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

16th
May 2016
Paul Martin
@Spazholio
May 16 2016 00:03
@danamlewis When you were saying that I’d want to enter carbs in order to use BWP (or was it AMA?) I assume you were talking about in the pump, and that I’d have to use the Bolus Wizard as opposed to just using the manual bolus feature, is that correct?
Dana Lewis
@danamlewis
May 16 2016 00:03
@tiaghose_twitter welcome! we did a small self-reported study, and we’re actually presenting details of the data at ADA Scientific Sessions in a few week. it’s embargoed, but you’ll generally find people saying they’re achieving ~1% reduction in A1C; time in range improvements, less highs/lows, better sleep, etc. (not sure if you’re asking about results of that to guide your decision, or you’re wondering about contributing data if you start looping. but we’ll have more ways soon for people to more formally donate or share their data for purposes of high level analyses like this)
@spazholio in order to use AMA, you need to enter carbs onsistently. You can set it so it pulls carbs from the pump using the bolus wizard OR from care portal in NS. But you’d need to do it in one place or the other.
Tia Ghose
@tiaghose_twitter
May 16 2016 00:05
Hi Dana, yes, I would definitely be interested in contributing data if we do decide to implement a loop. I'd also be happy to help out if data analysis is needed, though my coding skills are minimal.
Dana Lewis
@danamlewis
May 16 2016 00:06
@tiaghose_twitter doesn’t take much coding at this point in time to get a loop going :) but would welcome contributions in any of the areas you feel comfortable in!
also @tiaghose_twitter - animas ping is not a pump that’s compatible with openaps to date. it’s possible to communicate with it, but it does not allow remote temp basal setting, which is how openaps is designed to safely operate.
Paul Martin
@Spazholio
May 16 2016 00:08
@danamlewis I tend to use manual bolus since my ratio is 1:5 and it’s simple to do, but I could switch back to using the Wizard just as easily. This is only partly driven by my desire to see the pretty purple dots. =)
Dana Lewis
@danamlewis
May 16 2016 00:09
@Spazholio I feel you. I enter my stuff into DIYPS and use easy bolus ;) which is why I know it’s possible to use care portal (because my DIYPS entries then flow to care portal for NS, and thus into the loop)
it’s worth it, IMO, if you can bear to enter one way or the other for the beautiful (i’m so biased) purple line predictions :)
Paul Martin
@Spazholio
May 16 2016 00:09
@danamlewis Silly question - DIYPS = NS?
Dana Lewis
@danamlewis
May 16 2016 00:09
ps - check 1:1 message, had a Q for you @Spazholio
Scott Leibrand
@scottleibrand
May 16 2016 00:10
@tiaghose_twitter also, the biggest place we need lots more contributions is in improving the docs. so if you decide to set up a loop and want to pay it forward, please make note as you go along of the places where the docs are unclear or incomplete, so you can make edits to your own copy of the docs and submit the updates as a PR when you're ready.
Dana Lewis
@danamlewis
May 16 2016 00:10
@Spazholio not a silly Q, good q! no. DIYPS is my data entry and remote alerting system that I built before NS existed, and is the algorithm we first used to close the loop before we figured out how to do it open source with @bewest’s help!
Paul Martin
@Spazholio
May 16 2016 00:11
I know I kinda got BWP set up when I moved to dev, but @scottleibrand helped me deactivate it…I think. I may want to re-enable it. I like the idea of being a touch more aggressive with detecting ISF changes and keeping me more within my ranges. I’m having more lows than I’d like. I mean, they’re better than highs, but not as good as staying nice and stable.
Dana Lewis
@danamlewis
May 16 2016 00:11
@Spazholio for example, this is an old photo of the web interface..that’s so good (for me), it still looks just like this :) http://static3.businessinsider.com/image/55df85df9dd7cc19008b6826-2048-1536/diyps-prototype-with-pebble.jpg
Paul Martin
@Spazholio
May 16 2016 00:11
That link requires a u/p. =)
Dana Lewis
@danamlewis
May 16 2016 00:12
oh ha. hold on
Tia Ghose
@tiaghose_twitter
May 16 2016 00:12
Hi @danamlewis , yes, we're debating whether the lack of safety feature in Animas is a dealbreaker for us, and whether it's more dangerous than us just fumbling our way along with guesses as we are now. Right now very little of our son's TDD is from basal, so now sure if that makes the Animas implementation that Marius has done safer or less safer than us manually doing it.
Dana Lewis
@danamlewis
May 16 2016 00:12
@Spazholio fixed ;)
Tia Ghose
@tiaghose_twitter
May 16 2016 00:13
@scottleibrand I will keep that in mind. I am a writer by trade so am happy to help with documentation *if I am able to actually do that. I'm guessing it's hard to provide documentation for other people's code, especially when you're barely code literate yourself.
Scott Leibrand
@scottleibrand
May 16 2016 00:13
@Spazholio BWP is a NS feature. I think it might have been meal assist or autosens you're talking about.
Dana Lewis
@danamlewis
May 16 2016 00:13
@tiaghose_twitter the considerations are that you have to be a LOT more aware of whether the loop is working/not. you can’t fall back to normal scheduled basals, the way openaps does. it’s a lot more risky in the short term, in addition to having to engineer your own dosing algorithm based on combo microboluses.
@tiaghose_twitter and you’d be surprised, we get a TON of great feedback from people with little technical experience when working through the setup process. it’s not about the code, it’s about explaining in English what you do :)
Paul Martin
@Spazholio
May 16 2016 00:15
@scottleibrand Yeah, whatever was turned on when I went to the dev branch was what it was. I think it was meal assist. I’d be curious to look into autosense as well though. =)
Scott Leibrand
@scottleibrand
May 16 2016 00:15
@tiaghose_twitter what Dana said. we actually have done a fairly good job of the low-level technical documentation. it's the walkthroughs that could really use the attention of people with writing skills and/or more people who are actually running through the setup start to finish and can see the gaps that aren't obvious just looking at the docs.
@tiaghose_twitter and also :+1: on starting your loop with a Medtronic pump. Even if you're ok with the safety implications, the amount of technical work required to get an Animas loop running would be 5-10x what is required to do an OpenAPS instance with a Medtronic pump.
Paul Martin
@Spazholio
May 16 2016 00:17
I’m curious to know if people have a preference when it comes to a watch. I was considering getting one, but haven’t worn a watch it years. Went to look into them and some of the smart watches out there REALLY appeal to the geek in me. I know the Pebble is supported, but are any of the other ones, like the Gear S2 or any of those?
Scott Leibrand
@scottleibrand
May 16 2016 00:18
there are some good OpenAPS apps for the Pebble, but not for any other platforms yet. You'd need to write one from scratch or update one of the NS apps to show OpenAPS stuff.
Tia Ghose
@tiaghose_twitter
May 16 2016 00:19
@scottleibrand Yeah, that makes sense. Do you happen to know if you can use the same infusion sets as with the Animas pump? We'd be open if we can swap out the pumps when needed, but if implementing the APS means we'd have to do a site change, that might be a dealbreaker.
Jason Calabrese
@jasoncalabrese
May 16 2016 00:24
a lot of the clips on the sets can be swapped
that's what we did for a while, we used the inset-30 and then used the tubing from silhouettes
Scott Leibrand
@scottleibrand
May 16 2016 00:27
(you just end up with two reservoirs, each with their own tubing, so you can just unclip the site from one pump's tubing and swap on the other pump's tubing.)
Tia Ghose
@tiaghose_twitter
May 16 2016 00:28
@jasoncalabrese thanks, that is helpful.
if there’s anyone else who’s looped for 3+ days but hasn’t submitted a form…please do so in the next bit! :) Submitting a couple of slides for this week in the next two hours, I’m sure as soon as I hit “send” the number will be out of date again ;)
Jason Calabrese
@jasoncalabrese
May 16 2016 00:31
are the kids picking up any ground?
Dana Lewis
@danamlewis
May 16 2016 00:31
20 kids, 39 adults
1/3 is way more than I have been expecting
Jason Calabrese
@jasoncalabrese
May 16 2016 00:32
I think we may see a shift, I have some theories
for adults there's a better chance that they already had the pump
Tia Ghose
@tiaghose_twitter
May 16 2016 00:37
Honestly, I feel like this is extra needed for kids, especially the little ones who aren't verbal. Trying to make predictions about their behavior and their bodies day-to-day is just so hard, it's really hard to imagine an automated system that has the capability to adapt every 5 minutes would not be better than parents having to guess.
Dana Lewis
@danamlewis
May 16 2016 00:40
yea. @ddaniels1 just closed the loop for his 1yo, he might have some insight too specific to this situation
eyim
@eyim
May 16 2016 00:41
I think that now with 20 kids with the wide range of ages, it will give parents more confidence. I think it is always harder to test on your kids than yourself especially if you have to convince a spouse to go along with it
Jason Calabrese
@jasoncalabrese
May 16 2016 00:44
I saw a pattern the other day in the CGM group that I've seen before, one spouse tagging the other
think that translates into "we need to get this setup now"
I saw that a lot when NS was starting to grow fast
the other day @sulkaharo suggested I setup papertrail for my rigs, works really good and is so easy to setup
eyim
@eyim
May 16 2016 00:47
Yep - definitely feels like the momentum is growing
Jason Calabrese
@jasoncalabrese
May 16 2016 00:47
just got it setup on all my rigs
Screenshot 2016-05-15 17.48.59.png
you just need to edit 1 file at restart and restart rsyslog
eyim
@eyim
May 16 2016 00:51
very cool
Jason Calabrese
@jasoncalabrese
May 16 2016 00:52
with that even it a tunnel is down you can see what's going on
OpenAPS did pretty good on the camping trip, no BG alarms/checks or carb corrections all last night, even with a day tons of activity
AMA did seem to overshoot some during the day, the new minimum absorption rate should help with that
Dana Lewis
@danamlewis
May 16 2016 00:57
:+1:
eyim
@eyim
May 16 2016 00:58

Wow - congrats. Openaps worked really well for Alan at the tennis tournament. Picked up the only singles win against Team Canada and took the doubles match as well against some giants https://www.facebook.com/dan.willman/posts/10154131159427154?pnref=story

But was more impressed that he handled everything himself

Scott Leibrand
@scottleibrand
May 16 2016 01:00
:+1: yay for better tools giving kids more independence and self-confidence that they, too, can DIY. :)
Dana Lewis
@danamlewis
May 16 2016 01:00
Go Alan! :)
Jason Calabrese
@jasoncalabrese
May 16 2016 01:00
that's really great
I saw autosens hit 0.73, never saw that before
and was still needing extra carbs
Scott Leibrand
@scottleibrand
May 16 2016 01:03
:+1: Glad to see advanced features working well and making a difference.
eyim
@eyim
May 16 2016 01:03
The great thing especially for the teenage stage is that I think he realized for the first time a real benefit that he cares about - getting space from his parents. It got him a lot more interesting in the whole thing.
Dana Lewis
@danamlewis
May 16 2016 01:04
:)
Scott Leibrand
@scottleibrand
May 16 2016 01:07
Maybe you can even get him interested enough to start participating in building his own system like @inform880 does. :)
Jason Calabrese
@jasoncalabrese
May 16 2016 01:08
that's exactly what I was thinking
Scott Leibrand
@scottleibrand
May 16 2016 01:08
Nothing like taking over management of the system from Dad to instill some ownership and let him develop more independence. :)
I remember when my dad had me learn DOS 4.0 so I could take over managing the family's PC (and more importantly, get a modem). ;-)
Dana Lewis
@danamlewis
May 16 2016 01:10
Can I be the first to say that when I was a kid, we didn't have artificial pancreases? ;) (makes me feel better to get that one out of my system ;) :)
Jason Calabrese
@jasoncalabrese
May 16 2016 01:11
the way openaps is setup is a great way to learn too
Dana Lewis
@danamlewis
May 16 2016 01:12
:+1:
eyim
@eyim
May 16 2016 01:49
Yep - would love it. I'm not holding my breath though
Dana Lewis
@danamlewis
May 16 2016 02:01
Screen Shot 2016-05-15 at 7.01.22 PM.png
an updated contrast between # of loopers and their start dates; and the number of estimated total loop-hours from the community :)
(projecting >150,000 hours by June 1, and we’ll surpass that with any additional loopers who go live)
Bill Stackpole
@billstackpole
May 16 2016 02:05
@jasoncalabrese - did I see you say you are using Papertrail? Are you using it on the Pi? If so, what did you do to install it? (Pi is on arm processor - all papertrail senders seem to be for i386 or amd64...) Tnx for any insight you can provide
Ali Mazaheri
@amazaheri
May 16 2016 02:05
June is very special for us 😀
Bill Stackpole
@billstackpole
May 16 2016 02:06
(or are you not using the optional remote syslog2 for log aggregation?)
Dana Lewis
@danamlewis
May 16 2016 02:06
@amazaheri :)
Ali Mazaheri
@amazaheri
May 16 2016 02:06
our first openAPSriversary lol
Dana Lewis
@danamlewis
May 16 2016 02:06
hope there is GF cake to celebrate ;)
and kitten pics for @bewest!
Ali Mazaheri
@amazaheri
May 16 2016 02:07
lol
Dana Lewis
@danamlewis
May 16 2016 02:07
(or Kylo pics ;))
Ali Mazaheri
@amazaheri
May 16 2016 02:07
Kylo for sure GF yrs, kitten need to go to shelter 🐱💻
Scott Leibrand
@scottleibrand
May 16 2016 02:23
We're getting close to doing a 0.2.0 release of oref0 dev to master this evening (openaps/oref0#105). Dana still wants to add some stuff to the release notes. Does anyone else have anything else they think needs to be done before we release? @bewest @jasoncalabrese
Paul Martin
@Spazholio
May 16 2016 02:47
@scottleibrand You’ve been stupidly helpful these past few days, so I’m going to lead with thanking you for that. And then I’m going to mention that tomorrow, whenever you have a spare moment, I think I’d like to get AMA and auto-sense up and running. Or at least poke around with them a bit. I know it’s sorta dev-ish, but is there documentation on it? I may have missed it in the docs.
Dana Lewis
@danamlewis
May 16 2016 02:51
yes. there’s hints in the docs that point to other places that are .sh script like and not in English. we need people to turn them into English ;) so please take notes and plan to PR as you go
Scott Leibrand
@scottleibrand
May 16 2016 02:53
You might in fact have some experience with that particular .sh file. ;-)
Paul Martin
@Spazholio
May 16 2016 02:58

@danamlewis If you have any links to those docs that have hints and such, I’d love to see them. And I’d certainly be taking copious notes when implementing. =)

@scottleibrand You don’t say, eh?

Scott Leibrand
@scottleibrand
May 16 2016 02:59
No, I generally don't, unless I'm driving north. ;-)
Paul Martin
@Spazholio
May 16 2016 03:00
Oh, I live in Michigan and married a UP-er, so I’m fairly used to hearing it. =)
Dana Lewis
@danamlewis
May 16 2016 03:42
Matt Pressnall
@logichammer
May 16 2016 03:44
@scottleibrand, is now a decent time to hit you for AMA pointers?
If not, I have other days / nights that should work
Scott Leibrand
@scottleibrand
May 16 2016 03:44
@logichammer go for it
Matt Pressnall
@logichammer
May 16 2016 03:45
YES!
is there an AMA branch of openaps...because I am on dev and I don't see one in github...
Dana Lewis
@danamlewis
May 16 2016 03:45
@Spazholio (and watch as @logichammer works through it ;))
Matt Pressnall
@logichammer
May 16 2016 03:45
Ha...I'll take notes
Scott Leibrand
@scottleibrand
May 16 2016 03:45
the AMA branch is of oref0, not openaps
Dana Lewis
@danamlewis
May 16 2016 03:46
:+1: yes please. PRs desperately needed for AMA
Matt Pressnall
@logichammer
May 16 2016 03:47
aha
dana, which branch should I pull of the docs?
dev?
Dana Lewis
@danamlewis
May 16 2016 03:48
yea. not sure therre’s any major differences but dev to be safe
Scott Leibrand
@scottleibrand
May 16 2016 03:49
@bewest: you around this evening? I think we're ready to release oref0 dev to master and publish the updated packages. if you'll be around, I'd like to run it by you, but if not I'll give it a try myself.
Matt Pressnall
@logichammer
May 16 2016 03:53
OK...switched the branch on oref0...and rebooted...I can pull back model still...what now?
something about NPI or NMI?
Paul Martin
@Spazholio
May 16 2016 03:57
Oh, I’ll definitely be checking back here tomorrow in the morning. Thanks @danamlewis, @scottleibrand, @logichammer.
Tia Ghose
@tiaghose_twitter
May 16 2016 03:58
Hi folks, quick question: do all the Minimeds that are not the paradigm use infusion sets with the universal luer lock fittings found on many infusion sets?
Dana Lewis
@danamlewis
May 16 2016 04:01
I think all Medtronic pumps use the same infusion set types
Scott Leibrand
@scottleibrand
May 16 2016 04:01
@logichammer first thing I'd do now is try to run your loop alias and see what breaks
Matt Pressnall
@logichammer
May 16 2016 04:02
running it right now...so far looking good...but called via cron...so I should probably call it locally...so far nothing breaking
Scott Leibrand
@scottleibrand
May 16 2016 04:02
@tiaghose_twitter not sure if they might've done something different for the 640G in Europe and the upcoming 670G, but all the FDA-approved Medtronic pumps are the exact same physical form factor.
Matt Pressnall
@logichammer
May 16 2016 04:03
I'll call it locally and let you know
Scott Leibrand
@scottleibrand
May 16 2016 04:03
once that is done @logichammer you'll want to run the npm install steps and check again
Matt Pressnall
@logichammer
May 16 2016 04:04
Thanks so much! I was going to try to track down what that meant...
:)
I like the r
e
f
r
e
s
h
I
n
g
messages...
Dana Lewis
@danamlewis
May 16 2016 04:08
;)
Matt Pressnall
@logichammer
May 16 2016 04:08
Nice little ascii status indicators
Loop ran just fine...going to update npm
curious what these indicate as they happen:
>>>>>>>>>>>>>>>>>++--++->>>>>>>>++->+++++-+------+-+-+->>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>+-+->+>>+++++-------+-+>>>>>>>+>+++++-------+>++-+++>++->>>++--->>++-+-++-+-++++++-+-----+-----+---+-+++++++-+++++----++---+-+-+-+-+++++++++++++-+-++++++++-+++++++++++++ 38% of non-meal deviations negative (target 45%-50%)
Excess insulin resistance detected: ISF adjusted from 100 to 79.36507936507937
detect-sensitivity://text/shell/settings/autosens.json
reporting settings/autosens.json
Scott Leibrand
@scottleibrand
May 16 2016 04:11
I'm guessing nothing really changed before updating npm
you mean the > and + and - and = indicators?
Matt Pressnall
@logichammer
May 16 2016 04:12
yes
Scott Leibrand
@scottleibrand
May 16 2016 04:12
the > means deviation was high enough that we assume carbs were being absorbed, and disregard it for autosens purposes
(you're taking notes, right?) :-)
Matt Pressnall
@logichammer
May 16 2016 04:13
Damn...
Actually for this, I was going to include, yes...cool stuff
Scott Leibrand
@scottleibrand
May 16 2016 04:13
the + and - means the deviation was positive or negative
Matt Pressnall
@logichammer
May 16 2016 04:13
From expected
Scott Leibrand
@scottleibrand
May 16 2016 04:13
right. relative to BGI, which is BG impact of insulin
Matt Pressnall
@logichammer
May 16 2016 04:13
perfecto!
Love it
Scott Leibrand
@scottleibrand
May 16 2016 04:13
and if you see a =, it was exactly equal to BGI
Matt Pressnall
@logichammer
May 16 2016 04:14
Ha...I don't think we had any of those..
done updating
going to run the loop again
Scott Leibrand
@scottleibrand
May 16 2016 04:15
I think you'll probably only get = if you have time when the loop was down and not setting temps
then if there weren't any boluses either, BGI will be zero, so +0 deviations will be =
does that ISF of 79 seem legit? have you been running 20% resistant for the last 24h?
Matt Pressnall
@logichammer
May 16 2016 04:16
On AMA, can we start to use these reports to potentially tweak her values? I know you said on MA, that yo ushouldn't really do that because the data wasn't complete
Scott Leibrand
@scottleibrand
May 16 2016 04:16
the trick is knowing whether it's basals or ISF you need to adjust
Matt Pressnall
@logichammer
May 16 2016 04:17
I'm not sure it's been crazy this weekend and the AP has been saving our bacon but we need to review
Yes, we need to do some basal testing soon
Scott Leibrand
@scottleibrand
May 16 2016 04:17
without doing openaps/oref0#99 you can't really tell if your basals or your ISF is off
autosens just adjusts both at the same time, which is fine for day-to-day variations, but not specific enough for making changes to long-term settings
Matt Pressnall
@logichammer
May 16 2016 04:18
Hmmm...got to preflight fails
going to loop again
Battery at 1.32v == too low correct?
Scott Leibrand
@scottleibrand
May 16 2016 04:21
does the pump backlight come on?
if so, you can still loop
start drilling down on the aliases that are failing, and run their component reports individually
Matt Pressnall
@logichammer
May 16 2016 04:24
starting the debug process
battery issue...loop running
Scott Leibrand
@scottleibrand
May 16 2016 04:36
If no objections, merging openaps/oref0#105 and starting the release process for oref 0.2.0
Matt Pressnall
@logichammer
May 16 2016 04:39
was getting spotty mmtune performance so rebooted after updating NPM...we'll see
Dana Lewis
@danamlewis
May 16 2016 04:42
:+1: on release
Scott Leibrand
@scottleibrand
May 16 2016 04:43
merged. now to do the npm tagging and release stuff
Ben West
@bewest
May 16 2016 04:46
great
Matt Pressnall
@logichammer
May 16 2016 04:46
OK...I had grabbed a cheap battery figuring it would be ok for some quick testing...nope...new lithium and I can talk again...loop is running for the first time after updating
what do I need to add in to get AMA working?
Scott Leibrand
@scottleibrand
May 16 2016 04:47
@bewest did a npm version patch before figuring out I needed npm version major, then tried to roll that back to not create the extra patch release, but apparently that has already been pushed to npm...
Ben West
@bewest
May 16 2016 04:47
minor
npm version minor
Scott Leibrand
@scottleibrand
May 16 2016 04:47
sorry, yeah
Ben West
@bewest
May 16 2016 04:47
you can delete the tag
Scott Leibrand
@scottleibrand
May 16 2016 04:47
that's what I did
if I haven't pushed yet, and did a git reset, do the tags survive?
Ben West
@bewest
May 16 2016 04:48
nah, if you did a git tag -d badtag
and did reset to move the pointer back
Scott Leibrand
@scottleibrand
May 16 2016 04:48
k, trying that
Ben West
@bewest
May 16 2016 04:48
it should be gone
I like this thing a lot
Matt Pressnall
@logichammer
May 16 2016 04:49
Updated a NS site to .9 beta 2 and know I need to add tags and then openaps reports?
Ben West
@bewest
May 16 2016 04:49
this creates a release-x.y.z branch with the tag and everything
if that's ok/checks out I merge that immediately and use the tag that was created
this also includes a rollback before the tool applies the tag, similarly
Scott Leibrand
@scottleibrand
May 16 2016 04:49
we've been manually doing the steps in https://github.com/openaps/oref0/blob/master/release-master.sh
is the https://github.com/bewest/mmcsv/blob/master/prep_repo_release.sh better for this situation? looks like our tags are of the form v0.1.5
Ben West
@bewest
May 16 2016 04:50
yeah, I saw that
it's similar, yeah same form
I'm calling npm version, I just didn't like the order/way they apply the tag
I wanted the changelog imported in with the tag
Scott Leibrand
@scottleibrand
May 16 2016 04:51
speaking of which, where do you want the changelog / release notes besides the PR?
Ben West
@bewest
May 16 2016 04:51
that's what this adds, basically
well, that's kind of why I'm bringing it up, open to suggestions
this script generates it as Release.md
for cgm-remote-monitor, jason has done stellar job porting the release notes into the tag itself on github
so github's tag annotation contains full release notes
could save it in repo as a file, in wiki, in git tag, perhaps import it into a page on docs site
I'm fine with the using the github tag annotation, on github it's associated with releases anyway, and means one less file to maintain
Scott Leibrand
@scottleibrand
May 16 2016 04:53
you've lost me a long time ago
I don't care how we do it: tell me how you'd prefer and we'll do it that way this time
oh, I don't have a preference
Scott Leibrand
@scottleibrand
May 16 2016 04:55
well, you wrote up a bunch of release notes that you want to live somewhere
do we manually commit those to Release.md?
Ben West
@bewest
May 16 2016 04:56
I just thought it'd be nice to have some release notes and wondering what would happen if a stubbed out something that looked like release notes
Scott Leibrand
@scottleibrand
May 16 2016 04:56
we ran with that and wrote up some additional stuff. just not sure where to put it
Ben West
@bewest
May 16 2016 04:57
process wise, I ran out of opinions, yeah partly because I don't want to decide where it goes, partly because I don't envision myself being main maintainer/author of these docs if we decide we we'd like them
for this, what I had planned on doing as last resort was just to add it to the tag notes on github
Scott Leibrand
@scottleibrand
May 16 2016 04:57
let's consider these release notes to be a separate thing that lives just in oref0. we'll update openaps/docs separately
Ben West
@bewest
May 16 2016 04:57
and if someone wants to put some pressure to record it in Release.md or something, there are ways to do it
Scott Leibrand
@scottleibrand
May 16 2016 04:58
so I'm thinking I do your git changelog -t ${NEW_TAG} ${RELEASE_NOTES} thing and then manually put our summary at the top, then commit that as Release.md
Ben West
@bewest
May 16 2016 04:58
yeah, the previous comment is the output of running it
it drops you into an editor to preview/edit the ist
Scott Leibrand
@scottleibrand
May 16 2016 04:58
yeah, easier to re-run than copy-paste it. :)
Ben West
@bewest
May 16 2016 04:59
in practice, I find it easier to develop summary somewhere like that PR
Scott Leibrand
@scottleibrand
May 16 2016 04:59
I don't think anyone feels like editing the mess 'o commits this time
Ben West
@bewest
May 16 2016 04:59
yeah
Scott Leibrand
@scottleibrand
May 16 2016 05:00
git: 'changelog' is not a git command. See 'git --help'.
Ben West
@bewest
May 16 2016 05:00
ah
apt-get install git-extras
Scott Leibrand
@scottleibrand
May 16 2016 05:02
that works, but then it just generates an empty file with a v0.2.0 / 2016-05-15 header
Ben West
@bewest
May 16 2016 05:02
ah, because it's post release
that command takes differences between now and last release
to latest tag
if you already ran verion, it's too late to add the notes to the repo
Scott Leibrand
@scottleibrand
May 16 2016 05:03
I can delete tag and reset (again)
Ben West
@bewest
May 16 2016 05:03
that's my my thing runs the npm version to figure out what the new tag would be
then it backs it out, generates the changes with the new version name, and after then does the version thing again to keep it
Scott Leibrand
@scottleibrand
May 16 2016 05:04
ok, doing that manually
ok, got a committed Release.md with high level summary and the changelog, and ready to npm publish and git push --tags origin master - anything else I might be missing?
--- a/package.json
+++ b/package.json
@@ -1,6 +1,6 @@
 {
   "name": "oref0",
-  "version": "0.1.4",
+  "version": "0.2.0",
Matt Pressnall
@logichammer
May 16 2016 05:09
@scottleibrand, if you are busy with post release no worries...can you send me a list of things to add to my NS website and additions to reports / things to get AMA working and I can do it later...whenever you have a chance. I promise, it's going into documentation!
Scott Leibrand
@scottleibrand
May 16 2016 05:10
@bewest you mentioned earlier npm upload - is that done by npm publish?
@logichammer I think I can go back to multitasking now
Matt Pressnall
@logichammer
May 16 2016 05:10
cool...would love to get it banged out tonight if possible!
Scott Leibrand
@scottleibrand
May 16 2016 05:10
so your loop runs fine one the advanced-meal-assist branch?
Matt Pressnall
@logichammer
May 16 2016 05:10
yes
and I have a NS website .9 beta 2
so I should be set
Scott Leibrand
@scottleibrand
May 16 2016 05:11
cool. I thought it would break. maybe I don't have to do quite as much work to merge AMA to dev. :)
Matt Pressnall
@logichammer
May 16 2016 05:11
ha!
Scott Leibrand
@scottleibrand
May 16 2016 05:11
do you have meal-assist and/or autosens set up yet?
Matt Pressnall
@logichammer
May 16 2016 05:11
Yes, it is running just fine after a reboot
Using your loop and yes, based on the log it looks like both are working
autosense for sure
Scott Leibrand
@scottleibrand
May 16 2016 05:12
ok, and which openaps-sh branch did you start with, do you remember?
Matt Pressnall
@logichammer
May 16 2016 05:12
yes, the ama setup.sh
Scott Leibrand
@scottleibrand
May 16 2016 05:12
ah, ok. so that's probably why nothing broke when you did oref0 AMA branch
you have all the reports set up already
what does your meal.json look like?
Matt Pressnall
@logichammer
May 16 2016 05:13
all the ones from your setup...I haven't touch openaps other than to correct paths
{"carbs":60,"boluses":2.2,"mealCOB":1}
Scott Leibrand
@scottleibrand
May 16 2016 05:13
and your enact/suggested.json?
Matt Pressnall
@logichammer
May 16 2016 05:14
{"temp":"absolute","bg":189,"tick":"+5","eventualBG":213,"snoozeBG":213,"predBGs":{"IOB":[189,193,197,200,202,204,207,209,211,211,211,210,208,207,205,203,202,200,198,196,195,193,191,189,188,186,184,182,180,178,177,175,173,171,170,168,167,165,164,163,162,161,160,159,158,158,157,157,156],"aCOB":[189,186,183,179,175,172,170,169,168,166,165,163,161,159,158,156,154,153,151,149,147,146,144,142,140,138,137,135,133,131,129,127,126,124,122,121,119,118,117,115,114,113,112,112,111,110,110,109],"COB":[189,187,183,180,176,172,171,170,169,167,165,164,162,160,159,157,155,153,152,150,148,146,145,143,141,139,137,136,134,132,130,128,127,125,123,122,120,119,117,116,115,114,113,113,112,111,111,110]},"reason":"COB: 1, Dev: 53, BGI: -3.78, ISF: 84, Target: 115; Eventual BG 213>=120, adj. req. rate:2.7 to maxSafeBasal:1.0, temp 1 >~ req 1.0U/hr"}
Scott Leibrand
@scottleibrand
May 16 2016 05:14
ah, sweet
Matt Pressnall
@logichammer
May 16 2016 05:14
I like to hear that
Scott Leibrand
@scottleibrand
May 16 2016 05:14
those are the arrays that will turn into purple lines :)
Matt Pressnall
@logichammer
May 16 2016 05:14
beauty
Scott Leibrand
@scottleibrand
May 16 2016 05:14
have you done any of the ENABLE stuff on NS yet?
Matt Pressnall
@logichammer
May 16 2016 05:14
but does that mean the algo is already to AMA?
nope
Scott Leibrand
@scottleibrand
May 16 2016 05:14
yes
k
Matt Pressnall
@logichammer
May 16 2016 05:14
nice!
what do I need to add to NS?
Scott Leibrand
@scottleibrand
May 16 2016 05:15
env ENABLE="openaps pump basal pushover maker careportal iob bwp raw"
env SHOW_FORECAST="openaps ar2"
env DEVICESTATUS_ADVANCED=true
Ben West
@bewest
May 16 2016 05:15
yeah ,npm publish sounds right
Scott Leibrand
@scottleibrand
May 16 2016 05:15
not all of those are needed of course
@bewest ok, proceeding with npm publish && git push --tags origin master then
Matt Pressnall
@logichammer
May 16 2016 05:15
adding...thanks!
Scott Leibrand
@scottleibrand
May 16 2016 05:16
@logichammer important one to add is env SHOW_FORECAST="openaps ar2"
I think you'll already have the ENABLE="openaps" etc.
Matt Pressnall
@logichammer
May 16 2016 05:16
got it and restarting right now!
Scott Leibrand
@scottleibrand
May 16 2016 05:17
ok, oref0@0.2.0 published to npm and tags and release notes pushed to github
Dana Lewis
@danamlewis
May 16 2016 05:18
:+1:
Scott Leibrand
@scottleibrand
May 16 2016 05:18
lemme fix the paste error on the high-level summary indentation
Matt Pressnall
@logichammer
May 16 2016 05:19
YEHAW!!!!
Up and running!
Ben West
@bewest
May 16 2016 05:19
lot of work in that one
@logichammer pics (screenshot) or it didn't happen! ;-)
Ben West
@bewest
May 16 2016 05:21
yeah, that's one reason I like the little release script, it creates a unique release branch, so you can preview the branch together
it's basically a frozen branch, but it's trivial to burn and do again
Scott Leibrand
@scottleibrand
May 16 2016 05:21
if you want to get that set up to run next time, I'd have no objection
didn't want to add more plumbing at 10pm on a sunday night before releasing though :)
Ben West
@bewest
May 16 2016 05:21
yeah, definitely
Ben West
@bewest
May 16 2016 05:22
congrats to all
Matt Pressnall
@logichammer
May 16 2016 05:22
She went high while things were offline but happy to have this added in now!
:thumbsup:
Scott Leibrand
@scottleibrand
May 16 2016 05:22
@logichammer sweet! :)
Ben West
@bewest
May 16 2016 05:23
oh, one more thing
git branch -d dev
Scott Leibrand
@scottleibrand
May 16 2016 05:23
I'd rather not actually
Dana Lewis
@danamlewis
May 16 2016 05:23
Congrats @logichammer ! :)
Ben West
@bewest
May 16 2016 05:23
git checkout -b dev
Scott Leibrand
@scottleibrand
May 16 2016 05:23
that will break quick-src.sh
Ben West
@bewest
May 16 2016 05:23
no
we're fixing it
Scott Leibrand
@scottleibrand
May 16 2016 05:23
oh, ok
Matt Pressnall
@logichammer
May 16 2016 05:23
@scottleibrand , since I used this as my starting point:
Ben West
@bewest
May 16 2016 05:23
maybe git commit --allow-empty
Matt Pressnall
@logichammer
May 16 2016 05:23
Can I use that in documentation?
Scott Leibrand
@scottleibrand
May 16 2016 05:24
@bewest functionally equivalent to a merge of master to dev?
Matt Pressnall
@logichammer
May 16 2016 05:24
If not, what should I tell folks?
Ben West
@bewest
May 16 2016 05:24
yeah, cause branch was already deleted from github :-)
we just need a new dev based on master
Scott Leibrand
@scottleibrand
May 16 2016 05:24
@logichammer I'd rather not put that script in docs
Ben West
@bewest
May 16 2016 05:24
so delete your local version of dev, to match github
Dana Lewis
@danamlewis
May 16 2016 05:24
@logichammer needs to be translated to English like rest of docs
Ben West
@bewest
May 16 2016 05:24
git checkout -b dev master creates new dev from master
git commit --allow-empty creates new empty commit on dev
Matt Pressnall
@logichammer
May 16 2016 05:25
OK...there was magic for me by doing this but I'll run it by you guys /. try to reverse engineer what was added
Ben West
@bewest
May 16 2016 05:25
git push -u origin dev to create new dev branch on github
Scott Leibrand
@scottleibrand
May 16 2016 05:25
Ben West
@bewest
May 16 2016 05:26
oh, actually
Dana Lewis
@danamlewis
May 16 2016 05:26
@logichammer yea. Just need to break it down to step by step and explain what people are doing in English In addition to what they're running
Ben West
@bewest
May 16 2016 05:26
ok
I'll catch it up
Matt Pressnall
@logichammer
May 16 2016 05:26
And confirmation: if I am passing meal.json into enact/suggested.json, then I am going to be running AMA, correct?
Scott Leibrand
@scottleibrand
May 16 2016 05:26
if you're doing that on the AMA branch, yes
Matt Pressnall
@logichammer
May 16 2016 05:26
Gotcha, @danamlewis
Ben West
@bewest
May 16 2016 05:26
ok, done
Scott Leibrand
@scottleibrand
May 16 2016 05:27
thx. that was just a merge?
Ben West
@bewest
May 16 2016 05:27
it was just git checkout dev; git merge master
yeah
Scott Leibrand
@scottleibrand
May 16 2016 05:27
cool
Ben West
@bewest
May 16 2016 05:27
usually I like to delete dev altogether
Matt Pressnall
@logichammer
May 16 2016 05:27
Good...and how long does bolus snooze last or is it bypassed with AMA?
Ben West
@bewest
May 16 2016 05:27
then create new branch cleanly cut from dev
Scott Leibrand
@scottleibrand
May 16 2016 05:27
bolus snooze has been shorted to DIA/3 or DIA/4 in AMA
Matt Pressnall
@logichammer
May 16 2016 05:27
That's what I needed, Dana!
Ben West
@bewest
May 16 2016 05:27
probably doesn't matter too much, just makes me feel cleaner
Scott Leibrand
@scottleibrand
May 16 2016 05:28
yeah, I think they're the same in this case, so either way works for me
Matt Pressnall
@logichammer
May 16 2016 05:28
Good to hear
Ben West
@bewest
May 16 2016 05:28
now we stub out new release area based on dev
Scott Leibrand
@scottleibrand
May 16 2016 05:28
I never have understood that step
why do you stub out release when you haven't got anything to release yet?
Ben West
@bewest
May 16 2016 05:29
well, usually I change the version to say -dev
not sure if we can do that here
but that creates enough of a change to stub out a release
Scott Leibrand
@scottleibrand
May 16 2016 05:29
oh, is this so you can publish dev to npm?
Matt Pressnall
@logichammer
May 16 2016 05:29
And how do I set up a pump battery low warning alert? Is that ready / in the beta 2?
Ben West
@bewest
May 16 2016 05:29
in this case, it would mean having to edit it again to get the tool to work properly
so instead
Scott Leibrand
@scottleibrand
May 16 2016 05:30
@logichammer do you have a Pump pill in NS?
Matt Pressnall
@logichammer
May 16 2016 05:30
I do
Scott Leibrand
@scottleibrand
May 16 2016 05:30
and it shows battery level in the tooltip text?
Matt Pressnall
@logichammer
May 16 2016 05:30
will it show something when lowish magically?
yes
Ben West
@bewest
May 16 2016 05:30
git commit --allow-empty allows me to create an empty commit
which is again enough of a change to create a pr: openaps/oref0#115
Dana Lewis
@danamlewis
May 16 2016 05:31
Even better @logichammer, can also set it up to email or alert elsewhere when it starts dropping
Matt Pressnall
@logichammer
May 16 2016 05:31
How/ where?
sounds awesome!
I run into this too much
Scott Leibrand
@scottleibrand
May 16 2016 05:31
by default it will turn yellow when pump battery is low
based on a default voltage threshold
Matt Pressnall
@logichammer
May 16 2016 05:32
nice!
Scott Leibrand
@scottleibrand
May 16 2016 05:32
you can change that threshold with a variable, but I don't recall which one. @jasoncalabrese would remember...
or it might be searchable in gitter
Matt Pressnall
@logichammer
May 16 2016 05:32
I found something before in the docs
tracking it down now
Scott Leibrand
@scottleibrand
May 16 2016 05:32
@bewest so what, conceptually, is the point of opening that PR now, vs. when we're ready to start discussing another release?
Matt Pressnall
@logichammer
May 16 2016 05:33
Regardless of whether you use alkaline or lithium, you may want to consider a Nightscout alarm to alert you to when the battery is running low. You can do this by setting (in your Nightscout config vars) PUMP_WARN_BATT_V to 1.39 (if using lithium batteries, as is most common) and adding battery to your PUMP_FIELDS so that voltage is displayed on your Nightscout site.
Dana Lewis
@danamlewis
May 16 2016 05:33
:+1:
Matt Pressnall
@logichammer
May 16 2016 05:34
My goal is to have this written up by the end of the week...more to come folks and thanks so much!!!!
Scott Leibrand
@scottleibrand
May 16 2016 05:34
@bewest I could see having an always-open release PR as just an invitation for someone to accidentally hit the green button
so wondering what having one open buys us that makes it worthwhile
Matt Pressnall
@logichammer
May 16 2016 05:35
@jasoncalabrese , is there a way to set up email alerts for low batteries? I know about the pill changing color which is awesome!
Dana Lewis
@danamlewis
May 16 2016 05:35
Awesome @logichammer
Ben West
@bewest
May 16 2016 05:37
dev goes through a life cycle
it has a beginning, middle, end, I guess
this is the beginning
if dev is not headed towards release, one has to ask why it's in dev, I suppose
Scott Leibrand
@scottleibrand
May 16 2016 05:37
but why should a PR occur at the beginning? why not have a dev branch with no PR until it gets to the point someone things it should be released, and opens the PR to get comments on that idea?
Ben West
@bewest
May 16 2016 05:38
partly to help people point to the right branch
partly to be transparent about the process
Scott Leibrand
@scottleibrand
May 16 2016 05:38
does having a PR open change which branches allow you to open PRs against them?
Ben West
@bewest
May 16 2016 05:39
no, when I come across a new org/repo/project, I take a look at open issues/prs, lastest commits, etc
so when I see a dev branch with an active PR, I know they're using git flow
and a whole bunch of process knowledge basically comes along for the ride
I know without even reading the contrib guidelines, my PR should probably go there
it's kind of just planning central in general
Scott Leibrand
@scottleibrand
May 16 2016 05:40
k
Ben West
@bewest
May 16 2016 05:40
I can see where in the process someone is: if they just had a release and are at beginning of another cycle
or if they are closing one up, it might be good time to hurry to get something in
those all kind of give off smells that might attract or detract contributors
could put a hit list/roadmap for things we'd like in next cycle or plan on focusing on next so people know/can get oriented to contributing faster
in theory
Scott Leibrand
@scottleibrand
May 16 2016 05:43
so in this case, we have an entire new dev branch in waiting: advanced-meal-assist
Ben West
@bewest
May 16 2016 05:43
ah great
Scott Leibrand
@scottleibrand
May 16 2016 05:43
so we might want to open a PR from there and make a hitlist of things we need to fix before it can be merged to dev
Ben West
@bewest
May 16 2016 05:44
is there PR on dev? are there reasons it's should not be considered for next release?
yeah
Scott Leibrand
@scottleibrand
May 16 2016 05:44
it's not ready yet, but it will need to go to dev when it is, then get tested there before a release to master
Matt Pressnall
@logichammer
May 16 2016 05:44

Also, @jasoncalabrese , one thing I am confused about is that dev docs read:

Regardless of whether you use alkaline or lithium, you may want to consider a Nightscout alarm to alert you to when the battery is running low. You can do this by setting (in your Nightscout config vars) PUMP_WARN_BATT_V to 1.39 (if using lithium batteries, as is most common) and adding battery to your PUMP_FIELDS so that voltage is displayed on your Nightscout site.

Confused by the PUMP_FIELDS comment. I already have battery showing in my Pump Pill on NS (which in OpenAPS gets uploaded as part of the devicestatus upload) and my application variables in NS look like:
ENABLE="openaps pump basal pushover maker careportal iob bwp raw"
SHOW_FORECAST="openaps ar2"
DEVICESTATUS_ADVANCED=true

So I am NOT doing anything with a PUMP_FIELDS variable but getting this data shown to me...Just wondering why or if we can remove that line from the dev docs? Thoughts? And many thanks!!!

Ben West
@bewest
May 16 2016 05:44
ah ok
ok, added comment to that effect
affect?
Scott Leibrand
@scottleibrand
May 16 2016 05:47
Effect.
Thx. Will flesh that out later.
Ben West
@bewest
May 16 2016 05:49
nice, yeah, just going to add a bit to the tag as well
for extra google juice
Dana Lewis
@danamlewis
May 16 2016 05:53
:+1:
Sulka Haro
@sulkaharo
May 16 2016 07:14
Re Papertrail -> yeah it's surprisingly easy to setup and makes me comfortable to turn the local logging to almost zero. I should add it to the documentation as an optional step
Ben West
@bewest
May 16 2016 08:06
yeah, it's surprisingly easy :)
Ben West
@bewest
May 16 2016 10:02
instructions for papertrail are basically: sign up for papertrail, choose the rsyslog option, do what it says
John Sjolund
@sjolundjohn
May 16 2016 10:07
Having some trouble getting values from Dexcom, can anybody spot what I may be doing wrong?
pi@raspberrypi:~/openaps $ openaps use dexcom iter_glucose 1
Traceback (most recent call last):
  File "/usr/local/bin/openaps-use", line 4, in <module>
    __import__('pkg_resources').run_script('openaps==0.1.0', 'openaps-use')
  File "build/bdist.linux-armv7l/egg/pkg_resources/__init__.py", line 719, in run_script
  File "build/bdist.linux-armv7l/egg/pkg_resources/__init__.py", line 1504, in run_script
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.0-py2.7.egg/EGG-INFO/scripts/openaps-use", line 63, in <module>
    app( )
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.0-py2.7.egg/openaps/cli/__init__.py", line 51, in __call__
    self.run(self.args)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.0-py2.7.egg/EGG-INFO/scripts/openaps-use", line 57, in run
    output = app(args, self)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.0-py2.7.egg/openaps/uses/__init__.py", line 92, in __call__
    return self.method.selected(args)(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.0-py2.7.egg/openaps/uses/__init__.py", line 31, in __call__
    return self.method(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.0-py2.7.egg/openaps/uses/use.py", line 45, in __call__
    output = self.main(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.0-py2.7.egg/openaps/vendors/dexcom.py", line 271, in main
    for item in self.dexcom.iter_records(self.RECORD_TYPE):
AttributeError: 'NoneType' object has no attribute 'iter_records'
For clarity the device is called 'dexcom'
John Sjolund
@sjolundjohn
May 16 2016 10:35
Scratch that - poor USB cable I believe
Frank Best
@efbest
May 16 2016 10:35
my loop is interrupted and I can't figure out why: preflight is ok, 'suggested' is produced, but 'enact' throws an error - why?
determine-basal://text/shell/enact/suggested.json
{"delta":4,"glucose":196,"avgdelta":2.6666666666666665}
{"duration":7,"rate":8,"temp":"absolute"}
{"bolusiob":0,"iob":1.5805234482610229,"activity":0.06711572651515155}
{"max_iob":15,"type":"current","dia":4,"current_basal":2.2,"max_daily_basal":4,"max_basal":15,"min_bg":90,"max_bg":125,"sens":12}
reporting enact/suggested.json
MMT754://JSON/set_temp_basal/enact/enact.json
enact/enact.json  raised  byte must be in range(0, 256)
Traceback (most recent call last):
  File "/usr/local/bin/openaps-report", line 4, in <module>
    __import__('pkg_resources').run_script('openaps==0.1.0', 'openaps-report')
  File "/usr/local/lib/python2.7/dist-packages/setuptools-21.0.0-py2.7.egg/pkg_resources/__init__.py", line 719, in run_script
    self.require(requires)[0].run_script(script_name, ns)
  File "/usr/local/lib/python2.7/dist-packages/setuptools-21.0.0-py2.7.egg/pkg_resources/__init__.py", line 1504, in run_script
    exec(code, namespace, namespace)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.0-py2.7.egg/EGG-INFO/scripts/openaps-report", line 82, in <module>
    app( )
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.0-py2.7.egg/openaps/cli/__init__.py", line 51, in __call__
    self.run(self.args)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.0-py2.7.egg/EGG-INFO/scripts/openaps-report", line 75, in run
    output = app(args, self)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.0-py2.7.egg/openaps/cli/subcommand.py", line 52, in __call__
    return self.method.main(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.0-py2.7.egg/openaps/reports/invoke.py", line 40, in main
    output = task.method(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.0-py2.7.egg/openaps/uses/use.py", line 45, in __call__
    output = self.main(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.0-py2.7.egg/openaps/vendors/medtronic.py", line 317, in main
    results = self.upload_program(program)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.0-py2.7.egg/openaps/vendors/medtronic.py", line 410, in upload_program
    return self.pump.model.set_temp_basal(**program)
  File "/usr/local/lib/python2.7/dist-packages/decocare-0.0.21-py2.7.egg/decocare/models/__init__.py", line 188, in set_temp_basal
    result = self._set_temp_basal(**basals)
  File "/usr/local/lib/python2.7/dist-packages/decocare-0.0.21-py2.7.egg/decocare/models/__init__.py", line 25, in __call__
    self.response = inst.session.query(self.msg, **kwds)
  File "/usr/local/lib/python2.7/dist-packages/decocare-0.0.21-py2.7.egg/decocare/session.py", line 101, in query
    self.execute(command)
  File "/usr/local/lib/python2.7/dist-packages/decocare-0.0.21-py2.7.egg/decocare/session.py", line 98, in execute
    return super(type(self), self).execute(command)
  File "/usr/local/lib/python2.7/dist-packages/decocare-0.0.21-py2.7.egg/decocare/session.py", line 35, in execute
    self.transfer( )
  File "/usr/local/lib/python2.7/dist-packages/decocare-0.0.21-py2.7.egg/decocare/session.py", line 61, in transfer
    return self.stick.transmit_packet(self.command)
  File "/usr/local/lib/python2.7/dist-packages/decocare-0.0.21-py2.7.egg/decocare/stick.py", line 856, in transmit_packet
    result = self.process( )
  File "/usr/local/lib/python2.7/dist-packages/decocare-0.0.21-py2.7.egg/decocare/stick.py", line 485, in process
    raw = self.send_force_read( )
  File "/usr/local/lib/python2.7/dist-packages/decocare-0.0.21-py2.7.egg/decocare/stick.py", line 601, in send_force_read
    self.link.write(reader.format( ))
  File "/usr/local/lib/python2.7/dist-packages/decocare-0.0.21-py2.7.egg/decocare/stick.py", line 400, in format
    return bytearray(packet)
ValueError: byte must be in range(0, 256)
No CGM data
pi@raspberrypi ~/Frank_openaps $
John Sjolund
@sjolundjohn
May 16 2016 12:15

I am sitting here at the finish line and stuck which stinks. The output of enact/suggested.json looks fine, which prints to enact/enacted.json.

Looking at earlier comments all looks good.

I can see which command to sent tem_basal to pump. When I invoke reports all seems to work fine, but I can't confirm it actually gets sent to pump. IE basal rate on pump remains the same

I have troubleshooted every aspect, seems to me that either my command to send temp basal is wrong OR format is wrong. Though I get no errors.

Tips for how to figure out this very last part?

Seems I need to switch openaps use to openaps invoke to send to pump, but not working
John Males
@johnmales
May 16 2016 12:27
How do I disable meal assist in determine-basal?
Paul Martin
@Spazholio
May 16 2016 12:38
@sjolundjohn I'm no expert, but I believe the command that I use (from @scottleibrand's setup.sh script) is: enact = ! bash -c "rm enact/suggested.json; openaps invoke enact/suggested.json && cat enact/suggested.json && grep -q duration enact/suggested.json && ( openaps invoke enact/enacted.json && cat enact/enacted.json ) || echo No action required". Does that look like yours does?
Actually, that's the alias in my openaps.ini but I think you can figure out how that goes. =)
John Sjolund
@sjolundjohn
May 16 2016 12:39
Hi @Spazholio , I am actually a step before this, and the loop, it is getting the x3 values needed for set_temp_basal sent to pump
openaps use pump set_temp_basal enact/enacted.json
Seems like it should do it to me
Paul Martin
@Spazholio
May 16 2016 12:42
When I invoke my enact/enacted.json, here's what's in it:
[report "enact/enacted.json"]
device = pump
input = enact/suggested.json
use = set_temp_basal
reporter = JSON
So I think it's using enact/suggested.json and the enact/enacted.json is actually the output of that.
Looking at my openaps.ini to see what it's outputting here, just a moment.
(as an aside, I'd highly recommend taking a look at @scottleibrand's setup.sh script. It did a lot of the heavy lifting for me)
John Sjolund
@sjolundjohn
May 16 2016 12:46
@Spazholio Ok some work to do, here is what I get:
pi@raspberrypi:~/openaps $ cat enact/enacted.json
{
  "requested": {
    "duration": 30, 
    "rate": 1, 
    "temp": "absolute"
  }, 
  "bg": 123, 
  "temp": "percent", 
  "snoozeBG": 99, 
  "timestamp": "2016-05-16T14:45:37.278994", 
  "rate": 0, 
  "reason": "Eventual BG 99<100.8, setting 0.978U/hr", 
  "eventualBG": 99, 
  "recieved": false, 
  "duration": 0, 
  "mealAssist": "Off: Carbs: undefined Boluses: undefined Target: 101 Deviation: -16 BGI: -1.67", 
  "tick": -5
Paul Martin
@Spazholio
May 16 2016 12:47
Mine looks very similar to that. I don't have the requested section, but I think that's because my numbers are in range right now, so no temp basal is needed.
John Sjolund
@sjolundjohn
May 16 2016 12:47
I can't get that sent off to the pump which is the problem use 'use XXX set_temp_basal
Paul Martin
@Spazholio
May 16 2016 12:49
I think you're supposed to pass enact/suggested.json to the set_temp_basal command, not enact/enacted.json...
Now that I have a working loop, I need to whiteboard the entire flow so I can cement all the steps/files in my head. I'm constantly re-learning my openaps.ini file, and it'd be nice to know what each little piece does more thoroughly. =)
John Sjolund
@sjolundjohn
May 16 2016 12:52
Yes that is right, that is what goes into enacted.json. However when I invoke that report, nothing gets sent to pump
Paul Martin
@Spazholio
May 16 2016 12:53
What does your enact/enacted.json report look like in openaps.ini?
C-Ville
@C-Ville
May 16 2016 12:53
@sjolundjohn the report that @Spazholio posted already has use in there. The report seems to pull the relevant data off the enacted.json ans sends them to the pump
adding an enacted.json report using the
Openaps add gives you a better feel, what it is supposed to do
John Sjolund
@sjolundjohn
May 16 2016 12:56
@Spazholio
[report "enact/enacted.json"]
device = nightscout
input = enact/suggested.json
use = set_temp_basal
reporter = JSON
@Spazholio looks to be the same as yours right
Paul Martin
@Spazholio
May 16 2016 12:57
Yep. So what happens when you invoke the enact/suggested.json report?
John Sjolund
@sjolundjohn
May 16 2016 12:58
seems to run fine
Paul Martin
@Spazholio
May 16 2016 12:59
And then when you invoke the enact/enacted.json?
John Sjolund
@sjolundjohn
May 16 2016 13:00
pi@raspberrypi:~/openaps $ openaps report invoke enact/suggested.json
determine-basal://text/shell/enact/suggested.json
{"delta":-2,"glucose":73,"avgdelta":-1}
{"duration":0,"rate":0,"temp":"percent"}
{"netbasalinsulin":0,"activity":0,"basaliob":0,"hightempinsulin":0,"bolussnooze":0,"iob":0}
{"max_iob":6,"type":"current","dia":4,"current_basal":1.3,"max_daily_basal":1.3,"max_basal":5,"min_bg":100.8,"max_bg":100.8,"sens":32.4}
reporting enact/suggested.json
pi@raspberrypi:~/openaps $ openaps report invoke enact/enacted.json
nightscout://JSON/set_temp_basal/enact/enacted.json

All seems in order, also reviewing setup.sh script and it also seems I have things setup right, can't crack why it isn't be sent and/or received by pump.

Low blood sugar, off to eat :-)

John Sjolund
@sjolundjohn
May 16 2016 13:46
I haven't figured this one out so help is appreciated!"
JaysonEwer
@JaysonEwer
May 16 2016 13:48
@sjolundjohn do you have your temp basal type set correctly on your pump?
Act>basal>Temp basal type should be "Insulin Rate (U/H)"
John Sjolund
@sjolundjohn
May 16 2016 14:00
@JaysonEwer I did not no - thanks for the tip, will try now
JaysonEwer
@JaysonEwer
May 16 2016 14:03
Don't forget to take a picture of the first successful basal enacted by your loop. :-)
John Sjolund
@sjolundjohn
May 16 2016 14:06
@JaysonEwer yes sir that did it, working temp_basal sent
JaysonEwer
@JaysonEwer
May 16 2016 14:07
I'm pretty sure that gets everyone! 👏👍
C-Ville
@C-Ville
May 16 2016 14:07
@sjolundjohn congratulations
JaysonEwer
@JaysonEwer
May 16 2016 14:08
Congratulations!
Andrew Baugh
@baughaw
May 16 2016 14:08
welcome to the club!
John Sjolund
@sjolundjohn
May 16 2016 14:08

Shucks -now need to go and pickup my kids from daycare - so this will have to be for tomorrow to continue with.

I was however able to get through all last tweaks to get it all setup and uploading to NS.

Thanks so much for the help!
JaysonEwer
@JaysonEwer
May 16 2016 14:09
Excellent, good job!
You are very welcome. Thank the rock stars, not me, @bewest @scottleibrand @danamlewis @jasoncalabrese and all the other rock stars that helped you! :-)
Paul Martin
@Spazholio
May 16 2016 14:41
:thumbsup:
Bill Stackpole
@billstackpole
May 16 2016 15:29
got some weirdness going on w/OpenAPS v 0.0.10 using a TI stick with subg_rfspy 0.8 - when attempting a preflight message is ```
mmtune: Traceback (most recent call last):
  File "/usr/local/bin/openaps-use", line 4, in <module>
    __import__('pkg_resources').run_script('openaps==0.0.10', 'openaps-use')
  File "build/bdist.linux-armv7l/egg/pkg_resources/__init__.py", line 726, in run_script
  File "build/bdist.linux-armv7l/egg/pkg_resources/__init__.py", line 1484, in run_script
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.0.10-py2.7.egg/EGG-INFO/scripts/openaps-use", line 63, in <module>
    app( )
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.0.10-py2.7.egg/openaps/cli/__init__.py", line 44, in __call__
    self.run(self.args)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.0.10-py2.7.egg/EGG-INFO/scripts/openaps-use", line 57, in run
    output = app(args, self)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.0.10-py2.7.egg/openaps/uses/__init__.py", line 92, in __call__
    return self.method.selected(args)(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.0.10-py2.7.egg/openaps/uses/__init__.py", line 31, in __call__
    return self.method(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.0.10-py2.7.egg/openaps/uses/use.py", line 44, in __call__
    self.before_main(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.0.10-py2.7.egg/openaps/vendors/medtronic.py", line 61, in before_main
    self.setup_medtronic( )
  File "/home/pi/mmeowlink-source/mmeowlink/vendors/mmeowlink.py", line 83, in setup_medtronic
    setup_logging(self)
  File "/home/pi/mmeowlink-source/mmeowlink/vendors/mmeowlink.py", line 55, in setup_logging
    log.addHandler(logging.handlers.SysLogHandler(address=address))
  File "/usr/lib/python2.7/logging/handlers.py", line 760, in __init__
    self._connect_unixsocket(address)
  File "/usr/lib/python2.7/logging/handlers.py", line 788, in _connect_unixsocket
    self.socket.connect(address)
  File "/usr/lib/python2.7/socket.py", line 224, in meth
    return getattr(self._sock,name)(*args)
socket.error: [Errno 111] Connection refused
FAIL
looks like socket in use.
did a sudo reboot
to no avail
thoughts?
and if I try to "use pump read_status" I get mmeowlink.exceptions.InvalidPacketReceived"
Pete Schwamb
@ps2
May 16 2016 15:37
Looks like you’re pointing at a syslog server that isn’t available, or is firewalled.
Scott Leibrand
@scottleibrand
May 16 2016 15:39
@sjolundjohn can you find the best place to document what tripped you up so the next person won't get stuck there? Maybe in the troubleshooting FAQ and/or in the enact part of the docs?
Scott Leibrand
@scottleibrand
May 16 2016 15:50
Oh, ha! Just saw #168. Merged! :-)
Paul Martin
@Spazholio
May 16 2016 15:50
Question about upgrading NS - how do I do it? =) I'm currently running 0.9.0-dev, and tried doing a simple git pull and while that updated the code, it borked the install. I'm fairly green when it comes to node apps, so I wasn't really sure where to start debugging. Thankfully, I backed up the directory beforehand so I just reverted.
JaysonEwer
@JaysonEwer
May 16 2016 15:51
:-D :+1: @scottleibrand @sjolundjohn 168 :-D
Bill Stackpole
@billstackpole
May 16 2016 15:59
@ps2 - were you indicating to me the potential syslog server firewall issue? (I turned on Papertrail last night and it was working from home - now at work - confused about how the syslog could cause mmeowlink to bork...)
zaheria985
@zaheria985
May 16 2016 16:00
Hi, I am starting out on my openaps journey... just wondering how much people paid for their used minimed pumps... and has anyone bought one only to have it fail within a few months of purchase? Just want to kind of get an idea of what to expect.. though I realize it can be quite unexpected..
Bill Stackpole
@billstackpole
May 16 2016 16:01
used 723 recently purchased in the $750 range...
zaheria985
@zaheria985
May 16 2016 16:03
hmm, okay. did you find it on craigslist or elsewhere?
JaysonEwer
@JaysonEwer
May 16 2016 16:03
I was given a 515 early on, used it for a few weeks, ended up accidentally dropping it and it stopped working. Then I purchased two 522's, one from a private party for 250 USD and one for 450 USD from MedWOW. I bought the 250 one for backup and haven't actually tried using it yet. :-/ I should probably test it one of these days. :-D But the one I got for 450 has been working GREAT since mid January.
zaheria985
@zaheria985
May 16 2016 16:05
Nice. I was vaguely browsing Medwow, but I'm not really sure about buying something if I can't see it first, you know? Afraid I'll get stuck with something that's DOA.
Scott Leibrand
@scottleibrand
May 16 2016 16:06
@billstackpole from the traceback it looks like Python crashes mmeowlink when it can't write to syslog. Were you offline at the time? Wonder if that is an unintended side effect of papertrail. Might want to Google for how to properly use it for machines that go offline periodically...
JaysonEwer
@JaysonEwer
May 16 2016 16:06
I know, Its a little nerve racking!! :-D I closed my eyes and JUMPED!! :-P I'm happy I did!!
Bill Stackpole
@billstackpole
May 16 2016 16:06
Hmm. So if I turn papertrail off (remove the line in rsyslogd.conf) and reboot, should that restore my machine to former syslog defaults?
Scott Leibrand
@scottleibrand
May 16 2016 16:06
@zaheria985 you might ask the seller to send you a video of the bootup process, so you can see that it turns on successfully at least.
@billstackpole that's worth a try. I've never used papertrail before though.
Bill Stackpole
@billstackpole
May 16 2016 16:07
Well now a new error...
pi@baps:~/billopenaps $ openaps use pump read_status
subg_rfspy status: OK
Version: subg_rfspy 0.8
No handlers could be found for logger "mmeowlink.handlers.stick"
Traceback (most recent call last):
  File "/usr/local/bin/openaps-use", line 4, in <module>
    __import__('pkg_resources').run_script('openaps==0.0.10', 'openaps-use')
  File "build/bdist.linux-armv7l/egg/pkg_resources/__init__.py", line 726, in run_script
  File "build/bdist.linux-armv7l/egg/pkg_resources/__init__.py", line 1484, in run_script
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.0.10-py2.7.egg/EGG-INFO/scripts/openaps-use", line 63, in <module>
    app( )
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.0.10-py2.7.egg/openaps/cli/__init__.py", line 44, in __call__
    self.run(self.args)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.0.10-py2.7.egg/EGG-INFO/scripts/openaps-use", line 57, in run
    output = app(args, self)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.0.10-py2.7.egg/openaps/uses/__init__.py", line 92, in __call__
    return self.method.selected(args)(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.0.10-py2.7.egg/openaps/uses/__init__.py", line 31, in __call__
    return self.method(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.0.10-py2.7.egg/openaps/uses/use.py", line 44, in __call__
    self.before_main(args, app)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.0.10-py2.7.egg/openaps/vendors/medtronic.py", line 63, in before_main
    self.check_session(app)
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.0.10-py2.7.egg/openaps/vendors/medtronic.py", line 119, in check_session
    self.session = self.get_session_info( )
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.0.10-py2.7.egg/openaps/vendors/medtronic.py", line 85, in get_session_info
    fields = self.create_session( )
  File "/usr/local/lib/python2.7/dist-packages/openaps-0.0.10-py2.7.egg/openaps/vendors/medtronic.py", line 107, in create_session
    self.pump.power_control(minutes=minutes)
  File "/home/pi/mmeowlink-source/mmeowlink/handlers/stick.py", line 212, in power_control
    status = repeater(self.command, repetitions=500, ack_wait_seconds=20)
  File "/home/pi/mmeowlink-source/mmeowlink/handlers/stick.py", line 191, in __call__
    self.wait_for_ack()
  File "/home/pi/mmeowlink-source/mmeowlink/handlers/stick.py", line 84, in wait_for_ack
    resp = Packet.fromBuffer(buf)
  File "/home/pi/mmeowlink-source/mmeowlink/packets/rf.py", line 81, in fromBuffer
    rftype   = buf[0]
IndexError: bytearray index out of range
zaheria985
@zaheria985
May 16 2016 16:08
@JaysonEwer Haha, awesome. I found a few on there for 300-350. Maybe I should jump... @scottleibrand Good idea Scott, thanks.
Scott Leibrand
@scottleibrand
May 16 2016 16:08
IIRC that is a communications error.
JaysonEwer
@JaysonEwer
May 16 2016 16:08
Of course what @scottleibrand said is a bit better advice @zaheria985 I did have multiple conversations on the phone with the seller and ended up trusting them.
Bill Stackpole
@billstackpole
May 16 2016 16:09
well looks like I broke it by losing my network connection :-(
any ideas on a fix to get me back up and running?
Papertrail log returned to default, now getting errors decoding FourBySix packet...
Bill Stackpole
@billstackpole
May 16 2016 16:15
looks like all my logs stopped at 624 EDT today...
wtmp and Xorg.0.log updated around noontime..
reboot has no impact. Doing a hard shutdown and will restart with a cold boot next...
Bill Stackpole
@billstackpole
May 16 2016 16:21
K - so a hard reboot also doesn't do it. Socket.error - connection refused - to what socket is this referring? is this a network socket?
Paul Martin
@Spazholio
May 16 2016 16:32
If I had to guess, I'd think it was talking about the /dev/mmeowlink or the /dev/ttyACM0 devices. Do those exist still? Does anything relevant show up in dmesg?
Bill Stackpole
@billstackpole
May 16 2016 16:34
pi@baps:~ $ ls -la /dev/mmeowlink 
lrwxrwxrwx 1 root root 7 May 16 12:17 /dev/mmeowlink -> ttyACM0
pi@baps:~ $ ls -la /dev/ttyA
ttyACM0  ttyAMA0  
pi@baps:~ $ ls -la /dev/ttyA
ttyACM0  ttyAMA0  
pi@baps:~ $ ls -la /dev/ttyACM0 
crw-rw-r-- 1 root plugdev 166, 0 May 16 12:17 /dev/ttyACM0
both seem to exist...
dmesg output:
pi@baps:~ $ dmesg | grep mmeo
pi@baps:~ $ dmesg | grep tty
[    0.000000] Kernel command line: 8250.nr_uarts=1 dma.dmachans=0x7f35 bcm2708_fb.fbwidth=640 bcm2708_fb.fbheight=480 bcm2709.boardrev=0xa02082 bcm2709.serial=0xbac7fdf2 smsc95xx.macaddr=B8:27:EB:C7:FD:F2 bcm2708_fb.fbswap=1 bcm2709.uart_clock=48000000 vc_mem.mem_base=0x3dc00000 vc_mem.mem_size=0x3f000000  dwc_otg.lpm_enable=0 console=ttyS0,115200 console=tty1 root=/dev/mmcblk0p7 rootfstype=ext4 elevator=deadline fsck.repair=yes rootwait
[    0.001295] console [tty1] enabled
[    0.269460] console [ttyS0] disabled
[    0.271115] 3f215040.uart: ttyS0 at MMIO 0x3f215040 (irq = 29, base_baud = 50000000) is a 16550
[    1.470940] console [ttyS0] enabled
[    2.745162] 3f201000.uart: ttyAMA0 at MMIO 0x3f201000 (irq = 83, base_baud = 0) is a PL011 rev2
[    4.723517] systemd[1]: Expecting device dev-ttyS0.device...
[    5.165646] systemd[1]: Starting system-serial\x2dgetty.slice.
[    5.179329] systemd[1]: Created slice system-serial\x2dgetty.slice.
[    7.434621] cdc_acm 1-1.4:1.0: ttyACM0: USB ACM device
pi@baps:~ $
looks like it sees /dev/ttyACM0...
Scott Leibrand
@scottleibrand
May 16 2016 16:36
Connection refused is a network thing, which might mean you didn't properly disable papertrail.
Bill Stackpole
@billstackpole
May 16 2016 16:37
port 553?
or 514?
(neither show up in netstat -an)
(and lsof not installed on my pi - suggestions on how to grep for syslog open files or similar?)
or should I just install it?
<grin>
Paul Martin
@Spazholio
May 16 2016 16:41
Looks like it's 514. And...yeah, install it. =)
(I'm really surprised it's not installed by default - it's kinda useful)
Bill Stackpole
@billstackpole
May 16 2016 16:42
apt-get updating...
JaysonEwer
@JaysonEwer
May 16 2016 16:45
:+1: Kinda Useful....LSOF....:-D
Bill Stackpole
@billstackpole
May 16 2016 16:46
so lsof | grep log and lsof | grep sys yield only reference to run.log
wait
lemme sudo !!
Paul Martin
@Spazholio
May 16 2016 16:47
That is the most explosive sentence I've seen all day today!
Bill Stackpole
@billstackpole
May 16 2016 16:47
heh
JaysonEwer
@JaysonEwer
May 16 2016 16:48
:boom:
Bill Stackpole
@billstackpole
May 16 2016 16:48
Nothin (and coming your way in one sec...)
is there a way to put the text into a snippet like in slack?
(cause it's pretty big)
what's the lsof flag to show me network sockets so I can .... (forget it. If netstat doesn't show 514, then lsof wont either...)
hmmm.
I have a hard stop for 1 hr at 1p EDT - any ideas for the next 5 min?
<struggle struggle>
Scott Leibrand
@scottleibrand
May 16 2016 16:58
You're not gonna see an open connection for UDP.
What are you trying to figure out?
Bill Stackpole
@billstackpole
May 16 2016 17:00
how to kill the connection to PaperTrail...
Scott Leibrand
@scottleibrand
May 16 2016 17:01
I don't think that will help. You need to tell the system to quit trying to use papertrail and log to local syslog file again. Maybe just google for disabling papertrail?
Bill Stackpole
@billstackpole
May 16 2016 17:01
will do. Tnx - off to my 1p. Will let you know what I find later today..
Paul Martin
@Spazholio
May 16 2016 17:04
I'm not familiar with Papertrail (just remote syslog?) but that seems like a fairly nasty showstopper.
Dana Lewis
@danamlewis
May 16 2016 17:08
@zaheria985 one idea is to ask for a video of the pump - have them put a battery in, show it booting up (that also helps verify version) and pressing buttons to confirm it works.) doesn't fix for the "when will it break down the road" question but might make you feel better about buying it regardless
Paul Martin
@Spazholio
May 16 2016 17:21
I'm noticing a discrepancy on my NS install. On the OpenAPS pill, it'll show my current temp basal, but on the BASAL pill, it shows my last temp basal. Is this a known thing?
Scott Leibrand
@scottleibrand
May 16 2016 17:23
Your OpenAPS rig can't upload the latest treatments until it re-runs the pumphistory report. It usually waits until the next loop to do that, since it takes too long to do twice.
Also, not sure if basal pill shows temps or just scheduled basals.
Paul Martin
@Spazholio
May 16 2016 17:27
Well, the BASAL pill is consistently one behind, but that's fine. I'll just file that away in my head.
Also, did you know that if you type crontab -r instead of crontab -e, there's no prompt?
Spazholio @Spazholio cries
Scott Leibrand
@scottleibrand
May 16 2016 17:29
That is dumb.
(Don't run that. It deletes your crontab.)
Paul Martin
@Spazholio
May 16 2016 17:31
Oh yeah - that wasn't meant for people to verify. It will totally wipe crontab out, with no way to recover short of recreating from /var/log/syslog or something.
Paul Martin
@Spazholio
May 16 2016 17:36
Do you recommend a specific folder for the auto-sens.json file to live in?
Martin Haeberli
@mhaeberli
May 16 2016 17:39
can someone send me an example “good” and “bad” fw string from a MiniMed pump? I know FW 2.5 or later is “bad” as in can’t be controlled - I’m working on buying a used pump...
but have to explain to potential sellers what the string looks like
Paul Martin
@Spazholio
May 16 2016 17:42
Well, mine is a 722, and it's "good". If I hit the ESC button, and then scroll ALL the way to the bottom, I see: "VER 2.1A 1.1 0B 0B". I think that the "2.1A" is really all you need to worry about. Alternately, I think you can have them check for the "PC Connect" option - if it's there, the pump is "bad" for looping.
Martin Haeberli
@mhaeberli
May 16 2016 17:45
@Spazholio thx!
Paul Martin
@Spazholio
May 16 2016 17:46
np
@scottleibrand In the directions for autosens (https://github.com/openaps/docs/blob/master/docs/docs/walkthrough/phase-6/Configure-Automatic-Sensitivity-Mode.md) in Step 6, it says to set current-temps to [Your Path]/temp-basal-status.json. However, I already have something called temp_basal set to monitor/temp_basal.json. Are those two settings equivalent?
Or should I remove my current enact/suggested.json and replace it with the one in the docs?
Scott Leibrand
@scottleibrand
May 16 2016 17:49
I think the one you have is fine. Check the dev docs too (and the open PR): that page is under active revision.
Paul Martin
@Spazholio
May 16 2016 17:51
Hmmm. So should I assign current-temps to the same value/file as my current temp_basal?
(I'm checking both the dev and PR also)
Dana Lewis
@danamlewis
May 16 2016 17:54
@mhaeberli yes, checking for pc connect presence under Utilities menu is probably easiest. Or have them hit the esc button and scroll to bottom to find version (i.e. VER 2.4)
Paul Martin
@Spazholio
May 16 2016 17:59
Also, here's a possibly silly question - how does one know if one enabled Automatic Sensitivity Mode properly? I now have a settings/autosens.json file that contains {"ratio":1}. Is that...right?
Scott Leibrand
@scottleibrand
May 16 2016 18:02
If you run the autosens report manually it will give you lots more details, but that is normal if you're not running sensitive or resistant.
Paul Martin
@Spazholio
May 16 2016 18:03
Yeah, I tried that and got lots of info. Does the loop have to be modified to include invoking that specific report, or does it get pulled in somewhere during the standard gather/get-settings/enact routines?
Pete Schwamb
@ps2
May 16 2016 18:11
sigh rats have been cured of diabetes countless times, and now they even get interesting pump technology: http://www.iprecio.com/products/tabid/190/Default.aspx
RF controlled (915mhz!), tiny, battery lasts for weeks.
Dana Lewis
@danamlewis
May 16 2016 18:12
Lol @ps2
Scott Leibrand
@scottleibrand
May 16 2016 18:12
@Spazholio check your aliases: you may be invoking it already.
Martin Haeberli
@mhaeberli
May 16 2016 18:13
@danamlewis thx!
Matthias Granberry
@mgranberry
May 16 2016 18:13
@ps2 If it's for animals, it isn't regulated by the FDA, fish-antibiotics style. Too bad the reservoir volume is rodent-friendly.
Pete Schwamb
@ps2
May 16 2016 18:14
900ul = 90 units, no? Or is it 9 units?
Matthias Granberry
@mgranberry
May 16 2016 18:15
90 units. That's actually bigger than I would have expected.
Paul Martin
@Spazholio
May 16 2016 18:16
@scottleibrand I don't see it explicitly in any of my aliases, but I wasn't sure if any of the other reports that are already running pull it in. When I tail the log, I don't see the calculations/mean deviation stuff, so would I be correct in assuming that it's not running? If not, where's the best place to put it? I was thinking in the enact part of the loop, before the enact/suggested.json report gets run, since it looks like that report makes use of the settings/autosens.json report...
Pete Schwamb
@ps2
May 16 2016 18:16
Would be awesome to have something that small. Could actually be worn during gymnastics.
Paul Martin
@Spazholio
May 16 2016 18:22
Argh. Checked master and mmeowlink and totally forgot about that branch. My mistake.
So does it hurt anything to put it right before the suggested.json report gets run? I did it as a test, and it appears to work fine, but wasn't sure if there was a reason you had put it at the end.
Scott Leibrand
@scottleibrand
May 16 2016 18:29
it takes forever to run, so I didn't want it slowing down my enact
Paul Martin
@Spazholio
May 16 2016 18:31
Oh really? Hmm. Mine runs in something like 2s. Is that something I should be concerned with?
Scott Leibrand
@scottleibrand
May 16 2016 18:32
what rig hardware?
and how long is the ++--<><>b-a-start stuff? ;-)
Paul Martin
@Spazholio
May 16 2016 18:33
Pi3 + TI stick.
And the whole bunch of dots, plus all the p=0.60: 0.53, 0.00, 0.53 stuff is what only takes 2s.
Scott Leibrand
@scottleibrand
May 16 2016 18:35
can you paste the symbol line? sounds like you may be missing some data in the input files
Paul Martin
@Spazholio
May 16 2016 18:37
Sorry, can you elaborate on "symbol line"? Here's the entirety of the output when I invoke that report:
detect-sensitivity://text/shell/settings/autosens.json
.............................................................................................................................................................................................................................................................................................................
p=0.60: 0.53, 0.00, 0.53
p=0.58: 0.33, 0.00, 0.33
p=0.56: 0.33, 0.00, 0.33
p=0.54: 0.00, 0.00, 0.00
p=0.52: 0.00, 0.00, 0.00
p=0.50: 0.00, 0.00, 0.00
p=0.48: -0.17, 0.00, -0.17
p=0.46: -0.33, 0.00, -0.33
p=0.44: -0.33, 0.00, -0.33
p=0.42: -0.67, 0.00, -0.67
p=0.40: -0.67, 0.00, -0.67
p=0.38: -0.67, 0.00, -0.67
p=0.36: -1.00, 0.00, -1.00
p=0.34: -1.00, 0.00, -1.00
p=0.32: -1.00, 0.00, -1.00
p=0.30: -1.33, 0.00, -1.33
p=0.28: -1.33, 0.00, -1.33
p=0.26: -1.33, 0.00, -1.33
Mean deviation: 0.14
Sensitivity within normal ranges
Basal adjustment 0.00U/hr
Ratio: 100%: new ISF: 20.0mg/dL/U
reporting settings/autosens.json
Scott Leibrand
@scottleibrand
May 16 2016 18:38
you shouldn't be getting all ...'s
Paul Martin
@Spazholio
May 16 2016 18:38
(and I stand corrected, it was closer to 5s when I ran it this time)
Well that's troubling. Hmm.
Scott Leibrand
@scottleibrand
May 16 2016 18:38
check the files feeding the autosens report and see if they have 24h of data in them
Paul Martin
@Spazholio
May 16 2016 18:38
I was pretty sure to set 'em all to 27h, but I'll re-look. Standby.
Scott Leibrand
@scottleibrand
May 16 2016 18:38
should actually be 24h for glucose and 24h+DIA for pumphistory
Paul Martin
@Spazholio
May 16 2016 18:39
k
Scott Leibrand
@scottleibrand
May 16 2016 18:39
you don't want the glucose to be quite as long as the pumphistory, or it'll be calculating deviations for the first couple hours based on zero insulin activity
although once you get past 24h it won't affect the end result all that much
Eric Luhrs
@eluhrs
May 16 2016 18:44
I'm going to be in Cambridge, MA for a few nights this week. If there are any loopers who want to use my cc1111 programmer, soldering equipment, etc, I'd be happy to meet up.
Paul Martin
@Spazholio
May 16 2016 18:44
So...it appears as though my pumphistory-24h.json and pumphistory-24h-zoned.json files are quite old. And my reports are unsure whether or not they should go in settings or monitor. Peculiar - going to investigate.
Andrew Baugh
@baughaw
May 16 2016 18:45
Can someone tell me which pump will work with both the Minimed Connect system and OpenAPS? I have a 522 running openaps but it doesnt have the options for pairing bluetooth. Thanks!
Ben West
@bewest
May 16 2016 18:45
should is a funny word
Paul Martin
@Spazholio
May 16 2016 18:46
I imagine they should all go into monitor so that they can all be regenerated during the loop?
@bewest I agree, but I would think that I would have settled on one or the other, not both. =)
Dana Lewis
@danamlewis
May 16 2016 18:49
@ochenmiller ---> @eluhrs ^^
Ben West
@bewest
May 16 2016 18:52
actually, my recommendations are to create several directories
Scott Leibrand
@scottleibrand
May 16 2016 18:52
@Spazholio I put the 24h stuff in settings, because it's ok for it to get up to an hour out of date, and it doesn't necessarily get successfully refreshed on every loop
Ben West
@bewest
May 16 2016 18:52
recently I've been playing with two called raw-pump and raw-cgm
Paul Martin
@Spazholio
May 16 2016 18:52
No no no...the 24h variants go into the settings directory.
Ben West
@bewest
May 16 2016 18:52
so my advice is to put the stuff direct from device into a raw-directory
Paul Martin
@Spazholio
May 16 2016 18:52
Yeah, I'm realizing that. Ok, it's starting to crystalize for me now...
Scott Leibrand
@scottleibrand
May 16 2016 18:52
it doesn't really matter where it goes, but it shouldn't be in the monitor/ directory unless it's something that needs to be deleted every loop
I like @bewest's approach: we need to re-vamp the docs and scripts to reflect that. in the mean time, I would stick with the directories in the docs for an already-mostly-working setup
Ben West
@bewest
May 16 2016 18:53
then we can run one of the tools to move it from the raw directory, process it, and save it in the final place we expect it to be
yeah, it's something to keep aware of, there's no "should" here
Paul Martin
@Spazholio
May 16 2016 18:54
Yeah, I can definitely see what you mean. Both of you.
Scott Leibrand
@scottleibrand
May 16 2016 18:55
@baughaw the only pumps capable of doing both are 523s with older firmware. they're rare.
Ben West
@bewest
May 16 2016 18:55
there's an oref0 copy-fresher --since that should eliminate need to delete/remove files
saving it in a raw area, and then only copy fresher if it's new enough
Andrew Baugh
@baughaw
May 16 2016 18:56
thanks @scottleibrand
Scott Leibrand
@scottleibrand
May 16 2016 18:56
I wish someone (besides me) was willing to go through and re-work the way all this stuff fits together
Ben West
@bewest
May 16 2016 18:56
there are 554's that also work
Andrew Baugh
@baughaw
May 16 2016 18:58
I was an idiot 2 years ago and sold my 523 back to medtronic. Not sure which firmware it had. :(
Paul Martin
@Spazholio
May 16 2016 18:59
Ok, so I'm going to run through some numbers, and if you guys could tell me if they're sane, that'd be neat. My raw-cgm/glucose.json is set to use 25 hours (I think that was @bewest's suggestion). My monitor/cgm-glucose.json is set to 25 hours, my ns-glucose device is set to use 288 entries, and my monitor/glucose.json is set to count = 5. That last one feels wrong, but I don't trust my gut just yet...
Dana Lewis
@danamlewis
May 16 2016 19:00
@baughaw don’t beat yourself up. two years ago, most of us didn’t know this was possible.
Paul Martin
@Spazholio
May 16 2016 19:00
All of my pumphistory reports pull 27h.
Scott Leibrand
@scottleibrand
May 16 2016 19:02
count = 5 sounds wrong. the others sounds reasonable, although I would go with 24h instead of 25h.
Ben West
@bewest
May 16 2016 19:03
try oref0_glucose --hours 24 or similar
instead of iter_glucose
Paul Martin
@Spazholio
May 16 2016 19:04
Making those changes now...
(what's the character to comment things out in these files? #?)
Andrew Baugh
@baughaw
May 16 2016 19:08
@danamlewis Thanks, the 500 dollars sounded good at the time.
Paul Martin
@Spazholio
May 16 2016 19:09
@bewest Does oref0_glucose require the threshold entry to work properly?
Ben West
@bewest
May 16 2016 19:09
no
Paul Martin
@Spazholio
May 16 2016 19:09
Because it doesn't work if I omit it...
[report "monitor/cgm-glucose.json"]                                                                                                                                                                                                                                        
device = cgm 
hours = 25.0
use = oref0_glucose
reporter = JSON
That gives me an error.
Ben West
@bewest
May 16 2016 19:10
ah, yeah
I would just use the tools :-)
Paul Martin
@Spazholio
May 16 2016 19:10
If I include threshold = 100 (from another report) then it works fine.
Oh, so go back to iter_glucose?
Ben West
@bewest
May 16 2016 19:11
openaps use cgm oref0_glucose --hours 25
then save as report: replace use with report add mine.json JSON
yes, in the ini, a threshold = 5 or similar is required
100?
oh, maybe it's in seconds
should probably update the help on that sucker
Paul Martin
@Spazholio
May 16 2016 19:14
I think it auto-adds the threshold = 100...
Ben West
@bewest
May 16 2016 19:14
it's the default
Paul Martin
@Spazholio
May 16 2016 19:15
Should it be changed or is 100 ok?
Ben West
@bewest
May 16 2016 19:15
100 is fine
the use I suggested above uses 100, when it's added as a report, all the defaults that were implicit become explicit in the config so they never change
Paul Martin
@Spazholio
May 16 2016 19:17
Thanks. What about my monitor/glucose.json that has count = 5? That one seemed odd to me. There are clearly more than 5 entries in there, but that could be due to some CGM vs NS conditionals that are put in there...
Ben West
@bewest
May 16 2016 19:19
ok new help message for that:
  --threshold THRESHOLD
                        Merge EGV and Sensor records occuring within THRESHOLD
                        seconds of each other.
default is 100 seconds, (just under 2 minutes)
Paul Martin
@Spazholio
May 16 2016 19:21
I wonder if that even matters, since the monitor/glucose.json file is (apparently) either ns-glucose.json which is 288 entries, or cgm-glucose.json which is 25h...
I think it might just be vestigial.
@scottleibrand You said I shouldn't be getting all ..... when invoking the autosens.json report, right?
Scott Leibrand
@scottleibrand
May 16 2016 19:25
correct
hmm, unless the update to do > + and - isn't in the version of autosens you're running...
Paul Martin
@Spazholio
May 16 2016 19:26
Well, my 24h pumphistory reports are the same size as my regular ones, so I've made an error somewhere...
Scott Leibrand
@scottleibrand
May 16 2016 19:26
too many versions still to keep them all straight in my head. :)
Paul Martin
@Spazholio
May 16 2016 19:26
Hey, when the creator is having problems, it makes me feel a little better. =)
Scott Leibrand
@scottleibrand
May 16 2016 19:26
my stuff works great: I just stick with the AMA branch. :-p
it's trying to keep my head around what everyone else is using that is a bit more error-prone. ;-)
Paul Martin
@Spazholio
May 16 2016 19:27
hehehe
Ok, I was pulling 27h for both pumphistory and pumphistory-24h. My bad. Much better now.
Still only getting lots of .... for autosens.json. And it's throwing NaN in the mix now?
Stephan
@MosiGitHub
May 16 2016 19:30
@scottleibrand I just made you the PR concerning the "exchanges" carb_ratios unit, now Openaps should be able to work with both settings ... fyi
Paul Martin
@Spazholio
May 16 2016 19:30
(I followed the directions, I swear)
I think it's step 6 here that threw me: https://github.com/openaps/docs/blob/dev/docs/docs/walkthrough/phase-6/Configure-Automatic-Sensitivity-Mode.md. That states to have the pumphistory.json report pull 24+DIA hours. Should that be the pumphistory-24h.json report instead?
Scott Leibrand
@scottleibrand
May 16 2016 19:34
yeah, sounds like that needs fixed in the docs
well, maybe...
Paul Martin
@Spazholio
May 16 2016 19:34
Ok, I can do a PR when I get home. Just want...ok, maybe not?
Scott Leibrand
@scottleibrand
May 16 2016 19:34
do the docs suggest using pumphistory or pumphistory-24h.json as input to autosens?
Paul Martin
@Spazholio
May 16 2016 19:35
There's no mention of pumphistory-24h.json in the docs. But earlier, you had said something like it needs 24h of data to be accurate.
So...should the entire doc be changed to refer to pumphistory-24h.json?
Scott Leibrand
@scottleibrand
May 16 2016 19:36
so sounds like the docs are taking the simpler approach of doing it all as a single report
having two pumphistory reports is an optimization
it allows you to get the data you actually need every loop, and only get the 24h data when you're all done enacting
which means your loop is more likely to get to the enact step successfully (and a bit more quickly)
so probably don't need to update the docs at this point, just understand how they differ from what's in my setup script
Paul Martin
@Spazholio
May 16 2016 19:37
Ahhh, ok. So I can either change my pumphistory.json to be 24+DIA length, or I can change my autosens.json to look at my pumphistory-24h.json?
Scott Leibrand
@scottleibrand
May 16 2016 19:37
yeah
and if you do the latter, make sure the pumphistory-24h.json is getting refreshed appropriately
Paul Martin
@Spazholio
May 16 2016 19:38
Ok, cool. Pretty sure this thing is still pretty b0rked (I'm still getting all .... and lots of NaN in the autosens.json report) but I can look at that when I get home.
Scott Leibrand
@scottleibrand
May 16 2016 19:38
k
Peter Miller
@ochenmiller
May 16 2016 19:44
@eluhrs , As Dana mentioned, I'm in the Cambridge area. I'm all set for programmers, workstation, etc. But if you organize a get together, I'd gladly bring my goodfet, and other stuff along. Keep us posted!
Dana Lewis
@danamlewis
May 16 2016 19:48
:+1:
Scott Leibrand
@scottleibrand
May 16 2016 19:54
thx @MosiGitHub - I'll take a look in more detail later
zaheria985
@zaheria985
May 16 2016 19:58
hmm, where is a list of get togethers for open aps?
Dana Lewis
@danamlewis
May 16 2016 20:00
There's no one master list. What area are you in?
zaheria985
@zaheria985
May 16 2016 20:01
We're moving to Rochester, NY.
Dana Lewis
@danamlewis
May 16 2016 20:01
Ah! No meetup there but a few people in that neck of the woods, so ya'll could coordinate one.
zaheria985
@zaheria985
May 16 2016 20:02
I need to get one of the older style pumps first anyway :)
Dana Lewis
@danamlewis
May 16 2016 20:02
::bat signal:: @tghoward sarahhoward @billstackpole
Some other NY area people have also expressed interest, too
zaheria985
@zaheria985
May 16 2016 20:05
Thanks.
Dana Lewis
@danamlewis
May 16 2016 20:05
:+1:
jaylagorio
@jaylagorio
May 16 2016 20:07
Is there a DC-area meet up?
Dana Lewis
@danamlewis
May 16 2016 20:08
There's going to be one on June 1 in the evening! Possibly down near IAD/Herndon area
Will build a link for rsvp/info for that one today
If anyone decides to do a meetup in their area and wants to use whoozin (sample later), let me know and I can connect the event so all the openaps meet ups show together
jaylagorio
@jaylagorio
May 16 2016 20:14
That's a shame, on a week day that might be a bit far for me to get to from work. Hopefully next time though.
Scott Leibrand
@scottleibrand
May 16 2016 20:16
Dana and I are there for a day or two, so some people wanted to meet up while we were in town. There are at least three of you who live in the DC area, though, so you should plan another one.
jaylagorio
@jaylagorio
May 16 2016 20:16
That's pretty awesome, what are you in town for?
I don't get to monitor this channel at work, sorry if it's been talked about a lot
Ali Mazaheri
@amazaheri
May 16 2016 20:17
@Spazholio when I originally setup auto-sens back in February I went to one pump history but for now I have two reports like @scottleibrand mentioned now I have one report with 24+DIA running seperately.
Bill Stackpole
@billstackpole
May 16 2016 20:18
@danamlewis @zaheria985 - I'm in Rochester as is @itsn0tm3.
Dana Lewis
@danamlewis
May 16 2016 20:18
@jaylagorio no worries, it hasn’t really been discussed. I have been invited to participate in a combined MedicineX/White House workshop on patients as partners in research, so we’re in town for about 36 hours. (Know that a week night is not ideal, but hopefully ya’ll will have others and/or we’ll be in town with more flexibility next time :))
Ali Mazaheri
@amazaheri
May 16 2016 20:19
also there are some discrepancies on name of reports as they were originally based on my reports and hence seeing [your path] stuff. @ddaniels1 is working on revising the guide based on report naming and.structure suggested in the docs.
zaheria985
@zaheria985
May 16 2016 20:19
@billstackpole Maybe we should organize a get together?
jaylagorio
@jaylagorio
May 16 2016 20:20
@danamlewis That's pretty awesome! I too have done the 36 hour trip to other places and haven't had flexibility so I totally get it.
Dana Lewis
@danamlewis
May 16 2016 20:21
Here’s the “whoozin” example I was referencing. Nice for whoever leads the organization, easy to hold the address (if needed) and email those who say yes with the specific address, and easy to communicate without having to track a large email train. http://whoozin.com/HEE-9P9-CPDV
Bill Stackpole
@billstackpole
May 16 2016 20:21
@baughaw - what do you mean by "works with both"? I have a Minimed Connect and a 723 and am using the OpenAPS. 723 doesn't have BT, but I think my system needs a revamp. Currently my NS instance is not being fed data from the APS - it's being populated only by the Connect (due to the order in which I built things). Plan is to change the order so that I feed from the APS to NS, and use the Connect as a backup feed for NS. This should allow me to autopopulate the NS pills with input from my pump (using the Bolus Wizard). I don't quite know where to start doing this yet, but I'm pretty sure it's been done and documented by others.
you bet on the meetup. I have some work to do, some understanding to obtain, and some changes to make. Also have a place we can meet and labs in which we might work <smile>.
Dana Lewis
@danamlewis
May 16 2016 20:22
:+1:
zaheria985
@zaheria985
May 16 2016 20:24
@billstackpole Sounds good, but I am not a programmer and just now embarking on this quest, so if that can be tolerated... then sounds great. I felt empowered by getting nightscout set up for both the iphone and the droid tablet I have... but I know this is a lot more involved.
Andrew Baugh
@baughaw
May 16 2016 20:24
Hi Bill, My 522 does not support the Minimed connect. There is no way to pair the pump with the connect system. I would love to be able to send my glucose values from openaps to NS. Right now I am using the mmcommander which isnt the best.
Dana Lewis
@danamlewis
May 16 2016 20:27
@zaheria985 this is what I tell everyone who says “I am not a programmer” or “I am not a coder” or “I am not technical” or “I am not (label)”: https://diyps.org/2015/04/08/making-and-diying-continued/
it’s not just tolerated, it’s encouraged :)
Bill Stackpole
@billstackpole
May 16 2016 20:27
@zaheria985 - You'll learn tons - about yourself, diabetes, and about possibilities (plus a few things about unix and such <grin>.) @baughaw - ouch. I think I upgraded a few years back (2012?) cause the 722 (522's big sibling) had capability to connect to only one wireless device - in my case I had two wireless meters and a CGM. I hadn't considered the issues with the connect or APS (cause they didn't exist at the time)
and @danamlewis :clap:
@scottleibrand - haven't solved the PaperTrailapp.com problem yet - have set all rc.[1-5] scripts from Sxx to Kxx, rebooted, and sent an email to the folks there asking for pointers / links / suggestions. (will continue to advise till I fix it or get frustrated and completely rebuild <grin>)
Paul Martin
@Spazholio
May 16 2016 20:32
@amazaheri Very cool. So you have a separate cron to maintain the 24h stuff or…?
zaheria985
@zaheria985
May 16 2016 20:32
@danamlewis Haha, thanks! I AM a DIY-aholic ;) @billstackpole We are moving to Rochester for the first time on top of that.. should be interesting. :) I heard you guys talking about a Carelink USB and .... an alternative to that?
Maybe it was something else.
Bill Stackpole
@billstackpole
May 16 2016 20:36
I'm using the TI stick
have a programmer if you want to use it
mmeowlink
zaheria985
@zaheria985
May 16 2016 20:36
Did you say the TI stick has a longer range?
Bill Stackpole
@billstackpole
May 16 2016 20:36
yup. Lots better than carelink stick
Ali Mazaheri
@amazaheri
May 16 2016 20:38
same cron but at the end of my loop, after enact. if you wanna run a seperate cron you wanna make sure there is not gonna be a race condition in which both processes trying to access the pump.
zaheria985
@zaheria985
May 16 2016 20:38
nice.
Paul Martin
@Spazholio
May 16 2016 20:42
@amazaheri Yeah, I was wondering how you were resolving the race condition if you had 2 crons. =)
zaheria985
@zaheria985
May 16 2016 20:51
Why would you pick one of the compatible pumps ont he list over one of the other ones besides cost? I saw what Andrew said regarding minimed connect and the 522. Does that make it more difficult programming wise for the openaps?
Scott Leibrand
@scottleibrand
May 16 2016 20:53
523 with old firmware can do a few things the others can't (like CGM and Connect). 512 and older kinda suck because they're missing some commands so you have to hard code some settings. 522 is the most common and works quite well. 515 is good too. I don't know the European models very well.
zaheria985
@zaheria985
May 16 2016 20:57
Hmm.
Thanks.
Bill Stackpole
@billstackpole
May 16 2016 20:57
Scott - 7xx series same as 5xx series but bigger reservoir capacity - yes?
Scott Leibrand
@scottleibrand
May 16 2016 20:57
Right.
zaheria985
@zaheria985
May 16 2016 20:57
Aha! I didn't know that.
How big of a diff in the reservoir?
Bill Stackpole
@billstackpole
May 16 2016 20:57
think 300 units in a reservoir for 723
Dana Lewis
@danamlewis
May 16 2016 20:57
yep. 7xx are a little harder to find because not everyone needs that…but keep in mind if you find a reasonably prcied 7xx you can probably trade it down the line
yes. 300 vs 180 units
zaheria985
@zaheria985
May 16 2016 20:58
oh well then a 523 would be pretty good for me.
Scott Leibrand
@scottleibrand
May 16 2016 20:58
Remember most 523s have newer firmware and can't loop. So it all you know is model number and don't need CGM or Connect, 522 is easier to be sure will work.
Paul Martin
@Spazholio
May 16 2016 20:59
Are there any major differences between the 722 and the 723? I’ve always used a 723, but had to get a 722 to loop, and the changes I’ve noticed are mostly in terminology. Then again, that might just be firmware differences as well.
Dana Lewis
@danamlewis
May 16 2016 20:59
@Spazholio the x23 is mostly about making it easier to see IOB (which doesn’t matter anymore) and CGM capabilities compared to x22
Paul Martin
@Spazholio
May 16 2016 21:00
:thumbsup:
zaheria985
@zaheria985
May 16 2016 21:00
@danamlewis the CGM capabilities - are they mostly in relation to the medtronic's CGM or does that also go for the dexcom?
Dana Lewis
@danamlewis
May 16 2016 21:00
@zaheria985 just the medtronic CGM.
zaheria985
@zaheria985
May 16 2016 21:01
aha, okay. then 522 it is. thanks you all.
Dana Lewis
@danamlewis
May 16 2016 21:01
:+1:
Bill Stackpole
@billstackpole
May 16 2016 21:08
523
not sure of software version though.
you'll want to ask that directly.
zaheria985
@zaheria985
May 16 2016 21:09
Thanks. :)
Paul Martin
@Spazholio
May 16 2016 21:10
@bewest,@scottleibrand Ok, if you guys have a moment, I’m going to try to hammer this autosens thing out. My auto-sens.ini was using pumphistory as a field, so I switched that to pumphistory-24h instead. I think that’s what’s supposed to happen. Now, I’m running it via command line as such: oref0-detect-sensitivity monitor/glucose.json settings/pumphistory-24h.json settings/insulin_sensitivities.json settings/basal_profile.json settings/profile.json. I’m still getting the periods and NaN. The files are all recent, and appear to be populated properly. Any ideas?
Scott Leibrand
@scottleibrand
May 16 2016 21:13
you should be using pumphistory-24h-zoned I think
oref0-detect-sensitivity monitor/glucose.json settings/pumphistory-24h-zoned.json settings/insulin_sensitivities.json settings/basal_profile.json settings/profile.json
Paul Martin
@Spazholio
May 16 2016 21:14
Same result. I just changed that - should have clarified.
Scott Leibrand
@scottleibrand
May 16 2016 21:14
check the input files and make sure they're all complete
if that doesn't show you what's up, you might need to uncomment some debug lines in ~/src/oref0/bin/oref0-detect-sensitivity.js
the ... is fine, btw
Paul Martin
@Spazholio
May 16 2016 21:15
They are, I checked. =/
Scott Leibrand
@scottleibrand
May 16 2016 21:15
that's normal for the 0.2.0 version
Paul Martin
@Spazholio
May 16 2016 21:15
Oh, I thought you were saying….ah, ok.
Scott Leibrand
@scottleibrand
May 16 2016 21:15
the AMA branch is the one with all the fancy new symbols
Paul Martin
@Spazholio
May 16 2016 21:15
Still, the Not-A-Number thing probably isn’t correct?
Scott Leibrand
@scottleibrand
May 16 2016 21:16
now, if you're on AMA branch and still getting all . then you might need to do the npm install bits still
zaheria985
@zaheria985
May 16 2016 21:16
I found an odd soul on Medwow that lists like, dozens of pumps for sale.
Scott Leibrand
@scottleibrand
May 16 2016 21:16
I'm guessing NaN is bad, but I'd have to see the output to make any guesses about what it's doing
zaheria985
@zaheria985
May 16 2016 21:16
It's like a used car lot... but a used pump lot, lol
Paul Martin
@Spazholio
May 16 2016 21:17
I’m not 100% sure how I installed it, to be honest. I think it was via npm. I know @bewest was trying to get my gap fix thing going, so he was having me install lots of things, and I think that’s what we did...
.............................................................................................................................................................................................................................................................................................
p=0.60: NaN, -0.61, NaN
p=0.58: NaN, -0.65, NaN
p=0.56: NaN, -0.68, NaN
p=0.54: NaN, -0.71, NaN
p=0.52: NaN, -0.73, NaN
p=0.50: NaN, -0.78, NaN
p=0.48: NaN, -0.81, NaN
p=0.46: NaN, -0.87, NaN
p=0.44: -2.33, -0.89, NaN
p=0.42: NaN, -0.89, NaN
p=0.40: NaN, -0.90, NaN
p=0.38: NaN, -0.92, NaN
p=0.36: NaN, -0.94, NaN
p=0.34: NaN, -0.96, NaN
p=0.32: NaN, -0.99, NaN
p=0.30: NaN, -1.11, NaN
p=0.28: NaN, -1.15, NaN
p=0.26: NaN, -1.24, NaN
Mean deviation: NaN
Sensitivity within normal ranges
Basal adjustment 0.00U/hr
Ratio: 100%: new ISF: 20.0mg/dL/U
{"ratio":1}
If that’s what you meant by “output”.
Scott Leibrand
@scottleibrand
May 16 2016 21:17
yeah. which branch is your ~/src/oref0 on?
Paul Martin
@Spazholio
May 16 2016 21:18
dev
I can swich to AMA though.
Scott Leibrand
@scottleibrand
May 16 2016 21:19
I'd pick one or the other and run https://github.com/openaps/docs/blob/master/scripts/quick-src.sh#L35-L39 to make sure that version is installed consistently
not sure whether you have all the AMA reports set up yet. if not, might want to start with dev/master
Paul Martin
@Spazholio
May 16 2016 21:20
Is AMA ahead of dev with respect to the AMA stuff? I’d assume so, but I’d rather not assume lately. =)
Running the commands you linked to though.
Dana Lewis
@danamlewis
May 16 2016 21:20
Yes it is. although dev just got moved to master, so dev = master for all intents and purposes, and AMA will soon become dev
Paul Martin
@Spazholio
May 16 2016 21:21

@danamlewis Thanks.

@scottleibrand I’ve followed the steps here: https://github.com/openaps/docs/blob/dev/docs/docs/walkthrough/phase-6/Configure-Automatic-Sensitivity-Mode.md. Does that mean I’ve got all the requisite reports set up?

Nothing changed after running all the npm commands. If I swich to the AMA branch, and re-run those, is there any (realistic) chance that it’ll break things?
Ben West
@bewest
May 16 2016 21:31
AMA is still ahead of dev, yes
integrating AMA is one of priorities, we've been developing a roadmap for people that want to help: openaps/oref0#115
Paul Martin
@Spazholio
May 16 2016 21:33
Well, I apparently hate stability, so I’ve switched branches. Let’s see what happens. =)
Paul Martin
@Spazholio
May 16 2016 21:42
@bewest Alright, I’m getting an error invoking the monitor/meal.json. I can post the error, but if there are docs that I should read, I’ll happily RTFM before bugging you again.
Ah, heck with it, here’s the error:

/home/pi/src/oref0/bin/oref0-meal.js:75
    , basalprofile: basalprofile_data
                    ^
ReferenceError: basalprofile_data is not defined
    at Object.<anonymous> (/home/pi/src/oref0/bin/oref0-meal.js:75:21)
    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)
    at Function.Module.runMain (module.js:497:10)
    at startup (node.js:119:16)
    at node.js:906:3
reporting monitor/meal.json
Paul Martin
@Spazholio
May 16 2016 21:52
When I invoke settings/profile.json I see this. Related?
get-profile://text/shell/settings/profile.json
Could not sort temptargets_data.  Optional feature temporary targets disabled.
reporting settings/profile.json
Jason Calabrese
@jasoncalabrese
May 16 2016 22:09
@billstackpole I'm using papertrail on the pi, just added 1 line to the /etc/rsyslog.conf, I'm piping all cron output to syslog
@logichammer all the setting for pump battery logs are in the NS dev readme, if should work with IFTTT to be able able to send emails, but there's an issue with NS alerts that causes some alerts to be snoozed by others, I need to setup separate alert channels nightscout/cgm-remote-monitor#1391
Ali Mazaheri
@amazaheri
May 16 2016 22:19
@Spazholio in AMA that is an optional data you pass to use tempo target feature (you add the temp targets using NS care Portal and can pass it to your get-profiles
Scott Leibrand
@scottleibrand
May 16 2016 22:25
@Spazholio don't worry about temptargets right now. the first error is because the oref0-meal command now requires more arguments.
so you need to update that report something like this: https://github.com/scottleibrand/openaps-sh/blob/advanced-meal-assist/setup.sh#L120
notice the two extra arguments at the end that you don't have in yours
there'll be a few other things like that most likely. if you could document which things you have to add to get it working, that'll be quite useful to add to the docs for people who want to switch to AMA, and for us to work on so it fails more gracefully when people switch branches without changing their reports.
Ali Mazaheri
@amazaheri
May 16 2016 22:28
:+1:
Paul Martin
@Spazholio
May 16 2016 22:28
Looks like that takes a new report called carbhistory.json. Interesting.
zaheria985
@zaheria985
May 16 2016 22:50
Hey - so what if I find a pump I want to get that's in another city far away on craigslist? How can I arrange getting it?
Paul Martin
@Spazholio
May 16 2016 22:51
There’s no issues with shipping pumps, so if they’re willing to send it to you, shouldn’t be a problem.
Paul Martin
@Spazholio
May 16 2016 23:11
@scottleibrand Ok, there were too many issues, so I’ve reverted to dev of oref0. I’ve followed the docs on https://github.com/openaps/docs/blob/dev/docs/docs/walkthrough/phase-6/Configure-Automatic-Sensitivity-Mode.md, and now I’m getting a small issue when trying to invoke the auto-sense.json report:
auto-sens://text/shell/settings/auto-sens.json
settings/auto-sens.json  raised  'Namespace' object has no attribute 'pumphistory-24h’
Running the command manually seems fine, so I’m a little stumped.
Jarred Yaw
@jyaw
May 16 2016 23:39
@zaheria985 just my experience, never received a reply from anyone on medwow. I got responses from Craigslist listings within hours/minutes
Paul Martin
@Spazholio
May 16 2016 23:40
I got one reply, and boy, do they know all about OpenAPS and how much pumps are worth...
Ben West
@bewest
May 16 2016 23:41
lol
I just had a great chat with medtronic
need to keep calling in and asking :-)
Pete Schwamb
@ps2
May 16 2016 23:41
oh yeah?
Paul Martin
@Spazholio
May 16 2016 23:42
Yeah, do tell. =)
Ben West
@bewest
May 16 2016 23:43
well I plan to have more calls with them, for one thing
I gave them a menu of different types of questions, and they attempted to answer at least one
some of the terms they used changed a bit, this time they honestly wanted some help with a few issues
Paul Martin
@Spazholio
May 16 2016 23:44

“Quit blocking temp basal!”

“No!”

=)
Honestly though, it’s good to hear that they’re open to discussion. Before I even got involved with this, I wouldn’t have thought they’d be amenable to it.
Ben West
@bewest
May 16 2016 23:45
This message was deleted
anyway, they wanted to understand what I was talking about as well, to make sure we're all on the same page
so yes, I was sure to be quite explicit about how we're depending on access to that remote temp basal command to pull a lot of data from bunch of devices and then tweak the basal rates
Jason Calabrese
@jasoncalabrese
May 16 2016 23:46
almost impossible to see the details of that image
Ben West
@bewest
May 16 2016 23:47
I mentioned how it wasn't hard to imagine integrating with my gym/fitness-tracker/gps/calendar to increase the benefits
so covered a lot of ground, and the need to figure things out somehow
so one side is regulatory+security
other issue is libability
that's the major one
they don't know to alleviate themselves or potential contributors from liability of some extreme (and bad thing) from happening
and that's actually something everyone is struggling with, not just them
we also discussed how we need to create a process to start analyzing the risks and details of all these activities differently, and come up with something that makes sense
Jason Calabrese
@jasoncalabrese
May 16 2016 23:49
by removing the ability to set temps, and forcing people to use bolus and suspend they are increasing risk, maybe we could get them to see that as increasing their liability
Ben West
@bewest
May 16 2016 23:49
so he started to get idea of how write vs read could maybe be handled differently, and I continued a bit with how you can even have different suites of permisisons
how different kinds of writes might have different risks for different apps
so there's lots to talk about, they need to think about their appetite for different kinds of risks and process for classifying them as a business
I told them there was now intense pressure to get all the 522s
and maybe that wasn't the kind of story they would want to be a part of...
so, I had a few examples where companies considering developer community of some kind was concerned specifically about liability aside from regulatory
but this is clearly of interesting, if anyone has others
one of my go-to favorites is philip with the Hue