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

10th
May 2016
David Young
@wgs4
May 10 2016 00:50
ok I’m getting pretty stuck here. Walkthrough Phase 2 says I should already have my pump and cgm as openaps devices…. but when i do openaps use -h I only have ns-status and tz as devices. I’ve gone over and over the walkthrough doc to this point, and I’m just not seeing the steps for setting up the pump and cgm. Nor have I been successful in getting the OpenAPS pill to appear in NS. I have configured NS profile with acurate pump settings. I have worked ahead to the end of Phase 1 visualization steps. I even tried running the ns-uploader-setup.sh script, dev version.
Dana Lewis
@danamlewis
May 10 2016 00:51
this might be a things out of order issue. visualization was originally way later, so some of what you need may be further through phase 2 than where you are
David Young
@wgs4
May 10 2016 00:52
aggg.. Its so hard to follow directions and skip stuff. :-)
Scott Leibrand
@scottleibrand
May 10 2016 00:52
at this point it's waiting on someone like you to fix things to be in order again
Dana Lewis
@danamlewis
May 10 2016 00:52
there may be an overview page that’s incorrectly referencing what’s in the section
but i’m pretty sure phase 2 is correctly ordered, because that was in decent shape before.
Scott Leibrand
@scottleibrand
May 10 2016 00:53
we had to move visualization up a lot earlier because it's really important to get that right so you can start open looping before you start closing the loop...
Dana Lewis
@danamlewis
May 10 2016 00:53
what I find helpful is to skim the outline of all the content, that gives some clues about where things are. (or asking is totally fine)
but you might be reading an overly ambitious high level “this is what’s in the seection” that’s not quite jiving with what’s in there after the last re-org
so definitely keep making notes, because this is what’s tripping up other people, too
David Young
@wgs4
May 10 2016 00:54
can you look at what I did wrong with my image link on the PR that dana approved?
Dana Lewis
@danamlewis
May 10 2016 00:54
let me take a look
David Young
@wgs4
May 10 2016 00:54
the image viewed fine every step.. I don’t want to keep making same mistake as I correct docs
Dana Lewis
@danamlewis
May 10 2016 00:54
images are a pain to embed, image was there when PR approved, so let me check formatting of the embed
no, it’s embedded correctly
Screen Shot 2016-05-09 at 5.55.18 PM.png
@wgs4 where does it look broken for you?
Dana Lewis
@danamlewis
May 10 2016 00:59
ah, gotcha.
not sure how to render images on ReadTheDocs, it should be the same syntax. but as a workaround, instead of an embed, we could just link to the image so it works in both places
your call on which path to take ;)
David Young
@wgs4
May 10 2016 01:00
is ReadTheDocs how most follow the instructions? I have found it the easiest way to follow along up until this point
Scott Leibrand
@scottleibrand
May 10 2016 01:01
the RTD page is pointing to https://openaps.readthedocs.io/docs/docs/Images/Pull%20Request.png - you might be able to change that to be an absolute instead of a relative image link
Dana Lewis
@danamlewis
May 10 2016 01:01
it’s a really nice formatted version, but it’s pulling from the Github content so that’s the source of truth for any of the doc displays
Ben West
@bewest
May 10 2016 01:01
should be relative, not absolute
Scott Leibrand
@scottleibrand
May 10 2016 01:02
@bewest how does RTD handle relative links to images?
will they host a copy of the images on RTD?
Ben West
@bewest
May 10 2016 01:02
it's all relative to top level docs directory
same way the links themselves work
there's tons of relative links between pages
Ben West
@bewest
May 10 2016 01:02
works same way
that link doesn't appear to be relative to top docs directory
David Young
@wgs4
May 10 2016 01:03
what did i do wrong?
Ben West
@bewest
May 10 2016 01:05
not sure, I would try eliminating spaces from the filename
the other images seem to be working fine
David Young
@wgs4
May 10 2016 01:06
<img src="/docs/docs/Images/Pull Request.png”/> was my embed code
can I test this using RTD and direct it towards my own Fork to eliminate Dana having to approve my trial and error to fix this?
Dana Lewis
@danamlewis
May 10 2016 01:07
Yup!
You can generate your own RTD of your fork version
Ben West
@bewest
May 10 2016 01:07
yes
David Young
@wgs4
May 10 2016 01:07
hehe
Ben West
@bewest
May 10 2016 01:07
sign into RTD and create your project, point it to your fork
Dana Lewis
@danamlewis
May 10 2016 01:08
(but this also isn’t a big deal, that’s what dev is for :))
Ben West
@bewest
May 10 2016 01:08
can also run it on your mac/windows machine
David Young
@wgs4
May 10 2016 01:09
I will need more trial and error space down the road. I try to solve problems on my own when possible.. asking is too easy and wears on people. ;-)
Toby Canning
@TC2013
May 10 2016 01:14
@wgs4 No, ask any question you like any time! Chances are I need the answer more than you :)
Dana Lewis
@danamlewis
May 10 2016 01:14
or 50 other people :)
David Young
@wgs4
May 10 2016 01:16
how to post an image in gitter ?
Dana Lewis
@danamlewis
May 10 2016 01:16
that’s a pain. screenshot to your computer, then drag ’n drop, or paste a link to an image sharing site (that’s what you have to do if you’re on your phone)
Toby Canning
@TC2013
May 10 2016 01:16
What OS do you use?
Dana Lewis
@danamlewis
May 10 2016 01:16
so dropbox or the like
David Young
@wgs4
May 10 2016 01:16
RTD.tiff
The plus button wont click on my RTD attempt
stonewall after stonewall. lol
Toby Canning
@TC2013
May 10 2016 01:17
If you are using Windows, you can hit your print screen button, open Paint and paste it, select the portion to post, copy it, and paste it here using ctrl-v
Dana Lewis
@danamlewis
May 10 2016 01:17
builds character ;)
Scott Leibrand
@scottleibrand
May 10 2016 01:24
Or take a photo of your screen with the phone, then embed that in a word doc, and attach it to an email...
Or embed it in a spreadsheet and print it out for your endo... :-p
Dana Lewis
@danamlewis
May 10 2016 01:24
what are you going to suggest, faxing next? ;p
David Young
@wgs4
May 10 2016 01:36
LOL
JaysonEwer
@JaysonEwer
May 10 2016 03:19
@scottleibrand you forgot the Polaroid. Polaroid>picture with phone>embed in .xls>pdf>print on dot matrix for endo> then fax to Dana to do the post processing. 😜
A picture of the Polaroid that is.
Scott Leibrand
@scottleibrand
May 10 2016 03:25
:+1:
David Young
@wgs4
May 10 2016 04:10
ok, @danamlewis , please approve that PR i just made for the PR page. I’ve got the image situation fixed… I knew better than to have a space in a filename. also changed the way I did my link to be like ![Name](../Image/) which i think is better...
Dana Lewis
@danamlewis
May 10 2016 04:16
:+1:
merged. and then did an edit of removing spaces so it didn’t break the numbering (6 became 1, now should be fixed) https://github.com/openaps/docs/blob/dev/docs/docs/Resources/my-first-pr.md
fixed itself in RTD, too
Hristina Dimova
@xpucuto
May 10 2016 09:13
@mddub i have sucessfuly setup pancreabble from your instrcutions :) Thank you :) But how do i configure the pi to send notifications to the pebble ? can you give me an example? :blush:
Ben West
@bewest
May 10 2016 10:39
curses, I had coffee in afternoon
was against my rules, but I did it...
now... I'm awak
my pebbles are dead :-(
all my pebbled died
Hristina Dimova
@xpucuto
May 10 2016 10:41
Get a cup of hot milk with cocoa ;)
I rarely make that mistake but when I do I know you guys will be awake here ;)
JaysonEwer
@JaysonEwer
May 10 2016 16:33
@mm22dl I"m sorry for the delay in getting back to you regarding the kernel modules. You can use the following command to see what modules are loaded.
edison@Ewer-Edison-OpenAPS-31:~$ lsmod
Module                  Size  Used by
ftdi_sio               40121  0 
uvcvideo               71516  0 
videobuf2_vmalloc      13003  1 uvcvideo
videobuf2_memops       13001  1 videobuf2_vmalloc
videobuf2_core         37707  1 uvcvideo
usb_f_acm              14335  1 
u_serial               18582  6 usb_f_acm
g_multi                70813  0 
libcomposite           39245  2 usb_f_acm,g_multi
bcm_bt_lpm             13676  0 
bcm4334x              578947  0 
edison@Ewer-Edison-OpenAPS-31:~$
And here is an example from my pi:
pi@Ewer-Open-APS:~ $ lsmod
Module                  Size  Used by
cfg80211              407580  0 
rfkill                 16036  2 cfg80211
8192cu                519604  0 
cdc_acm                16033  0 
usbserial              21779  0 
snd_bcm2835            19802  0 
bcm2835_rng             1763  0 
bcm2835_gpiomem         2860  0 
snd_pcm                73474  1 snd_bcm2835
snd_timer              18848  1 snd_pcm
snd                    50779  3 snd_bcm2835,snd_timer,snd_pcm
uio_pdrv_genirq         2944  0 
uio                     7753  1 uio_pdrv_genirq
bcm2708_wdog            2838  0 
i2c_dev                 5671  0 
fuse                   80694  1 
ipv6                  338660  32 
pi@Ewer-Open-APS:~ $
mm22dl
@mm22dl
May 10 2016 16:53
@JaysonEwer so here is the output produced by my laptop, from which I am trying to read from the carelink stick
Module Size Used by
hid_generic 16384 0
usbhid 49152 0
rfcomm 69632 2
xt_CHECKSUM 16384 1
iptable_mangle 16384 1
ipt_MASQUERADE 16384 3
nf_nat_masquerade_ipv4 16384 1 ipt_MASQUERADE
iptable_nat 16384 1
nf_nat_ipv4 16384 1 iptable_nat
nf_nat 24576 2 nf_nat_ipv4,nf_nat_masquerade_ipv4
nf_conntrack_ipv4 16384 2
nf_defrag_ipv4 16384 1 nf_conntrack_ipv4
xt_conntrack 16384 1
nf_conntrack 106496 5 nf_nat,nf_nat_ipv4,xt_conntrack,nf_nat_masquerade_ipv4,nf_conntrack_ipv4
ipt_REJECT 16384 2
nf_reject_ipv4 16384 1 ipt_REJECT
xt_tcpudp 16384 6
bridge 126976 0
stp 16384 1 bridge
llc 16384 2 stp,bridge
ebtable_filter 16384 0
ebtables 36864 1 ebtable_filter
ip6table_filter 16384 0
ip6_tables 28672 1 ip6table_filter
iptable_filter 16384 1
ip_tables 28672 3 iptable_filter,iptable_mangle,iptable_nat
x_tables 36864 11 ip6table_filter,xt_CHECKSUM,ip_tables,xt_tcpudp,ipt_MASQUERADE,xt_conntrack,iptable_filter,ebtables,ipt_REJECT,iptable_mangle,ip6_tables
drbg 32768 1
ansi_cprng 16384 0
ctr 16384 2
ccm 20480 2
bnep 20480 2
arc4 16384 2
ath9k 143360 0
ath9k_common 36864 1 ath9k
snd_soc_rt5640 114688 0
ath9k_hw 466944 2 ath9k_common,ath9k
snd_hda_codec_realtek 81920 1
snd_hda_codec_hdmi 53248 1
snd_hda_codec_generic 77824 1 snd_hda_codec_realtek
uvcvideo 90112 0
intel_rapl 20480 0
snd_soc_rl6231 16384 1 snd_soc_rt5640
videobuf2_vmalloc 16384 1 uvcvideo
videobuf2_memops 16384 1 videobuf2_vmalloc
videobuf2_v4l2 28672 1 uvcvideo
videobuf2_core 36864 2 uvcvideo,videobuf2_v4l2
x86_pkg_temp_thermal 16384 0
ath 32768 3 ath9k_common,ath9k,ath9k_hw
v4l2_common 16384 1 videobuf2_v4l2
intel_powerclamp 16384 0
snd_soc_core 212992 1 snd_soc_rt5640
snd_hda_intel 36864 5
joydev 20480 0
snd_hda_codec 135168 4 snd_hda_codec_realtek,snd_hda_codec_hdmi,snd_hda_codec_generic,snd_hda_intel
videodev 176128 4 uvcvideo,v4l2_common,videobuf2_core,videobuf2_v4l2
mac80211 737280 1 ath9k
snd_hda_core 73728 5 snd_hda_codec_realtek,snd_hda_codec_hdmi,snd_hda_codec_generic,snd_hda_codec,snd_hda_intel
snd_compress 20480 1 snd_soc_core
coretemp 16384 0
ac97_bus 16384 1 snd_soc_core
acer_wmi 20480 0
snd_pcm_dmaengine 16384 1 snd_soc_core
sparse_keymap 16384 1 acer_wmi
hid_multitouch 20480 0
ath3k 20480 0
media 24576 2 uvcvideo,videodev
snd_hwdep 16384 1 snd_hda_codec
crct10dif_pclmul 16384 0
crc32_pclmul 16384 0
aesni_intel 167936 4
snd_pcm 106496 7 snd_soc_rt5640,snd_soc_core,snd_hda_codec_hdmi,snd_hda_codec,snd_hda_intel,snd_pcm_dmaengine,snd_hda_core
btusb 45056 0
cfg80211 565248 4 ath,ath9k_common,ath9k,mac80211
snd_seq_midi 16384 0
aes_x86_64 20480 1 aesni_intel
lrw 16384 1 aesni_intel
gf128mul 16384 1 lrw
btrtl 16384 1 btusb
btbcm 16384 1 btusb
snd_seq_midi_event 16384 1 snd_seq_midi
btintel 16384 1 btusb
glue_helper 16384 1 aesni_intel
snd_rawmidi 32768 1 snd_seq_midi
bluetooth 520192 30 bnep,ath3k,btbcm,btrtl,btusb,rfcomm,btintel
snd_seq 69632 2 snd_seq_midi_event,snd_seq_midi
ablk_helper 16384 1 aesni_intel
rtsx_pci_ms 20480 0
memstick 20480 1 rtsx_pci_ms
cryptd 20480 2 aesni_intel,ablk_helper
input_leds 16384 0
snd_seq_device 16384 3 snd_seq,snd_rawmidi,snd_seq_midi
kvm_intel 172032 0
serio_raw 16384 0
kvm 536576 1 kvm_intel
snd_timer 32768 2 snd_pcm,snd_seq
mei_me 36864 0
mei 98304 1 mei_me
snd 81920 23 snd_hda_codec_realtek,snd_soc_core,snd_hwdep,snd_timer,snd_hda_codec_hdmi,snd_pcm,snd_seq,snd_rawmidi,snd_hda_codec_generic,snd_hda_codec,snd_hda_intel,snd_seq_device,snd_compress
lpc_ich 24576 0
dw_dmac 16384 0
dw_dmac_core 24576 1 dw_dmac
shpchp 36864 0
snd_soc_sst_acpi 16384 0
spi_pxa2xx_platform 24576 0
8250_dw 16384 0
soundcore 16384 1 snd
i2c_designware_platform 16384 0
soc_button_array 16384 0
irqbypass 16384 1 kvm
mac_hid 16384 0
i2c_designware_core 20480 1 i2c_designware_platform
parport_pc 32768 0
ppdev 20480 0
lp 20480 0
parport 49152 3 lp,ppdev,parport_pc
autofs4 40960 2
rtsx_pci_sdmmc 24576 0
i915 1208320 4
i2c_algo_bit 16384 1 i915
drm_kms_helper 139264 1 i915
ahci 36864 2
syscopyarea 16384 1 drm_kms_helper
sysfillrect 16384 1 drm_kms_helper
libahci 32768 1 ahci
sysimgblt 16384 1 drm_kms_helper
fb_sys_fops 16384 1 drm_kms_helper
r8169 81920 0
drm 360448 5 i915,drm_kms_helper
mii 16384 1 r8169
rtsx_pci 53248 2 rtsx_pci_ms,rtsx_pci_sdmmc
sdhci_acpi 16384 0
wmi 20480 1 acer_wmi
video 40960 2 i915,acer_wmi
sdhci 45056 1 sdhci_acpi
i2c_hid 20480 0
hid 118784 4 i2c_hid,hid_multitouch,hid_generic,usbhid
fjes 28672 0
could it be that because I am using ubuntu and not raspbian, this problem happens?
Hristina Dimova
@xpucuto
May 10 2016 17:12
@danamlewis i am so unpatient for the meeting so wanted to ask you what gadjets will you bring ? i read about for most of the things but its hard to picture them . :cat:
ddaniels1
@ddaniels1
May 10 2016 17:39
Question: I have an open loop running now and notice that the it takes an extra cycle to change from looping to enacted, even when I run the complete loop alias including the upload command at the end. IS this normal? Here is my loop alias: 'loop ! bash -c "2>/dev/null && ( openaps preflight && openaps gather && openaps suggest && openaps status-upload && openaps upload-recent-treatments) || echo No CGM data."'.
Rob Kresha
@rkresha
May 10 2016 17:41
I'm going through and adding the devices to my Rpi and when I plug the dexcom receiver into the usb port, it causes wifi to drop and hangs up the Rpi. In the documentation it says to make sure the dexcom is fully charged, which it is, but still seeing the same issue. Any ideas?
JaysonEwer
@JaysonEwer
May 10 2016 18:04
@mm22dl I wouldn't think using ubuntu vs raspbian would cause any problem. But you have quite a few kernel modules being used. I would work through them to see if they are actually needed and rmmod the ones you don't need.
@rkresha Sounds like a power draw issue. What kind of power supply are you using?
Rob Kresha
@rkresha
May 10 2016 18:06
Canakit provided one
2.5a I believe
I'm going to try the max_usb_current=1 in the /boot/config.txt config file and give it a shot, apparently that doubles the mA output
mm22dl
@mm22dl
May 10 2016 18:16
@JaysonEwer what does "actually needed" mean? why would it be a problem that many modules be loaded? could this cause power issues? sorry for the newbie questions...
JaysonEwer
@JaysonEwer
May 10 2016 18:21
:-) I'm not an expert either, but yeah, kernel modules that are loaded into the kernel can use memory and can cause conflicts to occur and if you aren't using the kernel module, its a good idea to remove it. the one that jumped out at me was the "parport 49152 3 lp,ppdev,parport_pc" Are you using a parellel port printer? https://www.kernel.org/doc/Documentation/parport.txt
Dana Lewis
@danamlewis
May 10 2016 18:22
@xpucuto here are a couple of pics from Twitter. 1) here is a pic showing two pis (black box on right; and then the setup on top of the bag complete with battery) and the Edison rig is in the small case on bottom left: https://twitter.com/danamlewis/status/718960234836852736
and here is my Edison rig in that small case clipped to my pocket: https://twitter.com/danamlewis/status/718959056505806848
when I travel, I ususally have both the Edison rig for day use and also at least one Pi/TI rig to cover the hotel room.
JaysonEwer
@JaysonEwer
May 10 2016 18:24
@rkresha If you're using the Canakit provided one, that is a good Power supply.
Hristina Dimova
@xpucuto
May 10 2016 18:27
@danamlewis thank you . I suppose I will make a twitter account soon so I can follow :)
both look fabulous
Dana Lewis
@danamlewis
May 10 2016 18:27
and post! there are some great looper pics/stories being shared on the #OpenAPS hashtag
Hristina Dimova
@xpucuto
May 10 2016 18:37
I promise ;)
Dana Lewis
@danamlewis
May 10 2016 18:37
:)
mm22dl
@mm22dl
May 10 2016 18:43
@JaysonEwer not that i know of... I just freshly reinstalled ubuntu on my computer a couple weeks ago... i'll have a look at what kernel modules i'm using and which ones i'm not
Hristina Dimova
@xpucuto
May 10 2016 19:44
Just few questions from me again . How do you deal with ??? On the receiver ? Is it possible openaps to take CGM data from ns site and if there are no data to look for wired receiver or the opposite?
Has anyone setup Rpi as hotspot ?Sometimes when I am not at home I can't command the pi even thou we are on the same shared network
Dana Lewis
@danamlewis
May 10 2016 19:48
drink lots of water if you get ???. and yes, it’s possible to pull data from different places.
Paul Martin
@Spazholio
May 10 2016 20:28
So, the VA made a mistake and sent me a G4 rather than a G5. While I was on the phone attempting to rectify it, I started to think - do I really care all that much? Can anyone outline the pros/cons of the G4 vs the G5 with respect to looping?
jaylagorio
@jaylagorio
May 10 2016 20:33
@danamlewis Interesting on the ??? -> drink water thing - how does that work? I've only had the G4 for a month or so and I'm still figuring out the nuances.
The videos make a point of saying that Tylenol is my new mortal enemy but they don't mention anything about that
Scott Leibrand
@scottleibrand
May 10 2016 20:37
for whatever reason being dehydrated messes with the under-skin reaction, so getting properly hydrated often helps
@Spazholio I would stick with the G4 for now. we don't (yet) have as many options for getting BG data off the G5.
Dana Lewis
@danamlewis
May 10 2016 20:38
@spazholio I prefer G4. G5 means you can go without a receiver, but data to the phone is iffy, and I’ve heard lots of people having problems. up to you, but that’s the big difference is comfort level if you lose data. (it makes me cranky)
@jaylagorio yea, stay away from tylenol. i’m not sure why it’s not (as) common knowledge about dehydration being the biggest cause of ???
it’s not always the case, but 9/10 of my ??? are fixed by drinking, and the other times you can just see the sensor dying and know that’s what’s up and you should change it
Scott Leibrand
@scottleibrand
May 10 2016 20:40
@Spazholio G5 requires custom build of dexcom_reader to read BGs from plugged-in receiver, and openxshareble doesn't support getting the data directly off the G5 transmitter with Bluetooth (yet).
Dana Lewis
@danamlewis
May 10 2016 20:40
(usually well beyond a week or week and a half of using a sensor for me)
jaylagorio
@jaylagorio
May 10 2016 20:44
@scottleibrand @danamlewis That's a great tip, thank you! I'm sure this is something I would know if I knew another diabetic IRL.
Ben West
@bewest
May 10 2016 20:44
openxshareble probably will never support g5, FWIW
for g5, it'll be a new vendor, it's a completely different set of use cases, looks like
g5 actually is additional pressure on the daemon tools ;-)
Dana Lewis
@danamlewis
May 10 2016 20:45
@jaylagorio we’re real, I promise! even if I’m part robot, I’m real.
jaylagorio
@jaylagorio
May 10 2016 20:46
Touché, fair enough.
Dana Lewis
@danamlewis
May 10 2016 20:46
;) it is awesome to meet others physically in person, though
jaylagorio
@jaylagorio
May 10 2016 20:47
I only met @CrushingT1D in person while travelling for work, he was the one who gave me the CareLink USB stick to try and get it working with Windows
Stroke of luck that he was looking to give one away and I was going to the same area to be honest
ddaniels1
@ddaniels1
May 10 2016 20:47
Here is a screenshot of my loop. Wondering why it says "no temp" when there is a prior temp?
Paul Martin
@Spazholio
May 10 2016 20:49
Holy cow. The VA (inadvertently) got something right. Well ok, sounds like I'll be sticking with the G4. Thanks!
Dana Lewis
@danamlewis
May 10 2016 20:49
hehe. welcome.
ddaniels1
@ddaniels1
May 10 2016 20:50
ok, having trouble pasting a jpeg, how do I do that?
Dana Lewis
@danamlewis
May 10 2016 20:50
screenshot and drag and drop into gitter
or use dropbox or the like for a link
ddaniels1
@ddaniels1
May 10 2016 20:50
Screen shot 2016-05-10 at 11.56.11 AM.png
:)
Dana Lewis
@danamlewis
May 10 2016 20:51
there you go
what was the temp before, another zero?
it might have had a gap between the zeros if it was cancelling it
ddaniels1
@ddaniels1
May 10 2016 20:52
yes temp before was another zero for a good while, you can see from the basal graph
Dana Lewis
@danamlewis
May 10 2016 20:52
yea, looks long enough that it either expired and this new one was set a minute after that, or it cancelled and then was resetting
ddaniels1
@ddaniels1
May 10 2016 20:53
Also is it normal to have two entries on the bubble?
I am wondering if it matters the order I run the report aliases in the loop alias
?
Dana Lewis
@danamlewis
May 10 2016 20:53
do you have multiple rigs or is this single rig?
ddaniels1
@ddaniels1
May 10 2016 20:53
single rig
here is my loop alias
loop ! bash -c "2>/dev/null && ( openaps preflight && openaps gather && openaps suggest && openaps enact && openaps status-upload && openaps upload-recent-treatments) || echo No CGM data.".
look ok?
Dana Lewis
@danamlewis
May 10 2016 20:54
oh, yea that’s your status upload (enacted) and recent treatments
so just displaying both because you ran them both
ddaniels1
@ddaniels1
May 10 2016 20:55
so "enacted" is the recent treatment?
seems that is part of the status of the loop?
Dana Lewis
@danamlewis
May 10 2016 20:55
that looks like your status-upload
I’m just saying in the screenshot, it was showing the status as enacted
the first line in the bubble display
recent treatments is the second line with all the good info
Ben West
@bewest
May 10 2016 20:57
the set_temp_basal use does not generate historical treatment records
that's often a gotcha
that use in particular is a kind of pivot point, where we go from collecting and analyzing a lot of information, which culminates in a dosing change/suggestion, this is the pivot point that turns that suggestion into an actual change
ddaniels1
@ddaniels1
May 10 2016 20:59
when I go into my monitor directory and inspect the the temp basal.json it has the proper info, is it getting pulled from somewhere else?
Ben West
@bewest
May 10 2016 21:00
so the output is useful for status as "what was enacted/attempted?", which is very different from "what is contained in the pump's historical audit log"
it may be a difference that as a community we don't particularly care about, but I'm entering this design space carefully
been considering a tweak to set_temp_basal so that it will query the pump history and confirm the presence of the temp rate in the history, and outputting that instead of just the response from the command, or perhaps appending the history to an existing history file
ddaniels1
@ddaniels1
May 10 2016 21:06
I have been looking for the docs related to testing different scenarios and outputs from determine-basal. Found it once but now can't find it. Can you point me?
also what's the purpose of the initial part of this alias?
loop ! bash -c "2>/dev/null && ( openaps preflight && openaps gather && openaps suggest && openaps enact && openaps status-upload && openaps upload-recent-treatments) || echo No CGM data.".
Dana Lewis
@danamlewis
May 10 2016 21:07
this is good to bookmark so you can skim the outline of what’s in docs: https://openaps.readthedocs.io/en/latest/index.html
so under phase 3 you’ll find re: understanding determine-basal output: https://openaps.readthedocs.io/en/latest/docs/walkthrough/phase-3/Understand-determine-basal.html
ddaniels1
@ddaniels1
May 10 2016 21:07
I understand 2/dev/null&& dumps the output, but why is that needed here?
Thought that was suggested as part of the cron?
Also I took out the up front "monitor-cgm" because that is part of my gather alias. IS that ok or is there a reason you suggested it up front in the loop alias?
thanks @dana
I have found that page, but there was something else about contructing test files somewhere in order to do some bench testing of the algorithm and then being able to inspect the suggest.json output
Dana you mentioned it before
Ben West
@bewest
May 10 2016 21:51
I have some new-style recommendations that haven't quite made it into the docs yet
my examples are here: openaps/docs#132
in particular, the snippet 2> /dev/null throws away error messages, I don't necessarily recommend that, in general I prefer using syslog: my crontab looks like this:
*/5 * * * * (cd /home/bewest/Documents/foobar && time openaps do-everything) 2>&1 | logger -t openaps-loop
and I don't use any redirection of output in any aliases, what this does is "tie" the stderr stream to stdout, and send both combined to the system logger
all the output including errors (which are often useful and needed to solve a problem) shows up in /var/log/syslog although configuration changes to some files can make it appear elsewhere, including papertrail.com
Scott Leibrand
@scottleibrand
May 10 2016 21:57
the ones that throw away errors are generally designed to allow you to output the logs to a more "readable" log file. syslog is great for later debugging, but seeing full tracebacks for situations like "pump is out of range" makes it really hard to read the logs on an iphone screen, for example.
Ben West
@bewest
May 10 2016 21:57
a lot of these kinds of decisions come down to personal style/fit
Scott Leibrand
@scottleibrand
May 10 2016 21:58
yep. lots of different people do that part differently.
Ben West
@bewest
May 10 2016 21:58
for someone who needs to ask on gitter for some help, it will typically be important to include error messages
doesn't matter too much, it's a personal style thing
did you get to play with the yargs + determine-basal?
I've been using it all weekend, seems to perform same
that's last blocker for release, right?
Scott Leibrand
@scottleibrand
May 10 2016 22:08
I haven't had time to play with it yet. did two days of mother's day work parties this weekend. :)
Mark Wilson
@mddub
May 10 2016 22:12
@scottleibrand is oref0-dev meant to always have set-current-profile-basal-as-temp behavior?
Dana Lewis
@danamlewis
May 10 2016 22:26
@mddub yes
that’s a trick so you know if your loop has been reading running if you’re out of signal and
vs wondering if it thinks you don’t need an adjusted temp
Mark Wilson
@mddub
May 10 2016 22:27
hmm, seems i was on an older version of dev
i understand the rationale, but it seems like personal preference
i prefer not to have the beeping/buzzing unless absolutely necessary, and rely on other monitoring
Dana Lewis
@danamlewis
May 10 2016 22:28
a large number of persons are very unhappy and vocal when they go offline and have a hard time seeing if it’s running. but definitely can adjust if you want.
Mark Wilson
@mddub
May 10 2016 22:29
awesome, how do i turn it off?
Scott Leibrand
@scottleibrand
May 10 2016 22:46
it's not so simple, unfortunately
in dev and AMA, with autosens we often adjust the regular basal, so it sets the adjusted basal as temp whenever it doesn't need something else
there are actually quite a few places in the code that would need to be changed if you wanted to put in some logic for letting the current basal run instead
i see it repeated a lot
Scott Leibrand
@scottleibrand
May 10 2016 22:49
yep
would need to add some logic to check whether the basal variable matches profile.current_basal (or is "close enough")
which probably means pulling those 6 instances out into a new function
Mark Wilson
@mddub
May 10 2016 22:50
i think they should have already been their own function :P
Scott Leibrand
@scottleibrand
May 10 2016 22:51
unfortunately I didn't see a good way to actually reduce the repetition with a function
but I'm not exactly the best at refactoring code like that
and if we do go down this route, we need to figure out how to make it configurable, and what the different configurable behaviors should be
I like the simplicity of always setting a temp. that also means you can see from the color of the blue on the NS graph when the loop was running and when it wasn't.
Mark Wilson
@mddub
May 10 2016 22:54
yep, we all have our own monitoring preferences
for now i think i'll just do it at the level of basal-set-temp, something like this https://gist.github.com/mddub/4f4a8b135f8d734b1691ca2dcef659a7
Scott Leibrand
@scottleibrand
May 10 2016 22:55
yeah, that is fairly clean
if anyone else wants that, we can figure out how to make it configurable and merge it into dev
Mark Wilson
@mddub
May 10 2016 22:56
i could imagine the preference being a property on profile which could be pulled in by get-profile
Scott Leibrand
@scottleibrand
May 10 2016 22:56
there currently is no oref0-config.json, but we've talked about it a few times.
everything except max_iob.json comes from the pump
Mark Wilson
@mddub
May 10 2016 22:58
the max_iob.json argument could be a generic preferences file which includes max_iob
Scott Leibrand
@scottleibrand
May 10 2016 22:58
yep, that's the idea
Mark Wilson
@mddub
May 10 2016 22:58
keep it backwards-compatible
Scott Leibrand
@scottleibrand
May 10 2016 22:58
we've talked about using that for enabling several different optional features
most of those are enabled or disabled right now by the presence or absence of their input files
Mark Wilson
@mddub
May 10 2016 23:02
yeah, they should really be composed
get-profile should take the minimal pump data and make a profile json object
configure-meal-assist should take whatever args it needs and a profile json object, and extend it with extra properties
etc.
Scott Leibrand
@scottleibrand
May 10 2016 23:19
yeah, that sounds like a good approach. tagging @jasoncalabrese and @bewest as well.
Jason Calabrese
@jasoncalabrese
May 10 2016 23:23
the downside of that is each report creates a new file, so anywhere profile.json is used you'd need to change it to someting like ama-profile.json
I guess you could reverse it so you have profile-pump.json and profile.json that was processed for ama
Mark Wilson
@mddub
May 10 2016 23:27
ah, yeah, start with the optional stuff then combine at the end
what if profile.json were a report which just combines whatever input json files you give it
so you could give it only profile-pump.json and max_iob.json if you don't use ama
it's sort of like having an intermediate report for glucose from either ns or dex glucose, then a final glucose.json report
Jason Calabrese
@jasoncalabrese
May 10 2016 23:30
I like the idea of composing, might need to be supported in a general way, so you could pipe reports to the next
Mark Wilson
@mddub
May 10 2016 23:30
there's a pattern here
Jason Calabrese
@jasoncalabrese
May 10 2016 23:31
yeah, thats the other example I was thinking of
I could see if for zoning and converting from mmol, but having the intermediate reports does help with debugging
Mark Wilson
@mddub
May 10 2016 23:36
that feels like a slightly different problem, raw vs. processed
but they are similar in that you want to ensure something has been properly cleaned/converted/merged
for combining profile data from different steps i was thinking you'd want an merge-json use, native to openaps, like shell