My application [0] I used daily for 13 years stores the data in localStorage and I had several unpleasant data loss incidents - mostly me cleaning storage of the wrong browser tab. Thankfully, a simple JSON dump/import feature enables a good enough degree of reliability. Thanks to it, I didn't lose any data in 5 years, but now I have a questionable habit where I defensively export data before every major edit or deployment.

Simple localStorage replication has been an interest of mine for many years and I explore available solutions multiple times a year, but none of these ticks all the boxes - no server to maintain; free; optional for users; bring your own storage; the app stays a web page; cross-browser. I wish browsers had a file system API that allowed the app to read/write to specified directory or a file in background, which users may sync with their tool of choice, like Google Drive.

[0] https://github.com/Klaster1/timer-5

This made me think - is there a simple way to sync localStorage with my Firefox profile? If not, it doesn't sound like a WebExtension too difficult to create.

I actually implemented an export/restore function at one point and ended up taking it out.

I waffled heavily on that feature but in the end only losing the last recorded weight for each exercise didn’t feel like enough of a loss to warrant the cluttering up of the interface or reliance on external services.

It does have JavaScript, it's not just HTML.

Well, certainly some HTML ... but mostly vanilla javascript. Which in its own right is also beautiful.

You need to take that file and drop it in githack.com or just open the link in GitHub and change “github” to “githack”.

Like this: https://raw.githack.com/bmtwl/exerciseminimilism/13d4ee27deb...

I had a spreadsheet set up much like that in the past: it would bubble up the exercises I'd done the longest ago for a particular muscle group so I'd always get variety. It worked pretty good, but I found myself often avoiding unpleasant exercises when given choices and eventually the whole arrangement fell apart due to too many moving pieces. I also didn't see the kind of extra benefit that I thought I'd gain from performing so many variations. I also used to set up fully scheduled on/off timers for specific exercises in an app, but found that any interruption would throw the whole thing off, annoying me to no end. However, your last point was also my biggest enemy. If I get distracted online for long enough to cool down to the point that I'm worried about injury, the whole session tends to just blow straight out. That's the main driver I had for putting the simple timer in the footer: keep my device on one screen from beginning to end!

I’ve also been toying with a timer based app that incorporates some randomness, which I feel is usually missing or poorly handled in fitness apps. Some random exercises and variable-range durations should improve motivation and give your body a better workout.

To quote Arnie: “You have to shock the muscle, and you have to trick the muscle. The body will adapt to anything you do, so you have to keep it guessing. That’s how you grow.”

I actually built something similar last week

https://bestefforttools.com/workouts

I wanted something to create realistic yet slightly randomised warm ups and finishers - similar to CrossFit wods but more structured.

Behind the scenes I use a formula that estimates a total work capacity.

Feedback welcome!

I guess this should be easy to build on val.town or https://bolt.new/

Here's what I did:

- I wrote down my goals, equipment, and past workout history.

- Then I asked ChatGPT to generate a P/P/L workouts around that, in 8 separate chat instances

- I then wrote another prompt for a fitness expert adjudicator and asked it to compare two of the workouts, determine their strengths and weaknesses (based on above goals), and develop a third workout based on the strengths of the two provided

- I ran this in "tournament style" for all 8 workouts with three brackets. This ended up converging into an optimal workout

I actually took this final workout to my personal trainer and he said it was perfect for my goals and wouldn't change anything. I've been running it for 3 months now and I am quite happy.

Fitbod was pretty good about this. You can tell it what gym equipment you have and it will design workouts around that and the muscle groups that aren’t tired from the last session.

Thanks! I tried to use a multi-sport cross section of exercises and recovery routines to show some of the things you could do with it, but all the exercises are ones that are related to sports I do to keep it grounded in reality.

I was thinking it would be fun to do git branches with some more well-known and popular routines for different goals to give people a nice range of potential starting points.

As far as how long I let it grow: I used to keep years and years of data, but this app will not be my place for that going forward. This will be for a maintenance/marginal gains habit, and Golden Cheetah will be for running/cycling performance management where peaking actually matters.

Good luck with your own app! I love the idea of an offline capable local db.

Haha, I unironically considered this as well, but prefer the "always today, only today" clean automatic view and lack of a physical artifact to lose/forget. I travel for work often enough that being able to get to it on my phone consistently is also an advantage.

I was a strong user for years but I would now suggest https://www.hevyapp.com/. It has everything strong does but also a web app where you can create your programs on.

Most importantly for this crowd, you get a rest api to your data if you spring for the cheap premium plan. Csv export is free though I’m pretty sure