It's super cool and I love G Shock in general but the Casio app is straight up awful.

Have you tried this open-source app:

https://github.com/izivkov/gshock-smart-sync-webapp

Thanks for sharing this! I'm also looking for an alternative to the official app, since having to log in with an account and accepting a privacy policy every time I want to change the time on my watch is pretty ridiculous – unfortunately I think this doesn't quite fit my use case as usually I'll want to sync the time when I go to a different country, and I'll normally only have my iPhone with me, which won't support WebBluetooth. But this tool might be a starting point to fork and build a capacitor app with that can run on the phone I suppose.

You make it sound like rocket science or huge bother.

If you're a hacker/dev/tech nerd, that's trivial. You do similar things twice before breakfast without thinking about it.

You can use the app without hosting it yourself:

http://gshock.avmedia.org

I use this app on android : https://f-droid.org/en/packages/org.avmedia.gshockGoogleSync...

I mean... this is Hacker News. A lot of the audience would be excited about an self-hosted alternative.

Have you seen what vibe-coded slop it is? It's WILD.

https://github.com/izivkov/CasioGShockSmartSync/activity

You linked to the Android app there, not the experimental webapp linked above – I'm wondering what gave you that idea though? Those commit messages look pretty human to me, don't think an LLM would ever just commit 'WIP' or 'code cleanup', it's also 4 years old, before the height of vibe coding.

I can imagine the author may have used an LLM for this webapp version of it though

It is hard to reverse-engineer the Casio protocol by vibe-coding. Have you tried it?

Hot take: all government code should be open source.

This issue has the same vibes as the World of Warcraft forum on patch day.

Something strange is afoot with this post -- this is the third time it's showed up in the Newest RSS feed

Highly dubious of this.

I see zero reason for a person to care about the checkpoints.

And for agents, full sessions just needlessly fill context.

So not sure what is being solved by this.

With a 60M seed? I doubt it.

The most interesting part of this is that OpenAI are no longer in the lead but are playing catch-up.

Not exactly the same but somewhat similar is Atuin Desktop https://github.com/atuinsh/desktop

Thanks for the link! Atuin is a powerful tool, especially with their focus on command management and execution.

From my perspective, the main difference is the UI footprint and the intended use case. Sklad is intentionally 'passive' and limited to the system tray. It’s not meant to be a full window or a command runner/executor. I built it as a simple warehouse for static data — like passwords, specific IPs, or boilerplate templates — that I need to grab quickly and paste anywhere, whether it’s a terminal, a browser, or a GUI prompt.

I also focused heavily on the recursive folder hierarchy within the native tray menu itself to help with mental mapping, whereas Atuin (and its UI) feels more search-centric and terminal-focused. I think there’s definitely room for both depending on whether you need to execute a command or just grab a piece of secure data.

From the same site, the same stat for the US is 41% LOL

https://tradingeconomics.com/united-states/employment-rate

Edit: also you can't do math well -- it's 100 - 77 = 23 (not 33)

Jupyter notebooks also exist. Also see https://en.wikipedia.org/wiki/Literate_programming

I'm also a fan and heavy user of Jupyter notebooks and literate programming in general. I think the use case for runnable mardown files with AI tooling for automation applies to complementary cases.

Having said that, there are ad hoc automation tasks that I've traditionally used Jupyter notebooks to do that I'm finding are easier to get running using markdown files and Claude Code. It's early days and I still am getting a feel for this myself.

There are some comments from earlier with discussion of other literal program tools.