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

29th
Jan 2017
Dana Lewis
@danamlewis
Jan 29 2017 00:03
ok, it’s now in dev
so now we can get cracking on final testing of dev before we do a release to master..
katie disimone
@Kdisimone
Jan 29 2017 00:11
I really was just doing a copy&paste cheat sheet when you posted that
and i really had just added that line
Dana Lewis
@danamlewis
Jan 29 2017 00:12
heh, good timing :)
Tim Street
@tim2000s
Jan 29 2017 00:12
@scottleibrand Just left a comment on openaps/oref0#356
Scott Leibrand
@scottleibrand
Jan 29 2017 00:13
@tim2000s maybe with @cjo20's new mmtune-old stuff we can run mmtune more often without being as intrusive...
Chris Oattes
@cjo20
Jan 29 2017 00:13
@tim2000s Have you put more logging in pump-loop to work out why it takes 6 minutes yet?
Tim Street
@tim2000s
Jan 29 2017 00:20
No. I think that's more down to the number of errors Pre an mmtune and successful run meaning there is additional data to parse.
Chris Oattes
@cjo20
Jan 29 2017 00:21
huh?
Tim Street
@tim2000s
Jan 29 2017 00:21
I really need to add a timing point to the logging.
Essentially, going through the log I get no successful runs without a preceding mmtune and it does one of those every four attempts (more or less).
Chris Oattes
@cjo20
Jan 29 2017 00:25
@tim2000s sudo apt-get install locate && sudo updatedb && locate handlers/stick.py
Scott Leibrand
@scottleibrand
Jan 29 2017 00:45
jaylagorio
@jaylagorio
Jan 29 2017 00:57
Looks like g4-raw worked!
Dana Lewis
@danamlewis
Jan 29 2017 00:57
:+1: :clap:
jaylagorio
@jaylagorio
Jan 29 2017 00:59
Would be cool to get all that great stuff listed in that issue, I'd like to try BT tethering out now that I've broken it somehow
Dana Lewis
@danamlewis
Jan 29 2017 01:03
in g4, or in the dev to master release notes?
jaylagorio
@jaylagorio
Jan 29 2017 01:05
In general. I had it working a couple weeks ago with notes that I think were marked alpha and I've done something that broke it. I tried troubleshooting earlier today and then saw some chatter about new directions. I haven't tried those yet. All I know is that right now the device is paired with the phone and I can connect while in bluetoothctl but that's about it.
Dana Lewis
@danamlewis
Jan 29 2017 01:05
ok. try the new instructions (currently in phase 4 section). working well & reproduced successfully for many people, as of late last night.
that’s in dev, and is part of what will go to master in this release
Dana Lewis
@danamlewis
Jan 29 2017 01:07
yep
jaylagorio
@jaylagorio
Jan 29 2017 01:07
I'll give that a try!
Dana Lewis
@danamlewis
Jan 29 2017 01:07
:+1:
Eric Jensen
@elnjensen
Jan 29 2017 01:26
Hi all - I'm working openaps/oref0#327 (getting autotune to run on a Mac). I've gotten past the obvious errors (commands that just don't work, e.g. working around the different date implementation on OS X) and now I'm trying to understand whether my output makes any sense, or whether there are still underlying errors. I suspect the latter... For one, I get output like this:
Limiting ISF of -Infinity to 41.67 (which is pump ISF of 50 / 1.2 )
p50deviation: -0.505 p50BGI 0 p50ratios: -Infinity Old ISF: 50 fullNewISF: -Infinity newISF: 41.667
totalMealCarbs: 124 totalDeviations: 245.73999999999995 fullNewCSF: 1.98 newCSF: 4.698
I suspect that -Infinity is not, in fact, the correct ISF for me. :wink:
jaylagorio
@jaylagorio
Jan 29 2017 01:28
You're either cured, or the worst off of all of us. Numbers are hard so I don't know which one.
Eric Jensen
@elnjensen
Jan 29 2017 01:28
Has anyone posted anywhere the full output of an autotune run, e.g. a gist with all of the screen output from autotune (maybe only for a day or two)? That might help me see if my output makes sense as it's running. There are some funny-looking things in there, but I don't know what I should be seeing along the way. Is it normal to see lines like this:
start carb absorption
0 mealCOB: 45.0 mealCarbs: 15 basalBGI: -0.1 BGI: 0.0 at 2017-01-27T13:12:03.000Z dev: -1.50 avgDelta: -1.50 csf
0 mealCOB: 90.0 mealCarbs: 30 basalBGI: -0.1 BGI: 0.0 at 2017-01-27T13:17:03.000Z dev: -1.75 avgDelta: -1.75 csf
0 mealCOB: 120.0 mealCarbs: 30 basalBGI: -0.1 BGI: 0.0 at 2017-01-27T13:22:03.000Z dev: -1.76 avgDelta: -1.75 csf
0 mealCOB: 150.0 mealCarbs: 30 basalBGI: -0.1 BGI: 0.0 at 2017-01-27T13:27:03.000Z dev: -1.76 avgDelta: -1.75 csf
0 mealCOB: 180.0 mealCarbs: 30 basalBGI: -0.1 BGI: 0.0 at 2017-01-27T13:32:04.000Z dev: -2.76 avgDelta: -2.75 csf
0 mealCOB: 210.0 mealCarbs: 30 basalBGI: -0.1 BGI: 0.0 at 2017-01-27T13:37:03.000Z dev: -2.51 avgDelta: -2.50 csf
0 mealCOB: 240.0 mealCarbs: 30 basalBGI: -0.1 BGI: 0.0 at 2017-01-27T13:42:03.000Z dev: -0.76 avgDelta: -0.75 csf
1 mealCOB: 270.0 mealCarbs: 30 basalBGI: -0.1 BGI: 0.0 at 2017-01-27T13:47:03.000Z dev: 0.74 avgDelta: 0.75 csf
1 mealCOB: 300.0 mealCarbs: 30 basalBGI: -0.1 BGI: 0.0 at 2017-01-27T13:52:03.000Z dev: 2.24 avgDelta: 2.25 csf
1 mealCOB: 330.0 mealCarbs: 30 basalBGI: -0.1 BGI: 0.0 at 2017-01-27T13:57:04.000Z dev: 2.99 avgDelta: 3.00 csf
1 mealCOB: 360.0 mealCarbs: 30 basalBGI: -0.1 BGI: 0.0 at 2017-01-27T14:02:03.000Z dev: 1.99 avgDelta: 2.00 csf
1 mealCOB: 390.0 mealCarbs: 30 basalBGI: -0.1 BGI: 0.0 at 2017-01-27T14:07:03.000Z dev: 1.98 avgDelta: 2.00 csf
1 mealCOB: 420.0 mealCarbs: 30 basalBGI: -0.1 BGI: 0.0 at 2017-01-27T14:12:03.000Z dev: 1.73 avgDelta: 1.75 csf
1 mealCOB: 450.0 mealCarbs: 30 basalBGI: -0.1 BGI: 0.0 at 2017-01-27T14:17:03.000Z dev: 0.48 avgDelta: 0.50 csf
0 mealCOB: 480.0 mealCarbs: 30 basalBGI: -0.1 BGI: 0.0 at 2017-01-27T14:22:03.000Z dev: -0.77 avgDelta: -0.75 csf
@jaylagorio :smile: Yes, I couldn't tell whether -Infinity was good or bad...
Is it right that the mealCOB just keeps increasing there? How is it different from mealCarbs?
Dana Lewis
@danamlewis
Jan 29 2017 01:30
@elnjensen heh :smile: that's the first time I've seen it say that. Check the OpenAps/autotune Gitter channel - there's been some people pasting output, and someone who's online might be able to paste more for you to see
Eric Jensen
@elnjensen
Jan 29 2017 01:30
Thx @danamlewis - didn't realize there was a separate channel. I'll have a look there. Thanks!
Dana Lewis
@danamlewis
Jan 29 2017 01:31
:+1:
jaylagorio
@jaylagorio
Jan 29 2017 01:48
@danamlewis Got to the testing section and hit an IO error in BlueZ:
root@tinyaps:~# bt-pan client CC:F3:A5:7B:D2:CE                                                                                                          
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 196, in main                                                                                                        
    try: iface = net.Connect(opts.uuid)                                                                                                                  
  File "/usr/lib/python2.7/dist-packages/dbus/proxies.py", line 145, in __call__                                                                         
    **keywords)                                                                                                                                          
  File "/usr/lib/python2.7/dist-packages/dbus/connection.py", line 651, in call_blocking                                                                 
    message, timeout)                                                                                                                                    
dbus.exceptions.DBusException: org.bluez.Error.Failed: Input/output error
John Benjamin
@Bender1061
Jan 29 2017 01:50
@scottleibrand I think we made a mistake on that new MDT push, this is what I'm seeing
Profile less than 60m old
glucose.json newer than temp_basal.json
Temp refreshError, retrying
Starting pump-loop at Sat Jan 28 20:48:46 EST 2017:
Radio ok, Listening: .
it's just restarting after echoing retrying, it does not do the mmtune at all
@tim2000s I'm seeing that exact same type of error you are, I tried making some changes so I can see where the error is happening and it's when it calls the monitor-cgm
Eric
@ecc1
Jan 29 2017 01:54
@jaylagorio This is where I'm stuck, too. Curious whether you see a "Could not bring up bnep0" error message in /var/log/syslog
jaylagorio
@jaylagorio
Jan 29 2017 01:57
As a matter of fact, I can!
Jan 28 20:56:07 localhost bluetoothd[14035]: bnep: Could not bring up bnep0: Cannot assign requested address(99)                                         
Jan 28 20:56:07 localhost bluetoothd[14035]: connect failed Input/output error                                                                           
Jan 28 20:56:07 localhost avahi-daemon[370]: Withdrawing workstation service for bnep0.
Previously I've been able to have bnep0 get an IP address and talk to the Internet and everything, but I didn't have a working loop. Now I have a working loop and no bnep0. This is not an OpenAPS problem, this is a Me+*nix problem.
Eric
@ecc1
Jan 29 2017 01:59
So that's where it's failing, but I don't know why. It means that the Edison successfully handhakes with the phone and is all ready to create the BT PAN connection on its side, but then can't create the network interface.
jaylagorio
@jaylagorio
Jan 29 2017 02:00
Yeah, I've gotten bluetoothctl and the phone's Bluetooth interface to both say they're connected and happy, I just don't know what this error means except that I think it happens after the BT connection is established.
John Benjamin
@Bender1061
Jan 29 2017 02:03
@scottleibrand never mind, it does appear to be working, I'll keep watching though
Jarred Yaw
@jyaw
Jan 29 2017 02:10
I am now using a confirmed jubilinux mac (courtesy of @fowlerbenjamin , recent openaps user)... here's hoping it works for my edisons :smile:
kenstack
@kenstack
Jan 29 2017 02:26
Im rflashing a new edison to jubi and for some reason I cant get past A start job is running for LSB: Raise network interface …. it goes over 60 seconds and hangs, then reboots, then same thing … Ive let it reboot many times .. Ive seen this before with other edisons Ive flashed but usually after 1 or 2 reboots it gets past that line - anyone have a workaround? thanks
Jarred Yaw
@jyaw
Jan 29 2017 02:43
could you just modify the boot file that starts networking until you get it reflashed?
kenstack
@kenstack
Jan 29 2017 02:46
@jyaw houw would I do that? How do I edit files on the edison from the boot prompt? thanks!!
Jarred Yaw
@jyaw
Jan 29 2017 02:50
@kenstack thats a good question.... Just an idea. I was thinking you might already have an initial OS on there... and you'd edit like the rc.whatever files (maybe rc.local? this is the point where I google)
others may have seen that issue... I have seen all the other ones, but not that one :smile:
Scott Leibrand
@scottleibrand
Jan 29 2017 02:52
@kenstack not sure, but I would try re-flashing.
RedCyclist1
@RedCyclist1
Jan 29 2017 02:53
I tried to flash my edison board and it hung on me and never finished. Now it won't show up in my console window or device manager. Any ideas what to do next?
kenstack
@kenstack
Jan 29 2017 02:58
@scottleibrand ugh - 3rd flash was the charm apparently …. I dont get this flash process - so flaky … Ive used the same method for multiple edisons .. thanks @jyaw was a good idea but I have no idea how tro edit the files before they are installed from the img file - good idea though
Jarred Yaw
@jyaw
Jan 29 2017 03:03
1 of 2 flashed successfully! My imac at home is ancient and terrible. haha.
RedCyclist1
@RedCyclist1
Jan 29 2017 03:15
Does anyone know how to flash with recovery option?
RedCyclist1
@RedCyclist1
Jan 29 2017 03:32
@jyaw Is it common for these boards not to flash? What do you do if it doesn't work? Do I need to buy everything again since I can't get mine to work?
Jarred Yaw
@jyaw
Jan 29 2017 03:33
...the boards are fine. It should mostly have to do with the computer, equipment (cables, etc) and the steps you use to do the flash.
Brian Rabinovitz
@bdr1177
Jan 29 2017 03:45
@RedCyclist1 I had the same problem with the first Edison I tried to flash. I bought it from amazon and was able to return it for a replacement.
RedCyclist1
@RedCyclist1
Jan 29 2017 04:09
@bdr1177 Hey, how did you know it was the board?
Scott Leibrand
@scottleibrand
Jan 29 2017 04:12
I think he was talking about his Edison, not his Explorer board
RedCyclist1
@RedCyclist1
Jan 29 2017 04:16
Right, my bad, how did you know your edison was bad?
Aaron Michelson
@Aaybob
Jan 29 2017 06:20
@RedCyclist1 You open the command prompt, navigate to the Flash directory you've got to be there. You try different cable to the OTG connector. Most of the cables that come with Dexcom have worked well for me. Short is better than long. It can take a lot of retrys. see my video: https://www.youtube.com/watch?v=btU771DIpB8
Matthias
@ceben80
Jan 29 2017 06:29
@danamlewis Morning :wink: - regarding openaps/oref0#358 I just recognize it's been merged to dev, so nothing needs to be tested at this moment, right?
or it's sufficient give you a feedback when I'm running the setup the next time?
Aaron Michelson
@Aaybob
Jan 29 2017 06:36
@kenstack did you try removing power and letting it reboot until successful?. I remember that happening in my early trys. I just kept trying somehow and it started working. And I was sure at one point that it was a failed Edison.
Aaron Michelson
@Aaybob
Jan 29 2017 06:48
I'm not so sure I'm liking the resetting of my goal BG to 200+. What file do I edit to see what is going on?
That trace cut gets hardwired rig to run great.
Scott Leibrand
@scottleibrand
Jan 29 2017 06:53
@Aaybob "the resetting of my goal BG to 200+"? can you be more specific as to what you're seeing?
Aaron Michelson
@Aaybob
Jan 29 2017 06:54
blob
Scott Leibrand
@scottleibrand
Jan 29 2017 06:56
I'm not seeing anything there that would support that statement. can you scroll back to a point where you think it made the wrong decision, and mouse over the OpenAPS pill to see why it decided that?
Aaron Michelson
@Aaybob
Jan 29 2017 07:00
How does one "scroll back"?
Scott Leibrand
@scottleibrand
Jan 29 2017 07:01
click or drag along the little BG graph at the bottom
Aaron Michelson
@Aaybob
Jan 29 2017 07:06
Wow, that is informative!! I'm seeing good decisions.
How would I spend more time in range?
Scott Leibrand
@scottleibrand
Jan 29 2017 07:09
which advanced features do you have turned on?
Aaron Michelson
@Aaybob
Jan 29 2017 07:11
blob
Here's a wierd one. I have all the features turned on, I answered yes at install.
Scott Leibrand
@scottleibrand
Jan 29 2017 07:12
what do you see as weird there?
it looks like it was trying to set a maxSafeBasal temp at the time, but couldn't enact for some reason.
do you have autotune enabled?
and if so, have you looked at the recommendations?
Aaron Michelson
@Aaybob
Jan 29 2017 07:15
yes, autotune and there are supposed to be three values in a file....what file is "recommendations"?
Scott Leibrand
@scottleibrand
Jan 29 2017 07:16
something like egrep "\." ~/myopenaps/autotune/autotune_recommendations.log is a good way to display them
Aaron Michelson
@Aaybob
Jan 29 2017 07:18
ISF | 27.000 | 28.960
CSF | n/a | 2.160
Carb Ratio | 15.000 | 13.407
Scott Leibrand
@scottleibrand
Jan 29 2017 07:19
so it looks like it's trying to lower your carb ratio... do you think your carb ratio might be too high?
Aaron Michelson
@Aaybob
Jan 29 2017 07:20
I think those are great recommendations.
Scott Leibrand
@scottleibrand
Jan 29 2017 07:20
hmm, from your NS site it looks like your pump is actually using a carb ratio of 10...
Aaron Michelson
@Aaybob
Jan 29 2017 07:22
multiple ratios
Scott Leibrand
@scottleibrand
Jan 29 2017 07:22
ah, that could be a problem for autotune
it assumes you have a single ratio for the whole day, and tunes accordingly
Aaron Michelson
@Aaybob
Jan 29 2017 07:23
why is CSF na?
Scott Leibrand
@scottleibrand
Jan 29 2017 07:23
openaps/oref0#326 is the issue for supporting ISF and IC ratio schedules
CSF is not directly captured from the pump. it's instead calculated as the ratio of ISF to CR
but autotune also prints out the calculation, so you do see it in the output column
what does your CR schedule look like? does it start at 15 at midnight, and then go lower for breakfast?
Aaron Michelson
@Aaybob
Jan 29 2017 07:27
the doctor sets that for me....but I think I moved it to 10 in frustration....then when I'm active (all week) I ignore and take less. It was 15 at midnight then 10 at 5am. but I want to try 13 all day.
He had set four different time values but I update them as needed.
Scott Leibrand
@scottleibrand
Jan 29 2017 07:30
ok. whatever you set at midnight is what oref0 will use, and tune. the others will only be used by your pump
so if you think 13 is a good all day number, setting that at midnight (even if it changes right afterward) is a good start
Aaron Michelson
@Aaybob
Jan 29 2017 07:31
I think doctor had 29 and I moved it to 27. Guess I'll go with 29.
Oh, okay....thinking like a computer pancreas. :smile:
Scott Leibrand
@scottleibrand
Jan 29 2017 07:32
that one might not be as important to change just yet. it's a fairly minor difference (much less than the difference between 15 and 10 CR), and moving it up from 27 to 29 would actually make things less aggressive, which might not be what you want right now
but either way, 28.96 is what oref0 is using right now, so changing ISF on your pump wouldn't matter unless it's more than 20% different
Aaron Michelson
@Aaybob
Jan 29 2017 07:35
I'm so glad this is working and feel so much healthier even if I'm not 65 - 100 all the time.
Scott Leibrand
@scottleibrand
Jan 29 2017 07:35
what about the basal schedule? does autotune show much in the way of changes recommended there yet?
I'm guessing it's saying your basals are too low...
Aaron Michelson
@Aaybob
Jan 29 2017 07:38
00:00 | 0.900 | 0.810
01:00 | | 0.824
02:00 | | 0.691
03:00 | | 0.690
04:00 | | 0.839
05:00 | | 0.987
06:00 | | 1.080
07:00 | | 1.080
08:00 | | 0.964
09:00 | | 0.972
10:00 | | 0.868
11:00 | | 0.820
12:00 | | 0.630
13:00 | | 0.641
14:00 | | 0.849
15:00 | | 0.970
16:00 | | 0.962
17:00 | | 0.924
18:00 | | 0.838
19:00 | | 0.761
20:00 | | 0.747
21:00 | | 0.978
22:00 | | 1.080
23:00 | | 1.073
I just don't have a fixed schedule these days. There's little hope for a decent basal.
Scott Leibrand
@scottleibrand
Jan 29 2017 07:40
ok. looks like it wants to raise it at some times and lower it in others (quite a bit in a few cases).
Aaron Michelson
@Aaybob
Jan 29 2017 07:42
raise by .17, lower by 2.7, 13 lower, 11 higher
Scott Leibrand
@scottleibrand
Jan 29 2017 07:42
from your basal graph it looks like autosens is also detecting resistance and trying to compensate for it somewhat. what do you see in pump-loop.log for Adjusting basal, Adjusting target_bg, and Adjusting sens ?
Aaron Michelson
@Aaybob
Jan 29 2017 07:45
Adjusting basal from 0.99 to 1.2
Adjusting target_bg from 100 to 93
Adjusting sens from 25.286 to 21.1
Carb Impact: 15 mg/dL per 5m; CI Duration: 0 hours
Accel. Carb Impact: 10 mg/dL per 5m; ACI Duration: 0 hours
while BG is on an upswing from chocolate
without bolus
Scott Leibrand
@scottleibrand
Jan 29 2017 07:46
did you enter the carbs for the chocolate?
if you're not consistently entering your carbs, you'll want to start doing that, so AMA knows what's going on and can high-temp accordingly, and so autotune knows how to attribute the resulting BG deviations
you can put them into the bolus wizard, and just do a 0.1U bolus if you don't want to bolus for them for whatever reason. or you can enter them via Care portal, or Pebble, or IFTTT, or whatever
Aaron Michelson
@Aaybob
Jan 29 2017 07:48
what gets entered for hi protein meal?
Scott Leibrand
@scottleibrand
Jan 29 2017 07:49
if it's mixed, I would just enter the carbs. if it's almost all protein, then different people do different things, but they usually enter some carbs to stand in for the protein, but don't bolus for it, just let AMA high-temp based on the entered carbs and the BG rise.
Aaron Michelson
@Aaybob
Jan 29 2017 07:50
I keep remembering being told chocolate has so much fat that BG won't rise fast even though it doesn't work that way.
Scott Leibrand
@scottleibrand
Jan 29 2017 07:50
heh, we treat chocolate like straight sugar
maybe dark chocolate would be different, but milk chocolate is really sugary
Aaron Michelson
@Aaybob
Jan 29 2017 07:52
yup, but I quit eating dark chocolate because of cafine, details details
Scott Leibrand
@scottleibrand
Jan 29 2017 07:52
it looks like autosens is maxing out at 1.2x. I wonder if it might be worthwhile to raise your autosens_max to 1.3 or so.
given your highly variable schedule, and inability to dial in a basal rate that's correct for high vs. low activity level, you might want to give autosens a bit more flexibility to adjust, if you're comfortable with what it's doing so far.
Aaron Michelson
@Aaybob
Jan 29 2017 07:54
My doctor had to write me a letter to get a new job. He said, "aaron has had problems with low blood sugar....." I may not get the job.....soon will know
Scott Leibrand
@scottleibrand
Jan 29 2017 07:55
another thing you should consider is using Eating Soon temporary targets. basically, any time you're planning to eat within the next 60-90 minutes, or any time your BG is too high and you want oref0 to be more aggressive, you can set a temporary target to 80 and get some extra insulin in play faster.
Aaron Michelson
@Aaybob
Jan 29 2017 07:56
I'd thought of doing that. And I love when I remember before eating.
Scott Leibrand
@scottleibrand
Jan 29 2017 07:57
if you don't remember until closer to the meal, you can also do an eating soon bolus: a correction bolus targeting 80 instead of your normal target.
Aaron Michelson
@Aaybob
Jan 29 2017 07:58
that would be a head calculation?
Scott Leibrand
@scottleibrand
Jan 29 2017 07:59
you could use the bolus wizard and then add to it
there are a few ways to do it, depending. maybe the best approach would be to assume that oref0 is doing its thing and already giving you enough insulin to get down to 100-110 or so. if you want to get down to 80 instead, that would just be 1U of extra insulin
Aaron Michelson
@Aaybob
Jan 29 2017 08:01
love that write-up....doctor got into pre-bolus some time ago like as if it was new and foolproof.
Scott Leibrand
@scottleibrand
Jan 29 2017 08:02
so if you take that approach, the eating soon bolus for your ISF would be about 1U, given any time up to 60m before eating
then you can use your bolus wizard normally for the actual meal bolus
anyway, that's probably enough to try for now. let us know what works well, and what you're still having trouble with after a few days or a week or so, and we might have more ideas. :)
Aaron Michelson
@Aaybob
Jan 29 2017 08:03
:+1:
alternateal
@alternateal
Jan 29 2017 08:09
Has anybody worked out whether it is possible to take the G5 data used in Loop from an iPhone to an OpenAPS rig running standalone (probably with mLab and NS on it) to have an iPhone friendly off-line rig??
Scott Leibrand
@scottleibrand
Jan 29 2017 08:11
I know some folks (maybe @ceben80 or @Bender1061, as well as @SandraK82 and a few others, I can't recall) were working on setting up a NS instance to run on an offline rig, but I don't know if anyone has tried making Loop upload to it...
Matthias
@ceben80
Jan 29 2017 08:14
I think @C-Ville has good experience regards with ns running on a rig and looping offline
alternateal
@alternateal
Jan 29 2017 08:14
I haven't managed to make the leap to programming Swift so will keep my ears open then - all of the data elements are there so feels like a moving the data issue :smile:
Scott Leibrand
@scottleibrand
Jan 29 2017 08:17
and @j4cbo is also looking at doing some work to make it possible to talk to a G5 from a rig with BLE
alternateal
@alternateal
Jan 29 2017 08:18
👍
Matthias
@ceben80
Jan 29 2017 09:06
@scottleibrand @danamlewis I remember Dana posted here a rtd link what needs to be done exactly for 712 pumps, adjustment to gatheretc...
Would be happy you can point to me the link - Thanks!!!
Matthias
@ceben80
Jan 29 2017 09:09
blob
@scottleibrand :cry: ....sorry for this stupid q I didn't bookmark as she posted here
andyhandy
@andyhandy
Jan 29 2017 09:10
Hi All. Quick question. I only have openaps and nightscout running for my son at night but would like to have all historical CGM and pump data on nightscout. How would I upload all that data to nightscout when I set up openaps at night? Thanks
Scott Leibrand
@scottleibrand
Jan 29 2017 09:11
what kind of CGM?
andyhandy
@andyhandy
Jan 29 2017 09:11
G5
Scott Leibrand
@scottleibrand
Jan 29 2017 09:11
do you have the Share bridge set up inside NS?
or, how are you uploading the G5 data?
pump data should be automatically backfilled when you plug in. is it not getting the whole day?
andyhandy
@andyhandy
Jan 29 2017 09:12
yes i have share bridge set up
Scott Leibrand
@scottleibrand
Jan 29 2017 09:13
so you should already have 24x7 CGM data in NS then?
andyhandy
@andyhandy
Jan 29 2017 09:14
maybe i don't have it set up then
i did have share bridge set up on the old (manual) script
but then re-ran the new automated version of the setup script
andyhandy
@andyhandy
Jan 29 2017 09:15
maybe need to set it up again?
Scott Leibrand
@scottleibrand
Jan 29 2017 09:16
Share bridge is a NS thing, not at all related to OpenAPS.
andyhandy
@andyhandy
Jan 29 2017 09:17
yes ok but definitely only getting CGM data when openaps and nightscout are on overnight
during the day it is just the pump and G5 receiver and that data is not showing up on my nightscout site. I have to go back over bolus history and input manually into NS for some reason.
at night i have an iPhone acting as the G5 receiver
as well as the normal G5 receiver being plugged into the PI
Scott Leibrand
@scottleibrand
Jan 29 2017 09:19
During the day is G5 data being uploaded to Share? Can you see it with Follow?
Sounds like maybe no phone during the day?
I believe that oref0-setup only pulls like 4-6h of pumphistory for uploading to NS. it also pulls 24h for autosens though, so you could modify it to upload that.
andyhandy
@andyhandy
Jan 29 2017 09:21
yes no phone during the day, only at night
Scott Leibrand
@scottleibrand
Jan 29 2017 09:21
Ok that makes more sense. :smile:
andyhandy
@andyhandy
Jan 29 2017 09:22
yes that's how daycare was trained (on G4 with non-phone receiver) so we have stuck with it
Scott Leibrand
@scottleibrand
Jan 29 2017 09:23
I just merged oref0 G4 CGM uploading into dev. You could probably adapt that to also upload historical CGM data from the receiver, but I don't think it will work without some modifications to not pull raw.
andyhandy
@andyhandy
Jan 29 2017 09:24
ok i'll take a look thanks
Scott Leibrand
@scottleibrand
Jan 29 2017 09:24
Or you could send the phone to school with him and have them ignore it. :smile:
andyhandy
@andyhandy
Jan 29 2017 09:24
yes that is what i tried today on the weekend
and is probably the way to go
so that is the CGM issue taken care of
but what did you mean by"it also pulls 24h for autosens though, so you could modify it to upload that."
i have autosens running on the latest dev version
Scott Leibrand
@scottleibrand
Jan 29 2017 09:25
There is a pumphistory-24h-zoned.json file on your rig
You would just need to modify the upload reports to upload from that instead of from pumphistory-zoned.json I think.
andyhandy
@andyhandy
Jan 29 2017 09:26
oh ok i see
Scott Leibrand
@scottleibrand
Jan 29 2017 09:26
Or create a new upload report and alias
andyhandy
@andyhandy
Jan 29 2017 09:26
that makes sense
yes i'll try that
Scott Leibrand
@scottleibrand
Jan 29 2017 09:26
The 24h history doesn't get pulled as often
So if you just change it you won't see recent treatments for like 20m
andyhandy
@andyhandy
Jan 29 2017 09:27
right got it
thanks very much
Scott Leibrand
@scottleibrand
Jan 29 2017 09:27
Np
andyhandy
@andyhandy
Jan 29 2017 09:28
part of the reason i am interested in sorting it out is to use autotune properly which if i understand pulls data from NS?
Scott Leibrand
@scottleibrand
Jan 29 2017 09:35
Yes. And you really need complete data for it to do anything useful.
tepidjuice
@tepidjuice
Jan 29 2017 10:08
@ceben80 Do you have an offline version of openaps running with xdrip? I'd like to move to an offline set up and @danamlewis mentioned you had one running with xdrip.
Matthias
@ceben80
Jan 29 2017 10:10
@tepidjuice yes I have, just easy using dev and the parameter --cgm=xdripduring the setup, more infos here http://openaps.readthedocs.io/en/latest/docs/walkthrough/phase-1/offline-looping-and-monitoring.html?highlight=xdrip#xdripaps-for-offline-bgs
tepidjuice
@tepidjuice
Jan 29 2017 10:22
@ceben80 Looks great thanks :smile: one more question it says: "See xDrip instructions for seeing offline loop status (coming soon)" what are these and how soon are they coming?
"
As I don't own a pebble and knowning what openaps is doing is helpful.
but do not use httpswhen setting up REST upload, must be http :exclamation:
Sandra Keßler
@SandraK82
Jan 29 2017 10:26
@tepidjuice @scottleibrand I'm working on integrating a "smart proxy" for ns module in my libre app. It's not yet ready and i need to debug further before releasing.
tepidjuice
@tepidjuice
Jan 29 2017 10:28
@SandraK82 So this would allow for offline monitoring of the AP?
@ceben80 How do you know what basal decisions your rig is making?
Sandra Keßler
@SandraK82
Jan 29 2017 10:32
@tepidjuice it is more or less a full ns server without the GUI which syncs its data to a full ns Server once it gets a wifi connection. Therefore you can truly do offline looping and still get the whole data into ns
Matthias
@ceben80
Jan 29 2017 10:32
@tepidjuice http://openaps.readthedocs.io/en/latest/index.html should be read before looping, very good documentation....
btw....the loop log is located here regarding your q tail -F /var/log/openaps/pump-loop.log
tepidjuice
@tepidjuice
Jan 29 2017 10:37
@ceben80 Yes I've read those or at least a previous version. I have a working loop. Because I'm often without internet access I'd like to get it offline. What I'm after is something like the data visualisation given by NS for my online loop i.e. quickly seeing basal rates , IOB, COB... etc.
Matthias
@ceben80
Jan 29 2017 10:38
@tepidjuice okay misunderstood, your q regarding a offline loop :wink:
in case you have no internet I ssh to the corresponding log and make a cator head
Tim Street
@tim2000s
Jan 29 2017 10:40
@SandraK82 - don't you have a work offline NS install that could be used for offline monitoring? I know it doesn't upload to NS, but it might suit @tepidjuice
andyhandy
@andyhandy
Jan 29 2017 10:40
hi what does this symbol mean next to the BGL in nightscout? ⇼︎
Matthias
@ceben80
Jan 29 2017 10:40
@tepidjuice here a few examples.....
blob
it's just a workaround, but for me it's working fine only in the case no internet
Sandra Keßler
@SandraK82
Jan 29 2017 10:44
@tim2000s @tepidjuice yeah there is some ns local setup script in my github repo. Don't know how good it currently works as I merged various pr and had no time for proper testing
tepidjuice
@tepidjuice
Jan 29 2017 10:46
@ceben80 no problem. ssh is a bit of a work around but so long as I can review what's happening that's better than having no APS running while I'm offline. I guess writing a visualisation app could work.
Sandra Keßler
@SandraK82
Jan 29 2017 10:47
But be aware, even if the script runs smooth a pi2 will be very busy with running ns alone, you will need then a pi for ns and one for loop
@C-Ville runs such a local Setup on Edison
tepidjuice
@tepidjuice
Jan 29 2017 10:48
@SandraK82 thanks I might check it out after I get the offline rig going.
Sandra Keßler
@SandraK82
Jan 29 2017 10:49
@tepidjuice if I can release my module soon you should be able to connect to it via any ns follower app
tepidjuice
@tepidjuice
Jan 29 2017 10:53
Thanks for all the help and info everyone.
Oh sorry one more question @ceben80 what ssh app do you use?
Matthias
@ceben80
Jan 29 2017 10:56
JuiceSSH and regarding the above Raspi light (works for edision as well)
Tim Street
@tim2000s
Jan 29 2017 14:10
@Bender1061 I agree. I've had a chance this morning to monitor properly and the "use pump read_temp_basal" fails following a "use CGM iter_glucose_hours 24.0". It's something about the loop process though as when I run thos commands consecutively manually, they function with no problems.
As far as I can tell, given the way the aliases work, the latter shouldn't be crossing over with the former as they are in n serial.
C-Ville
@C-Ville
Jan 29 2017 14:39
@tim2000s @tepidjuice I do have NS and mongo both running on my edison together with OpenAPS. There is are caveats though.
a) the edison only has limited space and with mongodB and NS space is scarce. My mongodB currently has 450 MB after 4 months of looping.
b) I did not get NS 0.9.3 running on my Edison. I am still running 0.9.0 dev-3 a fairly old version
c) when switching between home and mobile (cell phone hotspot) the IP Adresses change, therefore I have in xdrip REST API while mobile and MondoDB REST configured as upload while at home.
While this is a neat setup, as NS is a nice place for checking status, Urchin only works for me either at home or mobile. The other problem is, that I do not know anyone else running this setup, which makes it more difficult if you need help. currently the most appealing option for looping offline seems the setup @colinlennon came up with
Chris Oattes
@cjo20
Jan 29 2017 16:09
@tim2000s sudo apt-get install locate && sudo updatedb && locate handlers/stick.py, please
Scorponok
@Scorponok
Jan 29 2017 16:11

I woke up this morning and found the raspberry pi that runs openaps giving the following error:

Starting pump-loop at Sun 29 Jan 09:22:18 EST 2017:
pump://JSON/read_temp_basal/monitor/temp_basal.json
reporting 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 52, in __call__
    self.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 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
Error, retrying
Starting pump-loop at Sun 29 Jan 09:22:26 EST 2017:

I tried restarting the pi and running oref0-reset-git, but that didn't change anything. Does anyone have any idea how to fix this? I commented out the call to create_git_commit but that doesn't sound like a good long-term solution.

Tim Street
@tim2000s
Jan 29 2017 17:29
@cjo20, already on the latest version of locate
sudo apt-get install locate && sudo updatedb && locate handlers/stick.py
Reading package lists... Done
Building dependency tree       
Reading state information... Done
locate is already the newest version.
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
/home/.rootfs/usr/local/lib/python2.7/dist-packages/mmeowlink/handlers/stick.py
/home/.rootfs/usr/local/lib/python2.7/dist-packages/mmeowlink/handlers/stick.pyc
/root/mmeowlink-source/mmeowlink/mmeowlink/handlers/stick.py
Chris Oattes
@cjo20
Jan 29 2017 17:53
sudo grep "=12" /usr/local/lib/python2.7/dist-packages/mmeowlink/handlers/stick.py
sudo grep "=12" /root/mmeowlink-source/mmeowlink/handlers/stick.py
Tim Street
@tim2000s
Jan 29 2017 17:55
~# sudo grep "=12" /usr/local/lib/python2.7/dist-packages/mmeowlink/handlers/stick.py
root@ed210:~# sudo grep "=12" /root/mmeowlink-source/mmeowlink/handlers/stick.py
grep: /root/mmeowlink-source/mmeowlink/handlers/stick.py: No such file or directory
Chris Oattes
@cjo20
Jan 29 2017 17:55
sudo grep "=12" /root/mmeowlink-source/mmeowlink/mmeowlink/handlers/stick.py
Tim Street
@tim2000s
Jan 29 2017 17:56
# sudo grep "=12" /root/mmeowlink-source/mmeowlink/mmeowlink/handlers/stick.py
root@ed210:~#
Chris Oattes
@cjo20
Jan 29 2017 18:00
@tim2000s I think it's been suggested a few times before, but you need to get the fixes I did to mmeowlink, which improves comms stability
Tim Street
@tim2000s
Jan 29 2017 18:01
I ran the install for them yeterday
*yesterday
Chris Oattes
@cjo20
Jan 29 2017 18:02
The changes I made includes a line going from ack_wait_seconds=20 to ack_wait_seconds=12 in handlers/stick.py. Neither of the copies of that file on your edison have the text =12 in them, which means that you haven't got the latest changes
Tim Street
@tim2000s
Jan 29 2017 18:02
More to the point, I deleted the mmeowlink source directory, re-cloned from Git and then ran python setup.py develop on it
According to git, your updates are in master but have not been included in the instal
Tim Street
@tim2000s
Jan 29 2017 18:08
interesting. It's been updated since yesterday.
Chris Oattes
@cjo20
Jan 29 2017 18:09
The latest release has. It's been in master for a week
Tim Street
@tim2000s
Jan 29 2017 18:09
Yes - I thought I'd pulled it yesterday
Chris Oattes
@cjo20
Jan 29 2017 18:09
did you pull from the main repo, or your clone of it
Tim Street
@tim2000s
Jan 29 2017 18:09
(after a conversation with @scottleibrand yesterday morning)
The main repo
Ran
cd ~
git clone https://github.com/oskarpearson/mmeowlink.git mmeowlink-source
cd mmeowlink-source
git checkout master
python setup.py develop
Tim Street
@tim2000s
Jan 29 2017 19:33
@cjo20 Yes - thanks - definitely fixed it.
Any ideas why my attempts to update yesterday might have failed?
Aaron Michelson
@Aaybob
Jan 29 2017 19:42
How do I fix "ntpd failed"? I rebooted. I replaced the pump battery.
Dana Lewis
@danamlewis
Jan 29 2017 19:44
@Aaybob by the way - did you know you can get email notifications from github? Just remembered seeing something In the my -first-pr changes that made me think you may not have those turned on. They're nice because when someone comments on your Pr, it emails it directly to you.
You can go into GitHub profile, settings, and change that
Aaron Michelson
@Aaybob
Jan 29 2017 19:50
Its an email account I don't often check....
I get tonnes of junk and it makes email pretty useless
Aaron Michelson
@Aaybob
Jan 29 2017 19:55
How has NTP stopped doing its job? Is there a minor install I can run?
Was it from a pump battery going dead? Its been replaced.
Dana Lewis
@danamlewis
Jan 29 2017 19:57
Ok. Just wanted to point that out.
@logichammer q when you get a chance...what did you ever end up doing re: ssh tunneling to remote rigs? Did you find tunnel as a service somewhere, or ?
Aaron Michelson
@Aaybob
Jan 29 2017 19:57
If I see it in email I still don't know how to respond.
Dana Lewis
@danamlewis
Jan 29 2017 19:59
@Aaybob depends on what it is. But there's a view-in-github link in every email so you can open it and decide if you need to take action. If someone comments on a pr and requests changes, it generally requires action before it can move forward. If the email just says the PR was merged, you don't have to do anything.
Aaron Michelson
@Aaybob
Jan 29 2017 20:02
I see email on iPhone...My most recent Git email said the serial link always works even after a failed flash. I had added text to say it doesn't because that has been my experience.
So I guess serial is blank until you flash again....maybe that would be helpful to write or maybe just more confusing...
PieterGit
@PieterGit
Jan 29 2017 20:05
My activity stream on gitter has not been updated since january 26th. (last update is taylorfowler forked). Do others have that problem too?
katie disimone
@Kdisimone
Jan 29 2017 20:21
root@edisonhost:~# tail -f /var/log/syslog
Jan 29 12:13:26 localhost systemd[444]: Starting Timers.
Jan 29 12:13:26 localhost systemd[444]: Reached target Timers.
Jan 29 12:13:26 localhost systemd[444]: Starting Sockets.
Jan 29 12:13:26 localhost systemd[444]: Reached target Sockets.
Jan 29 12:13:26 localhost systemd[444]: Starting Basic System.
Jan 29 12:13:26 localhost systemd[444]: Reached target Basic System.
Jan 29 12:13:26 localhost systemd[444]: Starting Default.
Jan 29 12:13:26 localhost systemd[444]: Reached target Default.
Jan 29 12:13:26 localhost systemd[444]: Startup finished in 95ms.
Jan 29 12:13:26 localhost systemd[1]: Started User Manager for UID 0.
Jan 29 12:17:01 localhost CRON[461]: (root) CMD (   cd / && run-parts --report /etc/cron.hourly)
Jan 29 12:18:14 localhost wpa_supplicant[275]: wlan0: WPA: Group rekeying completed with 00:15:ff:d6:96:c3 [GTK=CCMP]
my tail is showing this, is it stuck?
actually, i should say, it is stuck…why?
ok, looks like my cron is wrong...
RBJervey
@RBJervey
Jan 29 2017 20:34
Hey, so I'm in the process of setting up my openAPS but still very early (I've flashed the Edison and maybe managed to connect it to the internet). But my Dexcom just gave me a warning that my old G4 is dying. I need to call them tomorrow to order a new one. Is there any reason to prefer G5 to G4? I know G5 is slightly more accurate but the more frequent dropped numbers and the fact that it has a scheduled shut down bothers me. But I wanted to check with people who know / use openAPS, is there a good reason to go for the G5?
Tim Street
@tim2000s
Jan 29 2017 20:45
@RBJervey for me it's a stuff to carry benefit. With G5 you can use xDrip+ on your phone If using android and have your OpenAPS tethered to the phone.
Dana Lewis
@danamlewis
Jan 29 2017 20:45

@RBJervey glad you're asking here, too. For the reasons you have mentioned, I have stuck w G4 share and been very happy w accuracy for DIY closed looping with it now for 2+ years.

(Some people do like G5 because no receiver, but given the issues with dropped BGs, unless you are dying to ditch the receiver, I don't think it's worth it. Keeping G4 also gives you some more flexibility in options for offline looping, too.)

And high five for flashing your Edison! That's the hardest part :smile:
C-Ville
@C-Ville
Jan 29 2017 20:47
@RBJervey G4 with Xbridge and Xdrip on your cell is also pretty good and can be very tiny
Matthias
@ceben80
Jan 29 2017 20:52
@RBJervey The price of G4 is significantly higher than the G5
Dana Lewis
@danamlewis
Jan 29 2017 20:53
@ceben80 that depends on where you are. The price stacks almost equal if you truly compare 1 year of a G4 transmitter vs having to buy 2 sets of 6 months of G5 transmitter, etc
Matthias
@ceben80
Jan 29 2017 20:53
in Germany 4 times higher
jaylagorio
@jaylagorio
Jan 29 2017 20:54
@RBJervey I use a G4 with the receiver, since I can't have my phone on me at times the receiver is the only way to go. There's a lot of support for it in things other than OpenAPS like in xDrip on Android and there's a Nightscout uploader for Windows, both can talk to it via USB and Bluetooth.
Matthias
@ceben80
Jan 29 2017 20:55
....I love my G4 with the tiny micro xDrip box, u know it from twitter @danamlewis :wink:
jaylagorio
@jaylagorio
Jan 29 2017 20:55
@danamlewis's not kidding, flashing the Edison is the biggest hurdle. Congrats!
Tim Street
@tim2000s
Jan 29 2017 21:00
Using G5 with xDrip is 2x cost of G4 in U.K.
Matthias
@ceben80
Jan 29 2017 21:18
@tim2000s You can buy the G5 without sensors?
In Germany not possible :worried:
Tim Street
@tim2000s
Jan 29 2017 21:19
Yes. G5 can be bought on its own but you only get 1 transmitter and it die after 160 days using xDrip.
Dana Lewis
@danamlewis
Jan 29 2017 21:20
If you're calculating standalone price, make sure you're doing true apples to apples, and not one apple to five apples ;)
Matthias
@ceben80
Jan 29 2017 21:23
and due to the higher pricing most german insurances won't cover the G5
Aaron Michelson
@Aaybob
Jan 29 2017 21:26
:smile : mmeowlink.exceptions.CommsException: Could not get subg_rfspy state or version. Have you got the right port/device and radio_type?
..........
The last BG data was read at Sun Jan 29 2017 08:21:15 GMT-0800 (PST) but your system time currently is Sun Jan 29 2017 13:17:45 GMT-0800 (PST)
..........
Waiting for ntpd to synchronize... OK!
This keeps repeating....I'm going to try re-install dependencies on.... to see if that will stop
Aaron Michelson
@Aaybob
Jan 29 2017 21:31
The following packages will be upgraded:
libssl-dev libssl-doc libssl1.0.0 openssl tcpdump
5 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
1 not fully installed or removed.
Errors were encountered while processing:
bluez
E: Sub-process /usr/bin/dpkg returned an error code (1)
But that's Okay....
Aaron Michelson
@Aaybob
Jan 29 2017 21:40
From git://github.com/openaps/oref0
f6ce55a..42b71bc dev -> origin/dev
7c45386..e4dcfb1 autotune-mdi -> origin/autotune-mdi
  • [new branch] cgm-loop-git -> origin/cgm-loop-git
  • [new branch] mmtune-speedup -> origin/mmtune-speedup
  • [new branch] smb -> origin/smb
    Updating f6ce55a..42b71bc
    error: Your local changes to the following files would be overwritten by merge:
     bin/bt-pan
     bin/oref0-autotune-export-to-xlsx.py
     bin/oref0-bluetoothup.sh
     bin/oref0-dexusb-cgm-loop.py
     bin/oref0-find-ti.sh
     bin/oref0-subg-ww-radio-parameters-timeout.sh
     bin/oref0-subg-ww-radio-parameters.sh
     bin/peb-urchin-status.sh
    Please, commit your changes or stash them before you can merge.
    Aborting
    HELP what? how do I fix that?
Aaron Michelson
@Aaybob
Jan 29 2017 21:47
Tried this :smile:
root@red:~/myopenaps# git pull git://github.com/openaps/oref0.git
error: refs/heads/master does not point to a valid object!
error: refs/heads/master does not point to a valid object!
error: refs/heads/master does not point to a valid object!
remote: Counting objects: 5888, done.
remote: Compressing objects: 100% (4/4), done.
remote: Total 5888 (delta 0), reused 0 (delta 0), pack-reused 5884
Receiving objects: 100% (5888/5888), 993.14 KiB | 299.00 KiB/s, done.
Resolving deltas: 100% (3906/3906), done.
error: refs/heads/master does not point to a valid object!
fatal: bad object HEAD
error: git://github.com/openaps/oref0.git did not send all necessary objects
WTF? is wrong with GIT
Let me know if you read this and have an answer....I've got to start new from flash
Dana Lewis
@danamlewis
Jan 29 2017 21:55
Git stash
Tim Street
@tim2000s
Jan 29 2017 22:01
Also , make# sure your on the latest version of mmeowlink.
kenstack
@kenstack
Jan 29 2017 22:04
@scottleibrand been trying to get bluetooth tehering to work - it paris and trusts no problem, but when I try to get bnep0 up it fails -sudo bt-pan client aa:bb:cc:etc producesTraceback (most recent call last):
File "/usr/bin/bt-pan", line 228, in <module>
if name == 'main': sys.exit(main())
File "/usr/bin/bt-pan", line 196, in main
try: iface = net.Connect(opts.uuid)
File "/usr/lib/python2.7/dist-packages/dbus/proxies.py", line 145, in call
**keywords)
File "/usr/lib/python2.7/dist-packages/dbus/connection.py", line 651, in call_blocking
message, timeout)
dbus.exceptions.DBusException: org.bluez.Error.Failed: Input/output error
have you seen that before?
bnep0 is not up
Dana Lewis
@danamlewis
Jan 29 2017 22:07
@kenstack that was the problem we had constantly before we got it to work... Probably next try reboot, shut down blue tooth, etc. try all the testing instructions again
kenstack
@kenstack
Jan 29 2017 22:08
shut bluetooth on phone and edison or just edison - thanks!
Dana Lewis
@danamlewis
Jan 29 2017 22:08
All Edison
You can toggle Bluetooth on the phone if it makes you feel better too, but it's just the Edison side that really needs it. Also make sure your phone hotspot switch is on (but don't let anything connect to it) while you're at it
kenstack
@kenstack
Jan 29 2017 22:09
also Im using an iphone 6 … has anyone gotten it to run on a iphone 6 using ios 10.1.1? Just wondering if it could be a version issue - will reboot and killa nd retry a few times - for clarification what actually gets bnep0 up and running? the bt-pan?
thanks!
Dana Lewis
@danamlewis
Jan 29 2017 22:11
I'm in a 6s w older firmware, fwiw. Bt-pan creates bnep0, dhclient creates ip
kenstack
@kenstack
Jan 29 2017 22:12
ok cool thanks @danamlewis
Eric
@ecc1
Jan 29 2017 22:12
@kenstack I'm stuck on this too. bt-pan triggers the client PAN connection, but it's bluetoothd that actually tries to create the bnep0 interface (and is probably failing with "bnep: Could not bring up bnep0: Cannot assign requested address(99)" in syslog, right?)
kenstack
@kenstack
Jan 29 2017 22:14
@ecc1 note I had to add hciconfig hci0 up before did the sudo hciconfig hci0 name $HOSTNAME - wondering if Im doing soemthing wrng before then ...
@ecc1 interesing errors in my syslog Jan 29 17:01:53 localhost bluetoothd[2411]: a2dp-source profile connect failed for aa:bb:cc:dd:ee:ff: Protocol not available
Jan 29 17:01:54 localhost bluetoothd[2411]: bnep: Could not bring up bnep0: Cannot assign requested address(99)
Jan 29 17:01:54 localhost bluetoothd[2411]: connect failed Input/output error
Jan 29 17:01:54 localhost avahi-daemon[346]: Withdrawing workstation service for bnep0.
Dana Lewis
@danamlewis
Jan 29 2017 22:18
@kenstack assuming you're protecting your MAC address, but just checking to make sure you did on your side sub in iPhone info?
Eric
@ecc1
Jan 29 2017 22:18
Yes, I see the same. Pretty sure the a2dp is a red herring, and the avahi-daemon one comes from bnep0 being created and then destroyed again once the address assignment fails.
kenstack
@kenstack
Jan 29 2017 22:19
@danamlewis yes :) but always chec to make sure Im not an idiot .. its a fair concern ;)
OK from a different attempt @ecc1 Jan 29 16:59:24 localhost bluetoothd[2292]: Bluetooth daemon 5.37
Jan 29 16:59:24 localhost bluetoothd[2292]: Starting SDP server
Jan 29 16:59:24 localhost bluetoothd[2292]: Bluetooth management interface 1.3 initialized
Jan 29 16:59:24 localhost bluetoothd[2292]: Failed to obtain handles for "Service Changed" characteristic
Jan 29 16:59:24 localhost bluetoothd[2292]: Not enough free handles to register service
Jan 29 16:59:24 localhost bluetoothd[2292]: Error adding Link Loss service
Jan 29 16:59:24 localhost bluetoothd[2292]: Not enough free handles to register service
Jan 29 16:59:24 localhost bluetoothd[2292]: Not enough free handles to register service
Jan 29 16:59:24 localhost bluetoothd[2292]: Not enough free handles to register service
Jan 29 16:59:24 localhost bluetoothd[2292]: Current Time Service could not be registered
Jan 29 16:59:24 localhost bluetoothd[2292]: gatt-time-server: Input/output error (5)
Jan 29 16:59:24 localhost bluetoothd[2292]: Not enough free handles to register service
Jan 29 16:59:24 localhost bluetoothd[2292]: Not enough free handles to register service
Jan 29 16:59:24 localhost dbus[354]: [system] Activating via systemd: service name='org.freedesktop.hostname1' unit='dbus-org.freedesktop.hostname1.service'
Jan 29 16:59:24 localhost bluetoothd[2292]: Failed to read advertising features: Unknown Command (0x01)
Jan 29 16:59:24 localhost bluetoothd[2292]: hci0 Load Connection Parameters failed: Unknown Command (0x01)
Jan 29 16:59:24 localhost systemd1: Starting Hostname Service...
Jan 29 16:59:24 localhost systemd[2293]: Failed at step NAMESPACE spawning /lib/systemd/systemd-hostnamed: No such file or directory
Jan 29 16:59:24 localhost systemd1: systemd-hostnamed.service: main process exited, code=exited, status=226/NAMESPACE
Jan 29 16:59:24 localhost systemd1: Failed to start Hostname Service.
Jan 29 16:59:24 localhost systemd1: Unit systemd-hostnamed.service entered failed state.
Jan 29 16:59:49 localhost dbus[354]: [system] Failed to activate service 'org.freedesktop.hostname1': timed out
Eric
@ecc1
Jan 29 2017 22:22
Could someone who has it working please post the ifconfig output for their bnep0 interface? I wonder whether there's some weird conflict with other addresses in use on mine.
Dana Lewis
@danamlewis
Jan 29 2017 22:23
There's an example in the docs
Eric
@ecc1
Jan 29 2017 22:24
Got it, thanks. I'm using Android, but clearly this problem hits (some) users of both platforms, so that's unlikely to be it.
Dana Lewis
@danamlewis
Jan 29 2017 22:25
Hmm, k
kenstack
@kenstack
Jan 29 2017 22:34
@danamlewis that same issue above - not enough handles etc - happens immediately on a reboot as well - so its never really running I think
@ecc1 do you see the not enough handles problem when you start the bluetooth daemon?
Eric
@ecc1
Jan 29 2017 22:40
I see some of those too, but pretty sure the Bluetooth comms part itself is working fine. My rig is talking to my G4 Share receiver regularly over BLE, and the bnep0 issue happens after the Bluetooth-level protocol exchange occurs.
Aaron Michelson
@Aaybob
Jan 29 2017 22:40
is master now with ble and other new stuff?
kenstack
@kenstack
Jan 29 2017 22:41
@ecc1 huh ….
Eric
@ecc1
Jan 29 2017 22:41
I don't have the luxury of a "reflash and reinstall" since I'm not using oref0, but if you do, you might as well try it since it worked for Dana.
kenstack
@kenstack
Jan 29 2017 22:42
@ecc1 same boat ….. my loop is built on othr utlities ive written over the years ...
Eric
@ecc1
Jan 29 2017 22:42
I think the phrase "hoisted by our own petard" applies, then
kenstack
@kenstack
Jan 29 2017 22:42
@ecc1 yes :) well said :)
Dana Lewis
@danamlewis
Jan 29 2017 22:43
Hehehe
@Aaybob no, Bluetooth tethering is in dev. Hopefully to master this week perhaps...
kenstack
@kenstack
Jan 29 2017 22:51
@ecc1 here is something Ive noticed - if I run bt-pan in one sesion while also watching bluetoothctl I get a quick connected yes, followed about 2 sec later by a connected no and then I get the error from bt-pan - not sure if that helps
fezulin
@fezulin
Jan 29 2017 22:55
@scottleibrand You helped me with my Edi-cannot-reboot-after-flashing-Issue. Just for your information: I visited another Looper today: After some tests we found out, that my Edi does boot, if it is plugged on her MiniBreakout-Board. So I guess I destroyed the MiniBreakot-Board while I soldered the cables for connection to TI-Stick on it.
Dana Lewis
@danamlewis
Jan 29 2017 22:56
aw. bummer, but glad the Edison is still usable!
C-Ville
@C-Ville
Jan 29 2017 22:57
@danamlewis @fezulin Edisons seem to be able to take a lot :-)
Eric
@ecc1
Jan 29 2017 23:00
@kenstack Yes, I see the same. I'm going to try sniffing the BT exchange when it's successful (like from my desktop machine) versus the one from the Edison to see if I can figure out what's different.
kenstack
@kenstack
Jan 29 2017 23:02
I can force my ihpne to connect from its bluetooth menu, but even when I do the bt-pan still fails ...
jaylagorio
@jaylagorio
Jan 29 2017 23:14
Can confirm seeing the same behavior from my Edison and bt-pan. The syslog shows some errors when you try and make it work.
kenstack
@kenstack
Jan 29 2017 23:29
@jaylagorio @ecc1 I tried to bt-pan to a paired android as well - no dice - same result
I may try installing bluez again
Aaron Michelson
@Aaybob
Jan 29 2017 23:29
DEV install won't make a cron file......three tries...
Dana Lewis
@danamlewis
Jan 29 2017 23:30
@Aaybob “dev install won’t make a cron file” do you have an existing cron in that directory? or are you building a new directory with dev branch?
Aaron Michelson
@Aaybob
Jan 29 2017 23:30
all new edison from the start
Dana Lewis
@danamlewis
Jan 29 2017 23:30
If you can’t get it to work, switch back to master
ok. I would start w/ master to have a working rig to start
jaylagorio
@jaylagorio
Jan 29 2017 23:32
@kenstack I attempted to do that but it didn't seem to work. There's definitely the chance that I didn't remove everything correctly before trying to reinstall though.
kenstack
@kenstack
Jan 29 2017 23:32
how did you remove it?
jaylagorio
@jaylagorio
Jan 29 2017 23:34
I walked the manual process back by trying to find all the files that were in /etc/system, deleting the binaries where I could find them, and then rerunning oref0-setup with the MAC address of my phone
kenstack
@kenstack
Jan 29 2017 23:36
@jaylagorio thanks … ugh :)
Tim Street
@tim2000s
Jan 29 2017 23:38
Blimey @jaylagorio I think it may be quicker to reflash an Edison!
jaylagorio
@jaylagorio
Jan 29 2017 23:41
It may have been, but at this point I have everything except that working so that isn't a great option. Before I had everything else working I only had Bluetooth tethering working and didn't need to reflash! I don't know what I did to blow it up, but that's the way it goes I guess.
Dana Lewis
@danamlewis
Jan 29 2017 23:43
@jaylagorio remind me, when did you last flash your edison?
it’s newish, right?
jaylagorio
@jaylagorio
Jan 29 2017 23:46
It is newish, I got it on Christmas and flashed it maybe a week or two after?
It has the Jubilinux distro recommended in the docs
Dana Lewis
@danamlewis
Jan 29 2017 23:48
probably fine…but if nothing else cracks it (including following the new instructions for bluetooth 2-3 times), maybe try flashing one more time. I wasn’t able to get mine to work until I reflashed, and the new instrcutions. Admittedly with much older flashes of Edisons, though, so just don’t know when “recently flashed” matters for bluetooth
jaylagorio
@jaylagorio
Jan 29 2017 23:50
Sad days for me - I cut the trace on my Explorer board, so reflashing isn't an option anymore.
Not that I'm complaining, it works great offline and I don't mind putting my phone in Wi-Fi hotspot mode occiasionally.
Aaron Michelson
@Aaybob
Jan 29 2017 23:52
Is this? Could not parse input data: { [Error: Cannot find module '/root/myopenaps/monitor/glucose.json'] code: 'MODULE_NOT_FOUND' }
An unknown problem?
Dana Lewis
@danamlewis
Jan 29 2017 23:54
@jaylagorio ah, yes, that.
@Aaybob what step are you in the process? when are you getting that output, after running what command? completely out of context will not enable anyone to help troubleshoot with you.