I'm fairly far along on the data hoarding spectrum. This extends, for example, to over 200K photos from my life, categorized and indexed. It's trivial to look at "what happened on this day 20 years ago" or at least on the nearest date to that day that had pictures taken.
And here's the thing: As the memories fade, so does interest in the photos. 10 years ago you still remember clearly. 20 years ago is another era.
And that's photos. Fitness data? Do I really want to see how I declined over the last 20 years? I see that already from the average speed display on my bike computer. Fitness data is the ultimate "looking forward" thing. But maybe that's my age; it may look different from a younger perspective.
Agree with others that just use a platform that lets you export open-ish files. .gpx doesn't do heart rate, but .fit does, and gpsbabel can translate that to another format as needed.
As per photos you don’t need 11 pictures of those muffins you baked in 2003. One will probably suffice, if you even care about them 18 years later that is.
The same goes for health data. But overall trends over 20 years are really useful as you get older. Expected decline is inevitable but unexpected decline may be indicated earlier and result in some preventative action to improve your quality of life. There are a lot of long term metrics appearing in Apple health like walking stability on that front.
On photos, I’m not sure what I’d do with 200,000 photos. I have 7,623 which span 4 generations and 125 years at the moment. I spend a lot of time curating these photos, adding metadata, editing and pruning garbage. They are all managed with Apple’s native Photos app which is good enough for the job (non destructive, sqlite DB underneath it, easy to back up)
My point is, it seems like once you have put in place the habits and technological solutions necessary to store fitness data, it actually takes more effort to reduce the resolution of older data. I'd rather have data and not want it than want it and not have it, but then I am an almost pathological hoarder so maybe it's just me.
Some of them might be historically interesting enough to preserve on Wikimedia Commons, for example if they depict places that look very different today, or stuff that you don't see in everyday life anymore.
That's what makes them so interesting.
The photos from 15 years ago with my parents still alive and my kids as toddlers are more interesting than last year's holiday.
I have approx. 220k photos, of which 210k are in camera raw, 10k are ‘processed’, which typically means exposure corrected, cropped, etc. They are stored in two parallel directories (raw and processed), each of which is structured into years which are further subdivided into ‘shoot’ folders – one for each distinct upload from the camera. I use Adobe Bridge to do the uploads and to change the raw files names to support my indexing approach. The year folder names are simple (e.g. ‘2021’) and the shoot folders have names like ‘2021-010 garden birds young’ (where 010 is the sequence # of the shoot in the year). The folders thus automatically list themselves in chronological order when sorted alphabetically. A processed image might have a name like ‘20210523_010_my-IRL-name-here_0019.psd’ (so that I can immediately find the raw or processed shoot folder given a file name). I insert my name so that if I a send an image to someone I am identified as a creator (image metadata contains additional contact / copyright info). I’ll often extend the automatically generated file name to include the photo subject as I process specific images.
My indexing system relies on my memory and the fact that the relatively small number of processed pictures are indicative of the contents of the wider raw set – so I can browse the relatively small number of processed shots (using Bridge) if I am looking for something thematically. I usually manage to cull obvious fails from the raw set when I do an upload before doing any processing. Storage costs being what they are I have not yet had to further cull raw or processed images to make room on my hard-drives – currently 1.7Tb of images.
I have left it too late to add meaningful tags to this collection, and because I am the only librarian, I don’t have a need to help other users find particular images.
Why not? I have about 12 years worth because that's when I started tracking, no reason to ever delete it. Saving the 0.001% (actual value, I just checked) on disk space isn't a motivation.
It's fun to review old data. Occasionally useful, even. Mostly for fun.
> And here's the thing: As the memories fade, so does interest in the photos. 10 years ago you still remember clearly. 20 years ago is another era.
Wait another 30 years and those now-50 year old photos will be treasure.
I don't look much at my photos from 20 years ago, that's pretty recent so I remember it well. But photos from 50 years ago are quite the treasure. Even older ones (like when grandparents were toddlers) are invaluable. Wish there were tons more of those.
But yes that's photos. Fitness data is certainly less interesting, but still fun and costs approximately zero space to keep.
Just a few days ago I could find out the answer to a Q about when in the year I normally do my first cross country ski session, by looking at the graphs per year. So on average I will be going skiing next weekend 8)
> 200K photos from my life, categorized and indexed.
What app do you use to manage your photo library?
It doesn't seem like a good idea to tie a photo library to any app. Applications come and go, the photo library should last a lifetime (or more if you pass it on). You don't ever want to get caught with your photo data in some proprietary structure when the app that created it is no longer available for whatever reason. My photo library certainly has outlived many applications I've used to process the photos.
After processing I store them in just a directory hierarchy. Top level dirs for each year, event/topic subdirectories under each year. That's it.
I don't use most functionality, but steps, sleep time, heart rate, and notifications/alarms have all been useful to me.
Sounds like someone is trying to wrap this up, godspeed: https://spectrum.ieee.org/wearable-health-data-standards
Also, IMHO, I think healthcare is where protobuf could really win over JSON: a protobuf comes with the schema and any datatype can ride the message. Everything you need is there.
Is protobuf a heavier lift to get started? Yeah. So? Healthcare is too heavily regulated (and for good reasons) for this to be a significant issue. The barrier to entry is so much higher than the effort to handle protobuf, that it's lost in the wash.
It's just sad that FHIR has lots of issues as a standard. On one hand it suffers from trying to model the whole world, similar to the problems of the semantic web in the 90s. Even with the cop-out of extensions for things it doesn't support all the objects are very bloated.
A lot of things feel like ad-hoc solutions and there are tons of pitfalls. A primitive value can be there by the primitive value being missing, but an extension being there. FHIR XML cannot be translated to FHIR JSON without access to the resources' definitions, because elements that have a higher maximum cardinality than 1 are serialized as lists in JSON, but the same way in XML. Yet when you read a FHIR XML or JSON you do not know which resource definition applies because there is no way to know the FHIR version. Variant elements that can be different types do not have the element name als the XML element name, but the element name with the type name added to the end after capitalizing the first letter. Profiling the resources is absolutely necessary to ensure interoperability but it's extremely hard to get right.
In the end FHIR brings a whole lot of complexity, while it doesn't bring much of a benefit.
If FHIR doesn't bring much benefit then why is it so widely used? If you need to exchange healthcare data with other organizations then what's the alternative? The older HL7 standards weren't any better.
Without some overarching body, like a government regulator, I'm not optimistic there will be widespread adoption of interoperability.
GPX is described on its website as being an open standard and is very widely used, both for fitness data and other applications. The purpose of GPX is to describe GPS data; by itself it is not as suitable for fitness data as FIT or TCX as things like heart rate and cadence data are not part of the standard. However, GPX can be extended.
From an open web perspective, GPX with standard extensions for other fitness-related data would probably be best, but FIT and TCX are getting ever more popular. For now, the situation is fine because it's fairly easy to convert between them. The risk is that Garmin tries to close off the TCX and FIT standards in the future.
Of course this will only work in some jurisdictions, and even then it needs stronger enforcement to ensure that data controllers will comply (or even understand their obligations).
I can only hope it's a future in which we collectively realized that smartphones, smartwatches, smartglasses, smartypantstoilets, smartgloves, smartdiapers, etc, were a bloody well idiotic idea and that we've come to our collective senses and realized that they were never what they promised to be, but had, within a few years of launch, turned into just more ways of monopolizing attention for a quick venture capital and IPO payout. Perhaps not the hardware manufacturers, but at least the way the hardware was used.
More realistically, those are likely to be dim memories of the past, back before capitalism ran the planet into the ground, and took most of our technology stacks with it. Modern tech stacks are complexity, piled on complexity, built on a few stacks of complexity, built on broken hardware that's overly complex in an attempt to gain the performance required to run the teetering stacks on top of it (all the uarch vulns, DDR4 Rowhammer is a nice touch on top, etc).
You'll note that neither of those paths involve us running around with the FitBit 2050 gathering... whatever it is it thinks it ought to gather. The way things are going, probably nothing more useful than Covid tracking stuff. "You were in proximity with 1.7 potential individuals who may have been infected with Covid, please quarantine for the next two weeks. confetti"
It would be nice if we had some of our current technology, that open data standards had taken root and we could decide how we wanted to use the stuff, pay appropriately and be paid appropriately for our devices, attention, etc. But that's just not the way things are going, and I fear that's not the way things are going to go, under any reasonable future. There is far, far too much money and "Well, but, see, if you destroy the market you can monopolize it!" money floating around for the near to moderate future to see that future happen. I mean, a bunch of our current companies are literally based on "We will use the venture capital money to undercut everyone else, drive them out of business, and then be the monopoly!" Uber, Lyft, [insert the food delivery service of the week here], the various scooter services, and the list goes on. They're running into the brick wall of, "Wait... what do you mean, we're not going to accomplish that? You mean we have to turn a profit?" I've talked to people who have started taking taxis again, because at least they'll show up, unlike the Sharing Economy Rideshare App of The Future services that weren't remotely pandemic-proof.
We'll see. But I would predict that fitness trackers, in 20 years, are a big pile of ewaste, "... ugh, yeah, you know, it's not worth the hassle to get that data out...," and broken promises. And hopefully not being actively worn by then.
Sport will not perpetuate human existence. It was originally created to keep humans in shape for hard the graft of living off the land. Just look at any aboriginal culture, they knew they had done enough steps today when they had enough food.
The fitness tracker was created to make you feel better about being taken for a ride (in life). For those who needs affirmation from their peers that they arnt wasting there life away working for a whatever click bait porn floats onto there device next.
20 yrs from now? we already have big piles of ewaste everywhere.
I once interviewed for a startup making bluetooth diapers for old folks ~2013.
My recommended fitness tracker is to turn off all your devices and go outside into nature at every opportunity. Find a company who will let you be a remote developer and start living for today.*
I'm healthier and more active now because of them, and they've gotten me more in tune with my body, not less.
Most of the industry subsides on the aspirational user who doesn't make any lifestyle change and stops wearing the device after a year or less. The vague idea that it'll find some magic ingredient for improving your sleep or exercise.
Even Garmin is getting in on it trying to sell activity trackers for parents to put on their kids. When it's parents who are in full control of their kids' time and activity.
Of course it has a use for people training but that's a small minority of customers.
Here's the copy on a recent device:
"Meet our most advanced fitness & health tracker with tools like an on-wrist ECG app for heart health,* EDA Scan app for stress management and more. Get a 6-month membership of Fitbit Premium™** and optimize your routine with Daily Readiness Score.◆‡ Add a 6-month Premium membership for advanced insights & tools to improve your health"
Guaranteed most customers would benefit from finding exercise they enjoy, drinking less alcohol and worrying less about their sleep rather than being presented with Scores and Goals whose novelty and thus salience will quickly wear off. But that isn't something that can be sold so...
I think it'd be hard to say measuring something in order to improve on it is useless: and most of the smart devices allow one to extract exactly that value.
Now, I wouldn't be the one to discredit social and psychological value for any smart devices, because they can be very real too, but it's definitely obvious that we should look for alternative solutions to those problems at the same time.
