It isn’t just one thing. The big money wants to present this unified front to the public like LLMs are a single commodity anyone can use. In reality they’re a collection of complex tools that few can use " correctly" and whose utility is highly specialized for niches those few find valuable.

So you’re correct in a way. I’m sure model decoherence isn’t helping much either and isn’t as visible in those niche applications as it is for the general public.

Near as I understand it: years ago some dumb engineering decisions were made, acknowledged, and corrected. Is there some recent scandal I’m out of the loop on?

Sure! That’s an SMTP Relay. A lot of folks jumped on the poopoo wagon. It’s common wisdom in IT that you don’t do your own email. There are good reasons for that, and you should know why that sentiment exists, however; if you’re interested in running your own email: try it! Just don’t put all of your eggs in one basket. Keep your third party service until you’re quite sure you want to move it all in-house (after due diligence is satisfied and you’ve successfully completed at least a few months of testing and smtp reputation warming).

Email isn’t complex. It’s tough to get right at scale, a pain in the ass if it breaks, and not running afoul of spam filtering can be a challenge. It rarely makes sense for even a small business to roll their own email solution. For an individual approaching this investigatively it can make sense so long as you’re (a.) interested in learning about it, (b.) find the benefits outweigh the risks, and (c.) that the result is worth the ongoing investment (time and labor to set up, secure, update, maintain, etc).

What’ll get you in trouble regardless is being dependent on that in-house email but not making your solution robust enough to always fill its role. Say you host at home and your house burns down. How inconvenient is it that your self-hosted services burned with it? Can you recover quickly enough, while dealing with tragedy, that the loss of common utility doesn’t make navigating your new reality much more difficult?

That’s why it rarely makes sense for businesses. Email has become an essential gateway to other tooling and processes. It facilitates an incredible amount of our professional interactions. How many of your bills and bank statements and other important communication are delivered primarily by email? An unreliable email service is intolerable.

If you’re going to do it make sure you’re doing it right, respecting your future self’s reliance on what present-you builds, and taking it slow while you learn (and document!) how all the pieces fit together. If you can check all of those boxes with a smile then good luck and godspeed says I.

The features would break if they were built in.

You can’t know that and I can’t imagine it would be true. If the plugins many folks find essential were incorporated into GNOME itself then they’d be updated where necessary as a matter of course in developing a new release.

GNOME has clear philosophy and they work for themselves, not for you so they decide what features they care to invest time and what features they don’t care about.

You’re not wrong! This is an arrogant and common take produced in poor taste though. A holdover from the elitism that continues to plague so many projects. Design philosophy leads UX decision making and the proper first goal for any good and functional design is user accessibility. This is not limited to accomodations we deem worthy of our attention.

Good artists set ego aside to better serve their art. Engineers must set pet peeves aside to better serve their projects. If what they find irksome gets in the way of their ability to build functionally better bridges, homes, and software then it isn’t reality which has failed to live up to the Engineer’s standards. This is where GNOME, and many other projects, fall short. Defenders standing stalwart on the technical correctness of a volunteer’s lack of obligation to those whose needs they ostensibly labor for does not induce rightness. It exposes the masturbatory nature of the facade.

Engineers have every right to bake in options catering to their pet peeves (even making them the defaults). That’s not the issue. When those opinions disallow addressing the accessibility needs of those who like and use what they’ve built there is no justification other than naked pride. This is foolish.

Having a standardised method for plugins is in my opinion good enough, nobody forces you to use extensions. And if you don’t want extensions to break, then wait till the extensions are ready prior updating GNOME.

I agree! Having a standardized method for plugins is good, however; the argument which follows misses the point. GNOME lucked into a good pole position as one of the default GNU/Linux DEs and has enjoyed the benefit of that exposure. Continuing to ignore obvious failures in method elsewhere while enshrining chosen paradigms of tool use as sacrosanct alienates users for whom those paradigms are neither resonant nor useful.

No one will force Engineers to use accessibility features they don’t need. Not needing them doesn’t justify refusing the build them. Not building them as able is an abdication of social responsibility. If an engineer does not believe they have any social responsibility then they shouldn’t participate in projects whose published design philosophy includes language such as:

People are at the heart of GNOME design. Wherever possible, we seek to be as inclusive as possible. This means accommodating different physical abilities, cultures, and device form factors. Our software requires little specialist knowledge and technical ability.

Their walk isn’t matching their talk in a few areas and it is right and good to call them to task for it.

Post statement: This is coming from someone who drives Linux daily, mostly from the console, and prefers GNOME to KDE. All of the above is meant without vitriol or ire and sent in the spirit of progress and solidarity.

Whichever species, if any, rise to sapience after the age of mammals comes to its close.

Would you elaborate on this concern? I’m not sure I understand but I’d like to.

laughs in home lab

Not that I’d buy it but, if I did, that power button might get used twice a year. Likely less since I wouldn’t be able to upgrade or maintenance its hardware.

That’s a problem. Absolutely. It’s not the problem though. I’m not sure the problem can be summarized so succinctly. This is the way I’ve been putting it:

These are the top reasons humanity needs successful, decentralized, open social media platforms:

1. Collecting and selling user’s private data is dangerous and unethical.
2. Using that data to intentionally and directly manipulate user’s thinking is even worse.
3. All of the major centralized social media companies have been proven to either allow these illicit information campaigns or coordinate them directly. TikTok is the focus right now but Sophie Zhang exposed Facebook for doing exactly what TikTok has been exposed for recently. Can you recall any meaningful consequences for Facebook? Do you think Facebook is now safe to use?
4. It’s clear that most political leaders are either too ignorant, too corrupt, or too inept to meaningfully legislate against these problems.
5. The concerned public can’t shut Pandora’s box. No one is coming to save us from big tech or the monied interests and nation-states that wield it.
6. The concerned public can’t easily and legally audit the platforms big tech builds because they are closed and proprietary.
7. Personal choice is not enough. Not using centralized social media increases personal safety but does little to curb its influence otherwise.

These are listed by order of intuitive acceptance rather than importance. I find it aids the conversation.

The best reasonable answer to these problems I’ve seen proposed is for the public to create an open and decentralized alternative that’s easier to use and provides a better user experience.

Will that kind of alternative be a force for pure good? I’m not sure. To your point: I’m not convinced social media of any kind can be more than self-medication to cope with modernity. Then again I’ve had incredible and meaningful conversations with close friends after passing the bong around and spent time on Facebook/Reddit, and now Mastodon/Lemmy/etc, doing the same. Those interactions were uplifting and humanizing in ways that unified and encouraged all involved.

I think the truth lies somewhere in the middle. We need to take care of each other, refuse pure hedonism, and protect the vulnerable (and we’re all varying degrees of vulnerable). At the same time: humans aren’t happy in sterile viceless productivity prisons. Creating spaces for leisure which do no harm in the course of their use isn’t just a nice idea… It’s necessary for a functional and happy society.

Toss a message at Scott Reeder (Scott Prop and Roll). I’d bet money he either knows folks who worked that set or knows someone who knows someone. I’ve no idea if he’d respond but he seems chill like that.

That’s a fair take. Silver Blue is great and, in the spirit of the thread, if I were helping an interested but hesitant lifelong Windows/Intel/Nvidia user migrate to Linux today I would:

1. Buy them a new SSD or m.2 (a decent 1tb is ~$50 & a good one only ~$100).
2. Have them write down what applications, tools, games, sites, etc they use most often.
3. Swap their current Windows OS drive with the new drive and, if needed, show them how and why that works or provide an illustrated how-to (so this choice is not a one-way street paved with anxiety. If they want to swap back, or transfer files, or whatever else; they can. Easily). Storage drives are just diaries for computers. The user should know there’s nothing scary or mystical about them.
4. Install Fedora Kinoite on that new drive.
5. Swap them from Fedora’s custom Flatpak repository to Flathub proper. A decision that should be given to the user on install IMO but I digress.
6. Install their catalogue of goodies from step 2 so they’re not starting from scratch.
7. Install pika and configure a sane home directory backup cadence.
8. Ask them to kick the tires and test drive that Linux install for at least a month.

Kinoite is going to feel the most like Windows and, once configured, stay out of the way while being a safe, familiar, transparent gateway to the things the user wants to use.

My personal OS choices are driven by ideals, familiarity, design preferences, and a bank of good will / public trust.

I disagree with some of Red Hat’s business model. I fully support the approach SUSE takes. I’m also used to the OpenSUSE ecosystem, agree with most of their project’s design philosophies, and trust their intentions. I’m not a “fan” though and will happily recommend and install Silver Blue or any other FOSS system on someone’s computer if that’s what they want and it makes sense for them! Opinionated discussion can be productive and healthy. Zealotry facilitates neither.

That said: Aeon has been out of beta for a while. The latest release is Release Candidate 3 and they’re closing in on the first full release. Nvidia drivers work after a bit of fiddling. 🙂

I’m going to edit my previous post to add the Kinoite suggestion for posterity’s sake.

Check out Aeon and Fedora Silverblue. I’m installing Aeon on Desktops and MicroOS on Servers. My computer needs to be a reliable tool. Immutable distros make it exactly that.

The last thing I want to do in my free time or during my work day is be forced to fiddle with some poorly documented and/or implemented idiocy on my personal computer because I forgot to cast the correct incantation prior to updating something. I’m not a masochist.

EDIT To the hesitant but hopeful Windows+Nvidia user: give Fedora Kinoite a try. Check my reply to @independantiste@sh.itjust.works below for details.