mjaniczek

11 ms·

31.
by mjaniczek 2y ago
I suspect it happens because of your externally provided insulin (as in, you caused your hypoglycemia by injecting too much). IDK if hypoglycemias happen naturally in T1Ds in situations where they don't in healthy people. I assume that
32.
by mjaniczek 2y ago
You probably mean K. J is also up there, but more educational and less "let's make some money".
33.
by mjaniczek 2y ago
I saw this format in the other HN submissions so I assumed it's always supposed to be there. Is it optional when the article is current-year?
34.
by mjaniczek 2y ago
Hey, thank you for the correction! I am not keeping up to date with how are the closed loops progressing, and from quite a few of comments here it seems like the future is already here :) Maybe just not evenly distributed - I just need to w
35.
by mjaniczek 2y ago
99% TIR... that's crazy. Well done. You're an inspiration!
36.
by mjaniczek 2y ago
It's a very rough rule of thumb, take these numbers with a grain of salt. It could very well be that some food activates in 5 minutes etc. Different insulin types also have different curves. Anyways there are studies about the "in
37.
by mjaniczek 2y ago
A datapoint of one: T1D has definitely made me crave sweet stuff more. Perhaps due to being "forbidden fruit", etc. etc.
38.
by mjaniczek 2y ago
It will help, definitely! As will exercise itself etc. It is a valid argument and the doctor is right.
39.
by mjaniczek 2y ago
You're definitely making me reconsider it, thanks! Along with the person in another thread saying walks after meals help as well. With a fully remote work these risky "oops I injected but there's no food yet" situations
40.
by mjaniczek 2y ago
100kg+. It's a joke :)
41.
by mjaniczek 2y ago
The LibreLink app allows me to use different alarm tones for lows and highs, and I'm _still_ getting alert fatigue. To the point that my wife needs to ask "is that your phone telling you you have a hyper?" for me to actually
42.
by mjaniczek 2y ago
In my checks the calendar 15 years with the disease, I've thankfully only had a hypoglycemic coma once, at a summer camp. I was leading a bass guitar workshop and just suddenly started making less and less sense. It was the only time
43.
by mjaniczek 2y ago
Yeah, I do have some passing experience with both F# and C#, and since the example code provided by the university was in C#, I kept that code and built on it. But I would have felt better in F#.
44.
by mjaniczek 2y ago
You're talking about T2D (which is more about insulin resistance than not producing insulin at all) which can be somewhat mitigated with lifestyle. Healthly lifestyle (exercise, diet, ...) can help T1D management, but T1D is an autoimm
45.
by mjaniczek 2y ago
I haven't, but I am now! Thank you for the suggestion
46.
by mjaniczek 2y ago
Don't worry, I'm not using the app's suggestions blindly (or at all, currently). But yeah the SmartCGMS authors disclaimed as much - you're using it on your own risk. This is probably also why apps like LibreLink don&#x2
47.
by mjaniczek 2y ago
Out of curiosity, is your wife's closed loop solution official, or homebrew? (If official, which country do you live in, if I may ask?)
48.
Taking my diabetes treatment into my own hands(martin.janiczek.cz)
485 points by mjaniczek 2y ago|356 comments
49.
by mjaniczek 2y ago
My biggest issue with it is how everytime I open it, it asks me to install a new version. Like, I am glad they're working on the software, but I'm getting fatigued from the inevitable popup
50.
by mjaniczek 2y ago
There has been some research (can't find the paper right now but it was about F-metric or something?) suggesting uniform generators will trigger bugs more often than the edge-case-preferring ones. I'm still not too sure about whet
51.
by mjaniczek 2y ago
I've been talking with Edsko de Vries about Falsify when he was implementing it / writing the paper. Very cool stuff, although somewhat mind-bending :)
52.
by mjaniczek 2y ago
PBT libraries usually have some support for filtering ("if you generate a value x such that P(x), discard it and try again"), but it's frowned upon. You should only use it when you're relatively certain it's not goi
53.
by mjaniczek 2y ago
I've tried to replicate your issue in the Elm testing library I'm a maintainer of, which uses some of the core ideas from Hypothesis. It was able to generate values just fine ([1]), so your problems might have been just some issue
54.
by mjaniczek 2y ago
Re the talk: it's unfortunately one of the Haskell Exchange talks that got pulled from the internet when the organizer, SkillsMatter, imploded. The Haskell Foundation is trying to revive and reupload those recordings, but my particular
55.
by mjaniczek 2y ago
The author focuses on the state machine and parallel aspect of PBT, but there are other aspects that might have larger effect. One of these is coverage guided PBT, check eg. Dan Luu's article: https://danluu.com/testing
56.
by mjaniczek 2y ago
I'm pretty sure the impossibility advertised back then was also about just making a browser.
57.
by mjaniczek 2y ago
To clear a possible confusion: Elm runs on NodeJS as well. The reason it's not often done is that the (1st/3rd-party) published packages are heavily skewed towards frontend, because Elm is a language primarily focused on frontend
58.
Find PBT properties by relating multiple functions [video](youtube.com)
2 points by mjaniczek 2y ago|0 comments
59.
by mjaniczek 2y ago
This is unproven (and not a toy problem), but I imagine it's going to do pretty well at compilers. The amount of time I'm waiting at work, hypnotizing the tsc process that sits at 100% CPU, wishing it was parallel...
60.
by mjaniczek 2y ago
I think this is an unsolved tooling question right now. You could get some sense of the parallelism by using `/usr/bin/time` and dividing the wall time with the user time. You could look at the Task Manager / Activity Mo

More ›