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

24th
Sep 2016
garykidd
@garykidd
Sep 24 2016 00:24
This message was deleted
This message was deleted
garykidd
@garykidd
Sep 24 2016 00:30
So regarding eating soon mode...you just need to temp basal manually for a period of time or blind bolus ....and then in either of those cases follow up with a bolus wizard bolus?
Scott Leibrand
@scottleibrand
Sep 24 2016 00:31
definitely not blind bolus. :)
garykidd
@garykidd
Sep 24 2016 00:31
sorry---non bolus wizard bolus?
Scott Leibrand
@scottleibrand
Sep 24 2016 00:32
so the idea behind an eating soon early pre-bolus is to do the amount required to correct your BG down to 80 eventually (without any extra carbs)
so yes, you can do a non-bolus wizard bolus for that amount, up to an hour before you expect to eat
garykidd
@garykidd
Sep 24 2016 00:39
Thanks Dana read that. One clarifying question though...I'm assuming when you say you're bringing down your your temp target to lets say 80...that you're not actually adjuting any settings---you're just manually calculating and administering an adjustment bolus that should result in a lower BG after insulin has been fully absorbed...is that right?
garykidd
@garykidd
Sep 24 2016 00:45
I guess that's pretty similar to super bolus.....I guess an extra unit an hour prior probably would cut spike more or less by the amount equal to my ISF?
Eric Jensen
@elnjensen
Sep 24 2016 01:11
Just reading up on temporary targets on Nightscout - I see that the loop set up by the new oref0 setup script will pull targets from Nightscout if they have been entered recently (looks like in the past 6 hours). Looks very useful, so my question - how do people typically set these? I see it can be done in the NS web interface, but is there a slicker way, e.g. from a Pebble watch? I have the Careportal watch app which is great for quickly entering carbs, but doesn't look like targets are an option.
Lynne Butts
@flutelynne_twitter
Sep 24 2016 01:23
I use careportal to enter either the eating soon or activity. Sometimes I set activity for 6 hours. I do that before I leave for a gig so it is in wifi. If I forget, it is through the pump. For eating soon, I often start an hour or two before dinner and have it run for a while after. That seems to work well for me.
garykidd
@garykidd
Sep 24 2016 02:19
@elnjensen can you direct me further on your subject? How do you set temp target exactly via careportal...I only have these as available options in the careportal drop down box...
blob
blob
Lynne Butts
@flutelynne_twitter
Sep 24 2016 02:22
I noticed that today, too. I have a 9.1 version that shows temp target, but the dev version I use for Loop didn't show that option.
That was weird.
garykidd
@garykidd
Sep 24 2016 02:23
This message was deleted
Eric Jensen
@elnjensen
Sep 24 2016 02:23
@garykidd What version of cgm-remote-monitor are you using?
garykidd
@garykidd
Sep 24 2016 02:23
This message was deleted
jmoutria
@jmoutria
Sep 24 2016 02:26
Are you all using CareLink or the TI CC 1111? Is the range sufficiently better on the TI CC 1111?
Eric Jensen
@elnjensen
Sep 24 2016 02:26
@jmoutria Range is way better on the TI stick than Carelink.
What kind of rig are you using? Raspberry Pi?
jmoutria
@jmoutria
Sep 24 2016 02:27
RPi3
Eric Jensen
@elnjensen
Sep 24 2016 02:28
Cool - then I'd recommend considering buying one of these Slice of Radio boards.
jmoutria
@jmoutria
Sep 24 2016 02:29
I have it set up to the point of connecting the CGM and pump, which we just acquired through a gracious soul.
Eric Jensen
@elnjensen
Sep 24 2016 02:29
A lot cheaper than TI stick, and easy to install inside a Raspberry Pi. Core chip is the same, also has very good range.
jmoutria
@jmoutria
Sep 24 2016 02:29
Cool, I'll check it out. Thanks!
Eric Jensen
@elnjensen
Sep 24 2016 02:30
The chip has to be flashed, just as the TI stick does, and holding the wires in place is slightly trickier - but overall process isn't too different.
They have been discontinued (mfr went out of business) but you can still get one from the link above.
Instructions for setting up are here if you end up going that route.
@garykidd Here's what I see in Careportal - temp target entries are near the bottom:
careportal.jpg
I'm running 0.9.0 beta 3
garykidd
@garykidd
Sep 24 2016 02:41
Screen Shot 2016-09-23 at 10.13.20 PM.png
I didnt realize the Slice of Radio boards were that cheap---so much cheaper than TI stick---jut ordered 2...I'm sure they'll get used eventually
garykidd
@garykidd
Sep 24 2016 02:51
@elnjensen do I just need to add/enable a plugin for NS you think to get temporary target to show up?
Eric Jensen
@elnjensen
Sep 24 2016 02:52
Not sure. What version of cgm-remote-monitor are you running?
Do you have careportal and openaps plugins enabled?
I guess you must, given that you're showing the careportal dropdown there...
jmoutria
@jmoutria
Sep 24 2016 02:57
@garykidd That's hilarious because when I initially saw the boards there were only 14. I checked other sites and found a bunch of unavailable and backordered options. Went back and saw the number drop to 12. A sense of urgency came over me so I also ordered two!
garykidd
@garykidd
Sep 24 2016 03:00
not sure but Idid follow the automated update recently..how do I check cgm-remote-monitor version again
@jmoutria funny--sorry I rattled you...wonder how long it'll take now
jmoutria
@jmoutria
Sep 24 2016 03:05
5-15 more sleepless nights at our house I suppose. Thanks for the advice @elnjensen. I have no idea what I'm doing so it should be fun!
garykidd
@garykidd
Sep 24 2016 03:09
lloks like I amnot on dev...bur rather master...I'm assumeing @elnjensen that you are on dev for cgm-remote-monitor?
Eric Jensen
@elnjensen
Sep 24 2016 03:14
@garykidd @jmoutria As far as I know that site is the only place those are still available. Glad you got them!
garykidd
@garykidd
Sep 24 2016 03:16
@elnjensen do you know how I can revert to dev branch of cgm-remote-monitor....also wold you mind confirming if in fact you are on dev as well?
Eric Jensen
@elnjensen
Sep 24 2016 03:17
@garykidd I think that 0.9.0 beta 3 is a bit behind dev, but it's definitely ahead of master.
@scottleibrand told me earlier today that he was using 0.9.0 beta 3 as well, so that should work for you.
You can update by using the tool here: http://nightscout.github.io/pages/update-fork/
Hmm, but not sure that lets you change to a different fork, though.
garykidd
@garykidd
Sep 24 2016 03:30
yeah doesnt appear to
Eric Jensen
@elnjensen
Sep 24 2016 03:32
@garykidd I think this will let you disconnect your code and redeploy another branch - scroll down a bit to "disconnect and redeploy": http://www.nightscout.info/wiki/welcome/how-to-update-to-latest-cgm-remote-monitor-aka-cookie/update-my-fork-troubleshooting-part-2
OK, off to bed for me. :zzz:
garykidd
@garykidd
Sep 24 2016 03:44
ahhh....cant seem to select dev branch--only master is available...
Matthias
@ceben80
Sep 24 2016 03:45
After high fat food I was working with temp target 80/80, 180 min
Awesome flatline, yeah
blob
garykidd
@garykidd
Sep 24 2016 03:57
Can someone help me to update NS to dev 0.9.3...I screwed somethign up
dougtakeuti
@dougtakeuti
Sep 24 2016 04:08
@scottleibrand not sure if you've seen my previous comment, but when I try to use the oref0-setup.sh script I get this
Checking /home/pi/myopenaps: Traceback (most recent call last):
  File "/usr/local/bin/openaps", line 4, in <module>
    __import__('pkg_resources').require('openaps==0.1.6.dev0')
  File "build/bdist.linux-armv7l/egg/pkg_resources/__init__.py", line 2927, in <module>
  File "build/bdist.linux-armv7l/egg/pkg_resources/__init__.py", line 2913, in _call_aside
  File "build/bdist.linux-armv7l/egg/pkg_resources/__init__.py", line 2940, in _initialize_master_working_set
  File "build/bdist.linux-armv7l/egg/pkg_resources/__init__.py", line 635, in _build_master
  File "build/bdist.linux-armv7l/egg/pkg_resources/__init__.py", line 943, in require
  File "build/bdist.linux-armv7l/egg/pkg_resources/__init__.py", line 829, in resolve
pkg_resources.DistributionNotFound: The 'openaps==0.1.6.dev0' distribution was not found and is required by the application
Can't init /home/pi/myopenaps
Any clue?
dougtakeuti
@dougtakeuti
Sep 24 2016 04:45
Never mind, got it...
Saskia
@Missdazzle
Sep 24 2016 04:46
@ceben80 You can´t be cured though - otherwise it wouldn´t be such a perfect flatline, would it!? :)
Dana Lewis
@danamlewis
Sep 24 2016 04:57
@garykidd per earlier q. My first blog post there, yes it was manual adjustment and a manual eating soon bolus. Now that we have loop though I can set temp target and loop will do eating soon for me without needing to do the bolus.
@elnjensen yes, you can have watch enter temp targets :D (we use the ThisButton app to do so...)
Someone (@jasoncalabrese or others) may have gotten around to getting the care portal watch app to do that, though, but not sure
Touché @Missdazzle :)
dougtakeuti
@dougtakeuti
Sep 24 2016 05:02
@scottleibrand the oref0-setup.sh script fails here. Where cn I find the mdt-cgm.json file?
importing mdt-cgm file
cat: /home/pi/src/oref0/lib/oref0-setup/mdt-cgm.json: No such file or directory
Traceback (most recent call last):
  File "/usr/local/bin/openaps-import", line 6, in <module>
    exec(compile(open(__file__).read(), __file__, 'exec'))
  File "/home/pi/src/openaps/bin/openaps-import", line 89, in <module>
    app( )
  File "/home/pi/src/openaps/openaps/cli/__init__.py", line 51, in __call__
    self.run(self.args)
  File "/home/pi/src/openaps/bin/openaps-import", line 77, in run
    candidates = json.load(args.input)
  File "/usr/lib/python2.7/json/__init__.py", line 290, in load
    **kw)
  File "/usr/lib/python2.7/json/__init__.py", line 338, in loads
    return _default_decoder.decode(s)
  File "/usr/lib/python2.7/json/decoder.py", line 366, in decode
    obj, end = self.raw_decode(s, idx=_w(s, 0).end())
  File "/usr/lib/python2.7/json/decoder.py", line 384, in raw_decode
    raise ValueError("No JSON object could be decoded")
ValueError: No JSON object could be decoded
Could not import mdt-cgm.json
Did you pull the latest oref0-setup branch?
dougtakeuti
@dougtakeuti
Sep 24 2016 05:14
By running mkdir -p ~/src; cd ~/src && git clone -b dev git://github.com/openaps/oref0.git || (cd oref0 && git checkout dev && git pull) ?
ok, seems to work, onto testing now!
John Sjolund
@sjolundjohn
Sep 24 2016 05:26

@scottleibrand here is what the error sequence looks like

edison@edison2:~/openaps$ sudo tail -F /var/log/openaps/pump-loop.log
[sudo] password for edison: 
Listening: ....................................................................................................Error, retrying
Starting pump-loop at Sat Sep 24 06:59:17 CEST 2016:
Listening: ....................................................................................................Error, retrying
Starting pump-loop at Sat Sep 24 07:02:11 CEST 2016:
Listening: ....................................................................................................Error, retrying
Listening: ........................................................................................................................................................................................................Starting pump-loop at Sat Sep 24 07:10:40 CEST 2016:
Listening: ....................................................................................................Error, retrying
Listening: ........................................................................................................................................................................................................Starting pump-loop at Sat Sep 24 07:19:12 CEST 2016:
Listening: ..................Starting pump-loop at Sat Sep 24 07:21:04 CEST 2016:
Listening: ...................................................No comms detected
Old pumphistory: 
Error, retrying
Listening: .No comms detected
mmtune: "868.486", 5, -68 Starting pump-loop at Sat Sep 24 07:25:05 CEST 2016:
Listening: .

The only way to recover is to reboot the entire rig

Scott Leibrand
@scottleibrand
Sep 24 2016 05:27
@dougtakeuti no, that pulls the dev branch. You want oref0-setup branch instead for MDT CGM
@sjolundjohn I don't see any consistent sequence in that paste...
To troubleshoot further you can comment out the pump-loop in cron and run commands manually.
dougtakeuti
@dougtakeuti
Sep 24 2016 05:50
@scottleibrand I see. What do I run then, mkdir -p ~/src; cd ~/src && git clone -b dev git://github.com/openaps/oref0.git || (cd oref0 && git checkout oref0-setup && git pull)?
Scott Leibrand
@scottleibrand
Sep 24 2016 05:51
That would work. Or you could get rid of all the extra stuff and just cd oref0 && git checkout oref0-setup && git pull
dougtakeuti
@dougtakeuti
Sep 24 2016 06:03
@scottleibrand it does the setup for G4. Do I edit the script and change the defaults?
Scott Leibrand
@scottleibrand
Sep 24 2016 06:03
Read the usage text. Very first line it outputs.
Or pull it again. I just now merged openaps/oref0#222
dougtakeuti
@dougtakeuti
Sep 24 2016 06:06
ok
Tim Street
@tim2000s
Sep 24 2016 09:11
I have to admit I don't really understand temp target yet. Guess the option is play with it!
PieterGit
@PieterGit
Sep 24 2016 09:45
is it possible to use the ti stick with mmeowlink with european frequencies? the only way i get ti-stick to work is to change radio_locale to WW in https://github.com/oskarpearson/mmeowlink/blob/dev/mmeowlink/mmtune.py#L14
Saskia
@Missdazzle
Sep 24 2016 09:51
@pieterGit I don´t know if this will work for you, but after changing the line locale= from US to WW I changed this, too: 'WW': { 'start': 868.4, 'end': 868.6, 'default': 868.525 }
PieterGit
@PieterGit
Sep 24 2016 09:56
@Missdazzle : my mmtune reports "setFreq": 868.438, but i would rather configure the radio_locale to ww instead of changing the mmeowlink files
PieterGit
@PieterGit
Sep 24 2016 11:47
@scottleibrand : found some issues with the oref0-setup.sh script for pump that cannot enact. I would assume the purple suggestion line gets updated even if the enacting fails. It seems the purple line only updates when i run openaps report invoke enact/suggested.jsonfrom the CLI
small issue: mdt cgm is not listed in --help option
blob
PieterGit
@PieterGit
Sep 24 2016 12:03
might also be openaps enact that fixes the purple line again. I am not sure why the updating fails.
Eric Jensen
@elnjensen
Sep 24 2016 15:31
@PieterGit Are you asking whether it's possible to pass radio_locale as a command-line parameter rather than changing the default in the code? I'm not sure of the answer, but just trying to clarify if that's your question.
Scott Leibrand
@scottleibrand
Sep 24 2016 16:42
@PieterGit MDT is in the list of options in the oref0-setup branch: https://github.com/openaps/oref0/blob/oref0-setup/bin/oref0-setup.sh#L91
I still need someone to test https://github.com/openaps/oref0/pull/219#issuecomment-249089544 and comment before merging that to dev.
Jason Calabrese
@jasoncalabrese
Sep 24 2016 16:46
@garykidd saw you were having some problems updating NS to dev, I'm not sure if you'll see the github comments, but I think all you need to do is merge the PR you created awhile ago garykidd/cgm-remote-monitor#1
garykidd
@garykidd
Sep 24 2016 16:51
@jasoncalabrese thanks. I started to think I was merging something inadvertently into the dev for everyone and stopped there.
SO I merged---do I need to redeploy now?
Jason Calabrese
@jasoncalabrese
Sep 24 2016 16:52
it depends, where are you running NS?
the normal azure setup should auto redeploy, heroku needs some button pushing
garykidd
@garykidd
Sep 24 2016 16:53
azure
but I think I need to switch to something else here soon
Jason Calabrese
@jasoncalabrese
Sep 24 2016 16:54
when azure works it's fine, the main issue with it is hitting limits and having it shut off
but deploys are really intensive, something related to running linux and node on top of DOS...
garykidd
@garykidd
Sep 24 2016 16:55
yeah, I'm on the pay as you go right now and only just tarted looping so I have no idea what its gonna cost too.
Jason Calabrese
@jasoncalabrese
Sep 24 2016 16:56
theres a way to monitor the deploy in azure, but you'd have to click around to find it
if everything works you should be able to refresh NS is see the updated version
garykidd
@garykidd
Sep 24 2016 16:57
OK...looks like it may redply automatically based on how slow my NS site is loading
is this the latest---or did I mess something up again... 0.9.0-beta3
Jason Calabrese
@jasoncalabrese
Sep 24 2016 17:00
the name on dev is still beta3
garykidd
@garykidd
Sep 24 2016 17:01
so that's latest then--right?
Jason Calabrese
@jasoncalabrese
Sep 24 2016 17:02
yeah
garykidd
@garykidd
Sep 24 2016 17:11
cool. And just to confirm---I couldnt somehow mess up the cgm-remote-monitor for other by an inadvertent merge on my end--right? I have to go back and review git in general
Jason Calabrese
@jasoncalabrese
Sep 24 2016 17:14
yeah, right now you only have commit permissions to your own fork, not the main repo
garykidd
@garykidd
Sep 24 2016 17:16
That's what I thought but I was getting kind of punchy.........would it make sense to run this as part of cron? oref0-reset-git ?
Jason Calabrese
@jasoncalabrese
Sep 24 2016 17:17
yeah, thats to reset the local get repo used by openaps on your rig
git is used in multiple ways, that might be why it's confusing
every report you invoke with openaps does a git commit on your rig, but thats different repository, than the NS repo, oref0, etc
garykidd
@garykidd
Sep 24 2016 17:19
got it....and when it gets corrupted it's an issue locally on the rig and not actually ...I don't know somewhere on a git cloud or something?
Jason Calabrese
@jasoncalabrese
Sep 24 2016 17:20
right, locally losing power can corrupt the repo
you could push your local repo to github, but I don't think most people do that
but if you do it looks something like https://github.com/jasoncalabrese/indy-e1g
garykidd
@garykidd
Sep 24 2016 17:21
so can I just add it like this below in my cron job.......
Jason Calabrese
@jasoncalabrese
Sep 24 2016 17:22
yeah, would be 1 more line in the cron tab
garykidd
@garykidd
Sep 24 2016 17:23
SHELL=/bin/bash
PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/games:/usr/games
* * * * * /home/pi/bin/bt-pan client XX:XX:XX:XX:XX:XX
#* * * * * /home/pi/bin/check-and-connect-bt-pan.sh
*****oref0-reset-git
* * * * * sudo wpa_cli scan
*/5 * * * * (cd /home/pi/APS301 && time openaps do-everything) 2>&1 | logger -t openaps-loop
Jason Calabrese
@jasoncalabrese
Sep 24 2016 17:24
you need the cd
* * * * * cd /home/pi/APS301 && oref0-reset-git
garykidd
@garykidd
Sep 24 2016 17:24
how bout ...(cd /home/pi/APS301 && oref0-reset-git &&time openaps do-everything) 2>&1 | logger -t openaps-loop
Jason Calabrese
@jasoncalabrese
Sep 24 2016 17:25
I'd have it as a separate line, that runs every minute
garykidd
@garykidd
Sep 24 2016 17:25
got it.
thanks, Jason
Jason Calabrese
@jasoncalabrese
Sep 24 2016 17:26
:smile:
the reason you need the cd there, is that's the location of your local git repo
when you first did openaps init, that created the repo, and then every report invoke since then is a commit
so if things go wrong it's possible to dig through the history and find the exact inputs and reproduce the issue
garykidd
@garykidd
Sep 24 2016 17:28
SHELL=/bin/bash
PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/games:/usr/games
* * * * * /home/pi/bin/bt-pan client XX:XX:XX:XX:XX:XX
#* * * * * /home/pi/bin/check-and-connect-bt-pan.sh
* * * * * sudo wpa_cli scan
* * * * * cd home/pi/APS301 && oref0-reset-git
*/5 * * * * (cd /home/pi/APS301 && time openaps do-everything) 2>&1 | logger -t openaps-loop
Jason Calabrese
@jasoncalabrese
Sep 24 2016 17:29
that looks right, but I'm not sure about the wpa_cli there
garykidd
@garykidd
Sep 24 2016 17:30
would I be looing for a hypothetical "troublesome" commit within openaps-loop...assuming I was researching some sort of new problem?
Jason Calabrese
@jasoncalabrese
Sep 24 2016 17:32
I've used it once when there was an intermittent IOB bug, it's a little eaiser when the repo is pushed to github
found the old issue, the details don't matter now, but was really nice to be able really understand and reproduce the bug, then verify that it was fixed openaps/oref0#49
garykidd
@garykidd
Sep 24 2016 17:35
Very cool.
Jason Calabrese
@jasoncalabrese
Sep 24 2016 17:36
doing that gave me a lot of confidence in the system, not only did we find/fix the bug, but also saw that even with the wrong iob, oref0 behaved in safe way because of the other checks
garykidd
@garykidd
Sep 24 2016 17:38
nice work.
Jason Calabrese
@jasoncalabrese
Sep 24 2016 17:40
that was all the result of @bewest's design
garykidd
@garykidd
Sep 24 2016 17:40
I guess it probably doesnt make much sense to push to git less frequently...if you're gonna do it--do with each loop..or dont push at all---is that right?
Jason Calabrese
@jasoncalabrese
Sep 24 2016 17:40
I push every 5 mins, outside of the loop
garykidd
@garykidd
Sep 24 2016 17:40
Ithink you're referring to the fact that itbehaved in a safe way part
Jason Calabrese
@jasoncalabrese
Sep 24 2016 17:42
no that was @scottleibrand's desgin, I was referring to Ben's design of using git as the operations log for all actions
garykidd
@garykidd
Sep 24 2016 17:42
shofting gears for ascond...whats the thought behind needed to run mmtune periodically....that you may enter an rf environment that suits a different frequncy better?
Jason Calabrese
@jasoncalabrese
Sep 24 2016 17:44
lots of things can effect the RF, 1 is the temperature, so if you're sleeping on top of the pump and it get hotter the best freq is different
Scott Leibrand
@scottleibrand
Sep 24 2016 17:46
another is proximity to skin: your body becomes part of the antenna, so the frequency the pump ends up using shifts a bit
This message was deleted
I only mmtune on loop failures now
garykidd
@garykidd
Sep 24 2016 17:48
can mmtune actually change the frequency that pump transits on too?
transmits
Scott Leibrand
@scottleibrand
Sep 24 2016 17:53
no, it just makes the rig match whatever works best right now for talking/listening to the pump
garykidd
@garykidd
Sep 24 2016 18:01
OK. that's what I thought---which made me winder if you really ever had to switch frequencies ---or said differently that there might be one best frequency (at least absent the presence of certain other RF noise or other interference). Wast thinking that the carelink stick probably doesn't shift frequencies nor the rf enabled BG meter.
so different question (but related).....
Scott Leibrand
@scottleibrand
Sep 24 2016 18:02
the main reason for mmtune is different pumps have different base frequencies, due to manufacturing variations
garykidd
@garykidd
Sep 24 2016 18:03
hmmm...so then..doesnt that suggest that there should be one best (static) frequency for a given pump?
Scott Leibrand
@scottleibrand
Sep 24 2016 18:03
once you figure out the optimal frequency for your pump, you could make a script that skips the actual tuning and just sets that frequency every time, and it would almost always work. doing an actual mmtune will often get you somewhat better range though
no, there is one frequency that is best on average, but it's unlikely to be the absolute best at any given moment.
garykidd
@garykidd
Sep 24 2016 18:05
that makes a little more sense to me with that clarification. I guess if you had the pump right next to rig and assuming no other rf issues...that one frquency would probably always work...that's what I'm thinking anyway---does that make sense?
diabeticgonewild
@diabeticgonewild
Sep 24 2016 18:11
No, because direction plays a role.
Even close up
garykidd
@garykidd
Sep 24 2016 18:13
I'm sure there's something else going on with my rf issue (which I'm pretty sure you'll recall)....when we tweaked bin/mmtune and start playing with tuning after installing ruby...I'm not able to maintian that base frquency..for instance if I have to run mmtune again...it'll wipe out best frequency I had determined---right---so how do I lock in that frequncy permanently--is there a way to have that best frequency the default even if a sibsequent mmtune is run?
Scott Leibrand
@scottleibrand
Sep 24 2016 18:18
I think that would have to go deep in the firmware. easier to just run something to set it each time
garykidd
@garykidd
Sep 24 2016 18:23
hmmm so could you replace the mmtune that's created in your setup script with this other task? if so, like maybe cd ~/src/minimed_rf/ && ruby -I lib bin/mmtune /dev/ttyACM0 ######
right now the issue, i believe, is that the script ends up calling mmtune (and I can't know to press buttons)....and so it picks the generic default freq within the standard mmtune
garykidd
@garykidd
Sep 24 2016 18:41
I think I may need to strip the mmtune's from your setup script...well actually I think I did that by running with carelink instead of selecting ti stick option...right?
PieterGit
@PieterGit
Sep 24 2016 18:50
@elnjensen yes I would like to be able to pass radio_locale as a command-line parameter to either oref0-setup.sh or to openaps tools rather than changing the default in the mmeowlink/mmtune.py. I tried creating a pull request, but somehow my git config seems adding non existent mailadresses. See openaps/oref0#226
Eric Jensen
@elnjensen
Sep 24 2016 19:01
@garykidd @scottleibrand It's pretty easy to change the frequency that mmtune defaults to - just change it right here in the code on your installation. Note that (I think) it will only use the default frequency if mmtune fails completely, i.e. gets -99 on every frequency it tries.
The other thing you can do, which is less simple, but adds an extra level of robustness, is to build your own firmware and change the default frequency there, so that your setup is using that frequency right away from startup, before you run a tune. The code change is simple - just a couple of lines - but the only slightly tricky part is translate the frequency you want into the correct three hex values that the subg_rfspy firmware needs.
Eric Jensen
@elnjensen
Sep 24 2016 19:07
Here's the code I used to work that out and figure out the lines to change in the radio.c source file for building my own firmware. It's a bit of Python. If you have a freq you want the right values for, let me know and I can tell you what to change in that file.
Or I could just build some firmware for you with that frequency if you tell me the frequency and what type of chip.
Eric Jensen
@elnjensen
Sep 24 2016 19:18
@scottleibrand About to test my new installation from your setup script. If I'm not running autosens, is there any disadvantage to reducing the hours of glucose and pump history fetched? 24 hours is probably necessary for autosens (I'm guessing that's why it's in there) but overkill otherwise. Have done fine so far not having to pay anything for Azure and just want to be slightly mindful of how much data I'm transferring if I'm not actually using it.
@danamlewis Thanks for the clarification on triggering temp targets from the watch - I see that IFTTT is the key. That particular app (ThisButton) doesn't work for Pebble Time Round, but there are plenty of other IFTTT apps that do, so I'll give that a try when I have a chance. (So many things to spend time on! Hard to pick what to do next. Probably should prioritize AMA before I work on other things, I think.)
Jason Calabrese
@jasoncalabrese
Sep 24 2016 19:43
@sulkaharo has a fork of the careportal app that can do temp targets https://github.com/sulkaharo/PebbleCareportal/tree/wip/temptargets, but I don't know if it's published anywhere
Sulka Haro
@sulkaharo
Sep 24 2016 19:44
Nope, haven't put the pbw file up. I can!
I have some bizarre issue with the config page though that causes the setting to not work. If someone can help resolve that, that would be great
(Does temp targets and also consumes 5% of the ram of the original fork.)
Eric Jensen
@elnjensen
Sep 24 2016 19:56
@sulkaharo Great, I'll pull the code and try building it myself and see if I can figure out anything about the settings. I don't have much Pebble dev experience, but lately I've been working a little bit with a fork of Urchin to adapt it for Round watches as well, so I have the build environment set up. If anyone else out there is using a Pebble Time Round and wants to test Urchin, let me know - don't have everything worked out, but I do have a setup that has been working for me.
Thanks @jasoncalabrese for the pointer!
PieterGit
@PieterGit
Sep 24 2016 20:15
@oskarpearson or @scottleibrand can you help me with a review of openaps/oref0#226
Sulka Haro
@sulkaharo
Sep 24 2016 20:22
@elnjensen magically the app works on some pebbles and not on others
The issues seems to be with the iOS pebble app and it expecting something from the HTML page
You should be able to just clone, do pebble build and pebble install --phone=my.ip.is.this
@PieterGit that's a bit fragile way to implement the support
updating mmeowlink will overwrite the change
FWIW WW also works fine with ERF
Saskia
@Missdazzle
Sep 24 2016 20:27
Do I get the "temp targets" option (only) by going thru the new setup script?
PieterGit
@PieterGit
Sep 24 2016 20:29
@sulkaharo : i know it's very fragile, but I haven't got a clue how to be able to pass an attribute like radio_locale from oref0-setup.sh to openaps to mmeowlink. fortuately every time you run oref0-setup.sh mmewolink will be updated and the fix will be applied. without the fix a WW/European pump will not work with openaps and a TI stick afaict.
Jason Curry
@mccgm
Sep 24 2016 20:33
For an Edison rig, is the Jubilinux still the preferred image to start with?
Sulka Haro
@sulkaharo
Sep 24 2016 20:39
@PieterGit the "real" fix would be to patch openaps to read the data from the pump.ini file. Python...
What I'm doing is invoking a shells script to run mmtune
What I did with the openaps setup is changed tuning to call that instead of invoking openaps mmtune
That works across setup changes. Maybe we should pull that into oref0? oref0-euro-mmtune :)
PieterGit
@PieterGit
Sep 24 2016 21:32
am i right that the subg-ww-radio-parameters script sets the FREQ0 , FREQ1 and FREQ2 in the CC chip by using a serial connection?
garykidd
@garykidd
Sep 24 2016 21:35
@elnjensen its a cc1111. WIth 10 test though I got a lot of different frequencies. something very strange .
maybe 916.66...theres been alot of different numbers tough
garykidd
@garykidd
Sep 24 2016 22:26
Wondering if anyone can tell me why when I do this......
pi@raspberrypi:~/APS301 $ cd /home/pi/bin
pi@raspberrypi:~/bin $ bt-pan client [XX:XX:XX:XX:XX:XX]

my rig will instantly connect to my phone

but if I leave it to my cron to seek to connect....it doesnt work...
at least not unless I pull plug on rig and reboot it

here's the cron.....

PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/games:/usr/games
* * * * * cd /home/pi/bin/bt-pan client XX:XX:XX:XX:XX:XX
#* * * * * /home/pi/bin/check-and-connect-bt-pan.sh
* * * * * sudo wpa_cli scan
* * * * * cd /home/pi/APS301 && oref0-reset-git
*/5 * * * * (cd /home/pi/APS301 && time openaps do-everything) 2>&1 | logger -t openaps-loop
shouldnt that work with my cron too?
I waited like 10 mins and often nothing
Scott Leibrand
@scottleibrand
Sep 24 2016 22:40
do a which bt-pan
I suspect you're specifying the path to it wrong in cron.
Eric
@ecc1
Sep 24 2016 22:57
@garykidd You're missing a "&&" between the "cd ..." and the "client ..."
garykidd
@garykidd
Sep 24 2016 23:06
hmmmm.....
looks right ...no?
```
pi@raspberrypi:~ $ which bt-pan
/home/pi/bin/bt-pan
Scott Leibrand
@scottleibrand
Sep 24 2016 23:06
Ok, in that case you need to ditch the cd at the front
garykidd
@garykidd
Sep 24 2016 23:07
@scottleibrand yeah that what I had at first
pi@raspberrypi:~ $ crontab -l
SHELL=/bin/bash
PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/games:/usr/games
* * * * * /home/pi/bin && bt-pan client 11:22:33:11:22:33
#* * * * * /home/pi/bin/check-and-connect-bt-pan.sh
* * * * * sudo wpa_cli scan
* * * * * cd /home/pi/APS301 && oref0-reset-git
*/5 * * * * (cd /home/pi/APS301 && time openaps do-everything) 2>&1 | logger -t openaps-loop
strike the '&&' too?
Scott Leibrand
@scottleibrand
Sep 24 2016 23:13
No, * * * * * /home/pi/bin/bt-pan client
No cd, space, or &&
garykidd
@garykidd
Sep 24 2016 23:17
that is what I had originally----
but I did revert back....
and it works now (agian).
thhanks
still not sure why the commented out line below didnt work but I havent had any issues with it reconnecting---which I presume it is doing
every min
garykidd
@garykidd
Sep 24 2016 23:51
Just a general observation......seems like there are too many different things/sources that openaps depends on.....Pump and sensor (don't think you can really address that), but a typical setup might rely on azure, mongo, dex share, NS, BT, wifi, rf....any one of them could have an issue. I get some could be redundant but it just seems like there's a huge opportunity to simplify.
David
@davidkeddydb
Sep 24 2016 23:58
trying to use nightscout rather than having the cgm plugged into the pi. I am searching gitter here as well but figured I would just ask. What do I need to change this report to for it to work? oref0_glucose/raw-cgm/glucose-raw.json
raw-cgm/glucose-raw.json raised 'NoneType' object has no attribute 'iter_recor