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

15th
Jan 2017
Sgoldberg89
@Sgoldberg89
Jan 15 2017 00:04
Can't seem to find what this indicates in the trouble shooting page. I have tested the internet connection by pinging google, and I have verified that my NS is up and running. Does this indicate that the rig isn't communicating with the pump? "Starting pump-loop at Sat 14 Jan 15:58:19 PST 2017:
Radio ok, Listening: ...............................No comms detected
Old pumphistory:
Error, retrying"
seanneedle
@seanneedle
Jan 15 2017 00:04
@danamlewis I used the setup script once but didn't enable the cron loop. I wanted to use manual commands first but I'll try setting up the loop and checking the pump loop log and let you know. Thanks!
ejoe132
@ejoe132
Jan 15 2017 00:08
@danamlewis Im not sure what you mean by setup script? The issue I have is I dont have the port address for it so when I do the 0ref0 setup I dont have that value to enter
Dana Lewis
@danamlewis
Jan 15 2017 00:10
@ejoe132 rerun the setup script; look on that docs page under "be prepared to enter", it links to the mmeowlink stuff to figure out your port
@seanneedle :+1: yea, I think it will be easier to troubleshoot to go ahead and turn it on - I'm seeing a few bugs with people running manual commands that seem to be fine when actually looping but not sure what's changed to make manual commands be giving out those output when everything works fine otherwise
@Sgoldberg89 it means it's trying again, has too old of a history to loop on
Sgoldberg89
@Sgoldberg89
Jan 15 2017 00:14
@danamlewis But it keeps doing the same thing, is there anything I can do to provide new info?
Dana Lewis
@danamlewis
Jan 15 2017 00:16
Reboot, make sure you're in range, etc.
How long has it been giving that output?
Sgoldberg89
@Sgoldberg89
Jan 15 2017 00:16
an hour
Dana Lewis
@danamlewis
Jan 15 2017 00:16
Rebooted yet?
Sgoldberg89
@Sgoldberg89
Jan 15 2017 00:16
Yep, once
ejoe132
@ejoe132
Jan 15 2017 00:27
@danamlewis I found the "be prepared to enter" but it only says something for the explorer board then links to the mmeowlink page?
Dana Lewis
@danamlewis
Jan 15 2017 00:30
Yes - so the mmeowlink page should help you figure out the port for your setup
chrisharris751
@chrisharris751
Jan 15 2017 00:43
Hi, How do I change my openaps dependencies from dev back to master?
I tried running curl -s https://raw.githubusercontent.com/openaps/docs/master/scripts/quick-packages.sh | bash - but it seems to have just updated the dev version? Ta
jaylagorio
@jaylagorio
Jan 15 2017 00:47
Gitter search is failing me - does someone have the command line block to switch oref0 to the dev branch?
jaylagorio
@jaylagorio
Jan 15 2017 00:52
Nevermind, found it: https://github.com/openaps/docs/blob/dev/docs/docs/walkthrough/phase-0/setup-edison.md - mkdir -p ~/src; cd ~/src && git clone -b dev git://github.com/openaps/oref0.git || (cd oref0 && git checkout dev && git pull)
Toby Canning
@TC2013
Jan 15 2017 01:43
@chrisharris751 For me, going into the oref0 directory and running git checkout master usually works to revert from dev. I usually reinstall my Loop, but not sure that is required. You could run the npm commands again, doesn't seem to ever hurt anything.
chrisharris751
@chrisharris751
Jan 15 2017 02:34
@TC2013 , awesome thanks!
Aaron Michelson
@Aaybob
Jan 15 2017 02:45
@melodyandrews got you message from six hours ago. I saw the same error in my pump-loop.log file just now. OpenAPs had not updated for 1 hour. I'd been outside gardening. I had set up edison with copy and paste from the dev git from the section titled BTE optional. So I think it might be failing with bluetooth. I did the killall, shutdown command then unplugged and repowered edison. That got it to run again. If it hadn't I would have redone the git command from master instead of dev in the section titled Phase 2.
Aaron Michelson
@Aaybob
Jan 15 2017 02:51
@Sgoldberg89 I find that to establish connection to pump after I re-....boot, power, whatever...I have to put the pump close to the edison (inch or two).
@seanneedle @danamlewis command line not working definitly a bug
chrisharris751
@chrisharris751
Jan 15 2017 03:00

@TC2013 When I try that I get

error: Your local changes to the following files would be overwritten by checkout:
        bin/bt-pan
        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
Please, commit your changes or stash them before you can switch branches.
Aborting

Do I need to kill the cron jobs first?

Aaron Michelson
@Aaybob
Jan 15 2017 03:00
@Sgoldberg89 yes,
"Starting pump-loop at Sat 14 Jan 15:58:19 PST 2017:
Radio ok, Listening: ...............................No comms detected
Old pumphistory:
Error, retrying"
means it isn't communicating with the pump
I think the No comms detected mean it looked for other OPenaps rigs so it doesn't compete for resources so that's successful. "Error" is what it says
@chrisharris751 I've seen that too. No that has to do with GIT.
chrisharris751
@chrisharris751
Jan 15 2017 03:01
@Aaybob how did you over come it?
ejoe132
@ejoe132
Jan 15 2017 03:05
is there a way for me to find the port address for my ti stick? I set it all up with the cc debugger but when I ran my setup script i didnt know the port address for mmeowlink because I didnt know it. Can I look it up on the pi?
Dana Lewis
@danamlewis
Jan 15 2017 03:06
lsusb plug and unplug to see what disappears when you’re unplugged
ejoe132
@ejoe132
Jan 15 2017 03:07
do i type isusb into the pi using putty?
Terri Lyman
@tatigger
Jan 15 2017 03:15
Ok, husband left to go play music and now I can work for a little while without interruptions. Sheesh! Anyone who knows about IFTTT and This Button, I build an applet, got errors, so built another one at @danamlewis suggestion. That one did a little better. The trigger seemed to work (Congratulations! You've fired the %7Bcarb_correction%7D event), and the This Button gave a response of (Success), and when I click the Check Now button I get a response of (Applet Checked). However, no additional carbs I recorded in NS (should record 5 grams, my usual low correction now (woot! for OpenAPS!). And the Applet Activity Log returns (Applet Skipped), but only for the times I tried the This Button. Does anyone have any idea what I am doing wrong, how I can fix it, or where to find troubleshooting FAQs? I have googled, explored the IFTTT blogs, but can't really find any suggestions. Thanks guys!
Sacha M
@coolestkidsever
Jan 15 2017 03:16
Hi all, I had loop working this morning, then I worked through the pump Medtronic Vendor page to have a look at these, and since then (i think), i haven't been getting any pump data..
Log shows:
pi@raspberrypi:~ $ tail -F /var/log/openaps/pump-loop.log
File "/usr/local/lib/python2.7/dist-packages/decocare/stick.py", line 515, in product_info
return self.query(ProductInfo)
File "/usr/local/lib/python2.7/dist-packages/decocare/stick.py", line 509, in query
return self.process( )
File "/usr/local/lib/python2.7/dist-packages/decocare/stick.py", line 496, in process
ack, response = self.command.respond(raw)
File "/usr/local/lib/python2.7/dist-packages/decocare/stick.py", line 71, in respond
raise AckError("ACK is 0 bytes:\n%s" % lib.hexdump(raw))
decocare.errors.AckError: ACK is 0 bytes:
did i break it? :-)
ejoe132
@ejoe132
Jan 15 2017 03:16
so did it and the one it looks like says "bus 001 device 005: ID 1d50:9001 openmoko, inc" So in setup would it just be 1d50:9001?
@ejoe132 that was for setting up the ti stick
Tim Mecklem
@tmecklem
Jan 15 2017 03:18
Is there anything Edison specific in 915MHzEdisonExplorer_SW? I'm looking at using spi/mraa on a pi.
It looks like I could just flash the spi subg_rfspy, solder the spi pins on my srf, and I'd be off to the races?
Looks like it's referring to specific gpio pins in the code. I'd probably have to adjust those.
ejoe132
@ejoe132
Jan 15 2017 03:26
If I post my NS page could someone take a look and help me figure out why Im not high temping? I had been but the last day seems like it isnt going into high temping
Terri Lyman
@tatigger
Jan 15 2017 03:34
Is your Max IOB set to 0 in preferences.json? If you went back through the setup script, you reset all the preferences, @ejoe132
ejoe132
@ejoe132
Jan 15 2017 03:35
@tatigger Nope, I still have it set very high, I use a lot of insulin at meals so its high.
@tatigger It does say maxsafebasal rate: 13.5 no temp setting, setting 13.5u/hr, IOB -1.63, Basal IOB -1.63
So its like its saying what it wants to set is over the safemaxbasal (but its not its still under it) so it does nothing? why not just set something lower?
Terri Lyman
@tatigger
Jan 15 2017 03:43
You are hitting one of the safety settings, but I don't know which one. I think the settings are set very conservatively, and they all impact the rates that are set differently. Beyond that, I am still unsure how each one interacts with the settings. @scottleibrand or @danamlewis? I think @ejoe132's question is for you guys....
@ejoe132 I would post the "pill" with maxsafebasal in it, not the entire NS page. That I think is what will show one of those guys which thing you need to adjust
ejoe132
@ejoe132
Jan 15 2017 03:45
@tatigger how does one post the pill? I could take a picture and post it (need to know how to post a pic on here though cause I dont see an attachment icon
tepidjuice
@tepidjuice
Jan 15 2017 03:49
@coolestkidsever could it be a range problem? Are you using carelink? Try sitting the pump right next to the stick.
Sacha M
@coolestkidsever
Jan 15 2017 03:54
@tepidjuice don't think so. I am using carelink but it is siting right next to it. I just re-ran step 2 of setup and it seemed to pick it up once, but its still throwing these errors iin the log that I don't understand enough to troubleshoot :smile:
Terri Lyman
@tatigger
Jan 15 2017 03:55
I save it as a png file on the desktop and drag it to the message window.
Dana Lewis
@danamlewis
Jan 15 2017 03:57
@coolestkidsever try unplug replug carelink for the ack=0 to see if that resold first
@ejoe132 you may be running into the basal >6.3 issue, I think that requires dev branch for now...
Ps - you still need to meal bolus for meals. OpenAPS will not eliminate the need for a meal bolus, so even if you eat high carb and have a low I:c ratio and need a lot of insulin, you should still do meal bolus because otherwise OpenAPS won't be able to keep up given the realities of currently available insulin .
Scott Leibrand
@scottleibrand
Jan 15 2017 04:00
@chrisharris751 try what it suggests: git stash
Sacha M
@coolestkidsever
Jan 15 2017 04:03
How do I edit cron? i think i need to comment out the other times i've run the setup.
Scott Leibrand
@scottleibrand
Jan 15 2017 04:03
@ejoe132 what is your normal basal? If it's less than about 3U/hr, setting the max basal to 6.3U/hr might be the safest thing to start with. If you do Eating Soon mode and bolus for meals and still consistently see it hitting that limit, then you'll need the dev branch of openaps to go higher. One way to get that is switching to the dev branch of oref0 and re-running oref0-setup, or you can install the dev branch of the openaps toolkit directly.
@coolestkidsever just crontab -e
ejoe132
@ejoe132
Jan 15 2017 04:04
@danamlewis No worries, not using it to avoid meal boluses, just have high basals and need a lot sometimes as I jump at night.what is the >6.3 issue? I havent update my pi or anything so it should have changed?
Sacha M
@coolestkidsever
Jan 15 2017 04:05
@scottleibrand :+1:
ejoe132
@ejoe132
Jan 15 2017 04:06
@scottleibrand would that be maxbasal setting on my pump or in the nano preferences.json? I had had the pumps set at 9.55u and it had been working ok, then this issue happened and set it at like 15u i believe
@ejoe132 Ill try setting the setting on the pump to lower than 6.3. Highest basal right now is 2.35 currently
Scott Leibrand
@scottleibrand
Jan 15 2017 04:07
The problem was in decocare: it was only using the lowest 256 values (1 byte) of the pump's supported basal levels. Once we discovered the remaining bits and added them, that opened up support for 6.4U and higher temp basals.
6.3 setting of max basal on the pump should be fine. 6.4 or higher would hit this bug.
ejoe132
@ejoe132
Jan 15 2017 04:08
@scottleibrand so this is kinda odd, pump wont let me set max basal rate less than 9.55. and it goes up to 35u. Surprised it wont me set anything less than 9.55u
Dana Lewis
@danamlewis
Jan 15 2017 04:09
Max basal in pump should be set to 6.2; but I would also have max iob lower than that as you are just starting out. Work your way up from 0 with no more than 1u increments at a time.
Scott Leibrand
@scottleibrand
Jan 15 2017 04:09
Do you have a scheduled basal that high?
ejoe132
@ejoe132
Jan 15 2017 04:10
@scottleibrand Nope, my highest setting currently is 2.55u/h
Scott Leibrand
@scottleibrand
Jan 15 2017 04:11
The pump does basals in 0.025U/hr increments. It represents that as a 1. A full byte is 255, which would be 6.375U/hr. Rolling over to 6.4U/hr requires more than 1 byte, to represent 256 or higher.
Sacha M
@coolestkidsever
Jan 15 2017 04:12
@danamlewis I removed the carelink (and rebooted for good measure), (and commented out other instances in cron that were starting). I think Im back up
Scott Leibrand
@scottleibrand
Jan 15 2017 04:12
@ejoe132 that seems broken. Do you have any pattern A or B settings with higher basals than that?
ejoe132
@ejoe132
Jan 15 2017 04:13
@scottleibrand Good catch, when I got this it had a pattern B, it does have one of 9.55u. But why would it have been working fine for a few days then now start having a problem? Im removing the other pattern with high basal rate
Scott Leibrand
@scottleibrand
Jan 15 2017 04:14
@ejoe132 that bug wouldn't have affected you until openaps actually tried to set a temp higher than 6.375U/hr.
And then you wouldn't have noticed if it was transient, because it would just fail to set a temp until the recommendation came down below 6.4U/hr.
ejoe132
@ejoe132
Jan 15 2017 04:15
Pic 1.jpg
@scottleibrand That picture is what my openaps pill is showing
Scott Leibrand
@scottleibrand
Jan 15 2017 04:17
Is your ISF really 10?
After you change the max basal on the pump you'll have to wait for it to refresh pump settings to apply that change.
ejoe132
@ejoe132
Jan 15 2017 04:18
Yes, Ive had that setting since mid November, at a diabetic educator appointment
@scottleibrand I think it should have as openaps says 1m since last Loop right now
Scott Leibrand
@scottleibrand
Jan 15 2017 04:19
K. For now you might want to do a correction bolus if that's what you normally would do (maybe not quite as big as normal) while you wait for oref0 to pick up the new settings.
No, a loop doesn't necessarily imply a settings refresh.
Those are only every 10-15 minutes. You can see them in the pump-loop.log.
ejoe132
@ejoe132
Jan 15 2017 04:21
@scottleibrand It is showing in NS, says Note:ChangeMaxbasalRate 722
Scott Leibrand
@scottleibrand
Jan 15 2017 04:25
Crazy as it sounds, that doesn't mean oref0 knows about it yet.
That's the pumphistory, but it still has to read the actual settings and store those separately.
ejoe132
@ejoe132
Jan 15 2017 04:26
@scottleibrand Oh ok, its only been 6 minutes then so Ill wait, the I believe you said 10-15 min? and see what it gives me
Scott Leibrand
@scottleibrand
Jan 15 2017 04:28
Yeah, unless you want to go look at the contents of the logs or the json files on the rig.
chrisharris751
@chrisharris751
Jan 15 2017 04:29
@scottleibrand thanks Scott. Do I need to run the setup script again? Or will the loop pull from master now?
ejoe132
@ejoe132
Jan 15 2017 04:30
@scottleibrand That did it! I have a temp basal that came in running at 6.2u! So would this be the >6.3 issue?
Scott Leibrand
@scottleibrand
Jan 15 2017 04:31
Yep.
ejoe132
@ejoe132
Jan 15 2017 04:32
@scottleibrand So the fix for this would be to re do the 0ref0 install?
Scott Leibrand
@scottleibrand
Jan 15 2017 04:32
You might also want to make sure your max IOB isn't too high, just in case your ISF is too low or something else: you don't want the rig high-temping too much.
The openaps fix for >6.3 isn't released to master yet, so I would just leave it at 6.3 until then, or until you're comfortable enough to move to running dev code.
ejoe132
@ejoe132
Jan 15 2017 04:34
@scottleibrand Ok, I had it pretty high becasue I was having an issue with after eating a high carb meal it not wanting to do anything becasue of on board insulin then also casue of bolus snooze, Im assuming bolus snooze is reduced by increase the value ? so its set for 2 now so would going to 3 or 4 reduce it?
@scottleibrand Ok, Ill probably just wait for now, gonna be doing an edison/explorer set up when the board comes, also thinking autotune will be coming to master soon?
Scott Leibrand
@scottleibrand
Jan 15 2017 04:36
That is one option, but if you want it to high temp after meals, you probably want to enable advanced meal assist when you're ready.
I wouldn't say a dev to master release is imminent. Lots of flaky Bluetooth stuff people still need to work out.
ejoe132
@ejoe132
Jan 15 2017 04:36
@scottleibrand I have turned it on already but it still wouldnt, I believe becasue my meal boluses are often large
Scott Leibrand
@scottleibrand
Jan 15 2017 04:37
Sounds like maybe your ratios are off then.
If you bolus for the full meal and your I:C ratio is set properly, you shouldn't need any high temps for quite awhile.
You might want to set up a second rig running dev and try autotune on retrospective data and see what it recommends.
ejoe132
@ejoe132
Jan 15 2017 04:38
@scottleibrand oh ok. Also trying to move from carelink to ti stick but didnt have the port address before, Now I got what I think it is by using lsusb, came up with bus 001 device 006: ID 1d50:9001 OpenMoko, Inc. So would port address be the " 1d50:9001"?
Scott Leibrand
@scottleibrand
Jan 15 2017 04:39
No
Take another look at that section of the docs / mmeowlink wiki.
I believe one of those has instructions for determining it.
ejoe132
@ejoe132
Jan 15 2017 04:40
@scottleibrand That isnt the right item? I keep looking at that but not getting anywhere as I dont see where it explains to find it, says about setting up but pi tells me it cant casue its allready there, Ill try again to get the exact wording it uses
Terri Lyman
@tatigger
Jan 15 2017 04:41
So did anyone have any brilliant ideas re my IFTT issue?
Dana Lewis
@danamlewis
Jan 15 2017 04:42
@jyaw want to pr a link to your gist into the docs?
Scott Leibrand
@scottleibrand
Jan 15 2017 04:42
@ejoe132 That is correct device, and that is the USB identifier that would be used in constructing a udev rule. But I think the port you're talking about is something like /dev/something.
Dana Lewis
@danamlewis
Jan 15 2017 04:43
@tatigger not sure. But I'd try a temp target recipe next and see if that one works, or if you're still getting errors. If a temp target works, then that nails it down to how you're posting to NS
Terri Lyman
@tatigger
Jan 15 2017 04:45
Ok, sounds good. I have another question. How is the range set for max and min ISF in auto sense? Is it a percentage?
ejoe132
@ejoe132
Jan 15 2017 04:46
@scottleibrand its telling me " fatal: destination path "mmeowlink-source" already exists and is not an empty directory." the in grean says pi@raspberrypi:" the Blue symbol, the says "cd mmeowlink-source"
Scott Leibrand
@scottleibrand
Jan 15 2017 04:47
@tatigger It's a multiplier. 1.0 is no change, 1.2 is 20% more resistant (higher basals and lower ISF), and 0.7 is 30% more sensitive (lower basals and higher ISF).
ejoe132
@ejoe132
Jan 15 2017 04:47
then green stuff again, then in blue -'/mmeowlink-source" then in white says git cehckout master already on "master" (our branch is up to date with "origin/master" then says sudo pip install -pi@raspberrypi- $ cd-
Scott Leibrand
@scottleibrand
Jan 15 2017 04:48
@ejoe132 no idea what you're trying to do. You should figure out how to copy and paste to Gitter (with triple ` backticks etc.)
ejoe132
@ejoe132
Jan 15 2017 04:49
@scottleibrand I try to copy in putty but it just goes wild, do I do ''' in putty?
@scottleibrand cd ~
git clone https://github.com/oskarpearson/mmeowlink.git mmeowlink-source
cd mmeowlink-source
git checkout master
sudo pip install -e
@scottleibrand thats what im entering to try to start setting it up
Scott Leibrand
@scottleibrand
Jan 15 2017 04:54
You are doing the manual setup it looks like. That part would be done by oref0-setup.
ejoe132
@ejoe132
Jan 15 2017 04:55
@scottleibrand at one point it asks for the mmeowlink port as I did the 0ref0, I skipped cause I didnt know it and just plugged a carelink in and it worked great, just want more range
Scott Leibrand
@scottleibrand
Jan 15 2017 04:55
When pasting a big block into Gitter, precede the paste by three backticks (`) not quotes (') on their own line, and then follow the paste with the same.
@ejoe132 you just need to re-run oref0-setup and answer the mmeowlink question properly.
The only thing you need the mmeowlink wiki for is to find your mmeowlink device.
ejoe132
@ejoe132
Jan 15 2017 04:57
@scottleibrand is it "-(/dev/spidev5.1" even though Im using a pi3?
chrisharris751
@chrisharris751
Jan 15 2017 04:57
@scottleibrand After I switch my git back to master, do I need to run the setup script again?
Terri Lyman
@tatigger
Jan 15 2017 04:57
One of these days I need a whole tutorial on that ISF thing. Mine never seems to go below 54. I wondered if that was a set limit or just where my ISF alway stops. 54 to 74. My set ISF is 65.
Oh, and PS, I just spilled about 1/2 ounce mastisol across my bathroom. Toothbrush. Mirror. Q-tips. Soap dispenser, Yep.
Some days crime doesn't pay
ejoe132
@ejoe132
Jan 15 2017 05:14
@scottleibrand Guess I just dont know what to put in during the 0ref0 setup for the ti stick
Terri Lyman
@tatigger
Jan 15 2017 05:15
@danamlewis I did another app and am getting the same series of responses as with the carb correction.
Dana Lewis
@danamlewis
Jan 15 2017 05:18
@ejoe132 did you do lsusb as I suggested an hour or so ago?
That's what is in the mmeowlink wiki directions
@tatigger ok may want to check what you're putting in for NS URL and the post sections and whatnot next..
ejoe132
@ejoe132
Jan 15 2017 05:22
@danamlewis Yes, I found it to be listed as "Bus 001 Device 006: ID 1d50:8001 OpenMoko, Inc." So not sure what I should enter in during the 0ref0 setup?
Scott Leibrand
@scottleibrand
Jan 15 2017 05:24
@chrisharris751 what are you trying to disable? To fully roll back, you'd need to do npm run global-install from src/oref0, clear out your cron, and re-run oref0-setup.
Jarred Yaw
@jyaw
Jan 15 2017 05:24
@danamlewis will do!
Dana Lewis
@danamlewis
Jan 15 2017 05:26
@ejoe132 you should be getting /dev/something. Keep reading the mmeowlink docs to figure out what else you need to do to find it
Scott Leibrand
@scottleibrand
Jan 15 2017 05:26
@tatigger 65/54 is 1.2. So that's the limit of what autosens will adjust. If you're considering adjusting ISF, I would run autotune first on retrospective data to see what it recommends.
ejoe132
@ejoe132
Jan 15 2017 05:33
@danamlewis Is this the page to follow? https://github.com/oskarpearson/mmeowlink/wiki/Installing-MMeowlink I start doing the first think and get this each time ""cd ~
git clone https://github.com/oskarpearson/mmeowlink.git mmeowlink-source
cd mmeowlink-source
git checkout master
sudo pip install -e" then hit enter and it give me this"pi@raspberrypi:~ $ cd ~
pi@raspberrypi:~ $ git clone https://github.com/oskarpearson/mmeowlink.git mmeow link-source
fatal: destination path 'mmeowlink-source' already exists and is not an empty di rectory.
pi@raspberrypi:~ $ cd mmeowlink-source
pi@raspberrypi:~/mmeowlink-source $ git checkout master
Already on 'master'
Your branch is up-to-date with 'origin/master'.
pi@raspberrypi:~/mmeowlink-source $ sudo pip install -e"
Terri Lyman
@tatigger
Jan 15 2017 05:36
Thanks @scottleibrand I will try to get smart enough to run the autotune. My brain is now completely full Thanks for all your help. @danamlewis I will continue to troubleshoot. I thought I found the issue - I changed the Maker key, and didn't change that in my Heroku configurations, but even after I updated that I still got the same results. So more work on that later.
Scott Leibrand
@scottleibrand
Jan 15 2017 06:00
@ejoe132 looks like you already have mmeowlink, so you don't need to clone it again.
The backtick (`) is on the same key as ~, not the ' and " key.
ejoe132
@ejoe132
Jan 15 2017 06:04
@scottleibrand But when I put the TI stick in it doesnt connect to the pump, so how does one find the port to input in 0ref0 setup? i tried lsusb but sounds like that isnt correct
Scott Leibrand
@scottleibrand
Jan 15 2017 06:14
Are you actually reading the wiki directions? It doesn't just say to run lsusb. https://github.com/oskarpearson/mmeowlink/wiki/Installing-MMeowlink#using-udev-rules-to-make-meaningful-device-names
Matthias
@ceben80
Jan 15 2017 07:51
@ejoe132 look here, some tips for more readable postings in gitter :wink:
blob
chrisharris751
@chrisharris751
Jan 15 2017 08:09
@scottleibrand I'm having sporadic pump communication problems on dev, that I wasnt getting on master. The only reason I went to dev was BLE xdrip offline which with help I got working. I don't always need offline capabilities so I wanted to setup a master loop and only run the dev when offline is needed. Cheers
Tim Street
@tim2000s
Jan 15 2017 13:36
Here's an I've run the pre-requisites for bluetooth in Dev, all seems okay on the edison, but when I try and pair, it doesn't seem to be finding the iPhone:
[bluetooth]# pair B8:53:AC:74:5E:F5
Device B8:53:AC:74:5E:F5 not available
Scorponok
@Scorponok
Jan 15 2017 14:49
Hi everyone! I have a loop successfully set up on the raspberry pi, and the only thing I can't get working is bluetooth tethering to an iPhone. I followed the instructions in the openaps readthedocs, but they don't seem to do anything. I assume a new "bnep0" interface is meant to show up in ifconfig, but it never does - do I need to add it to /etc/network/interfaces, or is it meant to appear automatically?
Toby Canning
@TC2013
Jan 15 2017 15:55
Greetings. Anyone else here running an explorer board with on dev branch? If yes, what kind of battery life are you getting? I have a 2000mAh lipo. Yesterday it died after 12 hours or so (I wasn't tracking precisely).
Also, I'm using the default cron settings * * * * *.
Toby Canning
@TC2013
Jan 15 2017 16:01
@chrisharris751 Sorry, we have guests over so I'm slow to respond. For best results you may want to kill off current processes and stop pause you cron jobs by commenting them out with #. However, the issue you described should be resolved using git stash in the oref0 directory before you switch branches.
Matthias
@ceben80
Jan 15 2017 16:59
@TC2013 did the loop make a hard shutdown due to power shortage? Does your rig have EdisonVoltage implemented?
Tim Street
@tim2000s
Jan 15 2017 17:40
Baby steps forward. If I scan for devices from bluetoothctl, I can see the iPhone, then I can pair with it. I don't get the subsequent messages about confiirming passcode on the edison though, and it rapidly drops the connection while remaining paired
Tim Street
@tim2000s
Jan 15 2017 17:45
Hmmm. The steps I seem to have gone through don't reflect the docs. After I got it paired (without needing any passkeys), I've skipped straight to the trust AA:BB:CC:DD:EE:FF stage, and the device now connects with no issues, when I am in range of the iphone.
I'm not complaining, it just seems that different folks are getting different things to work for BLE set up?
Terri Lyman
@tatigger
Jan 15 2017 18:00
I have a quick question. To run Autotune as a one time run, do I have to update my ref0? I got this error when I tried to run it, so aside from any other errors I may have made to the command, autotune isn't recognized.
-bash: oref0-autotune: command not found
Tim Street
@tim2000s
Jan 15 2017 18:00
@tatigger Are you running the Dev branch of the code?
Terri Lyman
@tatigger
Jan 15 2017 18:01
No
Tim Street
@tim2000s
Jan 15 2017 18:04
That's probably why then
Terri Lyman
@tatigger
Jan 15 2017 18:08
Do I need to update to dev to just run the one-off? I don't want to run it continuously at this point
Scott Leibrand
@scottleibrand
Jan 15 2017 18:16
yes, you need oref0 dev on whichever rig (or VM) you want to run oref0-autotune on
if you want, you could update to dev so you could run oref0-autotune manually, but not re-run oref0-setup (or re-run it but don't choose any of the new optional features), and it shouldn't change anything for your existing setup. probably best to do that on a backup rig though.
Terri Lyman
@tatigger
Jan 15 2017 18:22
Ok, thanks @scottleibrand I was suspecting that was the case. :)
Tim Street
@tim2000s
Jan 15 2017 18:46
@scottleibrand . Something has gone spectacularly wrong with the BLE set up! I commented wifi out of wpa_supplicant.confand then when it ran oref0-online the Edison core-dumped....
I've not been able to get back onto the system to re-enable wifi and stop it picking up bluetooth.
Looking like a flashall is required.
PieterGit
@PieterGit
Jan 15 2017 18:56
I hit a ns-loop crash today. See openaps/oref0#334
anybody got a clue what went wrong?
C-Ville
@C-Ville
Jan 15 2017 19:07
does anybody have xdrip running as a BG source for the loop? for some reason my cellphone does not send the data to my edison. Anyone having an idea? I checked API_secret already
Dana Lewis
@danamlewis
Jan 15 2017 19:08
@tim2000s to answer an earlier Q, he's different things for different people for Bluetooth
Tim Street
@tim2000s
Jan 15 2017 19:09
Yes...
@danamlewis Have you notcied when you're running multiple edisons, they seem to interfere with each other - almost as though the random sleep period generator is synchonised?
Dana Lewis
@danamlewis
Jan 15 2017 19:16
No... What makes you think you're seeing that?
Tim Street
@tim2000s
Jan 15 2017 19:24
I'd noticed when there are two Edisons in the same place, one of them loses pump connectivity. It's happened consistently.
Tim Street
@tim2000s
Jan 15 2017 19:38
Looks like it's time for a new Edison. I can't even get flashall to work. Following the coredump, I can't write to the first partition on the Edison.
C-Ville
@C-Ville
Jan 15 2017 19:40
@tim2000s do you have an intel edison breakoutboard somewhere?
Eric
@ecc1
Jan 15 2017 19:41
Have you tried watching it boot via a serial console? I don't see how a "core dump" (really? what process dumped core?) could cause that.
C-Ville
@C-Ville
Jan 15 2017 19:42
I realized that flashing on an intel breakoutboard seems more reliable than flashing on the sparkfun board
Eric
@ecc1
Jan 15 2017 19:44
^^ agreed
Tim Street
@tim2000s
Jan 15 2017 19:49
@ecc1 I've watched it boot. It never gets to a command prompt and the display is not 100% clear.
Eric
@ecc1
Jan 15 2017 19:56
Where does it stop in the boot process? Can you post the output? I don't understand what you mean by the display not being clear.
Tim Street
@tim2000s
Jan 15 2017 19:56
Coming back to it after having powered down for 45mins, I'm not sure what happened. Run through boot and been able to get in okay. I've modified the wpa-supplicant back. It clearly got into an odd state. Wonder if it was too hot?
@ecc1 As it ran through boot, there'd be missing or overlayed characters.
Eric
@ecc1
Jan 15 2017 20:13
I'd try a different micro USB cable, maybe that connection is flaky.
Tim Street
@tim2000s
Jan 15 2017 20:29
If it was, I wouldn't expect it to be working properly now, which it is.
fezulin
@fezulin
Jan 15 2017 20:36

@ecc1 @C-Ville Just came here and read your last few posts. It seems I have a similar problem: I try to flash the Edison and it fails. I am connected with putty (com6, 11520), in the window I get lots of lines starting with [ ok ] and one line with [ failed ]:
[FAILED] Failed to start Load Kernel Modules.
after several further lines starting with [ ok ] booting is ending with
[ OK ] Reached target Rescue Mode.
When I try to restart "flashall.bat", it tells me I should "plug an reboot" the Edison. But "rebooting" seems to be what does not work at all.

I even installed "Intel Edsion Bord Configuration tools" - this does not bring any improvement: Automatic and manuall Flashing end up in the same way (described above). :worried:
Is there anybody who might give me a hint?

Eric
@ecc1
Jan 15 2017 20:36
You mean the boot output that was formerly garbled is now clear?
Toby Canning
@TC2013
Jan 15 2017 21:01
@ceben80 Thanks for the tip on Edison Voltage. I'll hunt that down and install it
Matthias
@ceben80
Jan 15 2017 21:05

@TC2013 afterwards, to prevent your rig from sudden shutdown due to power loss you can add this line to your cron:

* * * * * cd /home/edison/myopenaps && openaps report invoke monitor/edison-battery.json && awk '/batteryVoltage/ {if ($2<=3200)system("sudo shutdown -h now")}' ~/myopenaps/monitor/edison-battery.json

this will make a proper shutdown once 3200 mV is reached :wink:
Brian Rabinovitz
@bdr1177
Jan 15 2017 21:06
Can someone give an estimate of the approximate range of the explorer board from the pump?
Toby Canning
@TC2013
Jan 15 2017 21:06
Very nice
I see I have a edison-battery.json file, so it must be a part of dev
Aaron Michelson
@Aaybob
Jan 15 2017 21:06
@ceben80 I'm pretty sure the edison compute module does a clean shutdown upon low voltage.
probly saw in edison hardware doc
Dana Lewis
@danamlewis
Jan 15 2017 21:08
@bdr1177 varies based on where pump is on body and if pump is in line of range from board
Eric
@ecc1
Jan 15 2017 21:08
That's not a "clean shutdown" from the OS point of view; it's just powering off. You run the risk of filesystem and git repo corruption.
Matthias
@ceben80
Jan 15 2017 21:09
maybe but this cron line is a good approach in terms of low voltage :wink:
Eric
@ecc1
Jan 15 2017 21:09
(I was referring to @Aaybob's comment. I think your cronjob is a good idea.)
Matthias
@ceben80
Jan 15 2017 21:09
@ecc1 fully agree Eric
Brian Rabinovitz
@bdr1177
Jan 15 2017 21:09
@danamlewis I'm just trying to figure out if my daughter will have to wear the edison on her body at school, or if she can keep it in the bag that travels in the same room with her, which works fine for the dexcom?
Toby Canning
@TC2013
Jan 15 2017 21:10
We have done both ways @bdr1177
My son is 14 now and wants to minimize what he has to carry or show.
Brian Rabinovitz
@bdr1177
Jan 15 2017 21:11
@TC2013 That's helpful info. Thanks!
Toby Canning
@TC2013
Jan 15 2017 21:11
Currently, the only thing he keeps on his person is his phone for Dex G5 for BGs and the explorer rig--he doesn't carry any other supplies, but goes to the nurses if he needs them
I put the explorer rig in a zippo lighter pouch
I had to make some modifications though
I ground down the nuts on the Edison to be flush with the chip (worked well) and I took the battery connector off and soldered the battery on. That made it slim enough to fit more easily into the zippo belt clip
Dana Lewis
@danamlewis
Jan 15 2017 21:13
@bdr1177 you should do some tests and observe the mm tune to see how it does at various ranges
Tim Street
@tim2000s
Jan 15 2017 21:21
@ecc1 That's correct.
Brian Rabinovitz
@bdr1177
Jan 15 2017 21:21
@danamlewis What am I looking for in mmtune? Right now it says mmtune: "916.588", 5, -67. But I don't know how to interpret this.
I did look at the mmeowlink wiki, but didn't find an answer, unless I missed it.
Aaron Michelson
@Aaybob
Jan 15 2017 21:25
@ceben80 @ecc1 You convinced me. I have had some corruption problems.Going to ad that line in cron. Not a resource hog so every minute's okay?
Toby Canning
@TC2013
Jan 15 2017 21:25
@bdr1177 The 916.588 was the best frequency it found and the -67 rssi is really good
The 5 is the everything was perfect
0 would be "got nothing"
Brian Rabinovitz
@bdr1177
Jan 15 2017 21:26
@TC2013 Well that's good news since they're sitting right next to each other right now. Is there an acceptable range for each value?
Toby Canning
@TC2013
Jan 15 2017 21:27
Yes
It is working perfectly
Brian Rabinovitz
@bdr1177
Jan 15 2017 21:27
@TC2013 I meant as I test different distances, what ranges would be acceptable for each value.
Toby Canning
@TC2013
Jan 15 2017 21:28
Oh, :) sorry!
Yeah, really you're going to have to do a real world test
lab tests never pan out well for this in my experience
The radio interference is place specific
Brian Rabinovitz
@bdr1177
Jan 15 2017 21:29
So the measure will just be whether communication is occurring or not? There isn't anything I can tell from the numbers themselves?
Toby Canning
@TC2013
Jan 15 2017 21:29
Yes and no... one sec
The rssi doesn't matter if the loop completes. The information is in the logs, so when you go back to look if the loop completed, you can check to see how many failed. While you're at it, you an look at the rssi value anything over 90 will cause loop failure.
Most likely anyway.
Sometimes rssi values in the upper80s make the loop sometimes fail
My experience is that values lower than 85 generally mean the loop will run with few failures
Brian Rabinovitz
@bdr1177
Jan 15 2017 21:33
@TC2013 Got it. Thanks for the info!
Toby Canning
@TC2013
Jan 15 2017 21:34
You bet!
Dana Lewis
@danamlewis
Jan 15 2017 22:04
@bdr1177 and to your point...not sure we have that level of detail in the docs if you want to PR that in :smile: would be great
Brian Rabinovitz
@bdr1177
Jan 15 2017 22:05
@danamlewis will do
Dana Lewis
@danamlewis
Jan 15 2017 22:19
:+1: