mongo for nightscout data so I do t have that issue.
lukas-ondriga
@lukas-ondriga
Feb 21 2017 09:11
@tim2000s thanks for the mongo recommendation. I did not know that it is possible to setup it in this way.
Tim Street
@tim2000s
Feb 21 2017 09:43
:+1:
Paul Dickens
@thebookins
Feb 21 2017 10:27
@peterfulvi , @danimaniac I got my 3d printed case today :smile: It was a tight fit getting the PCB in past the screw bosses, but got there in the end
IMG_0012.JPG
also, the retainers on the lid don't really work as intended, which is why I have the tape on the case in the pic above. It will certainly do the job for now, but I'll try to fix those issues when I get to it next.
peterfulvi
@peterfulvi
Feb 21 2017 11:17
@thebookins Very nice! I'm going to order it today to get printed
dan lightner
@danimaniac
Feb 21 2017 14:04

@thebookins @peterfulvi That's great! love the clear resin. Could I trouble you to post a photo of the side with the miniUSB ports? Regarding the lid, my latest revisions include a location to screw the Explorer Board to the lid as well as the case divider. This should keep the case together but be easily taken apart with only one screw. I do not have my Explorer Board yet so all dimensions and measurements are subjest to change. I am getting close, though. I think I will still print one of yours after you make your revisions. Great job!

Edit: Where / what material are you guys using to print?

Diabetic_Dude
@diabetic_dude_twitter
Feb 21 2017 14:04
the Intel Edison Breakout Board Kit listed here, https://goo.gl/29kvYb (the one listed in the openAPS docs) I dont see where the battery connector is.
(I understand that one could get power through the micro USB ports, but I was under the impression it had a separate battery connector for a battery like this one https://goo.gl/8Yj5Ah
John Benjamin
@Bender1061
Feb 21 2017 14:45
2017-02-21 09.43.57.jpg
@diabetic_dude_twitter that's what I did
you can try to wire up a jst connector yourself (if you do, you will want to use one of the other spots to wire it in, so it will use the internal charging circuit) but I liked having the separate charger that would charge my battery faster, and was a bit more stable. and it matters if you are needing to power anything else you can get a bigger charger circuit that also has a usb output.
Diabetic_Dude
@diabetic_dude_twitter
Feb 21 2017 14:49
so this is my next purchase ? https://goo.gl/hMXyzE
John Benjamin
@Bender1061
Feb 21 2017 14:51
thats the bigger one that you can also attach the USB output, issue is this will always output 5V to the edison, which is fine, but will drain your battery faster. if you don't need the extra usb, I would get.. Just a sec, I'll find the part number
Diabetic_Dude
@diabetic_dude_twitter
Feb 21 2017 14:52
thanks yes I want to save the usb ports at all costs I ended up stripping out the connectors on my PI2 rig from plugging and removing the power so much
I use the one you pictured, to charge lipo batteries at my desk, and it does a great job as that, but the smaller one works perfectly, I do also solider the part to make it charge faster too
I use the explorer boards now, but I have several rigs made up just like the picture I have above
wow, just realized I have 8 rigs
Diabetic_Dude
@diabetic_dude_twitter
Feb 21 2017 14:56
woah
John Benjamin
@Bender1061
Feb 21 2017 14:56
it gets addicting
my solidering skills were eh before, now I'm real good
I recommend getting a good flux to use
Diabetic_Dude
@diabetic_dude_twitter
Feb 21 2017 14:57
I just ordered that piece, ]
WTF is a flux?
John Benjamin
@Bender1061
Feb 21 2017 15:01
2017-02-21 10.00.28.jpg
I use a little of that at the joint I'm soldiering at, there is a cemical process, but basically it helps the metal melt evenly
you can get resin cored solider, but I find this works way better
I buy it at Microcenter
Diabetic_Dude
@diabetic_dude_twitter
Feb 21 2017 15:04
where do you buy the wires? any hardware store?
John Benjamin
@Bender1061
Feb 21 2017 15:04
but it seriously makes it much easier, I know some people are scared of soldiering, and will use a solider glue, but that's not going to hold up near as well.
Chris Oattes
@cjo20
Feb 21 2017 15:04
(flux helps the solder flow, and also prevents oxidation of the hot metal)
Diabetic_Dude
@diabetic_dude_twitter
Feb 21 2017 15:05
do you use this instead or a soldering iron or in addition to?
Chris Oattes
@cjo20
Feb 21 2017 15:06
which? The glue or the flux?
John Benjamin
@Bender1061
Feb 21 2017 15:06
exactly as @cjo20 said, as for wire, anyplace that sells solider should have wire, I like single stranded wire, I find multi stranded tends to break over time
you need the flux with the Iron and solider.
Diabetic_Dude
@diabetic_dude_twitter
Feb 21 2017 15:07
cool thanks
John Benjamin
@Bender1061
Feb 21 2017 15:07
the glue (i don't know where to get it at cause I don't use it) you don't need any thing else, but like I said it's not going to last as long
Diabetic_Dude
@diabetic_dude_twitter
Feb 21 2017 15:08
Yeah and I want something stable
John Benjamin
@Bender1061
Feb 21 2017 15:08
and of course I'm just saying my personal preference, people find different ways all the time, and there is nothing wrong with that.
always more than one way to skin a cat
Chris Oattes
@cjo20
Feb 21 2017 15:09
You don't always need the flux with soldering. A lot of solder has flux in it already.
John Benjamin
@Bender1061
Feb 21 2017 15:10
i find it tends to be on the larger side and is hard to work with in small electronics, and I like that bit more flux than solder to start with
but total personal preference there
Diabetic_Dude
@diabetic_dude_twitter
Feb 21 2017 15:13
well I am sure I will be figuring all this out in about a week, thanks to both of you. is there somewhere in the docs that covers the soldering points that I missed? I am really trying to read through this all before asking questions that have already been answered in the docs
John Benjamin
@Bender1061
Feb 21 2017 15:15
the mmeowlink wiki is where the points that needs to be soldered are at for the TI stick https://github.com/oskarpearson/mmeowlink/wiki/TI-USB-Stick-over-Serial Be carefull it's very easy to solder in the wrong hole
Diabetic_Dude
@diabetic_dude_twitter
Feb 21 2017 15:17
so you use a TI?
John Benjamin
@Bender1061
Feb 21 2017 15:17
on the mini breakout board, yes
that's the smaller square on my picture, I remove the usb connection after I verify it works
I guess there is a way now, if I wire up a few extra wires, I can program it from the edison now, but I haven't done that yet
Diabetic_Dude
@diabetic_dude_twitter
Feb 21 2017 15:19
I have been considering TI because of the awesome range but not there yet ... Next go round I am sure
thanks again man I have to get to work (although I am sure I look and sound very busy now)
John Benjamin
@Bender1061
Feb 21 2017 15:19
what are you planning on using?
Diabetic_Dude
@diabetic_dude_twitter
Feb 21 2017 15:20
let me get you the link
John Benjamin
@Bender1061
Feb 21 2017 15:21
you might be better off ordering an explorer board, then no soldering required, and it uses the TI stick built in, so best of all worlds
you don't need to solder anything then
you don't even need the mini breakout board
you only need the edison itself
the stamp looking thing that plugs into the mini breakout board or the edison
Diabetic_Dude
@diabetic_dude_twitter
Feb 21 2017 15:22
yeah I know that but still no place to attach the battery
John Benjamin
@Bender1061
Feb 21 2017 15:22
the explorer has a spot to plug in the battery
Diabetic_Dude
@diabetic_dude_twitter
Feb 21 2017 15:23
ok I will review that when it actually comes in the mail at the end of the month
and save the other pieces for the next build
John Benjamin
@Bender1061
Feb 21 2017 15:24
This is all you need
2017-02-21 10.23.41.jpg
battery plugs into the white part of the explorer board
get the screw pack from sparkfun, that way you can attach the edison more securely to the explorer
Diabetic_Dude
@diabetic_dude_twitter
Feb 21 2017 15:27
ok gimme a sec and I will show you what I have
with the 900mhz-explorer-block on pre order
Diabetic_Dude
@diabetic_dude_twitter
Feb 21 2017 15:33
I also have the smaller battery that I plan to use
see the explorer board that I have does not have that connector
Chris Oattes
@cjo20
Feb 21 2017 15:35
That isn't an explorer board. Thats the intel breakout board
John Benjamin
@Bender1061
Feb 21 2017 15:35
@cjo20 beat me too it
you only needed the edison not the whole kit
but that's ok, when your explorer board that you ordered gets to you, you will have what you need. though I do suggest you get this too https://www.sparkfun.com/products/13187
the explorer does not come with the screws like the mini breakout board does
luckily the mini breakout is only like $20, so you did not buy that much extra that you did not need
Diabetic_Dude
@diabetic_dude_twitter
Feb 21 2017 15:41
it will all get saved and put to good use
If I do not use it I am sure someone in the community will need it
John Benjamin
@Bender1061
Feb 21 2017 15:42
:thumbsup:
Diabetic_Dude
@diabetic_dude_twitter
Feb 21 2017 15:43
ok I outta here thnx again
John Benjamin
@Bender1061
Feb 21 2017 15:44
anytime, glad we got that cleared up.
Matthew Williams
@mwilliams
Feb 21 2017 16:26
Did a big 'ol upgrade weekend on both OpenAPS and Nightscout. All went well aside from Nightscout it seems, OpenAPS is pushing glucose data to it just fine, but there's a disconnect and I'm not getting temp basal rates, and my OpenAPS pill is showing it has effectively no communication with Nightscout (with the OpenAPS plugin enabled). I'm running the OpenAPS dev branch, poking around to see if there's any known issues. Any thoughts otherwise?
(and I pulled in latest master of NS, I was running a dev branch release from late 2016)
Matthew Williams
@mwilliams
Feb 21 2017 16:40
Running autotune for the first time :sparkles:
katie disimone
@Kdisimone
Feb 21 2017 16:41

@scottleibrand expanding on the autotune discussion now that I've had awhile longer to think on it. Can you verify my understanding?

Assuming autotune is enabled to run in the rig. Autotune starts with pump profile settings. Takes one full day of data, and then at midnight, autotune compares original pump profile settings with the day's data and suggests new profile settings.

There's a cap on how far the new profile settings can go from the existing profile (which in the first run is the pump profile settings....subsequent runs would be based on the existing auto tuned profile for that day). Defaults are 1.2 multiplier increase and 0.7 multiplier decrease.

So unless you delete the autotune-generated profiles, autotune doesn't ever really go back to the pump profile as a starting point for the next day?

And what exactly would you have to do to get that "autotune reboot back to pump profile" going? Delete all the contents of the autotune folder? Just some of them? If you deleted all of the files, made updates to your pump settings (on the pump)....would that be sufficient? Because the rig would do an automatic reading of the pump profile later anyways?

thanks in advance for the discussion. This has been quite a good thought process for me

Oops. Sorry about the missed hashtag. My fat fingers.
Eric
@ecc1
Feb 21 2017 17:16
@david-t1 Did you try it with your original phone once you had uncommented the patchram line? That was what made it work for me, but it's possible that Android Marshmallow would still cause issues.
Tim Street
@tim2000s
Feb 21 2017 18:21
@Kdisimone The comments in oref0-Autotune.sh do a good job of explaining how it works.
Tim Street
@tim2000s
Feb 21 2017 18:29
If settings/autotune.json exists it moves that to become autotune/profile.json otherwise it starts from the pump profile. So if you delete settings/Autotune.json, it has to start from settings/profile.json. That should be all you need to do to reset.
So you could set up an os alias to do rm settings/autotune.json and you should he able to start from scratch.
Andy Sharrow
@dramageek
Feb 21 2017 18:43
@diabetic_dude_twitter You won't need the screw set from Sparkfun if you've already ordered the breakout board. The breakout board kit comes with screws included that work fine for the explorer board as well.
Scott Leibrand
@scottleibrand
Feb 21 2017 18:44
@Kdisimone @tim2000s close, but for one thing: autotune bases the 20-30% limits on the pumpprofile, not the current profile. So if you change the pumpprofile to be more than 20% more resistant than the current autotuned settings, the next autotune run will force autotune to adjust to within 20% of that.
You might need to delete settings/profile.json as well, as oref0-get-profile generates that using the autotune.json.
Luis Betancourt
@betluis
Feb 21 2017 18:48
@PieterGit I saw a message from you back in Oct about setting up Pancreable. you were asking about an error using the pebble notify command. I am going through the same process now and having the same question. I wonder if how you got over it?
@PieterGit this is the command and response: $ openaps use pebble notify "hello" "testing"
usage: openaps-use pebble [-h] USAGE ...
openaps-use pebble: error: argument USAGE: invalid choice: 'notify' (choose from 'shell')
Scott Leibrand
@scottleibrand
Feb 21 2017 18:51
@Kdisimone one other consideration: we should have to be manually overriding autotune much. If you find that autosens is maxing out at 20% routinely, and you're wanting it to be more aggressive than that, you might want to raise the 20% limit instead of manually changing profiles, so it can automatically normalize soon after sensitivity does.
katie disimone
@Kdisimone
Feb 21 2017 18:55
@scottleibrand I'm a little surprised to hear that it's not compounding the 20%. Because our isf hasn't been touched in a long time (set at 30 in pump) and it's been using 42 in autotune. Seems like there's some compounding
John Benjamin
@Bender1061
Feb 21 2017 18:57
@betluis have you installed with the latest oref0? if you install everything with the pebbles Mac address, it should setup everything then you should be able to test with the hello testing and it should work. as long as you've paired your watch correctly to your Edison
Scott Leibrand
@scottleibrand
Feb 21 2017 19:03
@Kdisimone autosens can do its 20% on top of autotune's 20%. But autotune can't do more than 20% from the pumpprofile on its own.
katie disimone
@Kdisimone
Feb 21 2017 19:04
Ahhh. That makes sense
Matthias
@ceben80
Feb 21 2017 19:14
@scottleibrand correct me if I'm wrong, at this stage, when autotune is rig enable it's running daily and a recommendation log is created, but not yet adjusted parameters?
What are the next steps in this great autotune solution to be performed :question:
katie disimone
@Kdisimone
Feb 21 2017 19:16
@ceben80 My autotune IS adjusting parameters for my closed loop, not just making recommendations.
Matthias
@ceben80
Feb 21 2017 19:17
is adjusting the basal dose in the pump?
katie disimone
@Kdisimone
Feb 21 2017 19:17
yes
ISF and basals
Matthias
@ceben80
Feb 21 2017 19:18
that's can be verified in the BolusExport settings e.g. ISF...I will have a look at this settings - thx
or it's adjusting parameters for the loop but not seen in the pump @Kdisimone ?
Dana Lewis
@danamlewis
Feb 21 2017 19:20
correct - it does not change the pump’s basic profile
Matthias
@ceben80
Feb 21 2017 19:21
or that's the reason for my confusion Dana :wink:
katie disimone
@Kdisimone
Feb 21 2017 19:22

I can see it when I got into my autotune profiles and in the pump loop.

Basically…the autotune starts with the pump profile at midnight. Makes a new profile to use for the rest of the day based on yesterday’s data. It can adjust from the pump profile up to the maximum adjustments you set (1.2 and 0.7 defaults). And then autosens can take that adjustment even further.

In the autotune folder, you can see the separate file for the pump profile it has started with and the new profiles it is using each day.
JELCRAWFORD
@JELCRAWFORD
Feb 21 2017 19:23
@scottleibrand - re: sorting change - we have had only 1occasion (we've observed) on Sunday where the COB went from zero to 23gms showing (Openaps COB). Otherwise, the sorting seems to be working well. I'm thinking the cob showing up is related to something else, though admittedly I don't know what... \
Matthias
@ceben80
Feb 21 2017 19:24
Okay, thx for clarification :clap:
katie disimone
@Kdisimone
Feb 21 2017 19:24
@JELCRAWFORD how old is your ns site? have you updated since Dec 18, 2016?
JELCRAWFORD
@JELCRAWFORD
Feb 21 2017 19:24
Good question...
I believe we have... lemme see..
katie disimone
@Kdisimone
Feb 21 2017 19:25
The COB pill display was fixed in a Dec 18th update so that it only displays what OpenAPS is giving. If your site hasn’t been updated since then, the COB pill sometimes displays careportal info and sometimes displays Openaps info…glitchy.
JELCRAWFORD
@JELCRAWFORD
Feb 21 2017 19:26
looks like version 0.9.2
@Kdisimone we used to be able to hover over and see either one - was handy when we thought the COB was wrong...
katie disimone
@Kdisimone
Feb 21 2017 19:27
Yes, you can still see it when hovering.
JELCRAWFORD
@JELCRAWFORD
Feb 21 2017 19:27
I think husband turned off NS display
katie disimone
@Kdisimone
Feb 21 2017 19:27
But, the fix makes it so that the static (not hovered over) display defaults to openaps always.
JELCRAWFORD
@JELCRAWFORD
Feb 21 2017 19:28
oops - hover wasn't working on downstairs computer which is slowly dying.. works fine on office computer. So yes, I think we are on current
@Kdisimone the weird remaining blip is the sudden 'appearance' of COB which was previously zero - we see this in the middle of the night (or at least our son's middle of the night! Usually after 10 or 11pm). Scott did a change to do some sorting of treatments which has fixed some other odd behaviour with COB. So we are part way fixed :)
Andy Sharrow
@dramageek
Feb 21 2017 19:52

Trying to track down a (system!) fatal bug on my rig. It's done it a few times now, and I'm stumped on where to look next.
The rig stops looping but I usually don't notice it til later, due to my backup rig taking over.
oref0-online logs go silent at 18:12:02. Syslog keeps running, but it's so full of minor errors I'm not sure what to look for. Pump-loop.log keeps running, but appears to be VERY intermittent- there's a few spots where there's an hour-and-a-half between loop runs.

Connecting via UART gets me a giberish login prompt- all the correct characters are there, but they're interspersed with static-looking giberish characters. Logging in and issuing a reboot works and resolves the giberish, so it's not a client-side connection thing, but a symptom. The reboot brings the system back up to full operation.

Any hints where to look?

dmdfreak
@dmdfreak
Feb 21 2017 19:53
Just setup my night scout site in anticipation of building my openAPS loop when my stuff all gets here. Is there a way to get the last 2 months of numbers imported from xdrip+ into the night scout site? And is there a good reason to do it if you can?
Dana Lewis
@danamlewis
Feb 21 2017 20:14
@dmdfreak yes because it'll be great for future research and useful for before/after comparisons of your looping
dmdfreak
@dmdfreak
Feb 21 2017 20:16
@danamlewis how do I import the old data in? It's only giving me readings from when I connected it
Dana Lewis
@danamlewis
Feb 21 2017 20:17
I don't know the answer for the how-to x drip part, just the second question :smile:
dmdfreak
@dmdfreak
Feb 21 2017 20:18
OK thanks. We are now in the same boat LOL
Lukasz Kaniowski
@lukasz-kaniowski
Feb 21 2017 20:26
hi guys, did anyone had this error after flashing jubilinux failed systemd-modules-load.service
?
David
@david-t1
Feb 21 2017 20:33
@ecc1 it has been a long way around to realize my phone won't work. I did try uncommenting the patchram line and I assume this is why bt tethering worked when I used my wife's Windows phone. As for my Samsung Galaxy S5, I cannot unrecommend it enough! I tried enabling tethering via the weblink you found for me but apparently this only enabled wifi tethering, not bt tethering. I tried rooting the phone but all attempts failed. Looks like Samsung + At&t put an encrypted bootloader that blocks any attempt to root or roll the software back from Marshmallow to Lollipop or KitKat. So it is back to openxshareble for me because I don't want to hotspot.
Matthias
@ceben80
Feb 21 2017 20:39
@danamlewis q to new profile json in autotune directory, the loop is using this file or it's copied after daily autotune run into /settings?
peterfulvi
@peterfulvi
Feb 21 2017 20:59

Some Bluetooth tehering questions. Last night i followed the docs and got my phone to paired to my rig, i trusted it, ran ifconfig bnep0 ip adrees showed up, copied the ip address, shut the wifi of on my phone and SSH'd to my rig, Yay!!Then i went for a walk to my house tried to access my rig through my phone, it said i wasn not connected to a network, then when i got a chance i SSH'D back into my rig and ran if config benp0' and it said it did not exist, now when i look for that adress its not there anymore, So went back to the drawing board and tried to make another network connectonsudo bt-pann client ect` and i keep gettig an error message which is this:

root@petersedison:~# sudo bt-pan client b6:e4:20:f7:1f:54
Traceback (most recent call last):
  File "/usr/local/bin/bt-pan", line 228, in <module>
    if __name__ == '__main__': sys.exit(main())
  File "/usr/local/bin/bt-pan", line 188, in main
    dev_remote = find_device(opts.remote_addr, devs.values()[0])
  File "/usr/local/bin/bt-pan", line 53, in find_device
    return find_device_in_objects(get_manager().GetManagedObjects(), device_address, adapter_pattern)
  File "/usr/local/bin/bt-pan", line 68, in find_device_in_objects
    raise BTError('Bluetooth device not found')
__main__.BTError: Bluetooth device not found

So any help would be great. Im going to have at it again tonight

sorry did not close the last code out sorry
John Benjamin
@Bender1061
Feb 21 2017 22:58
@peterfulvi when you got back home, the wifi connected again and oref0-online turned off the bluetooth. That's normal behavior. What you needed to find out, is when you go for your walk, are you still updating to nightscout? that's the big question. Figuring out how to ssh into your rig while on bluetooth is a little harder, but I believe the docs has the information on how to do that.
sorry, does not turn off bluetooth, it just turns off benp0. due to bt-pan turn off network.
katie disimone
@Kdisimone
Feb 21 2017 23:01
@peterfulvi we simply set up ssh app with multiple connection points. If rig is on home wifi, our phone is likely on home wifi too…so we setup that host. If rig is on phone’s BT, we have another host for that. I have noticed that every once in awhile, my rig gets assigned a new IP address from the home router and my old ssh won’t work anymore. I have to rescan and get the new IP address.
renegadeandy
@renegadeandy
Feb 21 2017 23:36
Hi everyone! I come here as a person seeking to use OpenAPS and Nightscout. I currently have a Medtronic Paradigm Veo 554. It has software version 3.1E 1.1 0B 0B. I believe in order to use OpenAPS I need to have firmware 2.6 (says 2.4 in docs but I have heard 2.6 in Europe works). I am a software engineer - Very new to this - but want to understand what attempts at making this work on 3.1 of my pump are, and if there was any hope of it working or if all routes have been exhausted. Can anyone help me find a suitable next step forward? Sorry for the long intro but had to set the scene!
katie disimone
@Kdisimone
Feb 21 2017 23:39
@renegadeandy Nobody has found a way to downgrade the firmware version on the pumps. The newer firmware have security features that prevent temp basals from being set by remote commands like our existing diy solutions provide. Sorry…but for the foreseeable future, you need to find a compatible pump like the models/firmware listed.
renegadeandy
@renegadeandy
Feb 21 2017 23:43
Do remote bolus still work? I know that's not how OpenAPS works but presumably that interface is still open?