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

16th
Apr 2018
Dana Lewis
@danamlewis
Apr 16 2018 00:39
Heads up for anyone looking to help out: I just made some notes about what to focus on testing from a fresh dev branch as we near a release candidate: https://github.com/openaps/oref0/pull/816#comment-381450794
Scott Leibrand
@scottleibrand
Apr 16 2018 00:40
I did some testing where we dripped insulin out of a reservoir that read empty into the sink, and measured how many units the pump registered per drop. Somewhere in the 6-8 U past empty range, the pump started registering more units for each drop, meaning that it wasn't pushing out as much as it thought (due to the rubber on the plunger hitting the angled end of the reservoir). somewhere in the 8-10U past empty range it actually registered a No Delivery error, but by then the pump thought it had delivered about 2U more than it actually had.
Dana Lewis
@danamlewis
Apr 16 2018 00:42
@Tornado-Tim when you say res status
Do you mean looking at the physical reservoir, or on the home screen of the pump?
Tim Gunn
@Tornado-Tim
Apr 16 2018 00:43
home screen , but the physical reservoir is inline with what shows on the home screen just not what is shown on OpenAPS or NS
Scott Leibrand
@scottleibrand
Apr 16 2018 00:44
the resolution on the pump screen is like 25%, so it won't go down from "1 bar" to "0 bars" until it gets to reading 0U
Tim Gunn
@Tornado-Tim
Apr 16 2018 00:45
and just to be clear when I mean home screen I mean the reservoir icon on the top left hand side of the screen
Scott Leibrand
@scottleibrand
Apr 16 2018 00:48
What Dana does is replace the reservoir (and usually does a site change as well) around when the pump gets to 0U. If she does a site change before it hits zero, she'll save that reservoir, so that if she gets to zero and isn't home to change it, she can swap in an almost-empty reservoir to make sure she doesn't get to No Delivery before doing a proper reservoir change when she gets home.
Tim Gunn
@Tornado-Tim
Apr 16 2018 00:59
Yea Im used to prepping another cartidge up and keeping it with me till it runs out then swap out when empty, I will probably have to change my Diabetes habits now im looping
Bulbul Ahmed
@bahmedraaz
Apr 16 2018 01:02
Hi, I am using openAPS. In pumphistory.json file, the format of date is like "_date": "4ff30c4a10". I was trying to understand about this representation of date. Can anyone please explain this? I appreciate your support.
alimhassam
@alimhassam
Apr 16 2018 01:34
@bahmedraaz that's the hex representation of the date value of that pump history record, i believe.
here is an excerpt of @ecc1 new go pump comm libs that should probably parse it:
// Decode a 5-byte timestamp from a pump history record.
func decodeTime(data []byte) Time {
  sec := int(data[0] & 0x3F)
  min := int(data[1] & 0x3F)
  hour := int(data[2] & 0x1F)
  day := int(data[3] & 0x1F)
  // The 4-bit month value is encoded in the high 2 bits of the first 2 bytes.
  month := time.Month(int(data[0]>>6)<<2 | int(data[1]>>6))
  year := 2000 + int(data[4]&0x7F)
  return Time(time.Date(year, month, day, hour, min, sec, 0, time.Local))
}
note that oref0 0.70 pumphistory won't contain this field anymore. Why are you trying to parse this field?
Bulbul Ahmed
@bahmedraaz
Apr 16 2018 04:55
Hi @alimhassam , Thank you very much for your support. Actually I am using openAPS for my research work. I don't have any pump, CGM or patient. I install openAPS in my linux machine. Now, I am trying to simulate the behavior of CGM, pump and patient. For this purpose, I need to understand this.
Scott Leibrand
@scottleibrand
Apr 16 2018 04:56
@bahmedraaz that "_date" field is not used by oref0: it's just debug output from the python-based openaps toolkit
and FWIW, we're in the process of deprecating the Python openaps toolkit and replacing it with newer Go-based pump communications libraries. you might want to use the oref0 0.7.0-dev branch for your research, as that simplifies a lot of things
Bulbul Ahmed
@bahmedraaz
Apr 16 2018 05:02
Hi @scottleibrand , Thank you so much. Any suggestion or link, how can I integrate Go-based libraries and also oref0 0.7.0-dev branch?
Scott Leibrand
@scottleibrand
Apr 16 2018 05:02
what's the goal of your research? what exactly are you trying to simulate?
(for anyone else following along, if you're looping and interested in helping test things, we need some folks to test and sign off on 0.6.1 dev first before moving on to testing 0.7.0-dev: https://github.com/openaps/oref0/pull/816#issuecomment-381450794)
Bulbul Ahmed
@bahmedraaz
Apr 16 2018 05:08
I am working on "integrating fault tolerance in medical device". For this, I chose Infusion pump. Meanwhile, while reading a research paper, I came to know about openAPS and I started working with this. Now I am trying to build an artificial pancreas first, then I will work to detect any fault and finally how this system can tolerate fault.
Thomas
@TH2100
Apr 16 2018 05:46
@scottleibrand Thank you! for grep sensitiv ~/myopenaps/settings/*: ISF actually is read from pump. pump-loop.log uses autotune.json values, which currently don’t fit with my always rapidly changing sensitivity … trying to understand more after work.
Scott Leibrand
@scottleibrand
Apr 16 2018 05:48
gotcha. do you use an ISF schedule on the pump? if so, be aware that autotune only tunes based on the first or last one of the day. so if you use a schedule for manual correction dosing, you might want to change your schedule so that the first and last ISF in the schedule (around midnight) are an "average" value for the day.
marafie
@marafie
Apr 16 2018 07:24
hi all, whats the name of the DIY device that gets the BG from Libree sensor ? i cant recall its name, can someone share the link of guthub of it
Samuchco
@Samuchco
Apr 16 2018 10:44
Tt
user1109
@user1109
Apr 16 2018 13:32
@mhaeberli Filesystem Size Used Avail Use% Mounted on
/dev/root 1.4G 874M 442M 67% /
devtmpfs 481M 0 481M 0% /dev
tmpfs 481M 0 481M 0% /dev/shm
tmpfs 481M 13M 468M 3% /run
tmpfs 5.0M 0 5.0M 0% /run/lock
tmpfs 481M 0 481M 0% /sys/fs/cgroup
tmpfs 481M 8.0K 481M 1% /tmp
/dev/mmcblk0p7 32M 4.8M 28M 16% /boot
/dev/mmcblk0p10 1.3G 796M 495M 62% /home
tmpfs 97M 0 97M 0% /run/user/0
ejoe132
@ejoe132
Apr 16 2018 13:33
Hi all, Looking at getting back on openaps after burning up my last edison a few months ago. Is the edison still the primary unit? I looking to buy another one but they are expensive now that their out of produciton
user1109
@user1109
Apr 16 2018 13:34
I nuked some files
user1109
@user1109
Apr 16 2018 13:39
image.png
image.png
Hi to all,
user1109
@user1109
Apr 16 2018 13:54
i can loop only 6-8 hours.
logrotate is not working. pump-loop.log grows up to 850 MB, the loop stops, no space left on device. I changed logrotate .conf from daily to hourly, and did all discussed here: https://gitter.im/nightscout/intend-to-bolus?at=5acd42db7c3a01610dcc1192, but nothing changed.
I suspect that all log files in var / log / openaps should have the extension log.1 after successful logrotate, isn´t it?
manual input: logrotate -f /etc/logrotate.conf is working.
Dana Lewis
@danamlewis
Apr 16 2018 14:17
@ejoe132 yes until Explorer hat's ship. Ping @Kdisimone, she was offering some for sale yesterday.
ejoe132
@ejoe132
Apr 16 2018 14:18
@danamlewis Ok thank you and Ill shoot her a pm
Martin Haeberli
@mhaeberli
Apr 16 2018 18:02
@bahmedraaz related to ‘fault tolerance’ - OpenAPS in conjunction at least w/Medtronic pumps relies on the pump’s default basal rates to ensure ongoing basal delivery if the loop controller (Openaps) or the CGM or whatever fails / prevents looping.
by all means NOT the whole story
statuur
@statuur
Apr 16 2018 18:17
@danamlewis you recently asked me to think about an infographic that explains the general concept of OpenAPS. Are there existing images or other things that i could use as input besides text?
Dana Lewis
@danamlewis
Apr 16 2018 18:20
@statuur the only one we ever had was this one (don't laugh!) to try to show components, so not a lot of existing images to pull from. But here's some context you might want to consider for input: http://openaps.readthedocs.io/en/latest/docs/Understanding%20OpenAPS-Overview/how-openaps-works-overview.html
katie disimone
@Kdisimone
Apr 16 2018 18:21
@scottleibrand @danamlewis looks like there’s a bit of code error here?
 Checking pump clock: "2018-04-16T19:31:34+02:00" is within 1m of current time: Mon Apr 16 19:34:33 CEST 2018
Apr 16 19:34:33 ALEXIAOPENAPS pump-loop.log: Pump clock is more than 55s off: attempting to reset it
Apr 16 19:34:34 ALEXIAOPENAPS pump-loop.log: Waiting for ntpd to synchronize...  OK!
Apr 16 19:34:34 ALEXIAOPENAPS pump-loop.log: Setting pump time to Mon Apr 16 19:34:34 CEST 2018
Apr 16 19:34:38 ALEXIAOPENAPS pump-loop.log: Setting CGM time to Mon Apr 16 19:34:38 CEST 2018
Apr 16 19:34:40 ALEXIAOPENAPS pump-loop.log: openaps-use cgm: error: argument USAGE: invalid choice: 'UpdateTime' (choose from 'bolus', 'config', 'filter_glucose_date', 'filter_isig_date', 'iter_glucose', 'iter_glucose_hours', 'iter_pump', 'iter_pump_hours', 'mmtune', 'model', 'mytest', 'press_keys', 'read_basal_profile_A', 'read_basal_profile_B', 'read_basal_profile_std', 'read_battery_status', 'read_bg_targets', 'read_carb_ratios', 'read_clock', 'read_current_glucose_pages', 'read_current_history_pages', 'read_glucose_data', 'read_history_data', 'read_insulin_sensitivies', 'read_insulin_sensitivities', 'read_selected_basal_profile', 'read_settings', 'read_status', 'read_temp_basal', 'reservoir', 'resume_pump', 'set_clock', 'set_temp_basal', 'settings', 'status', 'suspend_pump', 'test_oref0_compat_menu')
Apr 16 19:34:40 ALEXIAOPENAPS pump-loop.log: Error: pump clock refresh error / mismatch
Apr 16 19:34:43 ALEXIAOPENAPS pump-loop.log: oref0-pump-loop failed.
The longer story is that this rig keeps adding phantom carbs leading to lows. The user is entering in carbs in NS careportal, using NS bolus calculator to calulate the bolus (to account for IOB) and then manually giving bolus in pump (without carb entry). NS is picking up the bolus correctly. OpenAPS is correctly picking up the carb entry from NS (at least to begin with), but then there’s an error with the pump-loop (mostly looks like time issues)…and the rig digs farther back in NS entries and pulls old entries. So, instead of the 20g being active…openaps had 40g active. And then another loop later, it decayed it according to some (?) openaps logic and comes up with 25g active. So, this is happening night after night…the error in the pump loop is causing and error in OpenAPS’s ability to interpret careportal pulls…leading to errant COB. Tonight it was 5g extra in openaps cob, night before was 16g extra in openaps cob.
Dana Lewis
@danamlewis
Apr 16 2018 18:26
@Kdisimone does the bolus calculator record the carbs also into treatments somehow?
katie disimone
@Kdisimone
Apr 16 2018 18:26
Nope…I can see in papertrail that NS correctly gets just the original carb entry. nothing added by bolus calculator
Dana Lewis
@danamlewis
Apr 16 2018 18:27
also, what branch and what version?
can you check NS treatments though to confirm?
not papertrail, but actual NS?
I'd like to know if this is reproducible without the NS bolus wizard step to know if it's related to that or not
user1109
@user1109
Apr 16 2018 18:28
@mhaeberli
hello Martin,
i am sorry, i could not answer to you yesterday
if you have time, we can tackle my problem now
katie disimone
@Kdisimone
Apr 16 2018 18:30
@danamlewis yes…i have access to papertrail, NS, and NS reports. Looking at all of them directly
It’s real unfortunately
There’s several pump-loop errors going on periodically, and I think that’s finding the weakness.
Dana Lewis
@danamlewis
Apr 16 2018 18:32
ok. is this person using SMB? and what version of oref0? (I would suggest they turn off SMB right away, and/or consider lowering maxIOB until we can troubleshoot to completion)
katie disimone
@Kdisimone
Apr 16 2018 18:33
Apr 16 18:35:28 ALEXIAOPENAPS pump-loop.log: Settings refreshed; Refreshed pumphistoryWarning: clock input Invalid Date is unzoned; please pass clock-zoned.json instead
Apr 16 18:35:29 ALEXIAOPENAPS pump-loop.log: Warning: clock input Invalid Date is unzoned; please pass clock-zoned.json instead
Apr 16 18:35:30 ALEXIAOPENAPS pump-loop.log: Warning: clock input Invalid Date is unzoned; please pass clock-zoned.json instead
Apr 16 18:35:30 ALEXIAOPENAPS pump-loop.log:  and meal.json
not using smbs
offline, i’ll send you and scott the info
Martin Haeberli
@mhaeberli
Apr 16 2018 18:34
@user1109 I may have very intermittent time through the day.
Dana Lewis
@danamlewis
Apr 16 2018 18:34
@Kdisimone k, TY
Martin Haeberli
@mhaeberli
Apr 16 2018 18:35
but based on the table you sent earlier, looks like 400 Megabytes free, I think so hard to understand why it’s not working on the theory that logrotate / disk is full
katie disimone
@Kdisimone
Apr 16 2018 18:40
@scottleibrand @danamlewis it’s definitely not user error…I’ve been watching this for 2 days to make sure i could eliminate user error or NS weirdness. This is definitely on the OpenAPS side as far as how it seems to be dealing with recovery from pump-loop errors
statuur
@statuur
Apr 16 2018 18:42
@danamlewis haha.. sorry :-)
Martin Haeberli
@mhaeberli
Apr 16 2018 18:42
@user1109 - you might try uname -a and paste results which will help confirm what jubilinux version you’re running on your edison
Dana Lewis
@danamlewis
Apr 16 2018 18:42
k. will take a look this evening @Kdisimone, thanks for flagging
@statuur :D that's why we need people like you!
statuur
@statuur
Apr 16 2018 18:42
Thats a technical explanation of hardware. Is there something about the pump loop too?
katie disimone
@Kdisimone
Apr 16 2018 18:42
@danamlewis might want to put a timestamp on the convo so that you can find it in papertrail at the right time.
Martin Haeberli
@mhaeberli
Apr 16 2018 18:43
@Kdisimone - suggestion meanwhile - make sure ntp time can get set right when rig is network connected.
katie disimone
@Kdisimone
Apr 16 2018 18:44
@mhaeberli yes, i agree…getting time set right is important. My concern here is that IF/WHEN rig gets a failed loop…it’s not recovering from it well. Instead it appears to be adding carbs from previous entries in NS.
Martin Haeberli
@mhaeberli
Apr 16 2018 18:44
for example what happens with sudo ntpdate -s -b time.nist.gov
user1109
@user1109
Apr 16 2018 18:45
@mhaeberli
Linux myopenaps 3.10.98-jubilinux-edison #13 SMP PREEMPT Fri Jun 23 02:24:12 EDT 2017 i686 GNU/Linux
katie disimone
@Kdisimone
Apr 16 2018 19:00
@danamlewis @scottleibrand every time her MDT data is retrieved, this error message
Apr 16 19:52:58 ALEXIAOPENAPS pump-loop.log: MDT CGM data retrieved
Apr 16 19:53:09 ALEXIAOPENAPS pump-loop.log: MDT New cgm data reformatted
Apr 16 19:53:32 ALEXIAOPENAPS pump-loop.log: Profile less than 60m old; Profile valid. Refreshed pumphistoryWarning: clock input Invalid Date is unzoned; please pass clock-zoned.json instead
Apr 16 19:53:32 ALEXIAOPENAPS pump-loop.log: Warning: clock input Invalid Date is unzoned; please pass clock-zoned.json instead
Apr 16 19:53:33 ALEXIAOPENAPS pump-loop.log: Warning: clock input Invalid Date is unzoned; please pass clock-zoned.json instead
Apr 16 19:53:33 ALEXIAOPENAPS pump-loop.log:  and meal.json
Apr 16 19:53:40 ALEXIAOPENAPS pump-loop.log: IndexError: bytearray index out of range
Apr 16 19:53:40 ALEXIAOPENAPS pump-loop.log: Checking pump clock: "2018-04-16T19:53:17+02:00" is within 1m of current time: Mon Apr 16 19:53:40 CEST 2018
I have also confirmed that the rig is set properly with local time.
Dana Lewis
@danamlewis
Apr 16 2018 20:04
Reposting from last night: nearing release of 0.6.1, testers wanted. See my notes here and also the note at the top of the PR about what all is included. Thanks!
user1109
@user1109
Apr 16 2018 20:18
image.png
Scott Leibrand
@scottleibrand
Apr 16 2018 20:42
@user1109 so which log is filling up? what is it filling up with?
ls -larS shows you the files in the current working directory, sorted by size with the biggest at the bottom
Martin Haeberli
@mhaeberli
Apr 16 2018 21:57
@user1109 - FYI - I’m running 1.1.0 which shows up as Linux etghopenaps2 3.10.98-poky-edison #5 SMP PREEMPT Tue Apr 4 22:42:44 EDT 2017 i686 GNU/Linux- you are running a more recent jubilinux build. I’m NOT saying that is the root cause; just an observation ...
(I think, 1.1.0)
actually, jubilinux-v0.1.1.zip
@Kdisimone your observation leads me to ask: what pump model and firmware version, exactly? US or WW?
Martin Haeberli
@mhaeberli
Apr 16 2018 22:03
@Kdisimone also after
sudo ntpdate -s -b time.nist.gov
try
ntp-wait -n 1 -v
should get: Waiting for ntpd to synchronize... OK!
@Kdisimone the fact it is unzoned time seems an issue.. so - pump clock is good and in sync
katie disimone
@Kdisimone
Apr 16 2018 22:07
@mhaeberli it does synchronize
Martin Haeberli
@mhaeberli
Apr 16 2018 22:08
@Kdisimone has sudo dpkg-reconfigure tzdata been recently tried and re-validated?
katie disimone
@Kdisimone
Apr 16 2018 22:09
yes
i had her send me the results of that and date
Martin Haeberli
@mhaeberli
Apr 16 2018 22:10
sounds like she is in an unusual time zone ; i wonder if somehow that may relate to root cause?
I remember getting lost in zoned/unzoned times 18 months ago or so when I was first trying to figure some of this stuff out.
(hypothesis - there is somehow no rule for converting from UTC +2:00 to UTC or vice versa so the zoned-time thing fails?)
I mean well, but may be hurting more than helping
Scott Leibrand
@scottleibrand
Apr 16 2018 22:13
thanks. I think we probably have enough info to dig in and start debugging, just haven't had time to do so yet
Martin Haeberli
@mhaeberli
Apr 16 2018 22:13
:+1: will stand aside now...
Scott Leibrand
@scottleibrand
Apr 16 2018 22:14
thank you, though, for making sure we check the obvious. it's easy to miss stuff.
katie disimone
@Kdisimone
Apr 16 2018 22:14
Madrid.
Apr 17 00:12:10 ALEXIAOPENAPS pump-loop.log: Your instance of oref0 [0.6.0, master] is up-to-date.
I appreciate it, too @mhaeberli Probably most important to me isn’t necessarily that she has an error. Those are bound to happen. Lots of people get pump-loop errors. My concern is that this pump loop error is resulting in her OpenAPS algorithm having incorrect cob and therefore has potential for adverse event
Scott Leibrand
@scottleibrand
Apr 16 2018 22:19
agreed
katie disimone
@Kdisimone
Apr 16 2018 22:33
cob-mdt.pdf
image.png
Here’s the logs from where it went from reading correctly to reading incorrectly. The actual entry was NS carbs 20g into careportal. Between 19:28 and 19:39, the rig had a few errors the 20g cob went to 26g cob.
katie disimone
@Kdisimone
Apr 16 2018 22:39
30712104_10156325372549313_7361725643588894720_n.jpg
30723947_10156325344934313_1428194552837046272_n.jpg
30730077_10156325345039313_3903942559860260864_n.jpg
katie disimone
@Kdisimone
Apr 16 2018 22:47

The issue seems to be that the error causes the old careportal entries to somehow be “reused” since the carbs (not to be confused with mealCOB) went from 20 to 40, as did the NS cob.

18:21 carbs 50, NS carbs 50, mealCOB 0
18:34 carbs 50, NS carbs 50, mealCOB 0
19:18 carbs 50, NS carbs 50, mealCOB 0
19:23 carbs 20, NS carbs 20, mealCOB 20
19:28 carbs 20, NS carbs 20, mealCOB 20
19:39 carbs 40, NS carbs 40, mealCOB 26

John
@jhendersion_twitter
Apr 16 2018 22:55
I received my Intel Edison
it is connected to a board How do I find out if it is connected to the correct board to be used
Martin Haeberli
@mhaeberli
Apr 16 2018 22:57
@jhendersion_twitter to use in looping you need it connected to an explorer board
can you easily send a picture?
or a link to the exact product you ordered?
John
@jhendersion_twitter
Apr 16 2018 22:59
standby I can send a picture but dont see any numbers
20180416_185358.jpg
20180416_185344.jpg
Martin Haeberli
@mhaeberli
Apr 16 2018 23:07
@jhendersion_twitter so - you definitely need an Explorer board. do you have one?
John
@jhendersion_twitter
Apr 16 2018 23:07
no
Martin Haeberli
@mhaeberli
Apr 16 2018 23:07
with THIS board you can likely flash your Edison, but you won’t be able to do any of the cool looping stuff without
John
@jhendersion_twitter
Apr 16 2018 23:07
I guess that isn't an explorer board
anybody have one for sale
Martin Haeberli
@mhaeberli
Apr 16 2018 23:08
yes see enhanced radio devices explorer baord
hang on for link
never mind that it says pre order - looks like they are in stock
they also claim to have Edisons
which are pre-flashed
nope no longer
John
@jhendersion_twitter
Apr 16 2018 23:09
I already have the edison board
Martin Haeberli
@mhaeberli
Apr 16 2018 23:09
but yours can be flashed
so that’s ok
John
@jhendersion_twitter
Apr 16 2018 23:10
but I can't use it for closed looping
Should I go for the riley link setup and IOS
Martin Haeberli
@mhaeberli
Apr 16 2018 23:10
edison board + explorer + careful instruction following + loopable pump => closed looping
rileylink / iOS are also an excellent way to go.
if you already have an ios phone and a recent macintosh - and in either event, assuming you have a loopable pump
John
@jhendersion_twitter
Apr 16 2018 23:11
shit I might need to go this way and seel this board on ebay
I have a 722 pump with 2.4
Martin Haeberli
@mhaeberli
Apr 16 2018 23:11
the user experiences are -different- from each other.
so that should be loopable
John
@jhendersion_twitter
Apr 16 2018 23:12
I have android but can easly switch to apple phone
I have macbook that I haven't used in 2 years it is only 2 years old
Martin Haeberli
@mhaeberli
Apr 16 2018 23:13
so on the Mac side you should be good. Candidly, it’s an adventure any way you go. LoopKit/Loop channel is probably better for RileyLink/Loop users
it’s a WONDERFUL adventure, but not trivial...
as prerequisites, in my opinion, you should anyway be running Nightscout on Heroku (preferred) or on Azure, assuming you’re using a Dexcom CGM at least.
this goes both for openaps and for loop/rileylink
I have even recently heard some argument for what I’ll call ‘hybrid’ looping with both openaps and Loop … but I’m not in a position to explain that, and you’d have to get each running then.
John
@jhendersion_twitter
Apr 16 2018 23:18
I have netscout setup on Heroku I do have G5
Martin Haeberli
@mhaeberli
Apr 16 2018 23:18
great start then!
John
@jhendersion_twitter
Apr 16 2018 23:18
was looking to close loop with my samsung phone but if I can't get the board then I need to do soemthing elase
els
e
Martin Haeberli
@mhaeberli
Apr 16 2018 23:19
you can get the Explorer board
it’s just new Edisons that you can no longer get from EnhancedRadioDevices
depending on time/effort/money tradeoffs, if it was me, I would get an Explorer Board and a RileyLink ordered, along with an iPhone 7 or later (for G5 compatibility), and explore both paths in parallel.
John
@jhendersion_twitter
Apr 16 2018 23:22
so what you are saying is try both setup and decide which is better for me
Martin Haeberli
@mhaeberli
Apr 16 2018 23:26
yes - but remember it’s much more effortful. If you want something where (at the expense of learning a little linux command line stuff) you have waaay more configuration parameters you can easily set, and you can even manage remotely over the net if you want, then go with OpenAPS. If you want something that looks more like a turnkey solution but lacks lots of degrees of freedom, and you have an iPhone and a Mac, then go with Loop for now. [@danamlewis, @scottleibrand, @all, please weigh in to the extent I may have mis-characterized… ]. We tried Loop ~20 months ago, in its early days, didn’t quite fly for us. We’ve been looping on OpenAPS since maybe last May...
Scott Leibrand
@scottleibrand
Apr 16 2018 23:34
@jhendersion_twitter you have two pieces there: an Intel Edison, and a breakout board
the Edison is what you need. the breakout board is surplus, and should be replaced with the Explorer board
unless you have a Rileylink, you'll need to order something either way: either an Explorer board to work with your Edison, or a Rileylink to work with your iPhone.
John
@jhendersion_twitter
Apr 16 2018 23:58
where can I get the explorer board ?
Martin Haeberli
@mhaeberli
Apr 16 2018 23:59
see link above : https://enhanced-radio-devices.myshopify.com/products/900mhz-explorer-block-pre-order
it says pre-order, but it’s not - it looks available. i just ordered two more last week and they got here :-)
John
@jhendersion_twitter
Apr 16 2018 23:59
ok cool