Worth noting, if they are covered by the person's insurance, it may be worth looking into using a CGM (Continuous glucose monitor) and the self-hostable application Nightscout[0].
CGMs use the interstitial fluid in (usually, although people also use other spots) inside side of the upper arm to measure blood sugar. As long as they are placed in the right spot (i.e. in a fattier area instead of directly on top of muscle, hence why it is on the inside side of the upper arm and not the outside) and the sensor doesn't fail, they will measure blood sugar fairly accurately for 10-15 days depending on the CGM model (disclaimer: They do tend to run approximately 15 minutes behind the number that an actual finger poke would show you).
Nightscout is an awesome tool because it can be self-hosted so that you keep full control over your health data, but since it is a web application, you can still easily share it with other people too, like family or your endocrinologist.
On android, you can use the xdrip+ application to pair with most types of CGMs to completely keep your data out of the cloud, or you can also pair it with the API from Abbott or Dexcom (depending on the CGM model) to pull data from their 'share/link' applications, if you opt to pair with/use their official apps
I personally use Dexcom G7 15 day sensors (because it is what my insurance covers) and use xdrip+ to upload the data to a nightscout instance hosted in my homelab
Unfortunately, that’s not really an option for my target audience. In remote areas, the issue isn't only the cost. Its availability, access to the devices, internet connectivity, and the expertise needed to set them up can all be barriers. A basic glucometer and a phone are much more realistic.
Always follow the manufacturer’s instructions. Most CGMs are going on the back of the upper arm but some were put on the lower back, or other places. The manufacturer develops the CGM with a very specific placement in mind and they make sure to tell you very clearly.
My parents dog has diabetes and since they failed to draw blood from his ear he got the freestyle libre 2. The vet just put it near his shoulder. I'm not sure how accurate this is...
There is no way I'd track that info in an app. My meter keeps history. I can download it to my device. Why would I put another app in the middle of that process?
My use case is a bit different. I’m not trying to replace the meter or its history.
The problem I wanted to solve starts when the readings change. A doctor may ask, “What did you eat around that time?” With this app you can save a photo of the meal alongside the blood sugar reading, notes and medication.
So the goal is less about storing the number and more about keeping the context around the number and making that context easy to share with the doctor.
The most humble meters come free in a kit, since you're really buying the test strips, and they all have onboard memory and an optional app.
It's extremely unlikely the owners of those meters would even be able to use any other app. There's no standard for the device data sent over bluetooth to the app. There's no USB port. Beyond 90 days of data (many meters have 99), A1C is preferred. Reverse engineering to support alternative apps is also unlikely since there are always new devices to hack.
SugarTrack, and other apps like it, sit in a weird deadzone. The most likely users either have a nicer meter that already does offline export, or they have a crappy meter but prefer the official app or tracking their stats on paper anyway (in their food journal). Again, the app is always optional unless it's from a super sketchy brand that isn't even cheaper.
There's no room left for doing weird somewhat low-tech things in principle anymore. The market has moved on. The most common layperson user wants integration across all their health apps. The most common nerd user buys the nerd device. The manufacturers have known this for at least a decade.
> It's extremely unlikely that owners of these meters would be able to use any other app.
I've been using Xdrip+ for 2 years now with the Dexcom CGM instead of their proprietary app. Works great and gives me the ability to export my data or sync it with a self hosted database if I so please. Why would I trust a corporation with my private health data when they have continuously shown that they will sell it or lose it to a hack the first chance they get.
You might want to consider using a self-hosted database to store your data. Telemetry Blocklist (free) can help prevent any apps from phoning home, but it's not specific to this use case. If you're concerned about data security, you might want to explore other options for storing your health data offline.
"Kimi decompile this proprietary APK and find the communication channels used to communicate with this model GCM. Let me know what software you need to to install to achieve this "
The world is your oyster if you have some self motivation.
This is your honest recommendation? You might want to re-read how this thread started.
The issue isn't anyone's ability to reverse engineer a glucose meter nor their "self motivation" to do so.
The issue is how outdated the expectations of some devs can be. The typical consumer will never use these apps because the whole topic is washed up. Software maintenance is a much steeper hill to climb than it used to be. Throwing an LLM at that just makes it more Sisyphean than it already was.
You can have fun, but this isn't helping anyone else defend their privacy. That angle is starting to become as embarrassing as the advice to save the planet by riding a bike.
This looks really interesting. Thanks for sharing. I’ve already found some really good ideas here for the next version.
My focus is a bit different though. I’m trying to make the experience as simple and frictionless as possible, especially for non-technical amd old people.
Worth noting, if they are covered by the person's insurance, it may be worth looking into using a CGM (Continuous glucose monitor) and the self-hostable application Nightscout[0].
CGMs use the interstitial fluid in (usually, although people also use other spots) inside side of the upper arm to measure blood sugar. As long as they are placed in the right spot (i.e. in a fattier area instead of directly on top of muscle, hence why it is on the inside side of the upper arm and not the outside) and the sensor doesn't fail, they will measure blood sugar fairly accurately for 10-15 days depending on the CGM model (disclaimer: They do tend to run approximately 15 minutes behind the number that an actual finger poke would show you).
Nightscout is an awesome tool because it can be self-hosted so that you keep full control over your health data, but since it is a web application, you can still easily share it with other people too, like family or your endocrinologist.
On android, you can use the xdrip+ application to pair with most types of CGMs to completely keep your data out of the cloud, or you can also pair it with the API from Abbott or Dexcom (depending on the CGM model) to pull data from their 'share/link' applications, if you opt to pair with/use their official apps
[0]: https://nightscout.github.io/ [1]: https://github.com/nightscoutfoundation/xdrip
I personally use Dexcom G7 15 day sensors (because it is what my insurance covers) and use xdrip+ to upload the data to a nightscout instance hosted in my homelab
Unfortunately, that’s not really an option for my target audience. In remote areas, the issue isn't only the cost. Its availability, access to the devices, internet connectivity, and the expertise needed to set them up can all be barriers. A basic glucometer and a phone are much more realistic.
Interesting. My dad wears his on the back of the arm (posterior) not the inside (medial) - and when I asked him he pointed to the official instructions which do actually have it there https://www.freestyle.abbott/content/dam/adc/freestyle/count...
It must be a different kind of device.
No, he is correct. When I wrote this comment originally I was having trouble thinking of how exactly to describe the location
Always follow the manufacturer’s instructions. Most CGMs are going on the back of the upper arm but some were put on the lower back, or other places. The manufacturer develops the CGM with a very specific placement in mind and they make sure to tell you very clearly.
My parents dog has diabetes and since they failed to draw blood from his ear he got the freestyle libre 2. The vet just put it near his shoulder. I'm not sure how accurate this is...
There is no way I'd track that info in an app. My meter keeps history. I can download it to my device. Why would I put another app in the middle of that process?
My use case is a bit different. I’m not trying to replace the meter or its history.
The problem I wanted to solve starts when the readings change. A doctor may ask, “What did you eat around that time?” With this app you can save a photo of the meal alongside the blood sugar reading, notes and medication.
So the goal is less about storing the number and more about keeping the context around the number and making that context easy to share with the doctor.
For those who don’t have a meter that can export readings.
Which ones?
The most humble meters come free in a kit, since you're really buying the test strips, and they all have onboard memory and an optional app.
It's extremely unlikely the owners of those meters would even be able to use any other app. There's no standard for the device data sent over bluetooth to the app. There's no USB port. Beyond 90 days of data (many meters have 99), A1C is preferred. Reverse engineering to support alternative apps is also unlikely since there are always new devices to hack.
SugarTrack, and other apps like it, sit in a weird deadzone. The most likely users either have a nicer meter that already does offline export, or they have a crappy meter but prefer the official app or tracking their stats on paper anyway (in their food journal). Again, the app is always optional unless it's from a super sketchy brand that isn't even cheaper.
There's no room left for doing weird somewhat low-tech things in principle anymore. The market has moved on. The most common layperson user wants integration across all their health apps. The most common nerd user buys the nerd device. The manufacturers have known this for at least a decade.
> It's extremely unlikely that owners of these meters would be able to use any other app.
I've been using Xdrip+ for 2 years now with the Dexcom CGM instead of their proprietary app. Works great and gives me the ability to export my data or sync it with a self hosted database if I so please. Why would I trust a corporation with my private health data when they have continuously shown that they will sell it or lose it to a hack the first chance they get.
You might want to consider using a self-hosted database to store your data. Telemetry Blocklist (free) can help prevent any apps from phoning home, but it's not specific to this use case. If you're concerned about data security, you might want to explore other options for storing your health data offline.
I've used Xdrip before, and it's device support is not that great either.
I'm not arguing against the idea. I'm being realistic about what devs and users can do about this situation.
If you're lucky enough to find something that works, I'm not knocking that.
"Kimi decompile this proprietary APK and find the communication channels used to communicate with this model GCM. Let me know what software you need to to install to achieve this "
The world is your oyster if you have some self motivation.
This is your honest recommendation? You might want to re-read how this thread started.
The issue isn't anyone's ability to reverse engineer a glucose meter nor their "self motivation" to do so.
The issue is how outdated the expectations of some devs can be. The typical consumer will never use these apps because the whole topic is washed up. Software maintenance is a much steeper hill to climb than it used to be. Throwing an LLM at that just makes it more Sisyphean than it already was.
You can have fun, but this isn't helping anyone else defend their privacy. That angle is starting to become as embarrassing as the advice to save the planet by riding a bike.
There should be more simple, no login, apps like this
Check out https://codeberg.org/toz/MediLog
This looks really interesting. Thanks for sharing. I’ve already found some really good ideas here for the next version.
My focus is a bit different though. I’m trying to make the experience as simple and frictionless as possible, especially for non-technical amd old people.
True.