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

2nd
May 2015
Scott Leibrand
@scottleibrand
May 02 2015 01:07 UTC
thx @bewest for updating the readme. got an instance created, but am still having issues trying to use "openaps device". opened openaps/openaps#3
Ben West
@bewest
May 02 2015 01:56 UTC
ah
that's progress
Ben West
@bewest
May 02 2015 02:27 UTC
just discovered same bug in dexcom_reader also
my sources are updated
diabeticgonewild
@diabeticgonewild
May 02 2015 02:27 UTC
Cool
diabeticgonewild
@diabeticgonewild
May 02 2015 02:34 UTC
Just another day/night/whatever of me feeling crappy. I really want to get some work done...hopefully tomorrow. I'm tired of being this way.
diabeticgonewild
@diabeticgonewild
May 02 2015 02:46 UTC
It's a really sad thing for me to be this way.
diabeticgonewild
@diabeticgonewild
May 02 2015 03:44 UTC
Alright I took a shower and took some opium (prescribed to me for GI problems--went a day without taking it which was a mistake) and I feel slightly better...
Scott Leibrand
@scottleibrand
May 02 2015 04:13 UTC
@bewest anything I can do to help troubleshoot?
diabeticgonewild
@diabeticgonewild
May 02 2015 04:13 UTC
I got a W in my class already cause I took a complete session withdrawal, and my academic advisors encouraged me to take medical withdrawal on top of that. School is the least of my concerns right now...I don't even know how I am going to get through it feeling so ceappy
Ben West
@bewest
May 02 2015 04:13 UTC
@scottleibrand should be fixed if you pull in dexcom_reader from me and also decoding-carelink
diabeticgonewild
@diabeticgonewild
May 02 2015 04:13 UTC
crappy and stuff but maybe chemo will turn that around.
Ben West
@bewest
May 02 2015 04:13 UTC
it was same issue all 3 times
:-(
same mistake
diabeticgonewild
@diabeticgonewild
May 02 2015 04:14 UTC
I need to get a raspberry pi 2...soon
Ben West
@bewest
May 02 2015 04:14 UTC
@diabeticgonewild , yeah, sounds awful :-(
diabeticgonewild
@diabeticgonewild
May 02 2015 04:15 UTC
Hey that's how I live day to day...like every semester after my first since starting college it's been like "well I hope I can complete this semester"
Scott Leibrand
@scottleibrand
May 02 2015 04:16 UTC
ah, I had to do setup.py again too
Ben West
@bewest
May 02 2015 04:16 UTC
yes, you will have to do sudo python setup.py develop for dexcom_reader and decoding-carelink
also, don't run openaps in side the openaps repo
diabeticgonewild
@diabeticgonewild
May 02 2015 04:16 UTC
The school I go to now is infinitely times nicer than the one I was going to...btw doing laundry and cleaning...yay self care and dealing with things...
Scott Leibrand
@scottleibrand
May 02 2015 04:17 UTC
k
Dana Lewis
@danamlewis
May 02 2015 04:17 UTC
@bewest why not inside the repo?
Ben West
@bewest
May 02 2015 04:17 UTC
the repo is for the tool's source code
Scott Leibrand
@scottleibrand
May 02 2015 04:17 UTC
presumably because then it contaminates any commits you try to make with your myopenaps stuff
Ben West
@bewest
May 02 2015 04:18 UTC
the tool itself creates files and manages it's operation inside a new git repo that it manages itself
so using a tool to manipulate git repos inside of a git repo for another tool might lead to uh... interesting results
Scott Leibrand
@scottleibrand
May 02 2015 04:18 UTC
:)
Dana Lewis
@danamlewis
May 02 2015 04:18 UTC
Ha. Got it
Ben West
@bewest
May 02 2015 04:19 UTC
that's why openaps-example is literally the output of the tool
so we get meshing/backup/restore for free
your 4 boxen can remote to each other to catch up/sync up etc
Dana Lewis
@danamlewis
May 02 2015 04:19 UTC
Ah, makes sense
Ben West
@bewest
May 02 2015 04:20 UTC
also you can backup in private/public github repo and invite a friend to analyze things if something goes wrong
Scott Leibrand
@scottleibrand
May 02 2015 04:21 UTC
pi@raspberrypi3 ~/myopenaps $ openaps use pump status
PowerControl SERIAL 192788
Traceback (most recent call last):
File "/usr/local/bin/openaps-use", line 8, in <module>
execfile(file)
File "/home/pi/openaps/bin/openaps-use", line 61, in <module>
app( )
File "/home/pi/openaps/openaps/cli/init.py", line 41, in call
self.run(self.args)
File "/home/pi/openaps/bin/openaps-use", line 54, in run
output = app(args, self)
File "/home/pi/openaps/openaps/uses/init.py", line 71, in call
return self.method.selected(args)(args, app)
File "/home/pi/openaps/openaps/uses/init.py", line 31, in call
return self.method(args, app)
File "/home/pi/openaps/openaps/uses/use.py", line 23, in call
output = self.main(args, app)
File "/home/pi/openaps/openaps/vendors/medtronic.py", line 159, in main
return self.pump.model.read_status( )
AttributeError: 'ReadPumpModel' object has no attribute 'read_status'
pi@raspberrypi3 ~/myopenaps $
Ben West
@bewest
May 02 2015 04:22 UTC
hmm
progress
try model
oh
uh
hmmm
hmmm
Scott Leibrand
@scottleibrand
May 02 2015 04:23 UTC

pi@raspberrypi3 ~/myopenaps $ openaps use pump settings
PowerControl SERIAL 192788
No handlers could be found for logger "decocare.stick"
Traceback (most recent call last):

<snip>
File "/home/pi/decocare/decocare/stick.py", line 763, in download
len(results), len(data))))
TypeError: object of type 'NoneType' has no len()

Ben West
@bewest
May 02 2015 04:23 UTC
oh
oh
in decoding-carelink
sorry...
hmmm
you must be using master
Scott Leibrand
@scottleibrand
May 02 2015 04:23 UTC
should I be on decocare dev?
Ben West
@bewest
May 02 2015 04:24 UTC
give me a sec, I'll merge to master
Scott Leibrand
@scottleibrand
May 02 2015 04:25 UTC
k
Ben West
@bewest
May 02 2015 04:25 UTC
just merged
so if you pull master again you'll get the good stuff
Scott Leibrand
@scottleibrand
May 02 2015 04:26 UTC
pulled that, now:

pi@raspberrypi3 ~/myopenaps $ openaps use pump model
PowerControl SERIAL 192788
No handlers could be found for logger "decocare.stick"
Traceback (most recent call last):

<snip>
File "/home/pi/decocare/decocare/stick.py", line 231, in parse
raise AckError("readStatus: non-zero status: %02x" % self.ack)
decocare.errors.AckError: readStatus: non-zero status: 4a

Ben West
@bewest
May 02 2015 04:27 UTC
develop again
oh
oh
hmm
run python -m decocare.stick $(python -m decocare.scan)
Scott Leibrand
@scottleibrand
May 02 2015 04:27 UTC
k, ran develop again..
Ben West
@bewest
May 02 2015 04:27 UTC
counters should be ok
Scott Leibrand
@scottleibrand
May 02 2015 04:28 UTC
yeah, no dice. lemme troubleshoot with decocare directly first
Ben West
@bewest
May 02 2015 04:28 UTC
that python decocare command shoudl work anywhere
the mm- tools should be available anywhere also :-)
Scott Leibrand
@scottleibrand
May 02 2015 04:28 UTC
it errors out the same way, with decocare.errors.AckError: readStatus: non-zero status: 4a
Ben West
@bewest
May 02 2015 04:29 UTC
ok, unplug/replug stick I think
Scott Leibrand
@scottleibrand
May 02 2015 04:29 UTC
k
ok, fixed the usb comms and trying again
model works
Ben West
@bewest
May 02 2015 04:32 UTC
sweet, whole thing should work now
Scott Leibrand
@scottleibrand
May 02 2015 04:32 UTC
what does No handlers could be found for logger "decocare.stick" mean?
Ben West
@bewest
May 02 2015 04:32 UTC
right now, we just have ability to play around with desinging reports
it's the logger subsystem itself warning us that the logger is being used by has no place to go
this system tries to be quite a bit more silent than the others
Scott Leibrand
@scottleibrand
May 02 2015 04:33 UTC
so just not implemented yet?
Ben West
@bewest
May 02 2015 04:33 UTC
no, just ignore it
diabeticgonewild
@diabeticgonewild
May 02 2015 04:33 UTC
No dice! Lol that scene at fast times at ridgemont high! My physics teacher in college approached me and told me that he had been trying to figure out all semester who I reminded him of and he said he finally figured it out. Jeff Spicolli on Fast Times at Ridgemont High...but for some reason he emphasized the "female version" to try not to offend me...
Scott Leibrand
@scottleibrand
May 02 2015 04:34 UTC
funny, I have no idea where "no dice" originally come from. just something I picked up by osmosis. :)
Ben West
@bewest
May 02 2015 04:34 UTC
it's just a logging message
it's actually a meta-meta logging message
saying that the logging system isn't configured
but I know that and approve
I'll silence that warning at some point
diabeticgonewild
@diabeticgonewild
May 02 2015 04:35 UTC
Scott Leibrand
@scottleibrand
May 02 2015 04:35 UTC
ooh, fun, I broke it:
pi@raspberrypi3 ~/myopenaps $ openaps use pump iter_pump
No handlers could be found for logger "decocare.stick"
ERROR month must be in 1..12 0000 0x17 0x19 ..
ERROR month must be in 1..12 0000 0x17 0x19 ..
Traceback (most recent call last):
File "/home/pi/decocare/decocare/stick.py", line 307, in parse
assert len(data) == resLength
AssertionError
Ben West
@bewest
May 02 2015 04:36 UTC
oh, interesting
hmmm
Scott Leibrand
@scottleibrand
May 02 2015 04:36 UTC
I had just done some temp basals to populate the history of this recently-reset test pump
Ben West
@bewest
May 02 2015 04:36 UTC
ok
well
what's the model?
I'd like the data
I mean it could be you just have records I've never seen before...
Scott Leibrand
@scottleibrand
May 02 2015 04:37 UTC
model 515. it thinks it's 2004
Ben West
@bewest
May 02 2015 04:37 UTC
yeah, that's fine
get a page of data
Scott Leibrand
@scottleibrand
May 02 2015 04:38 UTC
what's the command for that again?
i just changed the date, lemme see if that fixed it first
Ben West
@bewest
May 02 2015 04:38 UTC
you can use mm-send-comm.py tweak ReadHistoryData --page 0 --save
export SERIAL
the date is not an issue
Scott Leibrand
@scottleibrand
May 02 2015 04:38 UTC
k
Ben West
@bewest
May 02 2015 04:38 UTC
those errors pop up when I don't know what the data is
Scott Leibrand
@scottleibrand
May 02 2015 04:38 UTC
still broke, getting data. where to put it?
Ben West
@bewest
May 02 2015 04:39 UTC
if I were you for now, I'd consider your private openaps?
do export SERIAL=...
Scott Leibrand
@scottleibrand
May 02 2015 04:39 UTC
yeah, already did that. isn't this really a decocare thing?
Ben West
@bewest
May 02 2015 04:39 UTC
mm-send-comm.py tweak ReadHistoryData --page 0 --save should save page of data
Scott Leibrand
@scottleibrand
May 02 2015 04:39 UTC
I could do an issue there
Ben West
@bewest
May 02 2015 04:40 UTC
I need the raw data
Scott Leibrand
@scottleibrand
May 02 2015 04:40 UTC
just asking where you want it sent/saved
Ben West
@bewest
May 02 2015 04:40 UTC
I'd put it in your private openaps repo
Scott Leibrand
@scottleibrand
May 02 2015 04:40 UTC
k, I renamed that to openaps-old
Ben West
@bewest
May 02 2015 04:40 UTC
since you have that set up and we have shared access
got it :-)
heh
Ben West
@bewest
May 02 2015 04:44 UTC
groan
it's the data...
Scott Leibrand
@scottleibrand
May 02 2015 04:44 UTC
this is that same new-pump problem as before?
Ben West
@bewest
May 02 2015 04:44 UTC
new-pump problem?
no, I knew about this sort of
are you able to get csv from carelink?
Scott Leibrand
@scottleibrand
May 02 2015 04:44 UTC
we had a bunch of unreadable data the last time I reset this pump I think
maybe, but it's a ton of work
Ben West
@bewest
May 02 2015 04:45 UTC
the records in particular
so at midnight or at some point
Scott Leibrand
@scottleibrand
May 02 2015 04:45 UTC
carelink's java crap is not fun.
Ben West
@bewest
May 02 2015 04:45 UTC
the pump writes several records about what happened all day
summaries
these are the problematic records
Scott Leibrand
@scottleibrand
May 02 2015 04:45 UTC
but if this would be valuable for other folks, I can probably do it
Ben West
@bewest
May 02 2015 04:45 UTC
it's a problem on all pumps
but especially 512, 515 I think
well
are we looking at 2008 on this pump
?
2005?
er
2004
I see battery and low reservoir for 2004
Scott Leibrand
@scottleibrand
May 02 2015 04:46 UTC
it was 2004 until I changed it, ya
Ben West
@bewest
May 02 2015 04:46 UTC
2004-01-21 ?
Scott Leibrand
@scottleibrand
May 02 2015 04:47 UTC
that was the default date when it reset. didn't bother changing it until just now
Ben West
@bewest
May 02 2015 04:47 UTC
it's fine
Scott Leibrand
@scottleibrand
May 02 2015 04:47 UTC
not 100% sure it was the 21st, but that sounds reasonable
I probably reset the pump 3 weeks ago
and have had it on a zero-basal pattern since
in other news, openaps use cgm glucose works, but openaps use cgm iter_glucose 3 gives me AttributeError: 'Dexcom' object has no attribute 'iter_records'
diabeticgonewild
@diabeticgonewild
May 02 2015 04:48 UTC
That sucks
Scott Leibrand
@scottleibrand
May 02 2015 04:49 UTC
not really, we're finding bugs and fixing them. this is normal development. :)
diabeticgonewild
@diabeticgonewild
May 02 2015 04:49 UTC
Ph
Oh
Scott Leibrand
@scottleibrand
May 02 2015 04:49 UTC
this is brand spanking new code
no one but ben has ever run it afaik. :)
Ben West
@bewest
May 02 2015 04:49 UTC
ah
ok
so you need my branch
Scott Leibrand
@scottleibrand
May 02 2015 04:50 UTC
k, which one?
diabeticgonewild
@diabeticgonewild
May 02 2015 04:50 UTC
True. I forked it but I need to get a raspberry pi 2 like this weekend or something...
Ben West
@bewest
May 02 2015 04:50 UTC
you can do: git remote add bewest git@github.com:bewest/dexcom_reader.git
inyour dexcom_reader repo
and then git fetch bewest
then git checkout bewest master
Scott Leibrand
@scottleibrand
May 02 2015 04:51 UTC
is that the same?
Ben West
@bewest
May 02 2015 04:51 UTC
oh
well
Scott Leibrand
@scottleibrand
May 02 2015 04:51 UTC
oh duh
Ben West
@bewest
May 02 2015 04:51 UTC
yeah
hehe
hmm
checking my repo
Scott Leibrand
@scottleibrand
May 02 2015 04:51 UTC
I'm on master though
Ben West
@bewest
May 02 2015 04:51 UTC
I'm clean
Scott Leibrand
@scottleibrand
May 02 2015 04:52 UTC
it says there is no bewest branch
Ben West
@bewest
May 02 2015 04:52 UTC
yeah
I meant myrepo
if I type git show
it says fri may 1, commit c1ccee .., packages=find_packages( )
Scott Leibrand
@scottleibrand
May 02 2015 04:53 UTC
actually you said dexcom_reader not decoding-dexcom
what's the diff?
Ben West
@bewest
May 02 2015 04:53 UTC
dexcom_reader is for dexcom
decoding-dexcom is nothing
Scott Leibrand
@scottleibrand
May 02 2015 04:53 UTC
ok, there's a dexcom_reader inside decoding-dexcom
but that's old?
Ben West
@bewest
May 02 2015 04:53 UTC
no
Ben West
@bewest
May 02 2015 04:54 UTC
decoding-dexcom is my bookmarking for the activity in general, and I found a library called dexcom_reader, that lives here https://github.com/compbrain/dexcom_reader
so I point to dexcom_reader in my pages
my fork/branch/repo is here: https://github.com/bewest/dexcom_reader
Scott Leibrand
@scottleibrand
May 02 2015 04:54 UTC
so should I clone https://github.com/bewest/dexcom_reader.git instead of decoding-dexcom?
Ben West
@bewest
May 02 2015 04:54 UTC
yes
Scott Leibrand
@scottleibrand
May 02 2015 04:54 UTC
k
Ben West
@bewest
May 02 2015 04:54 UTC
you do not need decoding-dexcom
Scott Leibrand
@scottleibrand
May 02 2015 04:54 UTC
k
Ben West
@bewest
May 02 2015 04:55 UTC
maybe I should make it less tempting somehow
that has confused me a few times
that's a good one
Scott Leibrand
@scottleibrand
May 02 2015 04:55 UTC
yay: pi@raspberrypi3 ~/myopenaps $ openaps use cgm iter_glucose 3
[{'trend_arrow': 'FLAT', 'system_time': '2015-05-02T04:49:15', 'display_time': '2015-05-01T20:51:52', 'glucose': 83}, {'trend_arrow': 'FLAT', 'system_time': '2015-05-02T04:44:15', 'display_time': '2015-05-01T20:46:52', 'glucose': 85}, {'trend_arrow': 'FLAT', 'system_time': '2015-05-02T04:39:15', 'display_time': '2015-05-01T20:41:52', 'glucose': 88}]
Ben West
@bewest
May 02 2015 04:55 UTC
hoohoooooohoo
w00t
muahahahahaha
happy dance
ok, you can "store" that config as a report
then whenever you invoke report it runs with those args
you can config it with json vs text output
JSON, sorry
Scott Leibrand
@scottleibrand
May 02 2015 04:57 UTC
where does it put it?
pi@raspberrypi3 ~/myopenaps $ openaps report add glucose.json JSON cgm glucose
app {}
added cgm://JSON/glucose/glucose.json
Ben West
@bewest
May 02 2015 04:57 UTC
put what
ok
Scott Leibrand
@scottleibrand
May 02 2015 04:57 UTC
what is cgm:// ?
Ben West
@bewest
May 02 2015 04:57 UTC
that goes in openaps.ini :-)
Scott Leibrand
@scottleibrand
May 02 2015 04:57 UTC
ah, then invoke
Ben West
@bewest
May 02 2015 04:57 UTC
that's an internal URI used to refer to this report's config
now you can openaps report invoke glucose.json
Dana Lewis
@danamlewis
May 02 2015 04:57 UTC
@bewest yes for making it less tempting
Ben West
@bewest
May 02 2015 04:58 UTC
openaps now knows how to "build" glucose.json
Dana Lewis
@danamlewis
May 02 2015 04:58 UTC
Yay! Now I know my BGs from another room ;)
Scott Leibrand
@scottleibrand
May 02 2015 04:58 UTC
we should make openaps report invoke glucose.json intelligently do an iter_glucose of just however much data is new
it looks like it's pulling everything each time and taking awhile to do it
Ben West
@bewest
May 02 2015 04:58 UTC
no
hmm
it should be fast
there was a bug I fixed earlier this afternoon
oh
uh
Scott Leibrand
@scottleibrand
May 02 2015 04:59 UTC

pi@raspberrypi3 ~/myopenaps $ time openaps report invoke glucose.json
app {u'glucose.json': <openaps.reports.report.Report object at 0x2491570>}
glucose <class 'openaps.vendors.dexcom.glucose'> <openaps.vendors.dexcom.glucose object at 0x247f650>
cgm://JSON/glucose/glucose.json
reporting glucose.json

real 0m24.069s
user 0m17.420s
sys 0m0.460s
pi@raspberrypi3 ~/myopenaps $

Ben West
@bewest
May 02 2015 04:59 UTC
oh
oh
ok
so the filename != equal use
that "report" is the entire db
you want the iter_glucose report, NOT the glucose report
sorry ,the use
Scott Leibrand
@scottleibrand
May 02 2015 05:00 UTC
not following. how do I pull the entire glucose history into the db, and then just incrementally update it?
Ben West
@bewest
May 02 2015 05:00 UTC
the report's filename is glucose.json, you should use the iter_glucose use/report to create it, not the glucose use
eg:
those are two different "reports"/uses
there is one use for getting entire db
another use for getting few records
you can configure both of them as reports
Scott Leibrand
@scottleibrand
May 02 2015 05:02 UTC
what is the report to get the glucose measurements since the last one in the db?
or is that level of smarts not built in yet?
Ben West
@bewest
May 02 2015 05:02 UTC
no, it doesn't work that way
right now, we have "last x records"
you want one that can do "since $x date"
that's 10-ish lines in dexcom.py for now
Scott Leibrand
@scottleibrand
May 02 2015 05:03 UTC
yeah, it'd be nice if it figured out what it needed, eventually
no big deal though for now
Ben West
@bewest
May 02 2015 05:03 UTC
you would also need
Scott Leibrand
@scottleibrand
May 02 2015 05:03 UTC
so would i do openaps report add glucose.json JSON cgm iter_glucose ?
Ben West
@bewest
May 02 2015 05:03 UTC
"this is the last cgm datetime in the db" use
yeah
use iter_glucose instead of glucose
also
it's the same sytnax there as use
so it's actually : iter_glucose 3
or however many you want
Scott Leibrand
@scottleibrand
May 02 2015 05:04 UTC
so that overrides the invoke thing on glucose.json
Ben West
@bewest
May 02 2015 05:04 UTC
invoke takes no args
just the repot name
Scott Leibrand
@scottleibrand
May 02 2015 05:05 UTC
yeah, but redefining glucose.json to be an iter_glucose report blew away all the history
is there a way to have it append?
not that I really need it for openaps purposes, just wondering how it works
Ben West
@bewest
May 02 2015 05:05 UTC
no, that's not the way it will work
this is "active RAM" of openaps
Scott Leibrand
@scottleibrand
May 02 2015 05:05 UTC
ok, so everything is just pulling latest, k
Ben West
@bewest
May 02 2015 05:05 UTC
archiving is different problem
yeah
we will optimize all this away eventually
Scott Leibrand
@scottleibrand
May 02 2015 05:06 UTC
that matches how I want the design to work, good
Ben West
@bewest
May 02 2015 05:06 UTC
yeah, this is only "the active stuff in openaps"
so the repots/config help to establish "what is current environment/context?"
and soon "what steps should we run in what order"
Scott Leibrand
@scottleibrand
May 02 2015 05:06 UTC
ok, so did you figure out the encoding of the weird pump records?
or do you still need a carelink csv?
Ben West
@bewest
May 02 2015 05:06 UTC
csv would help a lot
Scott Leibrand
@scottleibrand
May 02 2015 05:07 UTC
ok, let me fire up the PC
no guarantees tho
Ben West
@bewest
May 02 2015 05:07 UTC
parsing is definitely wrong at record 13
record 12 is "old6c" but looks wrong (I don't even know what old6c is)
hmmm
Scott Leibrand
@scottleibrand
May 02 2015 05:15 UTC
bah, have to wait for the infrequently-used PC to do its auto-updates before I can use it
Ben West
@bewest
May 02 2015 05:16 UTC
ok, wow
lots of little differences between models
i just put in big models subsystem to start making these distinctions
Ben West
@bewest
May 02 2015 05:28 UTC
ok, I tried a patch
@scottleibrand if you do git checkout bewest/dev in decoding-carelink it might work
mm-decode-history-page.py --model 515 ReadHistoryData-page-0.data works better now
Scott Leibrand
@scottleibrand
May 02 2015 05:41 UTC
do these various openaps commands have auto-init logic in them?
Ben West
@bewest
May 02 2015 05:42 UTC
for medtronic?
Scott Leibrand
@scottleibrand
May 02 2015 05:42 UTC
ya
Ben West
@bewest
May 02 2015 05:42 UTC
the way it works for medtronic, is it stores the time of the last known session expiry in openaps.ini
so session expiry is tracked in git automagically
for now, it will print POWER CONTROL if its creating new session
Scott Leibrand
@scottleibrand
May 02 2015 05:43 UTC
I haven't done anything with the pump recently and it's not doing POWERCONTROl
Ben West
@bewest
May 02 2015 05:43 UTC
otherwise it will try to re-use existing
Scott Leibrand
@scottleibrand
May 02 2015 05:43 UTC
expires = 2015-05-01T21:42:02.202564
Ben West
@bewest
May 02 2015 05:43 UTC
there's not a way to customize session life, right now it's 10 minutes
Scott Leibrand
@scottleibrand
May 02 2015 05:43 UTC
pi@raspberrypi3 ~/myopenaps $ date
Fri May 1 22:43:37 PDT 2015
Ben West
@bewest
May 02 2015 05:44 UTC
hmm
but if you don't do anything, it doesnt' need a new session
Scott Leibrand
@scottleibrand
May 02 2015 05:45 UTC
pi@raspberrypi3 ~/myopenaps $ openaps use pump status
File "/home/pi/decocare/decocare/stick.py", line 65, in respond
raise AckError("ACK is 0 bytes:\n%s" % lib.hexdump(raw))
decocare.errors.AckError: ACK is 0 bytes:
no attempt to init
Ben West
@bewest
May 02 2015 05:45 UTC
ok, that's different
weird... seem to be having stick issues tonight
does python -m decocare.stick $(python -m decocare.scan) print normal counters?
Scott Leibrand
@scottleibrand
May 02 2015 05:46 UTC
could be my normal decocare stuff running in the background in between what I'm doing
Ben West
@bewest
May 02 2015 05:46 UTC
oh?
Scott Leibrand
@scottleibrand
May 02 2015 05:46 UTC
I'm touching the lockfile before I use the stick, but I haven't shut it down
Ben West
@bewest
May 02 2015 05:46 UTC
you can't run things at same time
lockfile?
Scott Leibrand
@scottleibrand
May 02 2015 05:46 UTC
shouldn't be actually at the same time, I have a lockfile that prevents that
but let me kill the cron job completely and start over
Ben West
@bewest
May 02 2015 05:47 UTC
ah, that would be problem, yeah
you may need to replug stick
only one process can use it at a time or it gets mad
you can use more than one tick at same time though
Scott Leibrand
@scottleibrand
May 02 2015 05:48 UTC
ok, stick reinsert fixed it
probably conflicted at one point and then never got unstuck
Ben West
@bewest
May 02 2015 05:48 UTC
yeah, for stick ACK issues, that's generally the fix
Scott Leibrand
@scottleibrand
May 02 2015 05:49 UTC
still got ERROR month must be in 1..12 0000 0x17 0x19 ..
ERROR month must be in 1..12 0000 0x17 0x19 ..
but no crash, and lots of data
including a bunch of ugly stuff like {'_type': 'ChangeBasalProfile_old_profile', '_description': 'ChangeBasalProfile_old_profile 2004-01-01T23:45:35 head2, body[145] op[0x08]', 'date': 1073029535000.0, 'timestamp': '2004-01-01T23:45:35', '_body': '3f000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000', '_head': '0800', 'rate': 'offset', '_date': '236d370104'}
Ben West
@bewest
May 02 2015 05:50 UTC
can you get me page again
Scott Leibrand
@scottleibrand
May 02 2015 05:51 UTC
sure
Ben West
@bewest
May 02 2015 05:51 UTC
ChangeBasalProfile is big record
Scott Leibrand
@scottleibrand
May 02 2015 05:52 UTC
no changes in the record
I can do another temp if you want something to change
Ben West
@bewest
May 02 2015 05:52 UTC
hmm
--page 1 or --page 2
Scott Leibrand
@scottleibrand
May 02 2015 05:53 UTC
but I don't think anything has happened since the last one I gave you. git diff shows no changes
k
Ben West
@bewest
May 02 2015 05:53 UTC
fwiw, I recommend the iter_foo methods
oh
falso
Scott Leibrand
@scottleibrand
May 02 2015 05:54 UTC
pushed
Ben West
@bewest
May 02 2015 05:54 UTC
thanks
weir
not getting those errors
Scott Leibrand
@scottleibrand
May 02 2015 05:55 UTC
did you update anything other than decocare?
Ben West
@bewest
May 02 2015 05:57 UTC
nn
hmm I can decode those just fine
no errors about months
Scott Leibrand
@scottleibrand
May 02 2015 05:58 UTC
trying it after setup.py again. you'd think I'd have learned to do that every time I git pull by now. :)
iter_pump didn't PowerControl. status did.
Ben West
@bewest
May 02 2015 05:58 UTC
ah
Scott Leibrand
@scottleibrand
May 02 2015 05:59 UTC
still errors
Ben West
@bewest
May 02 2015 05:59 UTC
that's fine
Scott Leibrand
@scottleibrand
May 02 2015 05:59 UTC
even after setup.py
Ben West
@bewest
May 02 2015 05:59 UTC
status is erroring?
or iter_pump?
Scott Leibrand
@scottleibrand
May 02 2015 05:59 UTC
same iter_pump error:
actually lemme put it in git
pushed
Ben West
@bewest
May 02 2015 06:01 UTC
eesh
oh
hmm
that's fine
no
ok
let's run a test
:-)
git pull that same thing
run make
there should be no diff
Scott Leibrand
@scottleibrand
May 02 2015 06:09 UTC
sorry, git pull which thing?
Ben West
@bewest
May 02 2015 06:09 UTC
openaps-old
Scott Leibrand
@scottleibrand
May 02 2015 06:09 UTC
pi@raspberrypi3 ~/old/openaps $ make
make: * No targets specified and no makefile found. Stop.
make what?
Ben West
@bewest
May 02 2015 06:10 UTC
great
no diff
perfect
Scott Leibrand
@scottleibrand
May 02 2015 06:10 UTC
not sure what that told you
Ben West
@bewest
May 02 2015 06:10 UTC
oh, sorry
run rm *.markdown
then make again
Scott Leibrand
@scottleibrand
May 02 2015 06:11 UTC
there are no makefiles
what good does make do?
Ben West
@bewest
May 02 2015 06:11 UTC
in openaps-old?
your testdata one
Scott Leibrand
@scottleibrand
May 02 2015 06:11 UTC
this isn't your openaps repo, it's just data and py files
Ben West
@bewest
May 02 2015 06:11 UTC
ok
I forgot to push
Scott Leibrand
@scottleibrand
May 02 2015 06:11 UTC
ah
pi@raspberrypi3 ~/old/openaps $ make
make: Nothing to be done for 'all'.
there we go
Ben West
@bewest
May 02 2015 06:13 UTC
ok, third time's charm
pull again
make clean all
git diff should have nothing to say
but there should be lots of output
Scott Leibrand
@scottleibrand
May 02 2015 06:13 UTC
there are diffs
Ben West
@bewest
May 02 2015 06:13 UTC
aha
Scott Leibrand
@scottleibrand
May 02 2015 06:13 UTC
want me to push them?
Ben West
@bewest
May 02 2015 06:13 UTC
we're not using the same software
and that is the problem
doh
so sorry
ok
in decoding-carelink:
git fetch origin
git checkout origin bewest/dev
Scott Leibrand
@scottleibrand
May 02 2015 06:15 UTC
I just did git pull again and got
decocare/history.py | 8 +++++++-
decocare/models/init.py | 2 ++
Ben West
@bewest
May 02 2015 06:15 UTC
oh
beatiful
great
Scott Leibrand
@scottleibrand
May 02 2015 06:16 UTC
trying iter_pump again
Ben West
@bewest
May 02 2015 06:16 UTC
run that make test agian
ok
or that
Scott Leibrand
@scottleibrand
May 02 2015 06:16 UTC
no errors
Ben West
@bewest
May 02 2015 06:16 UTC
:-D
Scott Leibrand
@scottleibrand
May 02 2015 06:16 UTC
and git diff is clean
Ben West
@bewest
May 02 2015 06:16 UTC
nice
that's cool
Scott Leibrand
@scottleibrand
May 02 2015 06:17 UTC
ok, so are we good, or anything else?
Ben West
@bewest
May 02 2015 06:17 UTC
nah, play around with it :-)
kick the tires
Scott Leibrand
@scottleibrand
May 02 2015 06:17 UTC
yaup, will do tomorrow
Ben West
@bewest
May 02 2015 06:17 UTC
it'll store everything you do, and the repots in that git repo
Scott Leibrand
@scottleibrand
May 02 2015 06:17 UTC
now it's bed time: going hiking 4k vertical feet tomorrow. :)
Ben West
@bewest
May 02 2015 06:17 UTC
the "tasks" and stuff
oh wow
the tasks and stuff are all pretty sure, similar to what was going on in latest thing
Scott Leibrand
@scottleibrand
May 02 2015 06:17 UTC
@danamlewis' idea: she's crazy. :)
Ben West
@bewest
May 02 2015 06:17 UTC
but hopefully little simpler
Scott Leibrand
@scottleibrand
May 02 2015 06:18 UTC
I've done this hike before: it's tough, but we'll be fine if we take our time. :)
Ben West
@bewest
May 02 2015 06:18 UTC
sounds like quite an adventure
Scott Leibrand
@scottleibrand
May 02 2015 06:18 UTC
yeah, it looks pretty clean and modular
excited to actually start doing some stuff with it. :)
Ben West
@bewest
May 02 2015 06:18 UTC
yeah, there is very strict separation so far
decocare is wholly responsible for getting data out
openaps is just way to structure how which data goes into which file
and then way to do it reliably
representing some "active RAM"
KISS hopefully
thanks for hacking, hope the hike and hike prep goes well!
Scott Leibrand
@scottleibrand
May 02 2015 06:20 UTC
:) thx
will let you know how it goes when I get back to hacking later this weekend
Ben West
@bewest
May 02 2015 06:21 UTC
are you using mongo/nightscout locally on he rpi?
I don't care about nightscout per se
but I'm going to come up with "process" and "webservice" endor types
so you could create a "NS webservice" vendor
with uses, like store/archive/fetch latest records
I prefer that to mongo direct
Scott Leibrand
@scottleibrand
May 02 2015 06:23 UTC
for now I'm trying to make it work offline with no DB or Internet at all
Ben West
@bewest
May 02 2015 06:23 UTC
right
Scott Leibrand
@scottleibrand
May 02 2015 06:23 UTC
but eventually I want to have it upload pump and cgm data
no need for a local DB I don't think
Ben West
@bewest
May 02 2015 06:23 UTC
I'm also looking at sharding the data locally for archive
Scott Leibrand
@scottleibrand
May 02 2015 06:23 UTC
just needs to talk to nightscout with REST
(or do mongoimport, but that's less clean)
Ben West
@bewest
May 02 2015 06:24 UTC
just not sure which will happen first
probably need lot more reports from thse vendors first
Scott Leibrand
@scottleibrand
May 02 2015 06:26 UTC
Yeah, I'll let you know if I see any gaps.
'Nite.
Ben West
@bewest
May 02 2015 06:26 UTC
'night
diabeticgonewild
@diabeticgonewild
May 02 2015 22:43 UTC
FML went to the gym and exercised. Now lymph nodes throb, joints hurt, and neuropathic pain (not diabetes-related)....I seriously don't think my (expensive ) intravenous immunoglobulin infusions are working...this sucks! But I am going to get a pi 2 off of Amazon cause micro center wants too much for the same startup kit
So yay starting chemo in 2.5 weeks
Ben West
@bewest
May 02 2015 22:49 UTC
wow, that's really rough thing to look forward to :$
diabeticgonewild
@diabeticgonewild
May 02 2015 22:51 UTC
LOL it's not puking chemo at least...so I don't have to find things to get sick on👿
Ben West
@bewest
May 02 2015 22:52 UTC
bewest@hither:~/Documents/foo$ scons
scons: Reading SConscript files ...
scons: done reading SConscript files.
scons: Building targets ...
openaps report invoke status.json
pump://json/read_settings/status.json
reporting status.json
openaps report invoke clock.json
pump://JSON/read_clock/clock.json
reporting clock.json
openaps report invoke glucose.txt
cgm://text/iter_glucose/glucose.txt
reporting glucose.txt
openaps report invoke glucose.json
cgm://JSON/iter_glucose/glucose.json
reporting glucose.json
openaps report invoke medtronic-cgm.json
pump://JSON/read_glucose_data/medtronic-cgm.json
reporting medtronic-cgm.json
openaps report invoke pump-history.json
pump://JSON/iter_pump/pump-history.json
No handlers could be found for logger "decocare.stick"
ERROR day is out of range for month (2015, 4, 31, 0, 0, 0) 0000   0x5e 0x0f                                  ^.
ERROR day is out of range for month (2015, 4, 31, 0, 0, 0) 0000   0x5e 0x0f                                  ^.
ERROR day is out of range for month 0000   0x5e 0x0f                                  ^.
ERROR day is out of range for month 0000   0x5e 0x0f                                  ^.
reporting pump-history.json
echo monitor -- monitor.openaps
monitor -- monitor.openaps
scons: done building targets.
bewest@hither:~/Documents/foo$
diabeticgonewild
@diabeticgonewild
May 02 2015 22:52 UTC
Like IDK I can't eat gyros anymore and Greek food is pretty unappetizing to me now....as I had it for my last meal before puking a lot
diabeticgonewild
@diabeticgonewild
May 02 2015 23:07 UTC
I shouldn't be complaining...it's Saturday night and we will find out who wins the boxing match tonight