wltr

If so, I’d rather keep using it with like 10 different accounts (to overcome free limits, when needed) in Firefox containers.

Haven’t used one myself for years (close to a decade). Installed it for a relative about 5 years ago, never maintained it ever since.

What’s wrong with it?

It worked pretty well on an ancient PC which was running some Windows 7 if not XP. Can’t remember really. The relative is about 80 years old, so all he needs is a browser. So, Chrome OS came naturally. The hardest part was, for some really stupid reason Google wants Google account password to be entered upon booting, and not some other password. PIN code didn’t work for us for some reason. The solution I took is we changed the password to his birthday (perhaps with some A letter, if it wants at least one letter to be present). The password included dots, which was trivial to enter with a Numpad. Like A1945.09.05. But personally, I just hate it. There are use cases when you can allow a computer to have no password. Here, Google forced us to use less secure password, out of convenience. I’d prefer to have my Google account having stronger password, and forcing no password of my computer at all. The potential security risk is someone breaking into the house, and surely they’ll be very dumb to steal that computer, to have … what? YouTube history of some old fart? But that’s a bit of a different story anyway.

Me, I’d rather go with some very minimal distro and maybe even kiosk-mode browser, if necessary.

Still, what’s wrong with ChromeOS? Did I miss something important? Beyond Google dropping ‘don’t be evil’ obviously.

Take a look at the immutable distros like Fedora Silverblue. It would install updates automatically, and has the ability to always rollback to a working version. I haven’t used it long enough to have version upgrades tested. Perhaps it asks for user input. These upgrades happen twice a year.

If I was doing that these days with my current skills, I’d install some minimal version of Arch Linux and probably would remote into it once in a while to update, or invent some simple script to do the updates unattended. The lesser the packages the easier the whole task.

Also, don’t forget there’s Chrome OS which you can install on a regular PC. (It was called Chrome OS Flex last time I did that for a relative.) It’s the easiest I can remember right now. That’s for situations when all they need is actually just a browser. For those cases Chrome OS shines.

I keep repeating this for a decade. Buy a used old iPhone (like SE) and get a no fancy phone which is mostly fine for that use case.

While I have no respect for Windows people, it’s interesting to read through their failures. Yeah, do Windows instead of spending bits of your time to make an effort at learning something new.

I mean it, in a non-sarcastic way. You can start with Windows, and if you won’t give up on this hobby, I bet you’d come to some open source system instead at some point. After all, the entire self-hosting point is not in ditching Windows, but ditching proprietary thing corporations lure people to use, to farm their data and money too. And attention, not the least thing. It’s just that Windows is precisely the very thing a self-hoster would despise.

Having one to boot into ‘launch that game’ mode makes sense to some, but running it to run some services 24/7, makes little sense, if at all.

Well, or cheaper in some regions. Considering the RPi is sold with an extra fee in many places, while Mac mini is just an old outdated piece of hardware that cannot run its native OS any more. Unless you don’t care and use the outdated version. Most people with Macs have no idea Linux exists, and perhaps they aren’t aware Windows can be installed on those.

And saves battery, don’t forget.

A good thing in my book. Intel models are getting cheaper now. (Me eyeing a Mac mini, or a couple even.)

He is ham.

They understand that’s their position is by chance. In a parallel universe with the exactly sent things happening, there’s still a chance they’d have nothing they have here. They’re terrified of losing what they have, not understanding that would make them free, losing everything. Plus, I genuinely believe they’re simply dumb enough to not understand many simple things.

Gabe is just such an amazing pun, if we can call it that way.

I found the post to be useless. As a cherry on top, empty blog. I wonder who are the people who have nothing to say, then they post like one silly piece of word salad, then they don’t post for years again. I’d prefer reading someone who actually posts regularly, and is able toto communicate their mind clearly.

Here … I mean, why did you start so many projects, and why did you abandoned them? I too have an ADHD, like many others these days, I believe. Yet, for me, a GPT allows make progress when I’m stuck and don’t know where to start to continue working on my abandoned project. The massive improvement to me is the ability to communicate. To have my mind written, the problem clearly formulated. I have never paid for any subscription and I would never pay them. It would be less convenient when this shit won’t be given for free to folks like me. By that point, I expect I’d just run some less powerful local model. The value I get is from the ability to communicate and have feedback, sometimes (oftentimes) on topic. Yes, it’s slop. The longer the chat, the sloppier it is. But I usually have up to 10 messages conversations, and I use it only in my browser, for like an hour or two a day, tops. Most times, it’s much less than that. About 4 hours a week, summarised.

I usually hit the free limit with Claude and that’s where I stop. I almost never hit that limit with ChatGPT for some reason. Weirdly, Gemini was absolutely useless to me. I tried it thrice and never opened it again. I have no use for it.

I am not that good to juggle gazillion of projects and languages too. So I don’t. I work on one project, sometimes two or three, but they are highly related, they are almost the same, or solve the same thing from different angels.

I’d love to discuss it, but this particular piece has zero value. To me, there’s not much to discuss.

And, it’s irrelevant to my ADHD. Perhaps because I actually do something with it, and I have no Instagram, Facebook et al. I don’t doom scroll, so the dopamine is here only when I’m able to write some tiny helpful script with one prompt. When things go into some complex work, a GPT always produces incredible useless shit. So, I’ve learned there’s no point in using it that way. So, I use it with very small tasks only.

There screens simultaneously, I can only laugh at these delusional folks who believe they’re being efficient.

Does anyone have any notes on how to self host it and ideally some feedback on what it is to self-host it?

Ok, that’s the first thing I bookmarked. Whatever that bookmarks functionality is.

Wow, that’s impressive! Does it help you find any new music by the way? I thought whether it’s viable to replace a streaming service with a locally hosted music. With some scrobbler, like lastfm or similar.

Lucky you! I forgot even the name of my account! Not to say password. Heck, I won’t recall which email is there! Almost no chance to find it, as it was long before I started using not only password manager, but even a paper notebook. All my passwords were generated funny nonsense a teenager would make up. I juggled emails too, having some ‘I don’t care about this one, and that one too.’

You need to go to the extension settings. It’s not really obvious. There’s a checkbox for sync.

I may update my comment in a few hours when I’d be near the computer.

Always has been!

Literally the thing I cannot understand. They could just do their own electricity and even share back.

I searched for what HDMI-CEC but it’s not very clear to me. Does it mean that, say, if I have an HTPC, and if I run Kodi, I can control it with a regular TV remote? Should this thing be on a TV too? Would appreciate someone with the supported devices to comment how it works and how you use it.