I’m rather concerned about what I do when my Surface Pro 7 dies. I inherited a Chromebook from my dad, but that’s a poor substitute.

Do you need to use Windows? Because any old Thinkpad with Linux Mint will get you through a few years, and performance-wise should be able to handle anything you’re currently doing with your surface. Or hell, I can’t believe I’m saying this, a Macbook Neo.

83% of the searches that happen on PC are done with google and 95% of the searches done on mobile are done with google.

So 12% of users didn’t change the default from Edge with Bing?

As a physicist, my favorite referee comment ever was [That my claim was wrong] “should be obvious to anyone who has ever sat through an elementary electromagnetism course.” He was wrong BTW, and the paper was finally published in a different journal.

I am from a decidedly different field, so I don’t know if I can vouch for you in any meaningful way.

This! If you have good hardware that works, it’s good to keep it if possible.

My smartwatch/activity tracker is indeed a Garmin Instinct 2s with Gadgetbridge. It really does most of what the proprietary app does, and gives you near absolute control over your data.

Yes, that’s reasonable. That’s what e.g. mailbox.org does. And they publish periodic reports on how many requests they receive, how many they successfully reject, and how many they have to follow.

Stock has been continuously going down for at least 6 months.

It is AI, but it’s not GPT.

Not all AI is the same, and we should try not to throw everything in the same bin.

Still, not all machine learning is the same. LibreTranslate and Bergamot (the one used by Mozilla) can run very well on old low-spec PCs that wouldn’t manage to anything will LLM.

I think all machine translators not based on machine are basically dead. I promise that you haven’t used one in more than 10 years.

Do understand that there is a range of machine learning technologies that can be used, and they are not all the same.

DeepL, LibreTranslate or Bergamot (the one built into Firefox) use specialized models that are orders of magnitude more efficient than just using ChatGPT for it. I’d bet this is the same for Google translate, but can’t promise.

I personally use a combination of these three. DeepL I use mostly from FairEmail on my phone, because the app only supports that and it still seems like a mostly reasonable service, and I use the built-in Bergamot of Firefox mobile for surfing.

From my PC (8 year old laptop) it’s Firefox/Bergamot for the web, and I run LibreTranslate locally in a container for my emails (KMail has built-in support) as well as PDFs and documents in general.

LibreTranslate looks and behaves like your typical google translate, and you can try it out the website of the project.

If “never AI for anything” is a hill you’re willing to die on, you can take a look at Apertium. But ML-based translation is almost always way better, even locally on a fairly low-spec machine.

Yeah, Amazon being less than transparent, shocking, am I right?

I think what you’re finding in Amazon are offers from third party sellers. FP6 is 550 on the official site. Still, if the FP4 is 200€ new it is certainly a contender.

I love my FP4, but my manager just bought himself a FP6 and… I’m envious! It seems ti scratch almost every itch I have with FP4

(screen draws too much power and isn’t bright enough in summer, phone is a bit too large, etc.)

Pity my FP4 won’t stop working anytime soon.

(Actually joking, at almost 4 years, no phone has ever survived this long in my dangerous hands, and I’m so happy about that).

I’d give FP6 serious thought. On the other hand, maybe you can get some great bargain on a used FP4 (FP4 isn’t sold new anymore).

Bifteki!

First a technical thing which is not obvious to me.

I understand that the general, non-proprietary Android system service would uses a privacy preserving service like BeaconDB. From what I understand, Google offers an alternative, proprietary, location API in its Play Services. Is that one also prevented from giving your location to Google of you’re using Sandboxed Google Play?

It’s an honest question. I assumed that the provider option I had in MicroG was exactly for that purpose, but I could be wrong.

Next, a small rant.

Bloody hell, I really do appreciate your politeness, but how is it that every damn article about privacy starts with threat modeling, but every discussion about privacy ends with “yeah but if your threat model does not require QubesOS you’re doing it wrong”?

(I use Arch BTW)

I use Arch, BTW.

GrapheneOS requires I think a few other things, like the possibility to completely disable the data lines in the USB port, and a bunch of others.

The problem with Fairphone is that they have rather high demands (e.g. long term support for hardware, better production practices) but they are a rather small outfit, so the default answer from parts manufacturers is “talk to the hand”.

As they grow they’ll become more interesting

I would argue that Fairphone with /e/OS is a combination of a committed/sustainable manufacturer with a trustworthy OS. It does not actively spy or screw with you, and it tries to prevent snooping in many places.

Of course it’s not nearly as security hardened as GrapheneOS, so that may be an unacceptable compromise to some.

I wouldn’t say that it’s pretty similar.

They have rather different goals and feature sets.

Sent from my FP4 with /e/OS.

If you want to convince GOS users to occasionally drop their sense of superiority, it looks like you have your work cut out for you.

My wife’s previous phone was a Moto One Hyper. Pretty well made device, especially for the price.

Thank you for reading that, and for supporting the idea that these topics are worth discussing about, and different people can reach different conclusions.

Also notice a couple more useful posts in the responses to my post, courtesy of a users who decided to verbally disagree instead of just downvoting.