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

11th
Oct 2017
Steve Bernacki
@sbernacki
Oct 11 2017 02:25
Question regarding reading the logs.. I'm watching my wife's rig remotely (she's in Germany for business, I'm at home in New Hampshire USA) and was watching her BG drop. I noticed that pump-loop was reporting "no temp required", which I thought was strange since I thought it would set temp basal to 0, but NS reported no temp enacted and the pump was delivering her normal basal.
Oct 10 22:08:57 edison1 pump-loop.log: enact/smb-suggested.json: {"insulinReq":0,"bg":39,"reservoir":"32.7","temp":"absolute","snoozeBG":77,"predBGs":{"IOB":[39,39,39,39,39,40,42,44,46,49,52,55,59,62,66,69,72,74,77,79,81,82,84,85,86,87,88,88,89,89,90]},"minPredBG":999,"IOB":-1.242,"reason":"COB: 0, Dev: -36, BGI: 4.68, ISF: 64, Target: 111, minPredBG 77, IOBpredBG 90; BG 39<63, minDelta -1.33 < expectedDelta 7.2; 21m left and 0 ~ req 0U/hr: no temp required","COB":0,"eventualBG":65,"tick":"+0","deliverAt":"2017-10-11T02:08:56.116Z"}
Oct 10 22:09:01 edison1 pump-loop.log: No smb_enact needed. Temp refreshed: monitor/temp_basal.json: {"duration":21,"rate":0,"temp":"absolute"}
Steve Bernacki
@sbernacki
Oct 11 2017 02:33
(And also, NS is reporting a negative IOB of 0.90U, which seems interesting)
Steve Bernacki
@sbernacki
Oct 11 2017 03:21
And final question of the night: I was searching around to build a second rig for my wife, but it seems that Edisons are sold out everywhere at this point (Amazon, Adafruit, Hamshield etc). Now that they are discontinued, which direction is the community going in?
Dana Lewis
@danamlewis
Oct 11 2017 03:28
@sbernacki it looks like it did set a 0 temp there
Also, I heard adafruit has Edisons as part of a pack with the intel mini breakout board? Check and see if they are posted that way?
If not, hamshield probably still has a few preflashed left
Working on some Pihats for next gen hardware, but need to do some work to speed up loops for pi0 for the tiniest possible on the go rigs. Pi3 is fine out of the gate, but that's probably going t be more for stationary rigs
Steve Bernacki
@sbernacki
Oct 11 2017 03:30
@danamlewis Usually that's indicated by T0.000 in the basal pill though? And thanks, I'll check Adafruit again; I put my email in for stock updates on preflashed boards.
(Oh, score, you're right! https://www.adafruit.com/product/2111 ) Feel like I should buy a few and stuff them under our mattress :) )
garykidd
@garykidd
Oct 11 2017 03:39
@sbernacki that'll work with alittle more effort though'
Dana Lewis
@danamlewis
Oct 11 2017 04:25
@sbernacki OpenAPS pill and logs are source of truth. Some of the other pills can be delayed, wonky, or show things based on misentered stuff in NS profile, etc.
Scott Leibrand
@scottleibrand
Oct 11 2017 04:37
That "no temp required" is just saying "you already have a zero temp running with plenty of time left, no sense setting another one just yet".
LilDucky
@LilDucky
Oct 11 2017 04:43
@danamlewis I now have a Pi Zero and a Pi B test rigs with CC1110 radios up and going so let me know if you need any testing done. I will be adding RFM69 radios to them shortly. In the final stages of a compact rig case that will hold the Pi Zero, PHat and is powered by a replaceable 18650 LiPO cell. This design will be made freely available to 3D print.
Dana Lewis
@danamlewis
Oct 11 2017 04:45
Awesome. How long does it take your pi0 to run the full loop of everything? You may have some ideas to get a head start on speeding things up
LilDucky
@LilDucky
Oct 11 2017 05:04
will try and get some clean logs for you. I am still having issues on the zero with getting /dev/ttyAMA0 de-conflicted. going to try a totally clean build on the "B" rig then transfer it across. I will base the install on the current edison scripts and document any changes
am planning to use 0.6.0
Dana Lewis
@danamlewis
Oct 11 2017 05:13
:+1:
Scott Leibrand
@scottleibrand
Oct 11 2017 05:34
@LilDucky to start with we could use some thorough testing of oref0-setup and fixing any incompatibilities that have accumulated since we stopped using Pi's much.
It might also be worth getting openaps-install.sh working on Raspbian.
And then we'll need to start profiling oref0-pump-loop on the Pi0W and working on reducing the time spent launching Python and Node interpreters. @mgranberry had some ideas for symlinking to reduce some of that, but wasn't sure why it worked for him.
Tore Bjørndalen
@ToreBj
Oct 11 2017 05:56
@LilDucky is there any SW supporting the use of RFM69 to communicate with the pump? It certainly is a cheap option.
tepidjuice
@tepidjuice
Oct 11 2017 06:23
if someone could translate everything to python that'd cut loading node :P
LilDucky
@LilDucky
Oct 11 2017 06:44
@ToreBj not as yet but the new PHats will be using it. @ecc1 has written the code for the card in Go but it needs to be ported across, and this may be in progress.
@scottleibrand that's cool. I will focus on the install scripts for now. Which raspi a version are you aiming for? Jessie or stretch?
LilDucky
@LilDucky
Oct 11 2017 10:11
@scottleibrand I have now got on top of most of my serial port woes and will get started on the build tests tomorrow.
Mike
@libxmike
Oct 11 2017 10:12
@LilDucky the RFM69HCW has a good radio range, I did test with RFM69HCW and Pi last year with ecc1 go code, it worked but at the time not for production, did not touch it since then. I have maybe 5-6 RFM69HCW around, will make perfect rig with Pi Zero W, I tested last yes just with old Pi B (I think)
LilDucky
@LilDucky
Oct 11 2017 10:13
@libxmike great to hear.
Deborah Schmidt
@frauzufall
Oct 11 2017 10:41
hey all, I am happy to join your group! Does anyone know where I can order a fitting lipo for the rig from Germany? I found some, but there are no details about a protection circuit or the type of connector to be found.
sysyc
@sysyc
Oct 11 2017 14:18
blob
what's wrong above when install oref0?
@scottleibrand
jaylagorio
@jaylagorio
Oct 11 2017 14:25
Looks like your network connection timed out. Try installing again?
Reason I say that: ReadTimeoutError: HTTPSConnectionPool (host=pypi.python.org, port=443) Read timed out
sysyc
@sysyc
Oct 11 2017 14:33
@jaylagorio OK,I try again,thx
Scott Leibrand
@scottleibrand
Oct 11 2017 16:09
@sysyc yes, that's a problem with your Internet connection.
sysyc
@sysyc
Oct 11 2017 16:31
yes,my Internet connection is wrong today @scottleibrand
Brian Rabinovitz
@bdr1177
Oct 11 2017 16:35
We just had what looks like a problem with the insulin from a bolus not showing up at all. At 11:48 am, pump pill showed reservoir at 43.7U. At 11:53 am bolus of 0.6U was delivered via the bolus wizard for 26 carbs. Pump status screen shows, this bolus. However, the 26 carbs shows up in NS as having been entered through the pump/openaps as a carb correction with no insulin. Also at 11:53 am, pump pill showed 43.3U. At 11:59 am, pump pill continued to show 43.3U. By 12:00 pm, pump pill shows 43.1U, which seems to correspond to the 0.6U bolus, but openaps is still not acknowledging the additional bolus IOB.
Capture.JPG
3467056508551062197-account_id=4.jpg
Capture3.JPG
Capture4.JPG
Capture2.JPG
Dana Lewis
@danamlewis
Oct 11 2017 16:38
Let's set aside the pump pill. What did the OpenAPS pill and the OpenAPS logs say about iob?
Pump pill, cob pill etc can often be laggy or buggy, so checking OpenAPS pill and logs is important.
@bdr1177 ^
I'd like to separate what is a "Nightscout display on the graph or in separate pills is slow/buggy" from any problems with OpenAPS itself. Sometimes those get conflated.
Brian Rabinovitz
@bdr1177
Oct 11 2017 16:41
@danamlewis Bolus was at 11:53 am. At 11:59 am, OpenAPS pill was still showing IOB: -0.10, Basal IOB -0.10U
Capture9.JPG
It looks like OpenAPS did eventually recognize the the extra IOB. It categorized it as basal, but that is presumably because I'm not on the basal-vs-bolus branch right now.
Unfortunately, my daughter is at school right now and I'm looking at all of this remotely, so I can't look at the logs themselves to see what was going on under the hood.
cameronrenwick
@cameronrenwick
Oct 11 2017 16:48
wondering if anyone on 0.6.0 can comment for me. I've got my eSMB set to 75 min, have moved my IOB slightly higher than when I was on 0.5.3 (from 6.5 to 7.5) and have multiple IFTTT buttons for carbs; I'm on Fiasp and chosen the proper curve in pref. I find that the whole routine is working very well but I'd like it to be a bit more aggressive to control my peak BG. What is the suggested approach for this?
Dana Lewis
@danamlewis
Oct 11 2017 17:24
@bdr1177 do you have papertrail?
fidoman
@efidoman
Oct 11 2017 17:49
@cameronrenwick - I'm currently at the suggested maximum for eSMB minutes at 90 minutes and I don't have problems with lows. However, keep in mind that I'm on Novolog and using 'rapid-acting' curve. I think if you want to lower the post-meal peak a bit more then your best thing to try is going from 75 minutes to 90 minutes. I would only do that if you aren't having problems with post-meal lows though.
Brian Rabinovitz
@bdr1177
Oct 11 2017 17:53
@danamlewis I have papertrail, but it doesn't go through the school network. I get the whole school day as a single dump in papertrail as soon as my daughter gets home.
garykidd
@garykidd
Oct 11 2017 17:56
weird...manually entered a temp target via careportal at around 1130ET....in NS I'm showing two temp target bars (gray)....they are at 2 different times but indicte the 1130 Temp Target in both instances when you mouse over. Anyone else experience this?
cameronrenwick
@cameronrenwick
Oct 11 2017 17:57
I have seen this Gary but when I refresh my browser it only displays the actual one that's correct. You've tried a refresh?
garykidd
@garykidd
Oct 11 2017 17:58
that was the quickest fix ever--thanks!
Scott Leibrand
@scottleibrand
Oct 11 2017 18:01
@bdr1177 I took a look at your NS, and confirmed that OpenAPS did see the increase in IOB from -0.1 to 0.5U after the bolus, and that it was reflected in the IOB pill after the next loop run (at 12:00 noon EDT). So aside from the expected lag (as NS can't update until after the next pump-loop run and upload), it looks like the only problem is the actual NS treatments upload not showing the bolus. that is a known issue that we still need volunteers to work on reproducing and fixing.
yeah, temp target rendering in NS is buggy: I'm always having to refresh to fix it.
garykidd
@garykidd
Oct 11 2017 18:04
duly noted. I can say that my OpenAPS pill populates much more consistently after having updated NS to latest version. Previously it would hardly ever load initially and would require an immediate refresh.
soundwave42
@soundwave42
Oct 11 2017 18:09
Forgive me for not knowing how to search for this answer if it has already been posted.
I am wanting to start back on my insulin pump and was all set to get a Medtronic 670g but they do not have the cgm portion in stock and won't until "sometime" next year. I see no point in wasting the money on the setup without the cgm. I have a 722 and a minilink transmitter so I thought I would give openaps a try. I am familiar with Linux operating systems and the technical process seems fairly straightforward. My question is what is the plan for the intel Edison support now that is has been discontinued? Is there a specific non-jubilinux board I should buy? The one that is linked and already flashed is out of stock.
Scott Leibrand
@scottleibrand
Oct 11 2017 18:14
I would go ahead and order the already-flashed one, regardless of today's stock level. They re-stock frequently in small batches, so it probably won't be long until you get it.
There is work ongoing to begin making an Explorer HAT for the Pi (all types), but that won't be ready for a couple months. In the meantime, the Explorer Block and Edison are still available.
garykidd
@garykidd
Oct 11 2017 18:21
@scottleibrand does that mean a board similar to the current explorer would be able to plug directly into RP0 , for example? (when it's complete, of course)?
Scott Leibrand
@scottleibrand
Oct 11 2017 18:28
Yes. I have a prototype, and have tested that it works. There is still some work to be done, both on the hardware and software side. Among other things we need to speed up oref0 so it can do a full SMB pump-loop on the Pi0 in a reasonable amount of time. It works great on the Pi3 though. :)
daveklingler
@daveklingler
Oct 11 2017 18:29
I believe that the current explorer will plug into the Technexion boards, although the drivers need to be compiled.
If someone hasn't already done it.
garykidd
@garykidd
Oct 11 2017 18:31
could an optimizing compiler could aid in that endeavor?
daveklingler
@daveklingler
Oct 11 2017 18:32
It's on my list of things to try if I ever get my Edison rig working. I'm still trying to figure out what broke my Bluetooth pairing.
@garykidd, these are roughly $50 in a comparable configuration to the Edison:
garykidd
@garykidd
Oct 11 2017 18:37
that looks pretty cool.
daveklingler
@daveklingler
Oct 11 2017 18:39
@garykidd, I could be wrong, but I believe that most or all of oref0 is written in either javascript or python (I'm still poking through it), in which case I don't believe compilers could help. I've looked at rewriting the node scripts in python so that the whole thing could run on a pyboard or other micropython platform, but it's a pretty big task.
garykidd
@garykidd
Oct 11 2017 18:39
@daveklingler I've 2 issues with BT...first was rectified by deleting my rig from my iphone list of BT devices...the other was related to inadvertent and quite obscure change to my supplicant file....not sure if maybe one of those 2 things could've coincidentally affected you as well...
are there cross compilers for suchthings?
daveklingler
@daveklingler
Oct 11 2017 18:40
@garykidd, once you deleted your rig from your iPhone list, were you able to pair again with it?
garykidd
@garykidd
Oct 11 2017 18:41
yeah, but I had to go through setup again...I think. (while back)
daveklingler
@daveklingler
Oct 11 2017 18:46
@garykidd, There are javascript-python converters to help with the task. Adafruit makes two featherboards that run micropython. The major point of the exercise would be to get to a smaller, cheaper platform and also get the battery life up to at least a week. I'm not sure that oref0 can't do that already with just a few changes on Edison to let it hardware sleep, but the micropython featherboards are cheap!!! and the setup would become much simpler and more reliable.
@garykidd I just ran setup again, but I'm still not having any success. Do you remember what happened to your supplicant file?
garykidd
@garykidd
Oct 11 2017 18:49
yeah, I f'd it up by mistake...probably when editing via screen...sometimes my cables contribute to noisy screen interface and I think my cursor might not have been where I thought it was.....ended up being in the first 2 lines of that file though.
daveklingler
@daveklingler
Oct 11 2017 18:50
@garykidd Ah. I think my supplicant file's okay in that respect.
soundwave42
@soundwave42
Oct 11 2017 19:04
I figured the support for the Edison would stick around for a while at least. I wanted to buy the exact pieces that the docs were written for to maximize success.
A pi0 would be awesome to use in the future.
Does anyone know where I can get new sensors for my minilink? Is there another group I should post in for help with that?
garykidd
@garykidd
Oct 11 2017 19:06
@soundwave42 I may have some still laying around--could be a little dated--though I bet they'd be fine in any case.
Scott Leibrand
@scottleibrand
Oct 11 2017 19:08
@soundwave42 if you have any other extra supplies you could trade, there are a lot of diabetes supply pay-it-forward groups on FB and elsewhere as well...
garykidd
@garykidd
Oct 11 2017 19:08
I'll bet edison support will be around a long while still...w most of the users here all using edison w/ explorer boards and all...
Scott Leibrand
@scottleibrand
Oct 11 2017 19:09
yeah, and since the Pi0 is higher power consumption than the Edison, many people won't want to replace their mobile rig with a bigger one with a bigger battery.
soundwave42
@soundwave42
Oct 11 2017 19:10
@garykidd I was hoping to support this long term, especially if it works well. I can still buy sets for the 722 directly from Medtronic. At least I think the new ones will work with older pumps but I cannot determine is the newer sensors will work with the older minilink. I tried to ask my Medtronic sales guy but he hasn't responded.
daveklingler
@daveklingler
Oct 11 2017 19:11
@scottleibrand Scott, I've been thinking of making my Edison deep-sleep for 4 1/2 minutes, then wake up in time to catch the G5 transmission. Can you think of any reason that won't work? (I won't have a reason to try it until I figure out my BT problem).
garykidd
@garykidd
Oct 11 2017 19:12
they do...I had new (relatively) minilink w/ my 723 not that long ago...and the cool thing about it is that you can loop offline easily with the MM sensors...
soundwave42
@soundwave42
Oct 11 2017 19:12
As long as I can get the Edison and explorer board I think I am good to go. I was hoping to buy the preflashed one and not guess if the edisons on amazon were the right one.
garykidd
@garykidd
Oct 11 2017 19:13
using a mac the flashing instructions are spot on...pretty easy now...can't speak for PC...but I think they're well hashed through at this point.
Even if pre-flashed wasnt available---I wouldnt let that deter you.
soundwave42
@soundwave42
Oct 11 2017 19:15
@garykidd so the enlite sensors work with the minilink? Sorry for so many questions. It has been very hard to find any direct info from Medtronic.
Also what pump are you using now? I am guessing a dexcom cgm if you aren't using the minilink anymore.
Is most everyone using the auto loop successfully? I am pretty excited to try this especially with the small investment for the hardware needed.
jaylagorio
@jaylagorio
Oct 11 2017 19:17
Do you mean closed loop? I think there's something like 400+ of us worldwide who are.
Personally, the small investment for the hardware was worth not waking up once or twice every night after 8 years.
soundwave42
@soundwave42
Oct 11 2017 19:29
I am currently having nightly lows in the 40s and 50s. I am also dealing with highs during the day. I need better monitoring and management. Currently just finger sticks and injections. Years ago my insurance premiums and out of pocket went up drastically so I had to come off my pump and go back on shots. My numbers are all over the place using R and N insulins.
Can I do a closed loop with a Medtronic 722 and the minilink cgm?
jaylagorio
@jaylagorio
Oct 11 2017 19:30
What's the software version on the Medtronic 722?
daveklingler
@daveklingler
Oct 11 2017 19:37
@soundwave42 I'm getting a lot more sleep with my Edison running, even though I can't leave the house with it yet. My nighttime blood sugars have smoothed out a great deal. But a large part of that as well is that OpenAPS required me to get far more accurate with my carb estimates. Most of the credit goes to OpenAPS, but it comes with some extra education.
Strangely enough, I did better with R and N, many moons ago, than I've done in the past several years with Lantus or my pumps, probably due to higher levels of attention.
Scott Leibrand
@scottleibrand
Oct 11 2017 19:39
@jaylagorio as far as I know all 722s are loopable, regardless of firmware version
soundwave42
@soundwave42
Oct 11 2017 19:40
@jaylagorio version 2.4a
Scott Leibrand
@scottleibrand
Oct 11 2017 19:40
@soundwave42 if you can get your CGM data to show up on your 722, then yes, you can loop with it.
jaylagorio
@jaylagorio
Oct 11 2017 19:40
Oh, nice! Well there you go. Can't speak to whether you can use it's built-in CGM from personal experience but I believe that's part of it.
Good deals
Scott Leibrand
@scottleibrand
Oct 11 2017 19:41
@daveklingler the stuff oref0 does takes a lot longer than 30 seconds currently. and I don' t know of a way to deep-sleep the Edison. I couldn't even get it to put the CPU into power-save mode to run cooler.
soundwave42
@soundwave42
Oct 11 2017 19:44
With injections and the slower insulins I have issues where the insulin stacks up and causes lows and depending on when I stick my finger I might miss the trending high but even if I catch it, it takes a lot longer to bring it down with the slower insulins.
I could do more to try to get my numbers better daily but my A1C is good at 6.7 so that can be misleading with day to day life especially when you are hitting daily lows of 40 and 50 it brings your overall average down for A1C even though you have highs during the day.
Brian Rabinovitz
@bdr1177
Oct 11 2017 20:11
@scottleibrand Yes, I was also very relieved to see that OpenAPS did ultimately recognize the additional insulin. What I can't tell from NS, and maybe you can confirm, is whether took two or only one loop for the carbs and insulin to show up. I should be able to look back at this when I get home from work and can look at the logs from earlier. I'm also happy to help out testing if I can. Our only limitation is that school time is difficult for us to do any direct testing.
Diabetic_Dude
@diabetic_dude_twitter
Oct 11 2017 20:30
how can I pull old logs from my rig????
I would like to pull info from around 11 am until my current time 3:30 PM
trying to do information gathering
(looking into papertrail)
Scott Leibrand
@scottleibrand
Oct 11 2017 20:41
the l command just does a tail to show the last bit of the log file, but the actual pump-loop.log goes way back, so you can look at it with less or similar
or copy it off to your computer to read there
Diabetic_Dude
@diabetic_dude_twitter
Oct 11 2017 20:49
@scottleibrand where is the pump-loop.log stored?
PieterGit
@PieterGit
Oct 11 2017 20:55
@diabetic_dude_twitter : /var/log/openaps
Diabetic_Dude
@diabetic_dude_twitter
Oct 11 2017 20:59
thank you
Diadon81
@Diadon81
Oct 11 2017 21:01
I have an question. If I delete whole ~/myopenaps directory is it mean that I lose all autotune data, is it possible to restore it from NS, should
should I use a command from crontab?
manually but with wider time frame
thanks
PieterGit
@PieterGit
Oct 11 2017 21:10
autotune will download the cgm and bolus data from nightscout, but old profiles and reports are store in the ~/myopenaps/autotune directory
you can also move the dir to a new name, e.g. mv ~/myopenaps ~/myopenaps.20171011
Diadon81
@Diadon81
Oct 11 2017 21:14
Is that old info need for autotune to recalculate a sensibility or it always get all necessary information from NS?
PopcornPam
@PopcornPam_twitter
Oct 11 2017 21:14
Hey, I am trying to run autotune on my computer (have been using loop for a few days while on vacation), and when I follow the steps in the docs I get this error
Warning: API_SECRET is not set when calling oref0-autotune.sh
PieterGit
@PieterGit
Oct 11 2017 21:17
@Diadon81 it will get the info from NS and from the pump profile AFAICT
Diadon81
@Diadon81
Oct 11 2017 21:23
ok it is clear now, thank you
Scott Leibrand
@scottleibrand
Oct 11 2017 21:29
if you deleted your autotune directory and want to re-run autotune on previous days, you'll need to do that manually. by default autotune on the rig will just start tuning nightly from your current pump settings, and don't go back in time
@PopcornPam_twitter that's a warning, not an error: ignore it
PopcornPam
@PopcornPam_twitter
Oct 11 2017 21:30
ok
PopcornPam
@PopcornPam_twitter
Oct 11 2017 21:37
So I am still getting this message
Warning: API_SECRET is not set when calling oref0-autotune.sh
Option 2017-10-09 unknown
Nevermind.. found a stray space
Diabetic_Dude
@diabetic_dude_twitter
Oct 11 2017 21:43
@scottleibrand are you around? I would love to shoot you my pump log and a snapshot I am having problems with both UAM and SMB, I am on the latest dev version as of this past Tuesday I am having issues with spiking after meals weather I pre bolus or not. When I pre bolus I can keep my bg at around 200 if I try UAM it will go way too high sometimes 250 sometimes higher.
I was considering posting on the same post in looped but the issues I am having are the opposite of Christine.
Diabetic_Dude
@diabetic_dude_twitter
Oct 11 2017 21:48
(the looped facebook page that is)
Scott Leibrand
@scottleibrand
Oct 11 2017 22:24
@diabetic_dude_twitter what do you mean by "try UAM"?
Diabetic_Dude
@diabetic_dude_twitter
Oct 11 2017 22:47
@scottleibrand input my carbs into nightscout without a pre bolus
garykidd
@garykidd
Oct 11 2017 23:10
so what's the best way to examine past autotune files?
Diadon81
@Diadon81
Oct 11 2017 23:10
I am getting the following error in my pump-log:
File "/root/src/decocare/decocare/cgm/init.py", line 84, in init
raise DataTransferCorruptionError("Page size too short")
decocare.errors.DataTransferCorruptionError: Page size too short
any ideas?
there was a reason, why I reinstall myopenaps
it fix the problem but after one hour it raise again.
I am using openaps with MDT sensor
garykidd
@garykidd
Oct 11 2017 23:17
the autotune_recommendations.log is adjusted incrementally over time and re-written over any prior file to the extent that it exists..is that correct?
PopcornPam
@PopcornPam_twitter
Oct 11 2017 23:24
So trying to run autotune on computer versus rig, and getting these errors, and no recommendations

Autotune pump profile recommendations:

Recommendations Log File: /Users/pamelashwartz/myopenaps/autotune/autotune_recommendations.log

jq: error (at /Users/pamelashwartz/myopenaps/autotune/profile.json:1): Cannot iterate over null (null)
jq: error (at /Users/pamelashwartz/myopenaps/autotune/profile.json:1): Cannot iterate over null (null)
/usr/local/bin/oref0-autotune-recommends-report: line 71: printf: null: invalid number
/usr/local/bin/oref0-autotune-recommends-report: line 77: printf: null: invalid number

garykidd
@garykidd
Oct 11 2017 23:46
@scottleibrand mind taking a look at my autotune log and my proposed changes...just for a spot check?