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

8th
Jul 2016
Dana Lewis
@danamlewis
Jul 08 2016 00:00
Remember for everyone once they finish their third night with no major problems: https://github.com/openaps/docs/blob/master/docs/docs/walkthrough/phase-6/keeping-up-to-date.md
iananders
@iananders
Jul 08 2016 00:51
is it normal for the nightscout bubbles on the right side of the screen to lag the info from hovering over openaps status on the left?
(just finally got the closed loop running and connected to NS, thanks everyone!)
Scott Leibrand
@scottleibrand
Jul 08 2016 00:58
not sure what you mean by the bubbles on the right side
and yay! :clap: :confetti_ball:
iananders
@iananders
Jul 08 2016 00:58
specifically the basal bubble under the glucose delta and IOB
Dana Lewis
@danamlewis
Jul 08 2016 01:13
are you talking about the “pills” of info? might need a screenshot to make sure of which thing you’re referencing :)
GabeA23
@GabeA23
Jul 08 2016 01:35

does anyone else put the report-nightscout alias within their loop alias?

I ask because the sample loop alias given in the docs doesn't include the report-nightscout alias within it. So it wouldn't update nightscout about what's happening every time you enact a new temp-basal, right?

Here's the sample loop alias that's given in the docs:
loop = ! bash -c "( openaps preflight && openaps gather && openaps enact) || echo LOOP FAILED."

I don't think any of the aliases within loop have the ability to update nightscout
japrobo
@japrobo
Jul 08 2016 01:35
@iananders would it work to invoke pumphistory and pumphistory zoned after you enact and before you upload to NS? Might make your data less 'lag'
Dana Lewis
@danamlewis
Jul 08 2016 02:04
suhweet! @cjo20
Chris Oattes
@cjo20
Jul 08 2016 02:06
:)
Jason Calabrese
@jasoncalabrese
Jul 08 2016 02:06
it only works with the external charger right?
Chris Oattes
@cjo20
Jul 08 2016 02:07
No. I've got my lipo plugged directly in to my intel breakout board
Btw, I'd recommend sticking to the standalone code for now, that server thing is hacky and probably buggy. It shouldn't break anything, but the standalone code is probably more reliable
Jason Calabrese
@jasoncalabrese
Jul 08 2016 02:11
Cool, I didn't think that was possible
Where's the standalone code?
Chris Oattes
@cjo20
Jul 08 2016 02:12
in that repo, standalone.c
Jason Calabrese
@jasoncalabrese
Jul 08 2016 02:12
Ah
Chris Oattes
@cjo20
Jul 08 2016 02:12
it needs to be run as root though
And I don't think anyone thought it was possible to do. Which is why I'm so surprised that I managed it. everywhere seems to say it can't be done on ubilinux
Jason Calabrese
@jasoncalabrese
Jul 08 2016 02:13
Even since dropping xdrip for openxshareble I've been missing the battery level
I'm wondering if it should be handled different in NS than overloading uploader battery
Scott Leibrand
@scottleibrand
Jul 08 2016 02:14
I don't think sudo is any big deal: these are essentially single-user systems anyway, so passwordless sudo is fine
Chris Oattes
@cjo20
Jul 08 2016 02:15
Yeah, I was just trying to get something that could run as a plain user going as proof-of-concept, in case anyone starts making a fuss about security
Scott Leibrand
@scottleibrand
Jul 08 2016 02:16
any reason not to rename standalone to something like voltage?
Brett
@burgertime14
Jul 08 2016 02:16
@cjo20 nice job, glad @Spazholio wouldn't let me slide, just reading about how the quitters couldn't make it happen ;)
Chris Oattes
@cjo20
Jul 08 2016 02:16
@jasoncalabrese I think it should probably be done in the same way that pump battery voltage is, put in a pill. I just used the method I did to get something together quickly.
Martin Haeberli
@mhaeberli
Jul 08 2016 02:18
@bewest - are you around?
Chris Oattes
@cjo20
Jul 08 2016 02:19
@scottleibrand done
Jason Calabrese
@jasoncalabrese
Jul 08 2016 02:21
in my case the edison is the uploader, so it would make sense to treat it the same, but I know there are lots of other configurations
Chris Oattes
@cjo20
Jul 08 2016 02:23
@scottleibrand I'll take a look at that merge after work tomorrow, if that's ok?
I need to get some sleep :P
@jasoncalabrese I'm open to ideas, and to someone else implementing a proper "nightscout" way of handling the data
Jason Calabrese
@jasoncalabrese
Jul 08 2016 02:25
I think we could include it in ns-status
Chris Oattes
@cjo20
Jul 08 2016 02:27
I'll take a look over the weekend
GabeA23
@GabeA23
Jul 08 2016 02:29
Quick question for you all. I'm testing my loop and it seems to be working fine but I keep getting Errno 2 and is not an openaps command errors (one after the other) in my readout every time I run it. I copy-pasted a couple lines up from the errors:
reporting monitor/iob.json
determine-basal://text/shell/predict/oref0.json
{"delta":9,"glucose":124,"avgdelta":9}
{"duration":11,"rate":0,"temp":"absolute"}
{"netbasalinsulin":-0.3,"activity":0.0339,"basaliob":-0.295,"hightempinsulin":0,"bolussnooze":0,"iob":0.915}
{"max_iob":4,"type":"current","dia":3,"skip_neutral_temps":false,"current_basal":0.9,"max_daily_basal":0.95,"max_basal":35,"min_bg":90,"max_bg":120,"sens":45}
reporting predict/oref0.json
[Errno 2] No such file or directory
openaps: 'enactlib/determine-basal/determine-basal.js' is not an openaps command. See 'openaps --help'.

ns://JSON/shell/nightscout/preflight.json
reporting nightscout/preflight.json
and then it invokes a few more reports and ends without any problems
Chris Oattes
@cjo20
Jul 08 2016 02:31

@jasoncalabrese if you're planning on running this, I've got a script in sudo crontab -e that runs

echo `date +"%D %T"`,`/home/edison/cjo20_battery_voltage 1` >> /home/edison/battery_log.txt

every 5 minutes. I'd be interested to see something similar on your system, so I can try and tune the percentage.

cjo20 @cjo20 is 3 hours late for bed, disappears
Scott Leibrand
@scottleibrand
Jul 08 2016 02:45
@GabeA23 can you paste your crontab line and the openaps alias show output for any aliases you're running there?
iananders
@iananders
Jul 08 2016 02:55
blob
@danamlewis my bad. i am referring to the "pills" - see screenshot - a hover over the status sees openaps giving me 0.00 in basal, while the basal pill still shows my normal pattern. is that the way it's designed?
Paul Martin
@Spazholio
Jul 08 2016 03:00
@cjo20 I’m running the same thing that you asked @jasoncalabrese to run, in case you want a larger dataset.
clevens
@clevens
Jul 08 2016 03:23
Hi all! Can someone please help me find the firmware on a pump? I'm trying to buy a Minimed 723 and not sure if it meets the firmware requirements for OpenAPS
iananders
@iananders
Jul 08 2016 03:24
@clevens if you have the pump in hand (or someone can do it for you), hit ESC and scroll to the bottom - the firmware is on the last line
Dana Lewis
@danamlewis
Jul 08 2016 03:24
@clevens ask them if t has "PC connect" in the utilities menu somewhere. If that is present, it is not compatible. (Usually the easiest way, but you can also see the firmware version either when the pump turns on after a new battery, or if there's already a battery hit the back button and scroll to bottom to see it)
@iananders you must be missing an upload somewhere/somehow
clevens
@clevens
Jul 08 2016 03:25
Thanks so much @iananders and @danamlewis . Is there any way to do it possibly if it doesn't have a battery in it? It's a very old pump
(It is someone else's and they sent me a picture of the frnt and back)
Dana Lewis
@danamlewis
Jul 08 2016 03:27
@clevens nope. But it's good idea to ask them to do the battery anyway so you know it still turns on and works...
Paul Martin
@Spazholio
Jul 08 2016 03:30
So I’ve seen people with battery icons/percentages in their NS pills. I want that now. =) What needs to be present in which file to make that be a thing?
Dana Lewis
@danamlewis
Jul 08 2016 03:30
There's something in the docs :)
Paul Martin
@Spazholio
Jul 08 2016 03:31
I don’t have TIME for no fancy doc-readin’! I’ve got new features to stuff into NS!
clevens
@clevens
Jul 08 2016 03:31
On another note, I'm really excited to be part of this community! I'm curious, what time investment would you say goes into building an OpenAPS?
Dana Lewis
@danamlewis
Jul 08 2016 03:31
@Spazholio ;)
@clevens :) depends person to person basic on previous technical experience, plus how long you want to spend testing each stage.
Expect a few weeks probably
clevens
@clevens
Jul 08 2016 03:33
So to be honest I have very little technical experience, but lots of my friends are engineers, and I'm hoping to hire one of them LOL
Paul Martin
@Spazholio
Jul 08 2016 03:33
@clevens To wrap your head around it, absolutely a few weeks. I’ve been doing it for a few months, and I’m still learning stuff and picking things up.
May I offer a bit of advice? Hire them, but hire them to sit behind you and help, but YOU DO EVERYTHING. Seriously, it’ll help so much in the long run.
clevens
@clevens
Jul 08 2016 03:35
Thanks @Spazholio that is great advice!
Leo Clark
@LeodClark_twitter
Jul 08 2016 03:37
I can agree with the enormity of that, in consideration of building one for a four year old... knowing every detail is relevant.
Paul Martin
@Spazholio
Jul 08 2016 03:38
@danamlewis Can you point me towards a section that might illustrate the battery thing? I’m finding lots of references to battery.json but that’s the battery in the pump. I swear I’ve seen NS screenshots that show the battery level of the rig. I know I’ve seen it on some of the Pebble watchfaces...
Andrew Baugh
@baughaw
Jul 08 2016 03:41
blob
fjungman
@fjungman
Jul 08 2016 03:41
blob
You mean like that?
Dana Lewis
@danamlewis
Jul 08 2016 03:41
Oh, you want not the pump battery? Sorry, thought that's what you were talking about
Paul Martin
@Spazholio
Jul 08 2016 03:42
Yes, I want…not the pump battery.
(that hurt to type)
Dana Lewis
@danamlewis
Jul 08 2016 03:42
Heh
fjungman
@fjungman
Jul 08 2016 03:42
The 90% is the uploader battery.
That comes from NS, not OpenAPS, right Dana?
Dana Lewis
@danamlewis
Jul 08 2016 03:42
Uploder of what?
fjungman
@fjungman
Jul 08 2016 03:43
The phone.
The phone that uploads the data to NS.
Dana Lewis
@danamlewis
Jul 08 2016 03:43
Yea, if you're still using an android to send your BGs up to Ns
Paul Martin
@Spazholio
Jul 08 2016 03:43
I’ve seen something on a Nightscout watchface screenshot that showed something like, “Rig: 78%"
fjungman
@fjungman
Jul 08 2016 03:43
Oh, wait, I'm an xDrip user... Sorry...
Andrew Baugh
@baughaw
Jul 08 2016 03:43
battery power of the rig would be amazing
fjungman
@fjungman
Jul 08 2016 03:43
Yes, that shows on the pebble.
Dana Lewis
@danamlewis
Jul 08 2016 03:43
check out what @cjo20 has been working on
Paul Martin
@Spazholio
Jul 08 2016 03:43
I just don’t know what field I’d need to put into the pebble.json file to make it show...
Dana Lewis
@danamlewis
Jul 08 2016 03:44
A new way to see battery for an Edison rig, but I haven't tested it yet
Paul Martin
@Spazholio
Jul 08 2016 03:44
@cjo20 Is my new personal hero. =)
Dana Lewis
@danamlewis
Jul 08 2016 03:44
And that's not in the docs, so you can go back to blissfully ignoring them ;)
Paul Martin
@Spazholio
Jul 08 2016 03:45
I’ve been testing it, and it’s pretty sweet. Gonna run it overnight and log the output. My battery level reported is slowly going up - imagine that! - so I think it’s working as intended...
Scott Leibrand
@scottleibrand
Jul 08 2016 04:04
@iananders I don't think the basal pill is supposed to update when a temp basal is running.
cdaccarett
@cdaccarett
Jul 08 2016 04:27
I have a question is the "oref0-monitor/iob.json" same iob as "monitor/iob.json" im looking at the "New simpler method for Nightscout upload" steps
iananders
@iananders
Jul 08 2016 05:47
@scottleibrand hm interesting. it seems to change on the next 2nd upload of the same basal - is this the way it's supposed to react? it's not a big deal either way, just wondering
Dave Ewall
@daveewall
Jul 08 2016 07:34

I'm having an issue with Nightscout reports (version beta3) missing a lot of the sgv's on the graphs, often the entire morning, and sometimes other random parts of the day. These entries showed up fine in the normal Nightscout screen. I started digging into the mongo database and found that I have multiple records in the entries collection for the same date... sometimes as many as 178 records for the same sgv value with the same dateString. Here's a sample: (hopefully my quoting works properly)

> db.entries.find({"$where": "(this.date > 1467864180000)"}).sort({ date: 1 }).limit(5)
{ "_id" : ObjectId("577dd4a432cfa1114960c83c"), "packet_size" : 0, "name" : "DataEnd", "date" : 1467864480000, "date_type" : "none", "_tell" : 1, "op" : "0x01", "dateString" : "2016-07-07T00:08:00-0400", "type" : "pumpdata", "device" : "openaps://medtronic/pump/cgm" }
{ "_id" : ObjectId("577dd70332cfa1114960c960"), "name" : "GlucoseSensorData", "date_type" : "prevTimestamp", "_tell" : 3, "sgv" : 120, "date" : 1467864480000, "packet_size" : 0, "op" : 60, "dateString" : "2016-07-07T00:08:00-0400", "type" : "sgv", "device" : "openaps://medtronic/pump/cgm" }
{ "_id" : ObjectId("577dd95532cfa1114960ca84"), "name" : "GlucoseSensorData", "date_type" : "prevTimestamp", "_tell" : 5, "sgv" : 120, "date" : 1467864480000, "packet_size" : 0, "op" : 60, "dateString" : "2016-07-07T00:08:00-0400", "type" : "sgv", "device" : "openaps://medtronic/pump/cgm" }
{ "_id" : ObjectId("577dda8032cfa1114960cba5"), "name" : "GlucoseSensorData", "date_type" : "prevTimestamp", "_tell" : 6, "sgv" : 120, "date" : 1467864480000, "packet_size" : 0, "op" : 60, "dateString" : "2016-07-07T00:08:00-0400", "type" : "sgv", "device" : "openaps://medtronic/pump/cgm" }
{ "_id" : ObjectId("577ddbb032cfa1114960ccc9"), "name" : "GlucoseSensorData", "date_type" : "prevTimestamp", "_tell" : 7, "sgv" : 120, "date" : 1467864480000, "packet_size" : 0, "op" : 60, "dateString" : "2016-07-07T00:08:00-0400", "type" : "sgv", "device" : "openaps://medtronic/pump/cgm" }
...

I have 2 months of data in here, and in a month I go back to the endo, so I'm hoping there's a fix for this without having to blow away the data and start over. Is having duplicate entries normal like this? If so, maybe I'm barking up the wrong tree. What else would cause the reports to be missing glucose data? Everything else (treatments, basals, etc.) all seem to be reporting fine.

I did find a bunch of records with "undefined-0400" in the date field, and I blew those away, but there weren't many in there, and I see records for entries that should be showing in the report, but aren't.

Apologies in advance if there's a better forum for this question.

Chris Oattes
@cjo20
Jul 08 2016 10:53
My edison appears to be dead, so I guess I'm not doing any more battery voltage work until I can afford a new one
Paul Dickens
@thebookins
Jul 08 2016 12:08
Anyone know at what voltage the RF comms shut down on the 712? The loop has stopped working and all pump queries are failing. Voltage in NS says 1.35v but the pump says battery normal.
joeyede
@joeyede
Jul 08 2016 12:26
Hi. I am the the father of a 14 year old Type I and interested in closing the loop for her. I am a SW Eng. and have some experience as a maker. I have been reading up on OpenAPS and the big challenge seems to be getting a compatible pump. She is already on a UK Paradigm 754 minimed pump but the FW version seems to be 3.1E which seems incompatible. I was wondering where do things stand with either getting newer minimed pumps working or getting some other other pump to work? The news group seems to have some hints at omnipod working being ongoing but no clear status that I could find. Can anyone enlighten me?
Chris Oattes
@cjo20
Jul 08 2016 12:29
The blocking thing on 3.1E pumps is in the pump firmware itself, so unless someone manages to work out how to extract FW / flash pumps, it's not possible to work round
well, not using the default openaps anyway
Someone has managed to do something with hacking the remote that comes with the pump, but that doesn't allow you to implement a full closed loop
I'm not sure about other brands of pump
joeyede
@joeyede
Jul 08 2016 12:32
@cjo20 I noticed you said something about your Edison dying is that an Intel Edison?
Chris Oattes
@cjo20
Jul 08 2016 12:33
yeah, it is. Either something was wrong with it internally, I shorted something accidentally or I managed to plug in the micro-usb upside down. Not sure which, as it was dark. But I think I let the magic blue smoke out.
Now when I plug it in, it gets extremely hot to touch and smells incredibly bad
joeyede
@joeyede
Jul 08 2016 12:34
what are you using it for?
Chris Oattes
@cjo20
Jul 08 2016 12:34
I was using it to run the openaps software. It was connected to a TI CC1111 USB dev kit for communication with a pump
Paul Martin
@Spazholio
Jul 08 2016 12:35
Even with plugging in the battery wrong, I've only ever heard of it shorting the BO board, not the Edison itself. Maybe yours was faulty? And covered-under-warranty-worthy?
joeyede
@joeyede
Jul 08 2016 12:36
Where are you from? I think I have one at home I could live without for a while.
Chris Oattes
@cjo20
Jul 08 2016 12:36
@joeyede UK, near London. I'll probably just buy a new one at the weekend and go back to my sparkfun BO
joeyede
@joeyede
Jul 08 2016 12:38
I'm in London and would be more then happy to have somebody using mine rathern than it sitting in a box.
would also be happy to hear more about how you are using it. I was just about to go buy a raspberry Pi to use for for the OpenAPS SW
Taht's the setup I had
joeyede
@joeyede
Jul 08 2016 12:42
Is this documented someplace? How do you get the Dexcomm data?
Chris Oattes
@cjo20
Jul 08 2016 12:42
I don't use a dexcom, I've got the enlite by medtronic
joeyede
@joeyede
Jul 08 2016 12:42
so you pull the numbers from the pump?
Chris Oattes
@cjo20
Jul 08 2016 12:43
yes
https://github.com/oskarpearson/mmeowlink/wiki That's most of the hardware documentation
joeyede
@joeyede
Jul 08 2016 12:45
any advice on how to get older pump in UK?
Paul Martin
@Spazholio
Jul 08 2016 12:46
I use a Dexcom, and I get the data into my OpenAPS setup via Nightscout and Dexcom Share.
joeyede
@joeyede
Jul 08 2016 12:48
does the xdrip play well with OpenAPS?
Paul Martin
@Spazholio
Jul 08 2016 12:51
I don't use it, but I seem to recall people here stating that it does.
joeyede
@joeyede
Jul 08 2016 12:52
good. I have a fear of trying to hook anything up to the Dexcom monitor USB port. it FRAGILE!
Paul Martin
@Spazholio
Jul 08 2016 12:53
Oh? I honestly haven't noticed any specific fragility with mine. I'm not using it anymore like I used to (connected to my Pi3) but it always seemed fairly sturdy.
joeyede
@joeyede
Jul 08 2016 12:54
your lucky. before my daughter lost her's we could barely get it to charge anymore. Comm was totally plug and play with until it works.
jim-andrews
@jim-andrews
Jul 08 2016 12:58
@Spazholio Funny you should bring up the RPi3 and Dexcom. Has anyone looked at using the bluetooth on Dexcom to talk directly to the onboard bluetooth on the RPi3? I did a little experimenting with the pi the results of which are below. It would appear that the bluetooth works just fine, it's just a mater of connecting the dots. It's just software after all... lol
blob
Paul Martin
@Spazholio
Jul 08 2016 12:59
Yeah, there's a program called openxshareble. I can't get it to work with the onboard BT - had to use a dongle.
Alas, the G4/Share receiver can only pair with one thing at a time, so if you go that route, you can't continue to pair to your phone.
Chris Oattes
@cjo20
Jul 08 2016 12:59
@joeyede I think people tend to buy 2nd hand pumps wherever they can find them to get one that wokrs with openaps
jim-andrews
@jim-andrews
Jul 08 2016 12:59
Good point.
Sarah Hodson Grady
@shgrady
Jul 08 2016 13:18
@joeyede I am planning on using xDrip with my OpenAPS, mostly because I have been using it and Nightscout for a long time now (1yr plus, seems like a long time with new technology idk), and I think I am only going to loop at night in the beginning, and so the rig's CGM won't have many hours of data. It has issues related mostly to the phone I believe, but I feel I have the Troubleshooting down pretty well at this point. Plus I trialed keeping two Dexcom's calibrated, which was a pain.
David Cintron
@loudestnoise
Jul 08 2016 14:02
@iananders No, you’re right. Nightscout can be updated with the basal pill on the right side to reflect the temporary basals enacted by your loop. That’s how mine works. Sometimes I notice though it takes it a loop or two for it to update. You can see in this screenshot that OpenAPS pill says I have a 2.7u temp basal, but the Basal pill on the right says I have a 1.625u basal. Is that what you were asking about?
2016-07-08_09-00-09.jpg
Martin Haeberli
@mhaeberli
Jul 08 2016 14:07
@cjo20 please pm me re Edison
dougtakeuti
@dougtakeuti
Jul 08 2016 14:22
@johnmales @oskarpearson and other users of Enlite. I'm not sure if someone has figured out how to solve the issue of the pump not sending glucose data even though it displays the data on the pump screen, but I have one more piece of the puzzle: on two occasions I uploaded data to the MDT carelink website and the pump started to transmit glucose data right after that.
mm22dl
@mm22dl
Jul 08 2016 14:39

Hi all,

Just wanted to share my disgust with you... Like many of us I assume, I've been looking for an OpenAPS compatible pump. A couple of months ago, I was lucky and found a very decent 522, working flawlessly. Only thing is it has a 1.8mL reservoir, which is bugging me considering my daily insulin needs. I've thus been looking to exchange it or simply sell it and buy another one. I'm noticing that the prices are ridiculously high. One guy just sent me an e-mail, stating he's got all of the following pumps:

Black 522 Firmware 2.4 - $1000
Black 523 Firmware 2.5 - $750
Clear 523 Firmware 2.5 - $750
Blue 523 Firmware 3.1- $750
Black 722 Firmware 2.4 - $1000
Black 722 Firmware 2.4 - $1000
Clear 722 Firmware 2.4 - $1000
Blue 723 Firmware 2.5 - $750
Pink 751 (530g system) - $800

As you can see, the older ones, that is the ones working with OpenAPS, are clearly more expensive...

Can't believe people actually try to make money out of this... This is nasty...
Leo Clark
@LeodClark_twitter
Jul 08 2016 14:41
I am afraid that as this gains greater public attention, those prices will rise. On the other hand, I've only met the community a couple days and I've already seen someone donate one--free.
joeyede
@joeyede
Jul 08 2016 14:43
has anybody looked into flashing FW onto a pump and if they have protections against downgrading?
mm22dl
@mm22dl
Jul 08 2016 14:50
At some point I feel like it'd be easier to build our own open source pump 😆
John Males
@johnmales
Jul 08 2016 14:50
Interesting @dougtakeuti - it seems there a number of ways to get the data flowing again.
dougtakeuti
@dougtakeuti
Jul 08 2016 14:52
@johnmales what other methods do you know?
Dana Lewis
@danamlewis
Jul 08 2016 15:20
@joeyede people are looking into flashing firmware, but not success yet.
Chris Oattes
@cjo20
Jul 08 2016 15:21
mm22dl: that's a can of worms you probably don't want to open
mm22dl
@mm22dl
Jul 08 2016 15:22
Why?
John Males
@johnmales
Jul 08 2016 15:22
@dougtakeuti - pressing buttons on the pump, calibrating the pump
Chris Oattes
@cjo20
Jul 08 2016 15:23
@mm22dl I'm not sure about the exact rules, but I'd be surprised if the FDA would be a bit stricter about people sharing designs for homemade drug delivery methods
David Cintron
@loudestnoise
Jul 08 2016 15:23
There is that one dude in the UK who made his own pump
John Males
@johnmales
Jul 08 2016 15:23
@mm22dl despite the many grizzles we have with MDT and their changing firmware, most here (including me) are impressed with the quality and durability of the pumps, precision of dosing etc etc. Open source hardware isn’t something that I would find easy to trust
Chris Oattes
@cjo20
Jul 08 2016 15:26
If you were going to do that, you probably don't want OpenAPS people working on it anyway, or you'll get the same bugs in both systems
joeyede
@joeyede
Jul 08 2016 15:26
@danamlewis if you know who point them in my direction. I have friends who do that kind of thing for a living
Dana Lewis
@danamlewis
Jul 08 2016 15:26
@jaylagorio ^
David Cintron
@loudestnoise
Jul 08 2016 15:27
and I’ve never even heard of this before, but check this oout - https://www.wevolver.com/gerrit.niezen/openpump-an-open-source-hardware-syringe-pump/main/description/
Chris Oattes
@cjo20
Jul 08 2016 15:27
@joeyede not sure if you saw my message earlier, what sort of dev work do you do?
mm22dl
@mm22dl
Jul 08 2016 15:29
Indeed... I'm just thinking about the whole WeAreNotWaiting thing and I feel even more frustrated that those guys at Medtronic are doing everything they can to slow us down and release this black box of theirs they hope will work for every T1D as a closed-loop aka 670g... I'm sure you guys do share those frustrations too
Dana Lewis
@danamlewis
Jul 08 2016 15:29
They're not doing anything to slow us down.
Chris Oattes
@cjo20
Jul 08 2016 15:30
afaik the intent isn't to 'slow us down', it's mandated because of security concerns
Sarah Hodson Grady
@shgrady
Jul 08 2016 15:32
yeah, @johnmales though like the OpenInsulin Project, I think the idea is to put the passion towards the possibility, so that companies can pick up where they left off at competitive prices/or in this case, customizability, as overall through the FDA and safety and to market is reduced at the R&D level.
John Males
@johnmales
Jul 08 2016 15:34
Fair comments. Personally I’d be very happy with new pumps the same as the old firmware Medtronics, and super happy if we had a similar pump with AA batteries and better wireless range. That is all :)
mm22dl
@mm22dl
Jul 08 2016 15:51
They might not slow us down intently, but they aren't collaborating, as far as I know...
Dana Lewis
@danamlewis
Jul 08 2016 15:59
We have had several meetings and conversations with them. They're not able to give out pumps with old firmware, because of the FDA.
I get the frustration with big companies, but want to be fair and that includes recognizing that many have met and talked with us. Some of the co's are trying to figure out ways to support the community. But they also are continuing their business plans to get a commercial AP out. We all want them to continue to do that, too, while we are exploring other options quickly as a community.
Chris Oattes
@cjo20
Jul 08 2016 16:04
Anyone in the UK, I'd advise against buying things from coolcomponents. I received my Edison 2 weeks and 2 days ago. "Please accept our apologies - We really must be informed of problem within the first two weeks after you recieve the goods. Unfortunately after this period we're not able to do much due to the hackable nature of the products and the risk of damage"
Dana Lewis
@danamlewis
Jul 08 2016 16:05
:(
Chris Oattes
@cjo20
Jul 08 2016 16:07
I can see where they're coming from, but I'd expect better customer service from electronics companeis
David Cintron
@loudestnoise
Jul 08 2016 16:07
Booo
Ed Nykaza
@ed-nykaza
Jul 08 2016 16:07
I am getting super close to closing the loop. However, I was wondering what folks are doing to log the decisions and changes that are being made (i.e., the suggested.json and enacted.json files). Thanks!
David Cintron
@loudestnoise
Jul 08 2016 16:08
@ed-nykaza you can set in your cronjob where the log goes
most people either use /var/log/syslog or /var/log/openaps/loop.log
mm22dl
@mm22dl
Jul 08 2016 16:09
@danamlewis You are right... I don't mean to say the big companies are working against us, but it is difficult to see who is actually helping out there. :-(
David Cintron
@loudestnoise
Jul 08 2016 16:09
@ed-nykaza you would add something like | tee -a /var/log/openaps/loop.log at the end of the line
Chris Oattes
@cjo20
Jul 08 2016 16:09
it also adds the files to the git repository openaps sets up
so you can push those to github periodically
David Cintron
@loudestnoise
Jul 08 2016 16:10
I personally like using syslog because I send my logs to papertrail.com so I can trigger notifications when things happen
Ed Nykaza
@ed-nykaza
Jul 08 2016 16:11
thanks @loudestnoise and @cjo20 , this is just what I needed. Planning to go live tonight (fingers-crossed)
David Cintron
@loudestnoise
Jul 08 2016 16:11
:+1:
Dana Lewis
@danamlewis
Jul 08 2016 16:14
@mm22dl I hear you. I wrote this post (https://diyps.org/2015/10/29/building-diabetes-technology-is-like-building-a-mountain-bike/) last year after a meeting at Medtronic. Last paragraph is what’s pertinent. And after being at ADA Scientific Sessions and talking to so many companies alongside Ben and Scott, I feel like this is a sustained attitude.
mm22dl
@mm22dl
Jul 08 2016 16:14
Let me have a look at it
John Males
@johnmales
Jul 08 2016 16:17
@cjo20 do you have any consumer protection laws that can help you? 2 weeks seems awfully short.
Scott Leibrand
@scottleibrand
Jul 08 2016 16:19
@iananders sounds like you're looking at something that uses the same data source as the blue temp basal lines. Those can't be updated until your rig pills a new
until your rig pulls a new pumphistory after an enact and uploads treatments.
Since most people only pull pumphistory once per run, that means it won't update until the next run after the enact.
mm22dl
@mm22dl
Jul 08 2016 16:21
@danamlewis I hope you're right and companies are indeed increasingly willing to collaborate. If so, then we are on the right track. :-)
Btw, I love your articles. Didn't read all of them, but I really enjoyed the ones I did read!
Dana Lewis
@danamlewis
Jul 08 2016 16:22
@mm22dl thanks :)
mm22dl
@mm22dl
Jul 08 2016 16:24
I wish there was a quicker way to render the APS technology more open source... I mean, affordable pumps, CGM systems, open communication protocols... I guess this is just me day dreaming. :P
Scott Leibrand
@scottleibrand
Jul 08 2016 16:24
Companies are understandably focused on solutions that help thousands of people, not dozens. And of course on stuff that can legally be sold to recover costs and scale further to help more people.
The FDA is only just now starting down the road of encouraging and allowing interoperable devices to be approved separately, vs. requiring approval for the system as a whole. They think that may take 5-10 years (if we don't manage to accelerate it).
Dana Lewis
@danamlewis
Jul 08 2016 16:26
I’m not sure we can push price of pumps/CGMs, but we can on open comm protocols :)
in fact, @HowardLook was just with NIH and FDA yesterday I believe discussing these topics :)
joeyede
@joeyede
Jul 08 2016 16:26
@cjo20 I am a PM at Google in SRE
Scott Leibrand
@scottleibrand
Jul 08 2016 16:27
The cost of medical devices comes from the small quantities and all the fixed costs that are required to ensure safety and regulatory compliance (the traditional way).
So one thing we're trying to start working on is demonstrating new more agile methods of demonstrating safety and effectiveness to satisfy regulators. That is a much bigger project than "just" building OpenAPS though.
Terri Lyman
@tatigger
Jul 08 2016 16:31
@scottleibrand 5 to 10 years? Well, doesn't that sound a lot like the "cure" timeframe I've been quoted since ahhhh 1983.....
mm22dl
@mm22dl
Jul 08 2016 16:31
Haha the cure timeframe...
Dana Lewis
@danamlewis
Jul 08 2016 16:31
if you think about the development cycle…that actually makes a lot of sense
mm22dl
@mm22dl
Jul 08 2016 16:32
@scottleibrand I totally get that... that is just a hard reality check
Dana Lewis
@danamlewis
Jul 08 2016 16:32
and because we as a community have been working and advocating, I think it’ll be on the shorter timeframe rather than taking an extra cycle longer
Terri Lyman
@tatigger
Jul 08 2016 16:32
We are certainly contributing motivation to the process : )
FDA
Dana Lewis
@danamlewis
Jul 08 2016 16:34
this is fantastic. well done, @HowardLook :) and thanks for sharing the slides!
Pete Schwamb
@ps2
Jul 08 2016 16:42
:+1:
mm22dl
@mm22dl
Jul 08 2016 16:43
Awesome slides!
Howard Look
@HowardLook
Jul 08 2016 16:43
Thank you! It definitely generated a lively discussion with FDA and device makers. Def read the Torrance & von Hippel paper if you haven't already. They were really helpful.
Hannah Crabtree
@lollydaggle
Jul 08 2016 16:43
are the NIH events typically open to the public? I know FDA has a lot of public meetings. I live in the DC area and I want to be more involved. Plus I have more vacation time than I know what to do with :)
David Cintron
@loudestnoise
Jul 08 2016 16:43
:+1: :+1: :+1: :+1:
Howard Look
@HowardLook
Jul 08 2016 16:44
Close Concerns took good notes on the sessions that hopefully can be shared. Meeting was open to anyone.
Dragan Maksimovic
@dm61
Jul 08 2016 16:44
Great slides! :+1:
Chris Oattes
@cjo20
Jul 08 2016 16:45
johnmales: Theoretically, yes
@joeyede That sounds cool
John Males
@johnmales
Jul 08 2016 16:56
@HowardLook great slides :) do you have a link to the Torrance and Von Hippel paper?
Toby Canning
@TC2013
Jul 08 2016 16:57
I just did a new Edison/Ti Stick over serial build and used an install script that I've used a dozen times previously. For some reason temp basals aren't showing in NS. It seems to be uploading the temp basals in a different time format. I made sure the correct TZ is selected in linux and the TZ clock zoned appears to be running in openaps. Any thoughts?
Chris Oattes
@cjo20
Jul 08 2016 16:58
Are you zoning pumphistory?
Toby Canning
@TC2013
Jul 08 2016 16:58
I'll check that now, but should be
David Cintron
@loudestnoise
Jul 08 2016 16:59
I wish I understood how the zoning stuff worked
Chris Oattes
@cjo20
Jul 08 2016 16:59
@loudestnoise magic, mainly
@TC2013 I'd check that and make sure you're referring to the correct outputs (clock-zoned vs clock) in the appropriate places
David Cintron
@loudestnoise
Jul 08 2016 17:00
Yeah, I like Ben’s note
pi @ raspberrypi): ~/loop
⇢  openaps use tz -h
usage: openaps-use tz [-h] USAGE ...

optional arguments:
  -h, --help  show this help message and exit

## Device tz:
  vendor openapscontrib.timezones

  Timezones - manage timezones in diabetes data with ease.
“with ease"
if I knew how to use it
Paul Martin
@Spazholio
Jul 08 2016 17:00
:thumbsup:
David Cintron
@loudestnoise
Jul 08 2016 17:02
Like say this report add
openaps report add cgm/cgm-glucose.json JSON tz rezone --date display_time --date dateString raw-cgm/raw-entries.json
What would one do if there was no raw cgm because you’re using a G5 and don’t have raw
Paul Martin
@Spazholio
Jul 08 2016 17:03
Simply add the --g5 flag, of course. ;)
David Cintron
@loudestnoise
Jul 08 2016 17:03
Ha, if only.
Chris Oattes
@cjo20
Jul 08 2016 17:04
raw-cgm is just a folder for storing the reports directly from the pump afaik
Toby Canning
@TC2013
Jul 08 2016 17:04
Yup, pump-history is zoned
David Cintron
@loudestnoise
Jul 08 2016 17:05
@cjo20 Yeah, I just don’t understand the arguments I guess
why do I have two --date
Paul Martin
@Spazholio
Jul 08 2016 17:06
So you zone both fields.
The display_time field as well as the dateString one.
Chris Oattes
@cjo20
Jul 08 2016 17:07
loudestnoise: rezone works by looking for specific fields in the JSON and changing their timezone. --date is used to specify which fields it shoud look for
David Cintron
@loudestnoise
Jul 08 2016 17:07
Hmm, so I only have system_time and display_time
{
    "trend_arrow": "FLAT", 
    "system_time": "2016-07-08T17:15:46", 
    "display_time": "2016-07-08T11:19:05", 
    "glucose": 104
  },
Chris Oattes
@cjo20
Jul 08 2016 17:07
so in this case it'll look for "date": X and "dateString":"Y"
Paul Martin
@Spazholio
Jul 08 2016 17:08
What's that from?
David Cintron
@loudestnoise
Jul 08 2016 17:08
my monitor/cgm-glucose.json report
Toby Canning
@TC2013
Jul 08 2016 17:09
I think that's the difference. I was using xDrip before it might store as a datestring, whereas maybe Share is storing as a display_time
David Cintron
@loudestnoise
Jul 08 2016 17:09
which is iter_glucose_hours
Unfair, in my confusion you figured your problem out @TC2013 :laughing:
so I’m thinking, do I even need to zone, because looking at that entry it has display_time, which is already zoned?
Chris Oattes
@cjo20
Jul 08 2016 17:12
doesn't zoning add the "-05:00" part at the end?
Paul Martin
@Spazholio
Jul 08 2016 17:12
Yeah, that display_time isn't zoned.
Chris Oattes
@cjo20
Jul 08 2016 17:12
it needs to know which timezone it's in
David Cintron
@loudestnoise
Jul 08 2016 17:13
Ah ok, that makes sense
Martin Haeberli
@mhaeberli
Jul 08 2016 17:15
David Cintron
@loudestnoise
Jul 08 2016 17:20
so for a use example, what do I put after -z ?
e.g. openaps use tz glucose -z SOMETHING monitor/cgm-glucose.json
Paul Martin
@Spazholio
Jul 08 2016 17:23
To zone it? Hang on...
openaps report show --cli raw-cgm/glucose-zoned.json
Shows this: openaps use tz rezone --date timestamp --date dateString --date start_at --date end_at --date created_at --astimezone raw-cgm/glucose.json
You can probably ignore some of those --date entries. I haven't used this in forever.
David Cintron
@loudestnoise
Jul 08 2016 17:25
did not know you could do report show —cli for a single report! Handy
Paul Martin
@Spazholio
Jul 08 2016 17:25
But it's expecting to see a raw file from a connected Dexcom. At least, that's what mine was.
Yep, it'll take the format in the ini file and give you the command that it would take to re-create it.
David Cintron
@loudestnoise
Jul 08 2016 17:26
I always would do openaps report show --cli and have to find the one I’m looking for
Paul Martin
@Spazholio
Jul 08 2016 17:26
It works perfectly 99% of the time. =)
@bewest and I found a couple that were missing a default or something like that a while back.
David Cintron
@loudestnoise
Jul 08 2016 17:28
I’m gonna figure this out. I just need to stare at it more :wink:
and probably lunch
:meat_on_bone:
iananders
@iananders
Jul 08 2016 17:30
@loudestnoise @scottleibrand thanks guys, very helpful.
John Males
@johnmales
Jul 08 2016 17:55
Thanks @mhaeberli :)
Paul Martin
@Spazholio
Jul 08 2016 18:18
Those of you who are familiar with xDrip - what does the xDrip send to NS that will allow some watchfaces to indicate how much battery is left on the rig?
Oh, and how. How would be nice too. =)
David Cintron
@loudestnoise
Jul 08 2016 18:25
@Spazholio I know nothing about xDrip, but I'll speak anyway ha. I think the battery you're talking about is more used by old school NS users who have and Android phone connected to Dexcom receiver. That's what they would upload to NS I think via the NS Android app to show phone battery life as in that scenario phone + Dex = rig
Paul Martin
@Spazholio
Jul 08 2016 18:31
Right, but that means there's some field I can upload that will make my watchface indicate the battery level of SOMETHING. Now that @cjo20 has made it so we can get battery levels for Edisons running Ubilinux - I want that. I want that very much. =)
Scott Leibrand
@scottleibrand
Jul 08 2016 18:49
might ask in nightscout/public
Paul Martin
@Spazholio
Jul 08 2016 18:52
Good idea.
Dragan Maksimovic
@dm61
Jul 08 2016 20:15
what @cjo20 did with edison battery is amazing - thanks! I'd also be very interested in having that level displayed on NS; unfortunately, I am not familiar with xDrip or the old-school NS rig setup, but will let you know if I find out how they did that.
Jason Wittmer
@jbwittmer
Jul 08 2016 20:42
@HowardLook Enjoyed the slides. Thanks for sharing. Nice work.
jaylagorio
@jaylagorio
Jul 08 2016 21:06
@danamlewis Better update your n variable, last night was night #3 and I plan to use this all weekend!
Thanks to you and everyone else who has helped me along the way, especially @bewest and @CrushingT1D
Dana Lewis
@danamlewis
Jul 08 2016 21:07
@jaylagorio :confetti_ball: :clap: :sparkles:
jaylagorio
@jaylagorio
Jul 08 2016 21:08
What's the count at now?
Dana Lewis
@danamlewis
Jul 08 2016 21:09
;p have to do a few things later to roll the count
jaylagorio
@jaylagorio
Jul 08 2016 21:10
Great!
JaysonEwer
@JaysonEwer
Jul 08 2016 21:11
@jaylagorio NICE!
jaylagorio
@jaylagorio
Jul 08 2016 21:11
Thanks for your Linux help as well
Chris Oattes
@cjo20
Jul 08 2016 21:22
Does anyone with an edison running dev want to try out automatic battery percentage uploading?
I haven't been able to test it properly due to a lack of an edison
Dragan Maksimovic
@dm61
Jul 08 2016 21:31
@cjo20 I could give it a try later today - do not have the rig with me at the moment. I have edison/ubilinux on a sparkfun base block. Which oref0 version should I have?
Chris Oattes
@cjo20
Jul 08 2016 21:32
dev preferably
David Cintron
@loudestnoise
Jul 08 2016 21:32
Woah, this whole priming process is different with an x23 pump compared to the x22
Chris Oattes
@cjo20
Jul 08 2016 21:32
@dm61 I don't know whether it will work with anything other than the head of dev. How much later is later?
Dragan Maksimovic
@dm61
Jul 08 2016 21:37
I'll update to dev; do not know, probably around 9-10pm MDT
Chris Oattes
@cjo20
Jul 08 2016 21:40
if you're not on dev don't worry about it
that might be a bit late for me anyway, it's already 2240 here
Chris Oattes
@cjo20
Jul 08 2016 21:47
@scottleibrand wow that was quick :)
Scott Leibrand
@scottleibrand
Jul 08 2016 21:49
helps that I've seen that exact same pattern before :)
still need to get the merge-ama stuff figured out so we can get down to a single dev branch for all this stuff
Dana Lewis
@danamlewis
Jul 08 2016 21:51
:+1:
Chris Oattes
@cjo20
Jul 08 2016 21:52
@scottleibrand Only concern I have is the code comment an hour ago - it doesn't look like it should work, but it looks like it was like that in my original commit and I thought it worked then, so I don't know what I'm missing
Scott Leibrand
@scottleibrand
Jul 08 2016 21:54
probably worth fixing, at least for future clarity. if you want to test and push a fix to dev I'll pull it into merge-ama
also need to check on sulka's stuff and see if I left any of that out
Jim is CrushingT1D
@CrushingT1D
Jul 08 2016 21:59
@jaylagorio in the house! Congrats
Jeremy M Lucas (GetRileyLink.org)
@jlucasvt
Jul 08 2016 22:04
@jaylagorio :+1: welcome
Martin Haeberli
@mhaeberli
Jul 08 2016 22:14
@cjo20 which Edison OS is best to flash on the Edison ? UbiLinux?
ryanchen26
@ryanchen26
Jul 08 2016 22:29
General question: For those who are actively looping what has the response been like from your M.D. / healthcare provider? I think that Wall Street journal article a while back mentioned support by their doctor for the effort. Is that generally true? Anyone get a negative reaction?
Chris Oattes
@cjo20
Jul 08 2016 22:36
@mhaeberli I use ubilinux
Martin Haeberli
@mhaeberli
Jul 08 2016 22:39
@cjo20 thx - do you have an opinion re ubilinux (whose support I gather is vanishing) vs Yocto vs Brillo ? - I’ll go with ubilinux for now ...
Dana Lewis
@danamlewis
Jul 08 2016 22:40
@ryanchen26 most docs are supportive. mine was more indifferent, but I mostly go in for RXs so wasn’t totally unexpected.
Dana Lewis
@danamlewis
Jul 08 2016 22:50
This is awesome - someone who was at the NIH AP meeting yesterday showed data from the #OpenAPS poster from ADA :)
@jaylagorio ^ your requested count ;)

Re-posting here, for anyone who is not yet looping but IS logging their carbs & bolus data via CarePortal in NS and wants to donate some data:

I’ve been contacted by a researcher who’s looking for a set of basic diabetes data, and wanted to put the ask out there for someone who’s entering data into care portal (so it’ll be easy to pull down from NS): they want 24 hours of data in an excel spreadsheet of CGM data; basal rates; boluses; and carb data. Needs to be from someone who’s NOT looping :) which is why I can’t donate my data for this purpose.

If anyone’s interested in sharing a 24 hour download of the above data, feel free to ping me privately.

He’s hoping for 2-3 people to donate data so he can use it to test in a fuzzy logic controller they’re looking at.

Jason Wittmer
@jbwittmer
Jul 08 2016 23:25
@ryanchen26 Our pediatric endocrinologist was supportive. He said "go for it" and has been interested to hear the results. We're lucky to have a very progressive and open-minded doc
jaylagorio
@jaylagorio
Jul 08 2016 23:28
@danamlewis Awesome, thanks!
Dana Lewis
@danamlewis
Jul 08 2016 23:36
@cjo20 aren’t you supposed to be sleeping?? :)
Chris Oattes
@cjo20
Jul 08 2016 23:36
Probably. I've been getting 4.5 hours a night, I'll be fine :P
And I get the feeling that scott is quite keen on getting his merge done, I think that PR might fix the issue
cjo20 @cjo20 crosses his fingers
Dana Lewis
@danamlewis
Jul 08 2016 23:38
He is because I’m more keen on whacking things off the to-do list :) so sorry!
Scott Leibrand
@scottleibrand
Jul 08 2016 23:40
merged. time for bed? :)
Chris Oattes
@cjo20
Jul 08 2016 23:41
I've really enjoyed the coding on the stuff I've been doing the last couple of days. Compared to what I do at work, which is spend months adding a few small features, getting to write something actually useful and getting it finish is quite a big contrast.
Dana Lewis
@danamlewis
Jul 08 2016 23:41
:) :)
can’t wait to see more!
(which is another way of saying THANK YOU :))
Chris Oattes
@cjo20
Jul 08 2016 23:43
:D As I said, I enjoyed it. And it's nothing really compared to what has been done by others already. What's next on the todo list?
Dana Lewis
@danamlewis
Jul 08 2016 23:45
openaps/oref0#114 - switch over to merged-AMA branch and see if it breaks anything; if so, document what and help us fix those things next? :)
Chris Oattes
@cjo20
Jul 08 2016 23:48
I'll take a look over the weekend :)
G'night
Dana Lewis
@danamlewis
Jul 08 2016 23:48
night!
Scott Leibrand
@scottleibrand
Jul 08 2016 23:49
thx!
Jason Calabrese
@jasoncalabrese
Jul 08 2016 23:52
looking forward to setting up the uploader battery stuff, thanks @cjo20