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

1st
Jan 2016
schulzedj
@schulzedj
Jan 01 2016 00:57
Did some debugging and was able to get this output pi@raspberrypi:~/daneopenaps $ openaps use pump set_temp_basal enact/suggested.json { "bg": 240, "temp": "absolute", "snoozeBG": 226, "recieved": true, "rate": 0.0, "reason": "Eventual BG 226>113 but Delta +3 < BGI 0 / 2; cancel", "eventualBG": 226, "timestamp": "2015-12-31T16:21:35.934071", "duration": 0, "tick": "+3"
Where can I found out what all this means? I would like to see if this "suggested" basal is in fact where it should be.
Scott Leibrand
@scottleibrand
Jan 01 2016 01:28
your current BG is 240, with a tick/delta of +3 (so you were 237 5m ago). your BG is predicted to be 226 eventually. Looks like you're running an old version of oref0, so the "Delta +3 < BGI 0" part is incorrect, and it's not high-temping you the way the new "Expected Delta" logic would.
you should be able to upgrade oref0 by doing sudo npm install -g oref0: can you run that and paste the output?
diabeticgonewild
@diabeticgonewild
Jan 01 2016 03:47
In order to have a dedicated connection to a hotspot and a dedicated access point, you will likely need 2 Edimax USB wifi adapters. If you need a tutorial, please see: http://raspberrypi.stackexchange.com/questions/12276/how-to-make-pi-an-ap-with-2-wifi-adapters
diabeticgonewild
@diabeticgonewild
Jan 01 2016 04:26
So how do you deal with the read_current_history_pages command?
Like for example, right now I am using a basal "patterns" profile, so that would likely be needed for reading my current basal profile.
Or is it just read_basal_profile_std on the 522 pump?
no, it isn't
Scott Leibrand
@scottleibrand
Jan 01 2016 04:28
even better: there's a read_current_basal_profile or something like that
diabeticgonewild
@diabeticgonewild
Jan 01 2016 04:28
I didn't see that. Remember I don't really read...
Scott Leibrand
@scottleibrand
Jan 01 2016 04:29
no worries
diabeticgonewild
@diabeticgonewild
Jan 01 2016 04:29
it's read_selected_basal_profile
diabeticgonewild
@diabeticgonewild
Jan 01 2016 04:42
how do you get the battery profile? I tried openaps use pump status and openaps use pump read_status only to get the same results with no battery status...
Scott Leibrand
@scottleibrand
Jan 01 2016 04:42
I think there's just a battery use
$ openaps use pump -h | grep battery
read_battery_status
Check battery status.
diabeticgonewild
@diabeticgonewild
Jan 01 2016 04:44
oh
Rachel Sandlain
@audiefile
Jan 01 2016 04:44
openaps use pump read_battery_status
diabeticgonewild
@diabeticgonewild
Jan 01 2016 04:44
It wasn't in the openaps/openaps directory
yeah I will use grep...that's always useful
thanks
Scott Leibrand
@scottleibrand
Jan 01 2016 04:49
i think battery is a new requirement for the openaps nightscout pill. isn't required for basic loop
diabeticgonewild
@diabeticgonewild
Jan 01 2016 04:49
Yeah, I invoked the report. It's not hard to do...
Thanks though. I appreciate it.
Tony Zarro
@tzarro
Jan 01 2016 04:52
Hey guys, been plugging away at learning and starting my implementation and had a basic question: the easy_install pulls version 0.0.8, but it sounds like I should start with the latest version (0.1.2)...I'm not too familiar with RPi2 or bash commands, is there a recommended way of updating the version? Thanks!
Scott Leibrand
@scottleibrand
Jan 01 2016 04:53
0.0.8 is openaps
0.1.2 is oref0
pi@rpi3 ~/openaps-dev $ npm show oref0 | grep version:
version: '0.1.2',
pi@rpi3 ~/openaps-dev $ openaps --version
openaps 0.0.8
Tony Zarro
@tzarro
Jan 01 2016 04:54
ahh, gotcha...thanks!
diabeticgonewild
@diabeticgonewild
Jan 01 2016 04:55
What someone needs to do is make a "checklist" to reinforce what has been created/what hasn't, but with "blank spaces" so it's not "copy and paste"....I can eventually do that...
Peter Miller
@ochenmiller
Jan 01 2016 04:58
Sitting with my family tonight, each person telling the rest what the best thing was from 2015. OpenAPS. OpenAPS. OpenAPS. OpenAPS.
Dana Lewis
@danamlewis
Jan 01 2016 04:59
Aw. :) :) :) :) happy new year everyone!
Scott Leibrand
@scottleibrand
Jan 01 2016 04:59
:-D
Tony Zarro
@tzarro
Jan 01 2016 05:01
happy new year y'all :)
diabeticgonewild
@diabeticgonewild
Jan 01 2016 05:03
where is ISF in the commands? I tried openaps use pump -h | grep sensitivity to no avail
I'm trying ISF now....
didn't get anything for that...
Happy new year
I mean the ISF instructions are pretty inconsistent with the actual commands provided, and I am pretty sure of that.
I tried read_status and settings and I didn't get anything for that...
diabeticgonewild
@diabeticgonewild
Jan 01 2016 05:08
I really cannot find the command to retrieve ISF.
It's not apparent.
Does anyone know the command to retrieve ISF?
diabeticgonewild
@diabeticgonewild
Jan 01 2016 05:13
Or rather, the argument. Not the command.
diabeticgonewild
@diabeticgonewild
Jan 01 2016 05:20
I am surprised nobody can answer this.
it's not in the openaps/openaps readme
Does anyone know what the argument is in order to read ISF from the pump?
diabeticgonewild
@diabeticgonewild
Jan 01 2016 05:27
I mean, like I went through every single command on openaps/openaps readme directory and none of them fit the profile. I even opened up all of my JSON files to see if I had ISF data and I didn't. This is extremely annoying.
Could someone please answer this? Someone here has to know.
diabeticgonewild
@diabeticgonewild
Jan 01 2016 05:35
oh, it's read_insulin_sensitivities. It's not in the openaps/openaps readme, which I am using via screenreader. This is annoying as hell.
diabeticgonewild
@diabeticgonewild
Jan 01 2016 05:45
Whatever. I'm backing up my card and going to bed. Some housekeeping needs to be done on the writeup.
I'm not doing it tonight either.
Sulka Haro
@sulkaharo
Jan 01 2016 06:51
Oops @bewest the new years broke pump history reporting for 554. The year is reported as 2000. Something rolled over?
Sulka Haro
@sulkaharo
Jan 01 2016 06:57

Enacting the temp also thrown an exception, but the command succeeds:

Jan  1 06:49:28 raspberrypi openaps-loop: Traceback (most recent call last):
Jan  1 06:49:28 raspberrypi openaps-loop:   File "/usr/local/bin/openaps-report", line 6, in <module>
Jan  1 06:49:28 raspberrypi openaps-loop:     exec(compile(open(__file__).read(), __file__, 'exec'))
Jan  1 06:49:28 raspberrypi openaps-loop:   File "/home/pi/src/openaps/bin/openaps-report", line 82, in <module>
Jan  1 06:49:28 raspberrypi openaps-loop:     app( )
Jan  1 06:49:28 raspberrypi openaps-loop:   File "/home/pi/src/openaps/openaps/cli/__init__.py", line 44, in __call__
Jan  1 06:49:28 raspberrypi openaps-loop:     self.run(self.args)
Jan  1 06:49:28 raspberrypi openaps-loop:   File "/home/pi/src/openaps/bin/openaps-report", line 75, in run
Jan  1 06:49:28 raspberrypi openaps-loop:     output = app(args, self)
Jan  1 06:49:28 raspberrypi openaps-loop:   File "/home/pi/src/openaps/openaps/cli/subcommand.py", line 50, in __call__
Jan  1 06:49:28 raspberrypi openaps-loop:     return self.method.main(args, app)
Jan  1 06:49:28 raspberrypi openaps-loop:   File "/home/pi/src/openaps/openaps/reports/invoke.py", line 40, in main
Jan  1 06:49:28 raspberrypi openaps-loop:     output = task.method(args, app)
Jan  1 06:49:28 raspberrypi openaps-loop:   File "/home/pi/src/openaps/openaps/uses/use.py", line 45, in __call__
Jan  1 06:49:28 raspberrypi openaps-loop:     output = self.main(args, app)
Jan  1 06:49:28 raspberrypi openaps-loop:   File "/home/pi/src/openaps/openaps/vendors/medtronic.py", line 311, in main
Jan  1 06:49:28 raspberrypi openaps-loop:     results = self.upload_program(program)
Jan  1 06:49:28 raspberrypi openaps-loop:   File "/home/pi/src/openaps/openaps/vendors/medtronic.py", line 320, in upload_program
Jan  1 06:49:28 raspberrypi openaps-loop:     return self.pump.model.set_temp_basal(**program)
Jan  1 06:49:28 raspberrypi openaps-loop:   File "/home/pi/dev/decoding-carelink/decocare/models/__init__.py", line 188, in set_temp_basal
Jan  1 06:49:28 raspberrypi openaps-loop:     result = self._set_temp_basal(**basals)
Jan  1 06:49:28 raspberrypi openaps-loop:   File "/home/pi/dev/decoding-carelink/decocare/models/__init__.py", line 26, in __call__
Jan  1 06:49:28 raspberrypi openaps-loop:     self.validate( )
Jan  1 06:49:28 raspberrypi openaps-loop:   File "/home/pi/dev/decoding-carelink/decocare/models/__init__.py", line 17, in validate
Jan  1 06:49:28 raspberrypi openaps-loop:     data = self.response.getData( )
Jan  1 06:49:28 raspberrypi openaps-loop:   File "/home/pi/dev/decoding-carelink/decocare/commands.py", line 273, in getData
Jan  1 06:49:28 raspberrypi openaps-loop:     received = True if self.data[0] is 0 else False
Jan  1 06:49:28 raspberrypi openaps-loop: IndexError: bytearray index out of range

How can I dump the data to debug this?

Sulka Haro
@sulkaharo
Jan 01 2016 07:14
Ah, that issue causes openaps to think the action didn't happen, when it actually works fine, so the enacted action reporting doesn't work.
Ben West
@bewest
Jan 01 2016 08:36
mm-send-comm.py tweak ReadHistoryData --page 0 --save
or mm-temp-basals.py --set --rate 0 --duration 30
Jason Calabrese
@jasoncalabrese
Jan 01 2016 09:06
seeing something strange here, loop not seeing recent treatments
but they do show in pump history
times look right
Jason Calabrese
@jasoncalabrese
Jan 01 2016 09:13
no times aren't right
recent bolus is from 2000
@sulkaharo did you find a fix?
{
    "_type": "Bolus", 
    "_description": "Bolus 2000-01-01T00:54:25 head[4], body[0] op[0x01]", 
    "timestamp": "2000-01-01T00:54:25-08:00", 
    "_body": "", 
    "appended": [
      {
        "_type": "UnabsorbedInsulinBolus", 
        "_description": "UnabsorbedInsulinBolus unknown head[14], body[0] op[0x5c]", 
        "_body": "", 
        "_head": "5c0e5027c00c53d01867d04c7bd0", 
        "data": [
          {
            "amount": 2.0, 
            "age": 39
          }, 
          {
            "amount": 0.3, 
            "age": 339
          }, 
          {
            "amount": 0.6, 
            "age": 359
          }, 
          {
            "amount": 1.9, 
            "age": 379
          }
        ], 
        "_date": ""
      }
    ], 
    "programmed": 0.9, 
    "duration": 0, 
    "amount": 0.9, 
    "_head": "01090900", 
    "type": "normal", 
    "_date": "1976400110"
a broken loop and crazy high
yay 2016
Ben West
@bewest
Jan 01 2016 09:18
hmm
there's a bit missing
look at the hexdump with mm-send-comm
think @loudnate had a fix somewhere
Jason Calabrese
@jasoncalabrese
Jan 01 2016 09:21
pump does display year as 2016
Andrew Warrington
@andrew-warrington
Jan 01 2016 09:30
hi all, Happy New Year. I'm seeing the same issue as @sulkaharo
except additionally, the command is giving completely the wrong output
Sulka Haro
@sulkaharo
Jan 01 2016 09:31
I just chained a sed find/replace as a temp fix to the pump event reporting
Andrew Warrington
@andrew-warrington
Jan 01 2016 09:31
´´´
openaps use pump iter_glucose 1
[
{
"body": "",
"packet_size": 4,
"name": "SensorSync",
"raw": "0a-48-01-10",
"date": "2000-01-01T10:08:00",
"date_type": "minSpecific",
"_tell": 8,
"op": "0x0d"
},
{
"body": "",
"packet_size": 4,
"name": "BatteryChange",
"raw": "0a-48-21-10",
"date": "2000-01-01T10:08:00",
"date_type": "minSpecific",
"_tell": 13,
"op": "0x0a"
}
´´´
Sulka Haro
@sulkaharo
Jan 01 2016 09:33
As in, running this right after pulling the history from the pump
fix-history = ! bash -c "sed -i 's/2000-/2016-/g' monitor/pump-history.json"
Andrew Warrington
@andrew-warrington
Jan 01 2016 09:33
this started around 10pm local time for me
Sulka Haro
@sulkaharo
Jan 01 2016 09:36
Doesn't help with the enact failing to report success though
Jason Calabrese
@jasoncalabrese
Jan 01 2016 09:37
and I guess i need to do that before zoning
Sulka Haro
@sulkaharo
Jan 01 2016 09:38
yes
I split the history command to two parts, running that in the middle
Andrew Warrington
@andrew-warrington
Jan 01 2016 09:43
ok, in my case the output was completely wrong only because of the battery change i had just done
Andrew Warrington
@andrew-warrington
Jan 01 2016 10:14
@sulkaharo, thank for the fix-history line above! my loop is back up and running.
Jason Calabrese
@jasoncalabrese
Jan 01 2016 10:18
think enact is still broken, but I haven't seen it try since adding the fix-history hack
Andrew Warrington
@andrew-warrington
Jan 01 2016 10:19
i'm using a custom algorithm... not a pure OpenAPS implementation.
I think that's why I'm not #21. :)
in my case I'm not relying on enact. my pump is a 754, doesn't accept commands from the Carelink.
so... I've hacked together something using a Medtronic remote control
which has required writing a control loop, etc.
(just to explain why i'm not using enact)
Jason Calabrese
@jasoncalabrese
Jan 01 2016 10:23
oh
saw an enact work, no errors in the logs, temp came back in history and was uploaded to NS
Oskar Pearson
@oskarpearson
Jan 01 2016 12:12
@sulkaharo The error that you've listed about setting the temporary basal throwing an exception is being chatted about in the gitter oskarpearson/mmeowlink channel. It might be mmeowlink specific / related to interaction between mmeowlink and decocare. I don't get the same error on my side.
Either that or it's because you and the other reporters are on a newer version of decocare than I am.
It might be worth testing to see if the problem occurs with the carelink too.
Sulka Haro
@sulkaharo
Jan 01 2016 14:34
Ack. Will test. Now at mother-in-law's eating blini. ;)
Scott Leibrand
@scottleibrand
Jan 01 2016 15:17
Also seeing 2000- here, but the loop ran all night that way and kept @danamlewis completely in range despite not knowing about IOB. :-D
Scott Leibrand
@scottleibrand
Jan 01 2016 15:23
We're using Carelink and I believe decocare 0.0.19
enact is working fine for us.
Scott Leibrand
@scottleibrand
Jan 01 2016 15:29
@diabeticgonewild sorry we were all off celebrating new year's last night. You should be able to find commands like that with openaps pump use -h. Since they're documented there, we didn't list every single one of them in the web docs. If you don't want to read through the entire help output, one trick is to | less and then use / to search through that. Or you can use grep, but just do a more general (shorter) search, for something like "sens" in this case.
Scott Leibrand
@scottleibrand
Jan 01 2016 15:45
I love the fact that our algorithm still works so well even with complete amnesia and all its actions based solely on current BG. :-)
Scott Leibrand
@scottleibrand
Jan 01 2016 16:01
Pasteable version of @sulka’s fix for my loop setup:
openaps alias add fix-history '! bash -c "sed -i \"s/2000-/2016-/g\" monitor/pumphistory.json monitor/clock.json"'
openaps alias add fix-iob "report invoke monitor/pumphistory-zoned.json monitor/clock-zoned.json monitor/iob.json"
openaps alias add gather '! bash -c "rm monitor/*; ( openaps get-bg | grep reporting && echo -n Re && ( openaps monitor-pump && openaps fix-history && openaps fix-iob ) >/dev/null && echo -n fresh && ( openaps ns-meal-carbs; echo ed: ) && ls -C monitor/ ) 2>/dev/null"' || die "Can't add gather”
Pete Schwamb
@ps2
Jan 01 2016 16:19
I’ll see if I have a solution in minimed_rf. When I was working on decoding these records originally, I remember putting the pump date far into the future to see how many bits the various records held.
Scott Leibrand
@scottleibrand
Jan 01 2016 16:24
thx
Scott Leibrand
@scottleibrand
Jan 01 2016 16:33
@bustavo just opened bewest/decoding-carelink#159
Gustavo
@bustavo
Jan 01 2016 16:33
Y16K
@scottleibrand did you also get a surprise last night?
:worried:
Scott Leibrand
@scottleibrand
Jan 01 2016 16:35
read up: our loop dealt with the lack of IOB just fine. :)
so @danamlewis and I slept right through it, and I implemented the workaround in the morning after I saw the chatter here
Gustavo
@bustavo
Jan 01 2016 16:36
I just finished with the workaround on mine
Gustavo
@bustavo
Jan 01 2016 16:36
I am doing a Time.current.year imposition on the pump date
Scott Leibrand
@scottleibrand
Jan 01 2016 16:37
that's an oref0 loop with y2k16 IOB amnesia. almost indistinguishable from normal operation. :)
Gustavo
@bustavo
Jan 01 2016 16:39
just realized had missed the 2 on the title
Scott Leibrand
@scottleibrand
Jan 01 2016 16:40
the K you mean. :)
Gustavo
@bustavo
Jan 01 2016 16:41
Yes, sorry, It’s still really early for me
Scott Leibrand
@scottleibrand
Jan 01 2016 16:42
:)
I might've gone back to sleep if not for this bug, but I still got plenty. :)
(woke up without any alarm)
Gustavo
@bustavo
Jan 01 2016 16:45
I need to get coffee, just need to finish running the tests 🚨
Pete Schwamb
@ps2
Jan 01 2016 16:45
So minimed_rf does get the right dates: https://gist.github.com/ps2/91fde96a95b7b6b2f4e0 Need to find out where in decocare it decodes that.
Scott Leibrand
@scottleibrand
Jan 01 2016 16:46
:+1:
Gustavo
@bustavo
Jan 01 2016 16:47
Hmm, @scottleibrand did it also happen to you after a calibration?
Scott Leibrand
@scottleibrand
Jan 01 2016 16:52
@bustavo no, it's a straight-up y2k16 thing AFAICT.
Peter Miller
@ochenmiller
Jan 01 2016 17:07
Aha! It's Y2K for the hex crowd!
Andrew Warrington
@andrew-warrington
Jan 01 2016 17:13
anyone else having trouble with Azure in Europe at the moment?
my web app is running, but going to my Nightscout site I get a message that it isn't running.
just wondering if Azure is having a problem and I should just wait, or if it's specific to my implementation.
Scott Leibrand
@scottleibrand
Jan 01 2016 17:23
Might ask in nightscout/public or check CGM in the Cloud
a lot of us are on custom and/or non-Azure setups.
Sulka Haro
@sulkaharo
Jan 01 2016 17:31
My recommendation would be to get out of Azure. I've helped a lot of people with issues on Azure, where stuff that just breaks for no apparent reason. Heroku OTOH has been rock solid so far.
Dragan Maksimovic
@dm61
Jan 01 2016 18:41
Hi everyone - just a quick intro: over the last couple of days, I've managed to close the loop; had it working all night for the first time last night. When I woke up everything looked great, but then I discovered the pump history json was all messed up. I see above that there is a Y2K2016-type issue... amazingly, I could not tell there was a problem just from looking at CGM and openAPS actions. Anyway, thanks everyone who is contributing to openAPS - it is pretty amazing, and I am very happy I have a chance to play with it a bit.
Scott Leibrand
@scottleibrand
Jan 01 2016 18:42
:+1:
Scott Leibrand
@scottleibrand
Jan 01 2016 19:00
there is now a fix for the y2k16 issue in decoding-carelink. if you're running from source, you can upgrade to the fixed version by running cd ~/src/decoding-carelink/ && git checkout dev && git pull && sudo python setup.py develop. thx @loudnate.
Pete Schwamb
@ps2
Jan 01 2016 19:25
The carelink stick seems to be active for a short period of time after using it with decocare. It will continue to send “ack” reply packets to packets that the pump sends out, even if it’s done doing its commands. This causes interference if something else is trying to talk to the pump within some time window after the carelink stick does. Just something to consider for those of you running multiple device setups.
I’m not sure how long the window is. Seems like it’s less than a minute.
Scott Leibrand
@scottleibrand
Jan 01 2016 19:27
interesting
I've had pretty decent success with our two-Pi setup using a 120-second sleep after any preflight fail
I figured they only interfered while actively attempting to send commands, but that is another reason to do so
Pete Schwamb
@ps2
Jan 01 2016 19:31
I was running a carelink dump history, followed by a RL dump history, and was seeing bizarre stuff in the sdr. I first thought I had some signal generation issues on the RL. But I saw some isolated responses that were clearly not the RL.
Ben West
@bewest
Jan 01 2016 19:35
@loudnate thanks for your patch
I need to release hotfix now?
dev and master are even?
err, or up to date with respect to this?
Ben West
@bewest
Jan 01 2016 19:51
version 0.0.20 of decocare is released
fixes y2k bug
Scott Leibrand
@scottleibrand
Jan 01 2016 19:53
Thx
Ben West
@bewest
Jan 01 2016 19:57
version 0.0.21 of decocare is also released bewest/decoding-carelink#163
0.0.22 is in progress
Ben West
@bewest
Jan 01 2016 20:08
ok, version 0.0.9 of openaps has been released also
it depends on version 0.0.20 of decocare by default
version 0.0.10 is in progress now
openaps/openaps#75
Scott Leibrand
@scottleibrand
Jan 01 2016 20:11
seeing lots of this:
Traceback (most recent call last):
  File "/usr/local/bin/openaps", line 6, in <module>
    exec(compile(open(__file__).read(), __file__, 'exec'))
  File "/home/pi/src/openaps/bin/openaps", line 167, in <module>
    app( )
  File "/home/pi/src/openaps/openaps/cli/__init__.py", line 44, in __call__
    self.run(self.args)
  File "/home/pi/src/openaps/bin/openaps", line 153, in run
    builtins.dispatch(args, self)
  File "/home/pi/src/openaps/openaps/builtins.py", line 48, in dispatch
    get_alias(args.command, app)(args)
  File "/home/pi/src/openaps/openaps/builtins.py", line 27, in __call__
    spec_command = shlex.split(self.spec.fields.get('command'))
  File "/usr/lib/python2.7/shlex.py", line 279, in split
    return list(lex)
  File "/usr/lib/python2.7/shlex.py", line 269, in next
    token = self.get_token()
  File "/usr/lib/python2.7/shlex.py", line 96, in get_token
    raw = self.read_token()
  File "/usr/lib/python2.7/shlex.py", line 172, in read_token
    raise ValueError, "No closing quotation"
ValueError: No closing quotation
Ben West
@bewest
Jan 01 2016 20:11
sounds like an issue in your ini
Scott Leibrand
@scottleibrand
Jan 01 2016 20:23
it's intermittent, and usually works
Scott Leibrand
@scottleibrand
Jan 01 2016 20:28
how do I tell where's it's complaining about the missing quote?
Ben West
@bewest
Jan 01 2016 20:56
Im guessing it's the alias section
pretty sure
90%
you can create a new blank openaps instance somewhere and put only the alias section in the ini
should error same way
Scott Leibrand
@scottleibrand
Jan 01 2016 20:57
ok, I'll run my aliases individually and see if any of them error
Tony Zarro
@tzarro
Jan 01 2016 21:00
Quick update and a question: I've been using the ns-uploader-setup script as a guide and have setup all of my reports, aliases and virtual devices manually...I have been testing each component individually, and everything is working great so far...except for the following:
...continued...(1) monitor/ns-glucose.json is generating a file, but it is empty. (2) when running the enact command, I get the following error (Cannot find module '/home/pi/openaps/monitor/iob.json'). I see much talk today of issues with iob, but I have installed the latest openaps and decocare, so perhaps I'm having a different issue. Setting up my aliases as they were in the ns-uploader-setup file, I can't actually see where the monitor/iob.json report would actually be invoked...is there a recommended way to invoke it as part of one of those aliases?
oh, if I invoke monitor/iob.json manually, then the enact command works fine.
Scott Leibrand
@scottleibrand
Jan 01 2016 21:10
@bewest: weird: openaps alias add loop '! bash -c "openaps preflight && openaps gather && openaps get-settings >/dev/null && openaps enact; openaps upload"' works, but openaps alias add loop '! bash -c "openaps preflight && openaps gather && openaps get-settings >/dev/null && openaps enact ; openaps upload"' doesn't. difference is the space before the ;
@tzarro: Cannot find module '/home/pi/openaps/monitor/iob.json' means that iob.json doesn't exist when something tries to reference it as input
so sounds like you just need to invoke monitor/iob.json in one of your aliases before you try to reference it
for ns-glucose.json, check to make sure you have the NIGHTSCOUT_HOST and API_SECRET values set properly in your crontab (if you're running it from there) or in your shell
if it's not just that, go ahead and start running individual aliases / reports individually and see what's failing, and then run the command manually of whichever one is failing to do what you want.
Tony Zarro
@tzarro
Jan 01 2016 21:15
@scottleibrand OK, that's what I was thinking about iob.json, just wasn't sure...for ns-glucose.json, I haven't started working with crontab yet, so I'll move to that next...thanks!
Ben West
@bewest
Jan 01 2016 21:15
hmm maybe need an oref0-tutorial tool
Scott Leibrand
@scottleibrand
Jan 01 2016 21:15
for running it in shell, you'll need to export those variables, like export NIGHTSCOUT_HOST=https://yoursite.com
probably easier to add this kind of stuff to documentation
Ben West
@bewest
Jan 01 2016 21:16
something that can call all the --config switches and walk use through
walk through the setup
it can validate as it goes
Tony Zarro
@tzarro
Jan 01 2016 21:16
@scottleibrand makes sense...haven't worked with variables yet :)
Ben West
@bewest
Jan 01 2016 21:16
you can put it in args as well
I think it's simpler that way
Scott Leibrand
@scottleibrand
Jan 01 2016 21:17
if you're going to automate setup, why not just fully automate it? if goal is go make it a learning experience, simpler isn't necessarily the goal.
Ben West
@bewest
Jan 01 2016 21:17
some things need to be provided
sI'm saying have a wizard walk kthrough that
like ns host, api secret etc
serial
setting up and validating or3f0
Scott Leibrand
@scottleibrand
Jan 01 2016 21:18
good docs should make it very straightforward for someone who learns/understands what they're doing, but look very daunting to someone who's not prepared to really DIY. wizards often have the opposite effect
Ben West
@bewest
Jan 01 2016 21:18
oref0 can take different inputs
people are having ok time coming up with the inputs
it's the subsequent steps... en ensuring actually invoking reports to pull in data
with ble transports, it'll be slifghtly more complicated
Scott Leibrand
@scottleibrand
Jan 01 2016 21:20
would love to get some more input from @tzarro @dm61 and everyone else who's new to openaps and setting up for the first time: what do you think is unnecessarily difficult, in a way that's not really teaching you anything and should just be automated?
yeah, better troubleshooting tools would be good
Ben West
@bewest
Jan 01 2016 21:20
this is how git and lots of other tools evolve
right now we still have pretty low level tools
need to keep augmenting with more and more higher level abstractions
scottleibrand @scottleibrand nods
Ben West
@bewest
Jan 01 2016 21:21
as we do that, itll be easier as well, so long as it's still easy to debug at every level
with git you can manipulate heads directly, but you probably just want to rebase or branch or merge, or something high level
Tony Zarro
@tzarro
Jan 01 2016 21:27
@scottleibrand I'll be honest...it has actually turned out to be easier than I had feared...I wasn't sure I'd get anywhere close, but I kept reading and trying simple things in my shell connection, and learned how things work...I've never had any experience with linux systems before, no coding experience, and only cursory experience trying to read code, but your guides were VERY constructive in my learning...I was pessimistic for the first few days trying to learn, but as I made more "connections" regarding how things worked, I kept moving faster and faster...I'll think about what could have been done easier and with less time , I've got some notes I can review from setting things up :)
Dana Lewis
@danamlewis
Jan 01 2016 21:27
@tzarro Awesome. Would love PRs and more thoughts as you reflect back.
Tony Zarro
@tzarro
Jan 01 2016 21:28
@danamlewis sure thing...what's a PR? :)
Dana Lewis
@danamlewis
Jan 01 2016 21:29
Pull request :) I.e. Adding to or editing the docs in Github
Tony Zarro
@tzarro
Jan 01 2016 21:30
@danamlewis ah, gotcha! will do :)
Scott Leibrand
@scottleibrand
Jan 01 2016 21:31
:+1: glad to hear they worked so well for you @tzarro, and looking forward to your contributions. :-)
Aaron Michelson
@Aaybob
Jan 01 2016 21:41
Introducing myself. I just ordered the parts list. Ordered pi kit through vilros on amazon. Medtronic guest account wouldn't place order for two dongles so I used personal account. Put new battery in old 522, setting for five to ten years. Nothing. Removed and put back in and worked great. Needed the tip of holding finger on reservoir plunger to exit refill screen.
Dana Lewis
@danamlewis
Jan 01 2016 21:44
Welcome @Aaybob
eszcloud
@eszcloud
Jan 01 2016 21:53
can someone remind me the command to check which version of openaps is running?
openaps --version
Ben West
@bewest
Jan 01 2016 22:06
sudo easy_install -ZU openaps should update everything
eszcloud
@eszcloud
Jan 01 2016 22:10
with that, I'm getting AttributeError: 'module' object has no attribute 'packaging'
Dragan Maksimovic
@dm61
Jan 01 2016 22:18
@scottleibrand , I'd concur with @tzarro. My background is in engineering, but not in software development. Nevertheless the process of setting up a working closed loop was not too difficult - it took me about half a day to put the system together and get to the point when I can use the pump and the CGM, and another day to close the loop. The most annoying parts for me were related to the MM CareLink stick: the range is small, and it apparently hangs up randomly. I had to spend some time searching for how to reset the USB ports (this is never mentioned in the documents). Would be good to have a reference for what various variables that show up in json files mean. I think I now understand most of them, but certainly not all. There is almost no error reporting at all - for a novice, it is very difficult to tell when and why something failed. When cron is running the loop, I do not know how to tell what is going on, or even whether the system is running or it's stuck (other than opening various json files, or my log file that collects all error/outputs, but even then it is not clear what the status of the system really is). But, overall, it is impressive, amazing work! I am yet to try to upload data to Nightscout (that's also not even mentioned in the docs), ... work in progress.
Dana Lewis
@danamlewis
Jan 01 2016 22:21
@dm61 There's a new add to the docs about Nightscout upload, @eyim added a few days ago. So that should help you get started, but you can definitely add to it!
Ben West
@bewest
Jan 01 2016 22:22
hmm
Scott Leibrand
@scottleibrand
Jan 01 2016 22:23
Good feedback. I'll try to write something up on interpreting the variables and other output.
Ns upload will really help with the visibility. Maybe should mention something about setting loop up to log output too so you can tail that log for debugging.
Dana Lewis
@danamlewis
Jan 01 2016 22:25
Thinking: would love for people to note as an issue in the docs repo when they find good things like this that are not documented yet. Then we can work through the list to resolve. (I will also add this to the docs :) so people see it as they go through)
I might make an issue for each thing we know is not yet in docs - example for Medtronic cgm that we know works but is not well documented - so once it's done we can check it off.
Ben West
@bewest
Jan 01 2016 22:34
:+1: on issues too
good idea
Sulka Haro
@sulkaharo
Jan 01 2016 22:37
I'm still intrigued by the "Sara6E" event from the pump log every midnight. Is Sara a little elf who lives inside the pump?
Alin Artiom Kenibasov
@ktomy
Jan 01 2016 22:39
Hello everyone and happy new year! For openaps toolkit automated install script (https://raw.githubusercontent.com/openaps/docs/master/scripts/quick-packages.sh), I think we don't really need apt-get -y upgrade
As we afterwards install latest versions of needed packages and they will bring dependencies after them
Oskar Pearson
@oskarpearson
Jan 01 2016 22:42
If we don't upgrade it doesn't necessarily include all security fixes
Alin Artiom Kenibasov
@ktomy
Jan 01 2016 22:42
on other hand apt-get upgrade is dangerous by itself as if by chance it contains kernel update it may porentially bring system down
Oskar Pearson
@oskarpearson
Jan 01 2016 22:42
Kernel upgrades are normally only done when there's a security hole in the kernel which needs fixing. Or a serious flaw :)
Welcome, by the way!
And happy new year too
Dragan Maksimovic
@dm61
Jan 01 2016 22:43
@danamlewis will do, although I think I would first need to get a little better idea about github; earlier I was about to ask what PR was :smile:
Oskar Pearson
@oskarpearson
Jan 01 2016 22:43
I'm pro requiring security upgrades!
it ensures continuous upgrade
Didn't find in docs any info about why RPI2 and notTPi1 , is the readon only about 4xUSB?
And unfortunately my English is too bad to try to contribute to documentation :(
Dana Lewis
@danamlewis
Jan 01 2016 22:51
@dm61 sounds good. The OpenAPS docs repo is a great place to learn about Github :) and practice submitting PRs (pull requests).
Oskar Pearson
@oskarpearson
Jan 01 2016 22:52
So far as I know any Pi is fine. Most people use the Pi2 because it's the fastest and easily available. But there's no specific requirement for the Pi2. You'll need enough ports to connect things, but some people are using an A+ for better battery life
Rachel Sandlain
@audiefile
Jan 01 2016 22:52
I agree with @dm61 's asessment of set up and the docs
schulzedj
@schulzedj
Jan 01 2016 22:52
I am with @tzarro in the overall documentation. Would value being able to provide feedback on this process, where I had questions, etc.
Alin Artiom Kenibasov
@ktomy
Jan 01 2016 22:53
Also maybe a glossary would help
Dana Lewis
@danamlewis
Jan 01 2016 22:54
There is a glossary in the Openaps docs, please add to it :) even as a PR with the terms that we need to fill in
Rachel Sandlain
@audiefile
Jan 01 2016 22:54
:+1:
Dana Lewis
@danamlewis
Jan 01 2016 22:55
@schulzedj See above about adding an issue to the docs if there is an area that needs to be fleshed out; or start a PR targeting a particular section with your feedback. That would be great!
diabeticgonewild
@diabeticgonewild
Jan 01 2016 22:58
What I would like to eventually see is an upgrade from docs, like a migration to something more "interactive". Like the Knewton platform, where we can guide people more.
knewton.jpg
Alin Artiom Kenibasov
@ktomy
Jan 01 2016 23:01
Ahaaa, wat was mesleading for me is that there is a docs directory in docs :) https://github.com/openaps/docs/tree/master/docs and there is no readme in it :) will try to fork it and add a readme with a short text for "go 1 level up for readme :) good exercise in PRs :)
Dana Lewis
@danamlewis
Jan 01 2016 23:01
Sounds great @ktomy :)
diabeticgonewild
@diabeticgonewild
Jan 01 2016 23:02
I mean Knewton would be more guided and interactive...but I think that is at least 3 months away as a possibility.
You could use it and it would be more straight to the point.
eszcloud
@eszcloud
Jan 01 2016 23:03
Popped in a new sd card with a fresh install and sudo easy_install -ZU openaps worked. openaps is now up to date (e.g. normalize_history shows correct year).
Any idea why I got AttributeError: 'module' object has no attribute 'packaging' earlier with everything on the old setup?
Dana Lewis
@danamlewis
Jan 01 2016 23:03
Here is the summary/outline of what's in the docs, too, if it helps (you'll find it included in various places as an attempt to help with the navigation): https://github.com/openaps/docs/blob/master/SUMMARY.md and glossary here for anyone who wants to start there :) https://github.com/openaps/docs/blob/master/docs/Resources/glossary.md
eszcloud
@eszcloud
Jan 01 2016 23:03
clearly not critical here but I'm curious about the difference
Ben West
@bewest
Jan 01 2016 23:22
sara6e was renamed in latest decocare
diabeticgonewild
@diabeticgonewild
Jan 01 2016 23:24
OK I put in openaps report add monitor/glucose.json JSON <cgm> iter_glucose 5 and I tried to do openaps report invoke monitor/glucose.json and I get an error, cause the monitor directory does not exist. Does anyone know what is going on with the first command, with respect to where glucose.json is going?
eszcloud
@eszcloud
Jan 01 2016 23:25
what does sara6e do?
Oskar Pearson
@oskarpearson
Jan 01 2016 23:25
@diabeticgonewild doing a 'mkdir monitor' should sort that out
I don't think openaps will automatically create the parent directory for you
diabeticgonewild
@diabeticgonewild
Jan 01 2016 23:26
weird. Thanks
Scott Leibrand
@scottleibrand
Jan 01 2016 23:48
thx @ktomy. I merged your PR to dev, verified it looks good, and then merged to master as well.
Alin Artiom Kenibasov
@ktomy
Jan 01 2016 23:49
thank you
Toby Canning
@TC2013
Jan 01 2016 23:50
I upgraded today and though the loop is running well, on upload I get this error: <cr> Upload
module.js:340
throw err;
Error: Cannot find module '/home/pi/src/openaps-dev/enact/enacted.json'
at Function.Module._resolveFilename (module.js:338:15)
at Function.Module._load (module.js:280:25)
at Module.require (module.js:364:17)
at require (module.js:380:17)
at requireWithTimestamp (/usr/local/lib/node_modules/oref0/bin/ns-status.js:22:18)
at Object.<anonymous> (/usr/local/lib/node_modules/oref0/bin/ns-status.js:52:20)
at Module._compile (module.js:456:26)
at Object.Module._extensions..js (module.js:474:10)
at Module.load (module.js:356:32)
at Function.Module._load (module.js:312:12)
module.js:340
throw err;
Scott Leibrand
@scottleibrand
Jan 01 2016 23:55
have you enacted any temps?
Toby Canning
@TC2013
Jan 01 2016 23:56
No, none have been needed since trying, but I can manually set one
Scott Leibrand
@scottleibrand
Jan 01 2016 23:56
Cannot find module '/home/pi/src/openaps-dev/enact/enacted.json' is the main error there
that upload script assumes you don't delete old enacted.json
so the very first time you run it (before the first enact) it'll error like that
Toby Canning
@TC2013
Jan 01 2016 23:57
ahhh... let me touch that
Scott Leibrand
@scottleibrand
Jan 01 2016 23:57
but afterward should be fine if you don't delete it
Toby Canning
@TC2013
Jan 01 2016 23:57
Wow, which means we've need one every other time
Scott Leibrand
@scottleibrand
Jan 01 2016 23:58
it's expected to be an old one if no action was required recently
Toby Canning
@TC2013
Jan 01 2016 23:58
I'm really really impressed by how well your APS logic functions Scott
We are getting unbelievably good results
diabeticgonewild
@diabeticgonewild
Jan 01 2016 23:58
Yeah I am at the enact/enacted.json error too
Scott Leibrand
@scottleibrand
Jan 01 2016 23:58
:-D
Dana Lewis
@danamlewis
Jan 01 2016 23:58
:)
I'm happy it worked so well even last night without knowledge of iob due to the bug
Scott Leibrand
@scottleibrand
Jan 01 2016 23:59
we probably should update that uploader script to check if enact exists and just warn if not
diabeticgonewild
@diabeticgonewild
Jan 01 2016 23:59
I did openaps loop and I still got the enact/enacted.json error, even though I ran it twice
Dana Lewis
@danamlewis
Jan 01 2016 23:59
means our uber safety approach, even with new layers of features like meal-assist in dev, is sticking
Scott Leibrand
@scottleibrand
Jan 01 2016 23:59
loop running doesn't necessarily mean it needed to enact a temp
if there was nothing to do, it won't create one
diabeticgonewild
@diabeticgonewild
Jan 01 2016 23:59
Oh...
Well I am 262 right now...