No? Most app versions of web frontends are a wrapper for a pwa (teams, zoom).

Ff dumped pwa support a while back, and it has only just come back.

That is an unrelated issue, clearly.

THEY ALREADY HAD THE MONEY FROM GOOGLE.

I see a lot of people shit on Firefox for integrating LLM-related features. What the blog post points out and people fail to recognize is that if Mozilla stopped maintaining Firefox, almost all of its forks would die.

You’re misreading why people are shitting on FF. It would have been nice if Firefox had actually worked on literally any other features in the last decade, like not trashing support for PWAs. Instead, we got nothing but stupid partnerships with companies for features not one wanted, and a succession of failures in the CEO seat of Firefox.

So, yeah, I guess I don’t think it"s realistic for Firefox to have ignored the LLM craze entirely, but man, they could have managed that Google money better.

I think you should look up the definition of “rube”.

I collect GPS tracks (whenever I go somewhere I haven’t been, hiking included), and I also use wanderer for basic editing.

However, Wanderer has been trouble to upgrade and recently I’ve decided to just back up my Comaps tracks and export individual ones for editing on something like https://gpx.studio/. What I really need is a local version of comaps for big-screen visualization and planning.

How is Wanderer helping you?

Classless interdomain routing is not the point of this post, but basically we started giving people more options for network sizes, and while complex in implementation, it was philosophically virtually uncontroversial.

You’re talking about 2 different things here. Bit masking part of what is now CIDR was implementable since the mid 90s on network segments, it’s the logic to route packets based on CIDR that was integrated into RIP, GRIP, and BGP.

Also, CIDR can’t alleviate multi-hop routing address exhaustion.

They are saying wireguard alone, no reverse proxy. No exposing any services at all to the outside, just tunnel in with WG and access locally.

Yeah, enabling a bootctl flag is doable in any android version, it’s a prop available since Android 7. Again, it’s undocumented APIs at fault for the MTE not working in the Pixel 11.

It’s not an on/off kind of thing, MTE is a set of ARM extensions. Either the api to use them is clearly documented or, as in the pixel 11’s case, Google started to not document their implementation.

We know this matches their plan to obscure the AOSP code to eventually lock out anyone else from using their base.

GOS is not in a position to reverse-engineer api calls on what was formally documented hardware.

Instead, Google rewrote critical core Android system daemons using Rust. Like the keystore, Bluetooth stack, DNS resolver, and others. I believe using Rust essentially makes MTE unnecessary.

No. Rust being “memory safe” doesn’t mean rust apps can’t abuse speculation or buffer under/overruns.

The point is that the post makes it sound like using a Pixel 11 under Android results in less security than before, but for most people there is no change since they never used MTE anyway.

Read the BlueSky posts, Google didn’t force app devs to opt into MTE, so no one did.

Google probably decided that since hardly anyone used MTE anyway on earlier Pixel devices, they could use the space in the chip for other uses.

“Space in the chips”? Again, see the Bluesky posts.

Google is very clearly shifting over to a completely walled garden, because, same as Apple, it isn’t about the hardware performance and capabilities, it’s about control.

Yeah, I was pretty militant about it before, but the horse is out of the barn and no amount of head-in-sand will make it go away.

In favour of what? The kernel itself is now worked on with ai tooling. I don’t see how you can get away from it…

That was part of the quote from the article, so I’m not sure why you’re responding to me about it.

These articles are meant to be read, that’s the premise of posting them.

Flatpak is a mature project and the best option that exists on GNU/Linux today, but its security and sandboxing features still trail behind those of well-funded proprietary platforms such as Android and iOS.

without traffic, that’s just doesn’t work well enouhg

Yeah, that’s fair enough for most of us.

Openstreetmap data isn’t nearly as complete as google maps, to be sure. However, like a number of other FOSS and community-driven apps, there’s always something I need to either live without (and in my current rural setting, that is traffic data), or I work around it.

One thing comaps is really good at is recording GPS tracks out-of-service, which google maps SUCKS at.

Ordnance Survey looks great! Too bad I’m not in Britain.

deleted by creator

Traffic information? Did you not read the article?

This was a response to an earthquake. Not only no traffic, but no infrastructure whatsoever, including live data.

It does seem that we have hit diminishing returns when it comes to a lot of technologies.

But that isn’t a function of technology itself, but of our approach to it once it becomes widely adopted.

Yeah, really.

This person doesn’t realize how public pressure made nuclear disarmament and transparent inspections happen.

“Within their legal toolbox to take them” and “practical ability to take them” are wildly different.