Plaid was a security nightmare for years before they were basically forced at gunpoint to get it together and even then it just seems from the outside that they are better. I'm not sure I would trust any company by the founder.

I would love to know what alternative phone OSes exist. I would happily throw money and my engineering weight behind a promising project.

Sailfish has a Linux-based OS with some Android compatibility.

Ubuntu Touch has left behind UBPorts which uses Android's hardware layer to provide a usable UI on existing Android devices.

There's the Librem phone with an open source stack, though its hardware doesn't come close to a cheap Android phone these days.

There are projects like PostmarketOS which work to get mainline Linux on phones. From there, you can run Linux on phones through desktop environments such as Phosh and Plasma Mobile which are touch optimized.

FirefoxOS died but was forked into KaiOS, though the modern iteration of KaiOS is an Android fork.

Samsung has TizenOS but I don't think any phones run it anymore. It's probably the closest equivalent to Android in the way it has been developed.

If you want to throw money at something, Sailfish may be the project to keep an eye on. All the other volunteer-run operating systems are worth donating to, of course, but Sailfish is actually trying to be a real third option rather than an open source proof of concept or alternative for shitty vendors stopping updates.

> There are projects like PostmarketOS which work to get mainline Linux on phones. From there, you can run Linux on phones through desktop environments such as Phosh and Plasma Mobile which are touch optimized.

Gnome is undertaking some mobile work in-house (no need for a different DE like Phosh or Plasma). My ideal would be a "Fedora Mobile" spin (based on ARM Silverblue).

https://www.youtube.com/watch?v=T6vCWFleBHk

It's looking good! Wish this would get the momentum it deserves.

Plenty. The ones you want to donate to are postmarketOS, Alpine Linux (the distro postmarketOS is based on) and Plasma Mobile (the main DE used by postmarketOS and other distros like Mobian).

In terms of usability, SailfishOS is the most usable, but it has some proprietary components and is commercially backed by Jolla, so there's no need to donate to them. I do encourage you to buy their devices if you can though. Speaking of devices, Furilabs and Volla also make practical, decent spec Linux phones (based on Debian and Ubuntu Touch), so they're worth checking out as well.

I've been saying that for a while. We need a viable third option.

But it is not easy, and it takes a huge amount of effort. As others have said, theres SailfishOS, Ubuntu Touch, PostmarketOS and a few others.

I'm not sure which one is in a better position to become that third alternative we need, but this year I'm experimenting with all of them on different devices, to try to understand where they stand.

"posted from an Ubuntu Touch mobile device"

It's like PC parts being modular. IBMs greatest mistake and the industry wants to make sure it never happens again.

When IBM tried that with PS/2 it didn’t go so well.

Apple seems to be having more success.

Ridiculous question and implication. Having a utility for something you do over and over with the same steps is automation 101.

The laws are for you and me not companies like Anthropic and Meta.

I want this but for ePaper tablets like Remarkable.

This is for epaper tablets, just tiny ones.

Remarkable looks like it has a pretty active hacking community: https://github.com/rehackable/awesome-remarkable#launchers

I had a similar thought but thinking it through, would probably lose the near-zero-lag stylus stuff. That's kinda rM's whole thing, not ebook reading.

This is back to the propaganda that people need to recycle because big companies dont want to clean up after themselves.

Tokens per second. The difference between 8B and something like 16B is not as big as you might think in practical usage and 8B is a lot faster and interactive than 16B but there are certain things where it is useful to farm it out to the large model.

Exactly this.

Creating conversation titles and parsing HTML/JSON don't benefit from 27B models.

The B70 can run both models comfortably side-by-side so it makes better use of time and resources.

Agree. For local coding help, latency often matters more than raw benchmark quality. A slightly weaker model that answers immediately changes how often you reach for it.

What SKG is asking for is not only technical reality but easy to do if you plan for it from the start.

I have worked at several tech companies that deal with data compliance. Most audits are jokes and most of your data is not secure.