Hacker Newsnew | past | comments | ask | show | jobs | submit | asdfsa32's commentslogin

Models have no understanding.

That's a bold assertion. What do you mean by "understanding"? This paper suggests exactly a symbolic of grasping how something works, which is pretty much the definition of understanding.

Models have an understanding that will surpass any human form of intelligence or emotional state

There are some browsers but nothing with JIT, Apple doesn't want to allow JIT because it makes it easy to bypass the operating system restrictions.

They don't want to allow JIT because it lessens the monopoly of the App Store.

For all the hacker news users talking about apple's monopoly, please first talk about Google's monopoly controls of the web. After all apple is at least not monopoly on the entire smart phone market while Google is the web and half of the smart phone market

But hacker news users won't do that because they complain about Apple not because they are close or monopoly, but they make squeezing money from web users a little bit harder


I don't disagree with you, but that goes both ways here too.

Chrome isn't exclusive to one platform, and conversely Android doesn't force you to use it across the entire operating system.

If Chrome is forcing Web standards even in places where no one wants them, the equal and opposite problem is that Safari has a long history of holding back Web standards that are inconvenient to them, and making the entire industry bend around them because their users don't have the option to just install something else.

Part of the problem is actually that Apple also won't really compete here. They want the vertically-integrated everything, and so WebKit everywhere other than their devices is a second class citizen. The support for its underlying stack anywhere else -- from Linux-based browsers like GNOME Web to runtimes like Bun using JavaScriptCore -- has come pretty much entirely from third parties, and there doesn't really seem to be any sort of toolkit analogous to the CEF that could be used to embed it into other applications.

If anything, the only viable effort to make WebKit available across platforms was Chrome.


> If Chrome is forcing Web standards even in places where no one wants them

Chrome isn't forcing web standards. Chrome is forcing Google's own APIs, and calling them Web Standards


You're not wrong, but think of it this way:

As long as this is the current state of things with WebKit, the counterweight to that is that Apple ultimately gets veto power over any standard the rest of the industry does agree on, and random HTML and CSS features will sit functionally unusable for years because a single OS vendor doesn't see them as a priority.

The problem that we're both ultimately getting at here is that one vendor shouldn't get to have a chokehold on the spec. With Safari it just happens to be closer to what was wrong with IE.


>the rest of the industry does agree on

The "rest of the industry" is functionality just Google. The only reason Google pretends to care about "standards" is because of the large amount of iOS users on WebKit. That evaporates when Google, via Chromium has complete control of the web platform.


> The only reason Google pretends to care about "standards"

Honestly, they are already past caring. web.dev often posts celebratory articles about "standards" already enabled in Chrome as fait accompli.


> As long as this is the current state of things with WebKit

Firefox also rejects the Chrome-only non-standards

> ultimately gets veto power over any standard the rest of the industry does agree on

Where "rest of industry" is invariably "we build from Chromium source code with very little changes".

Apple very rarely vetoes standards that non-Chromium browsers (aka Firefox) also agrees on.

Edit:

> and random HTML and CSS features will sit functionally unusable

HTML and CSS features Apple is usually very much behind (as in: support them fully), and ships them very much on time with everyone else. They are often held back by Apple's own inane "release yearly" schedule.

See interop 2025 vs 2026: https://wpt.fyi/interop-2025?stable vs https://wpt.fyi/interop-2026?stable and then click to "experimental"


The poor adoption of PWAs is part of that disease.

I mean, Web Workers is something only Google could love. They could have provided an API that would let you give commands to the cache layer of the browser and it would have the speed and reliability of native code behind it and if the system wants to evict one piece of data from the cache to deal with storage pressure... it can. Instead they invented a Rube Goldberg machine which doesn't really do anything and you have to write another Rube Goldberg machine inside the first Rube Goldberg to get "offline operation as a service." And you've got to deal with all the async programming difficulties and unreliability and I know for years all the storage operations you could back it with are awful.

Like I'm the guy who signs off on the ticket being done and I've been the guy who answers the pager when it goes off in 2 in the morning and even when I'm getting a stage demo ready for somebody who has the improv acting and stage magic skills to make a 20% done app look like... magic I have to get that 20% functionality working 100% in some certain sense and the Web Workers API screams "you're going to mess around with this forever and it's never really going to work right" whether the product is under or over baked or just right.

So the discussion about PWAs is another one of those discussions that SV has with itself without involving the other 99% of software developers and web users. Google might have had a lot more luck pushing it had they involved a focus group of people in Minnesota or Dublin somewhere.

And of course that 1/3-baked notifications API just raises my hackles, like the first thing I am going to do with a new web browser is turn it off. I mean, every web site is already trying to spam me every way it can possibly spam me, the last things I need are another pop-up menu that covers content, another dialogue of any kind to close, and the possibility that I'll click the wrong button and get spammed.

(And in the "iOS sucks" department, notifications are a big part. I've found that if I have any kind of filtering or groupings I lose important notifications, I will find out i got a text message from my son a week after I get it, etc. To have a decent experience I have to turn filtering/grouping off and give notifications permissions to the few apps that I need notifications from)


It's possible to be critical about more than one thing at a type. But to hand it to google, chrome won because it was better for users. Webkit only wins because of structural control. Same for android really, you can run a non google OS on most (all?) android phones, for now at least. the consumer still has choice. Being a hacker is about liberty, and a critical part of liberty is the freedom to make choices (even not the best ones, looking at chrome lol).

> you can run a non google OS on most (all?) android phones, for now at least

No, you can't, at least on most phones. Recently bought a new Android phone and it was sad to see how many vendors don't allow unlocking the bootloader anymore.

I've been running LineageOs without gApps for many years now and it is perfectly clear, that all big actors are actively working against that choice:

- Google forbidding installation of apps outside the play store - Apps refusing to run without Play services, e.g. some banking apps - Vendors forbidding to unlock the bootloader, e.g. Xiaomi in recent years.

It is getting worse on all fronts.


This isn't about Google, it's about Apple. And it's not just about browsers, it's about JIT too.

Consider that if the web was created today, it would never be allowed to run on iOS. The next web-like thing won't be allowed run on iOS without new lawsuits.


Apple makes hardware and software. And in comparison to the competition, they do both fairly well. Google only cares about one thing: collecting information. so that they can sell to their real customers.

If Apple did not make their own hardware, well, obviously they wouldn’t be in the position they’re in. if someone out there starts making their own hardware to go with their own operating system, they also can be in the same position. Its very simple, if you are a vertical computer company, you get to call the shots.

Apple would not be here today. If they had not made the decision to design their own hardware, just that simple, and it will be simple for someone else in the future who does the same thing because it will be the only way you can carve out a market and survive there used to be many vertical computer companies in the 1980s and Wintel put an end to them.

I wish SGI, Sun, Next, Digital, Acorn and Commodore Amiga, were still around some of them, because I can now afford to buy them.


>For all the hacker news users talking about apple's monopoly, please first talk about Google's monopoly controls of the web.

No, Google does not control "the web". You're trying to divert the conversation to Google instead of the topic of Apple's abusive business practices.

Apple doesn't even make Safari for any other platform anymore. They used to but they decided to stop. They could compete with Google's web browser, but they don't. They prefer to stay in their closed ecosystem, not even trying to compete and forcing other browser makers to use their browser engine.

Maybe we should talk about why Apple doesn't make Safari for other platforms and their abusive business tactics, instead of whatever Google does.


Google does control most of the web as an app platform and that's how most hackernews users use web to make money. Safari is one of the last reason google does not control all. and no matter how abusive apple is, they are not monopoly on a critical part of human being's daily life. People can just buy an android (or even linux) phone and call it a day. Hence a person can avoid the whole apple ecosystem easily. That's not true for google. If you cannot see that then there's not point to continue the discussion.

edit: let me rephrase: apple is easy to avoid, google is not. Hackernews users disproportionately discuss about apple's "monopoly" and closeness not because they think it's morally wrong, but because it reduce the easiness to squeeze money from users.


> and no matter how abusive apple is, they are not monopoly on a critical part of human being's daily life.

I can't think of a single way that Google has made life harder or worse for me.

And Google is very easy to avoid. It hardly takes any effort at all. You just have to buy an iPhone where Apple is the monopoly, forcing you to use their browser engine whether you want to or not. Apple also stifles innovation in the W3C, actively making people's lives worse just for their own greed.

>If you cannot see that then there's not point to continue the discussion.

This pointless internet interaction is over.


Can we talk about the mating habits of spotted owls in Nebraska instead? I think it's more relevant to the topic at hand.

You say that like we can't walk and chew gum at the same time.

This article is about Apple, so we'll yeah talk Apple. Other articles are about Google so we yeah talk Google. It's not some conspiracy, it's just how topics work.


No, we can't. Talking about apple's monopoly without goolge's is hypocrisy.

I'm not sure why JIT is okay on desktop, but not okay on mobile. Is the desktop any less important than mobile? Does Apple not care if JIT is used to exploit desktop computers? My mobile phone is way more likely to get lost or stolen than any other device I own, so I don't keep anything of value on it. I simply don't worry about JIT exploiting it as much as I do someone at a coffee shop walking away with it.

> Does Apple not care if JIT is used to exploit desktop computers?

Desktop has a looser security model due to inertia, it also has much more RAM to spare so the overhead of running multiple browser engines is usually deemed acceptable; iOS still supports devices with 4GiB RAM.


Why is this down voted? Zen looks interesting on the surface, specially as Firefox is the last frontier of "Open Web" with Chrome and Safari being neutered by their respective corporate overlords and their closed gardens interests.

Zen has some controversy a year or so ago related to security and privacy. Can remember the outcome but a lot of people here weren't happy with the dev response to it.

Why bother with a low effort comment like this, all it does it stokes unfounded fears? You've done nothing to share the actual controversy (if it can be called that).

https://github.com/zen-browser/desktop/issues/5947 is this what you're referring to?


Are you calling the comment I was replying to 'low effort' or my comment?

If it's the parent I don't understand what fears it's stoking.

If it's mine, I still don't understand what fears it's stoking. And who are you asking regarding bothering with my 'low effort' comment? Yourself? In which case why are you bothering? Or me? In which case are you asking for a source?


It was directed at yours, because saying "I recall a privacy issue...I don't remember what or how it resolved- but people weren't happy about it" is vague privacy fear-stoking. To rectify that, I posted a plausible privacy issue that was brought up and asked for clarification if that was what you were referring to.

I assume it is downvoted, because the author explicitly mentioned Zen and described why it didn't suit them?

Eh, Firefox can also feel pretty neutered at times. In some aspects, Safari has even done better on the ‘Open Web’ front recently; I guess having several competing corporate overlords is better than having just one.

That said, I fully agree on the importance of Firefox. Avoiding Webkit/Blink monoculture is crucial.


Yes. Firefox is the least reliable browser still. It has the most crashes, seems slower, and chokes on certain content. It’s better than being spied on but not by much. I know I’ll get downvoted for disagreeing with the Firefox glazing hive mind but I’ve been developing for web browsers for almost 20 years. I’m not just hating on Firefox. It’s okay. Not terrible. Not great.

I don’t experience any such problems with Firefox. It works, it’s pretty fast and I don’t remember it crahsing on me in years.

Last time I used Chromium it rendered my entire desktop unusable after a reboot. I had to spend a while tracking the issue, only to find out it somehow messed up the font cache, which made a lot of other programs crash.

Safari, I have no personal experience with. Technically, I think all of the major browsers are comparable. It’s the care for privacy, freedom, and the open web which make Firefox the obvious choice, in spite of Mozilla’s recent direction.


Basically, bit-rot-ly missed up by deleting old link names when deleting accounts; people now register old links and redirect them to wherever they want.

Good memories, but I am glad to see that X11 is now well and thoroughly legacy. Wayland and Pipewire has been thoroughly amazing by just breaking clean with the past. iw has done the same to a large degree. While systemd has its short comings, systemd-networkd, journald, and so on are now providing a cohesive system that is well aligned, even if not perfectly aligned with ones' expectation, I don't have to tinker to make systemd-networkd to work with journactl or the likes.

We just need someone to break away clean from the cancer that is Gnome/GTK and friends. Qt and KDE is does not count, they too are very encumbered with legacy. Yes, it will be a grind to get anything working without those, but I could live with a gtk/kde shim a la xwayland.


X11 is absolutely NOT legacy. It's still better in any relevant way than wayland. Freedesktop just wanted to kill it for weird reasons. But Xorg forks like Xlibre are very active and there are even new implementations like YServer. I still use X11 for every new setup.

The great thing about Wayland is that if someone comes up with something better, this time around they can just implement that without the friction coming from X Window System's custom driver layer.

The Wayland ecosystem just runs on top of standard kernel interfaces and that's a really huge win for any future protocols.


If. They weren't able to do anything better.

> Freedesktop just wanted to kill it for weird reasons.

I wouldn't wish to defend them, but by all accounts the biggest reason is because Xorg's code quality is quite poor, which is (if true) somewhat fair. I wish they tried replacing it without breaking everything and shipping the terrible design that is Wayland, but the goal of replacing Xorg is reasonable.


No, that is a straw man argument; metux already debunked this if you go through all the comments he made in the last ~5 years, before Red Hat banned him:

https://github.com/X11Libre/xserver

The reality is much simpler. The parent corporations, mostly Red Hat really back in the days, wanted to have more control. And, in some ways, wayland is indeed simpler as a protocol.

Also, you write "replacing Xorg is reasonable". First, I don't think it is, but even more importantly than that, wayland is NOT a replacement. It is another display server and protocol. They do not have as goal to be compatible to Xorg at all, so "replacing" is the wrong term here. They indeed broke a ton of things, or rather, required of downstream applications to be rewritten - but the statement that they aim to replace Xorg is not the case. They are simply another display server + protocol at the end of the day. And they won't be a full "replacement" feature-wise either, that's why using wayland always feels like a dumbed down variant of xorg-server. Admittedly wayland also has a few new features such as the easy restart, but this may come to the new xserver too eventually. One has to wait and see. (The easy restart variant is something I like in wayland; hopefully xserver will have that too eventually. It would be technically possible. Also, xserver already is different to xorg-server, in many ways better too. Neither wayland nor xserver aim for 100% compatibility with the old, venerable xorg-server anymore.)


poor code quality has something to do with it, but is a bad reason to kill the whole protocol and ecosystem. And now that whe have LLMs it's easier then ever to clean that up. (and wayland has still not caught up after all those years)

The protocol was dead about 2-3 decades ago, no one actually did X11 for a longggg time, the only few real X11 apps were xeyes and some weird apps like that. No one did X11. Mutter, KWin, Picom, Compiz and so on did the painting, X11 just provided an insecure input and display protocol.

People who say these things about X11 and anything non-systemd sound like out of touch newcomers unaware of the BSDs or mainline Linux distributions that haven't gone along with Red Hat's directions. There is really no issue with X11 that isn't solved with a few options set to the X executable or using something other than startx to initialize, but I think OpenBSD made the right call just forking it.

Every single thing under the Sun going into systemd is going to cause issues (and probably already has with age verification laws and such) with compliance where a company is a singular target (Red Hat under IBM) to compel to make undesired changes contrary to our interests as users.

Red Hat is probably an even bigger problem than GNOME.

On the topic of WMs, some of these classics are really great. Openbox is a favorite of mine among the floating, but bspwm has been my overall favorite for 11 years.


I don't think I'll switch to Wayland for a few years, until it's usable. They don't have a portable way to script it yet, like wmctrl or something. Can't get anything to work there.

I don't see wayland ever being really usable. Wayland is now as old as Xorg was when wayland started because X11 is "old and crufty" and still not usable. And the "clean" protocol is simply lacking so again we started stacking whatever shit on top like compositor dependant extensions for screenshots or screensharing, desktop portals etc. and are STILL missing features like control over windows placement....

> Wayland is now as old as Xorg was when wayland started because X11 is "old and crufty" and still not usable.

How is this comparison supposed to make any sense? When Xorg started it was a fork of the existing XFree86 codebase (due to XFree86 changing license to one not deemed open source compatible), not starting from a clean sheet of paper. And XFree86 in turn wasn't a clean sheet implementation of the X protocols either, it IIRC started as a bunch of patches on top of the reference X11 codebase which goes back to the 1980'ies.


this is counting the xfree86 legacy. That one was started in 1991 and 17-18 years old when wayland was startet in 2008 which is now 18 years ago. Counting the time before xfree86 doesn't make much sense. While initial implementations of X where older than that, those early implementations don't really have much in common with todays Xorg. And the problem is, that wayland is still much less mature than Xorg was when wayland was started.

[different person] That's even worse than I thought. I assumed XFree86 was 1986, and I don't think I heard of Wayland until right around a decade ago. I know I didn't think it started until around 2011/2012.

The big problem with Wayland I see is that every small app needs to do everything (fonts, ...) which can be a lot of startup overhead...

(I'm 100% not a fan of using a single process for all terminals).


tbf in reality nobody uses X11 corefonts anymore. Even on X11 all typical toolkits like GTK or Qt have been doing their own client side font rendering for years. (but X11 still supports server side fonts if you want to make trivial clients)

I don't miss the font mismatch, partial Unicode, and the wrong mapping of ÅÄÖåäö of X11 servers...

There's so many irritations with using Wayland that I continue to strip it out of installs. Not being able to correctly use push-to-talk in voice chat apps because of window security is infuriating, and one app crashing still takes down the entire desktop.

I'm afraid you vastly underestimate what it takes to produce a competent GUI toolkit. Performance, portability, layout control, theming, keyboard control support, accessibility for things like screen readers, and, of course, a sane API.

And I don't mean the code, I mean the design / architecture / UX expertise.


In what sense am I underestimating it?

> Yes, it will be a grind to get anything working without those

Those being the GUI toolkits like Qt and KDE one.

But yes, I agree, I have done UIs in low level (Skia) and high level (web, android) and I know it is a lottttt of work.


> In what sense am I underestimating it?

Given the design complexity you're aware of, the need for practical testing to iron out the kinks, and the burden of converting anything important to a new GUI toolkit, I find the chances of this happening rather slim.

On one hand, AI tools can help develop the toolkit faster and easier. On the other hand, the same tools just make the task of dealing with GTK and Qt easy enough for many not to bother, while enjoying the network effects of relying on a large and established project.


> I'm afraid you vastly underestimate what it takes to produce a competent GUI toolkit.

WxWindows, FLTK, Motif, Athena, XT. Even QT and GTK were "competent GUI toolkits" but they reinvent themselves every other year.


So much emotion here: I'm not sure why Wayland enthusiast have to be so negatively disposed towards X11. I enjoy new ideas as much as anyone, but I also respect what works. Build something better, and people will switch.

As for X11, its network transparency is something I used a lot in the past and relied on. I miss it.


You can use waypipe.

Wayland was developed by people who got sick of X11. No one who actually dealt with X11 is missing it, most people who complain about wayland and yarn for X11 just miss features that took a while to get right.


Waypipe is quite janky, compared to forwarded X11. On local-only network (host <-> container), Waypipe feels like it's in a different region, whereas forwarded X feels instant, as if the network simply weren't there.

I'm currently writing an audio service and pipewire is definitely not amazing. Its an improvement to be sure. But audio on Linux is still a shit show compared to basically any other os

> We just need someone to break away clean from the cancer that is Gnome/GTK

Can you explain why Gnome specifically? Genuine question. I know Gnome is very opinionated, many people don't like it, and these days there are super polished projects like Niri. But in what way is Gnome holding us back?


Gnome has a history of not playing ball with the rest of the ecosystem. They often hold back very simple things hostage (server-side decorations as an example).

Ok but in 2026, it’s pretty easy to avoid Gnome if that’s not your cup of tea. Doesn’t mean it doesn’t have to exist.

Barely, we still have a bit to go for a GTK free system, but it is approaching it.

Isny GTK way less opinionated than Gnome though?

Since GTK is mainly (only?) developed by Gnome it's become more and more tied to its design philosophies.

It is a specially bad with libadwaita.

I don't wish GTK free, except Gnome specific parts (client side decorations disabled, at least by default, etc).

You don't have to use or make software that uses server-side decorations if you don't want to. You can make software that won't work on Gnome if you want to.

> But in what way is Gnome holding us back?

I can't count the number of times I've watched people experience Linux for the first time via Gnome, and become convinced that Linux is weird and unworkable. Thankfully, those days have mostly passed, and vastly more friendly, conventional and appealing desktops are now often the first experience someone is likely to have.

I'm not a Gnome hater. I've deliberately chosen to use it at times. But I've been at all of this a while, and I understand what I'm dealing with, so the Gnome papercuts and "opinionated" behavior are things I can cope with. "Coping," being exactly what I feel like I'm doing when using Gnome.

Anyhow, more on theme with regard to lightweight DE's, my current go-to is openbox, which I use on auxiliary displays for game development. It starts faster than I can lift my finger from the enter key and neatly manages windows. That's about all I want from it and it pleases me.


> I can't count the number of times I've watched people experience Linux for the first time via Gnome, and become convinced that Linux is weird and unworkable.

Do you have a suggestion for a modern distro/DE that's still non-surprising to new users?


Mint Cinnamon?

My wife switched from Windows 10 to Gnome 3 without too much fuss when her computer died witohut warning. She needed a laptop that day, the only one I had available was my spare thinkpad I just created an account, a new luks passphrase and she just started using it. Sure at first she had to learn a few things but nothing that lead to swears. My daughters have been using gnome 3 since they were of age of using a computer and like it much more than the windows at their mother's place.

Grumpy luddites gonna be grumpy luddites but I am pretty sure that Gnome is more approcheable than windows and macOs for the newer generations who are more accustomed to mobile interfaces since they are kids.


My gut is that it's both "opinionated" AND pushed in a "lopsided" way, much like Omarchy?

Which is, there's this tendency in free/open source to presume a level playing field, and may the best piece of software win, but that's not how it works today; both Gnome and Omarchy are pulling in a lot of money and that gives them eyeball leverage, as I understand it.


> GNOME 3 is what happens when you follow the Apple designer-is-always-right philosophy without Apple-level talent.

https://news.ycombinator.com/item?id=49833842

But to explain further, Gnome is run by people who have the attention span of a dead goldfish. They're constantly changing things, a good example is the control centre where it used to be just like other systems but in Gnome 3 they made it into a wall of incoherent text and disorganised categories.

They often tend to ditch things halfway and create new ones, Evince -> Papers, Totem -> showtime, EOG to Loupe, Cheese to snapshot, and so on and so on.

And for decades, they kept removing much used features and configurations and broke things and everytime you search for a fix, there was 10 version and a new one each week because Gnome keeps changing things.

In short, Gnome is what happens when you think you're Apple but have the design knowhow and depth of a paddle you couldn't drown a fruit fly in.


It is interesting you make the observation about GNOME and Apple. In 1999, GNOME had the involvement, and financial resources, of some of those Apple people due to Eazel. Andy Hertzfeld, Bud Trimble, Susan Kare, Mike Boich and more all gave improving GNOME and Linux some work.

This wasn’t a hostile takeover attempt; we all tried very hard to be team players, many times to what I think was our detriment. A very nice version of Nautilus was rejected because it was in C++ and we ended up doing all the work in GTK.

We had some very good ideas about distributed objects but they were pushed aside and we ended up using CORBA/ORBit/Bonobo and those distributed object systems ended up being replaced.

The dotcom crash caused a lot of Eazel people to end up at Apple. Nat and Miguel moved onto their dot NET Xamarin effort and many others ended up at Red Hat. There was an experience vacuum and this is one reason why the GNOME UX had so many issues IMHO.


- We just need someone to break away clean from the cancer that is Gnome/GTK and friends. Qt and KDE is does not count ...

There's a very good alternative in development - check out V OS - https://v-os.dev/ - it is attempting to port the Be / Haiku GUI and SDK on to Linux, and even though it's still an alpha-beta release, the experience is really good. Unlike GTK's "clunkiness", Be/Haiku GUI feels so buttery smooth, and consumes so little computing resource (compared to GTK or Qt). As they aren't re-inventing anything new, I have high hopes for it. It just needs more support from all of us to thrive.


I am going to sound very ungrateful at sight of anything new given the state of affairs, but I will bite. I think UI kits are as much a matter of taste as it is a technical concern and the screenshots provided does not "do it for me", not in that it is missing what I want but it has things that just won't fly for me, for example those hanging title bar controls, I could probably remove them, but it is not a good start.

That's the thing - this is neither new nor just a "UI kit". This is a 30+ years old technology (see Be OS - https://en.wikipedia.org/wiki/BeOS and Haiku OS - https://www.haiku-os.org/ ) ! It is a full-fledged integrated Desktop offering. It has its own display server and does not use Xorg or Wayland. (As for the windows title, it can cover the whole window too - it's configurable). All the extra bells and whistles that we want (and do not want) will eventually come once the product is fully developed ... right now, they are working on porting it and ironing out all the bugs.

Niiice, thanks for the pointer. Rebasing Haiku on top of Linux looks like a solid approach.

Loving Gnome since like 15 years.

Switched for Unity for a few years when I was erring on Ubuntu side, where I like the minimalism style.

Gnome 3 was a shock but nothing any normal human being can easily overcome. The hardest was to live without a taskbar, but I'm free now.

Would never touch KDE-let-s-make-this-pixel-also-clickable, but that's the power of freedom: you can choose what you like.


You should try KDE just out of curiosity. It’s gone pretty far from what it was 15 years ago (which, iirc, was the sad KDE 4 era).

Nothing is forcing you, but you may be surprised.


I really want to like KDE. Every time I set it up on a system I have non-stop issues. Probably mostly tied to Wayland, but either way it's just not reliable enough for me to use. Even tried KDE Neon to see if being on the latest version would help, and that eventually killed its entire install by itself.

> It’s gone pretty far from what it was 15 years ago

GNOME 3.0 was released 15 years ago last April.

https://mail.gnome.org/archives/gnome-announce-list/2011-Apr...

It is exactly the same desktop today. The most visible changes are that the strange floating word "Activities" is now a strange floating pill instead, whose meaning only becomes apparent once you switch virtual desktops a few times, and that the virtual desktop switcher has moved to become horizontal (i.e. worse and less space-efficient on a widescreen) when it was vertical.

I often hear "try it again, it's really changed" but I see no difference. I powerfully disliked it in 2011 and nothing has changed. It's a bit faster, but there are more hamburger menus now, to the extent that apps like Disks has two of them. It's still horrible.

I am all for UI innovation, but it needs to be informed by what's gone before.

I would like the GNOME development team to spend a while learning to use a few pre-Windows desktop environments efficiently and well, to learn how seriously different layouts can work.

Classic MacOS was sublime. 8.1 was a good snapshot.

RISC OS is a great example.

Original Psion System 3MX EPOC is another.

Psion NetBook EPOC32 is another, and very different.

Some Amiga-inspired environment such as AROS or MorphOS.


There is no "GNOME development team", there are just people that are working together to make an DE (and other stuff). You can do the same if you want to.

I don't have numbers, but hasn't modern Gnome been like 90% funded by Redhat and other corps?

Yeah, if you don't know you can always make stuff up!

What do you use from Gnome?

Do you use NetworkManager or systemd-networkd or something else? What is your login manager? Do you use Gnome for email or anything else?

I am asking this in earnest because I don't believe Gnome has left anything useful beyond GTK as the library (if you consider that useful) and maybe nautilus. Everything else is just bolted on and weird and broken.


I'm very happy running Linux Mint with X, and I don't see why it would be legacy or even what Wayland does better. In fact I've read about many problems, and that's why Mint still doesn't use Wayland by default.

For most commercial software vendors that go out of their way to support Linux, it will account for less than 1% of their userbase. It already doesn't make any commercial sense for them to support it -- in a lot of cases it will be the project one or two passionate programmers that ported in their free time, and vouched to support it themselves.

They already have to deal with the fact that if they're shipping binaries that they actually want to work, they have to compile their own compiler, libc, libc++, Xlib, static link everything, don't touch anything the system provides, dynamically load the things you can't get around, test multiple .so paths for different distros, etc, etc, etc...

Now, on top of that, they're being asked to support a second completely different Window System layer, that an even smaller percentage of people use (and who mind you all also run Xwayland).Unlike X11, Wayland doesn't have one de-facto standard implementation, and the spec doesn't define anything - some common functionality is provided through different incompatible extensions, and some common functionality is not supported at all.

I was trying to set up my VR headset for Linux a few week back and I saw [1]. Feature disparity is so bad that Valve, the biggest corporate proponent of the Linux desktop, concluded they just can't support SteamVR on the biggest desktop environment, and basically tells you to switch to X or another compositor.

That's without going into the technical insanity of the protocol, like using sockets and a client-server architecture for something that can't actually run over the network, because Linux developers just can't envision a form of IPC that doesn't pass through the kernel and take 2ms, or the fact that the standard client library can't be static linked by default without you rewriting its build system.

If one day X11 dies, it will be because Linux has settled on the Win32 API, not Wayland.

[1] https://help.steampowered.com/en/faqs/view/18A4-1E10-8A94-3D...


I have recently switched to https://duxlabs.dev/ and my experience is very good.

Would love to, but it seems HDR10 can't be supported?

Anyone who wants me to install my DE via a curl | sh is insane. Next :)

You have other options, in case you didn't look further: https://duxlabs.dev/docs/dms/installation/

You can probably install it via package manager that installs it via curl. If you want

But why would I need a package manager?

I tend to compile everything from source on Linux. I don't want to be tied down to any distribution. So from my point of view I'd agree - I understand why curl/sh is used, it is super convenient, I also use it for installing Rust. But I also dislike it. (Naturally for Rust I need rust for the bootstrapping, so I understand that solution. But I still dislike it too.)


What package manager installs things via curl|sh?

If you want something to happen you can do it yourself.

However, if you are the type of person who calls people who work on Free Software that you are totally free to use or not "cancer" you probably will not get many people who want to work with you.


Pipewire even supports almost everything Jackd does while Wayland barely got remote display and screen recording. What breaking with the past are you referring to in the Pipewire case?

I have no idea what you write, but clearly what you write is absolutely incorrect.

See: https://github.com/X11Libre/xserver

Also I don't understand why you write that pipewire broke with the past. That's also totally incorrect.

Wayland I agree with your statement but that's a corporate-controlled project by the GNOMEy overlords, so this is not a surprise (GTK5 will support wayland only; that was one reason why I stopped writing more GTK code).

As for your systemd comment: again I have no idea what you write. It does not make any sense. And the word "cohesive" is odd: all the systemd components, of course, form a cohesive unit WITHIN ITSELF from the get go - that was the whole point for the systemd developers to make themselves more important. systemd was never a "replace init only" attempt, they deliberately took over the ecosystem here to leverage more control. And that also, kind of, makes sense for those controlling many computers in a, say, university campus or so; administrators find that some jobs were made easier by that. For me, and many others, systemd brought nothing to the table that was needed. They always built strawmen such as "you may abandon your legacy shell scripts booting your system" - I don't use shell scripts in general. All logic I have is done via ruby + yaml files mostly (or, data files that describe the logic in a DSL like manner), so I never had any use case for systemd. As for your claim that you do not have to tinker with systemd: again, I have no idea why you write this. It sounds like a very poor and one-sided attempt to promote systemd. You never had an issue? I had numerous issues whenever I would use systemd, starting from smaller to larger issues. For instance, I remember that my local clock was not well-aligned so I changed it. Then I tested again and noticed the clock did not change. It took me a while to figure out that systemd inserted itself here and ignored my attempt to fix the local clock. Past that point I treated systemd as an invasive virus that attempts to ignore my commands, so off it went. That's just one example of many more issues with systemd. The philosophy the systemd devs use is mega-aggressive and incompatible with many other things, including the old UNIX pipe idea as well as "everything is a file".

> We just need someone to break away clean from the cancer that is Gnome/GTK and friends.

I would not use the term cancer, but I would agree that we could need more flexibility. We simply lack viable options here. Corporations de-facto control most of the code here. Or AI. Not a good future really.

> Qt and KDE is does not count, they too are very encumbered with legacy. Yes, it will be a grind to get anything working without those, but I could live with a gtk/kde shim a la xwayland.

Again, no clue what you write here, but let's also be realistic: which alternatives to GTK or QT do you see coming up here? Plus, we then have more code for a toolkit. What we need is a meta-toolkit, that can be used in its own right (e. g. without qt and KDE) but also be used in way to allow people to create applications in qt and kde at the same time. One toolkit to rule them all. And it must be as flexible as possible. In many ways it is kind of HTML/CSS/JavaScript, de-facto really; would be nice to have a way to have something that could target all of the above, but I think the boat has sailed. HTML/CSS/JavaScript kind of won, and perhaps via webassembly we could at the least use something else other than JavaScript.


So this is how it is going to be? AI written everywhere?

> I read that driver and copied the protocol: which registers to write, in which order, to wake the sensor up and start streaming.

> legocam is GPL-2.0, like the driver it learned from.

Please write yourself.


Well, it makes sense that the author of the driver should write about it.

Not the first time I've seen this, but this one really hit me for some reason.

I genuinely don't understand why someone would want to work like this. There's obviously no practical use in having a 26-year-old novelty webcam work at 5fps. What made it interesting was the process of figuring it out and writing it up, which others could then read and learn from.

Instead it looks like this 'author' just watched an LLM code it and write up for them, and published it unedited. What is anyone getting out of this completely soulless process?


While I'd never let an LLM write a blog post for me, I do "get" having the LLM write the code for a project like this, especially for automated debug cycles.

It's the kind of thing I'd do, and enjoy doing, even as someone who'd previously do it by hand. It's fun to make something work, even if Claude is doing the heavy lifting.

But I find myself in a weird spot now. I've completed a ton of cool projects recently, and I want to write them up in a compelling way but... "Claude did x, then Claude did y" does not make an interesting writeup, and I don't want to pretend it was all me.

I think the effort of understanding a project to the extent that I can write about it well is almost equivalent to just doing it myself in the first place.


I’m in the same place!

An example: https://www.blog.montgomerie.net/posts/2026-11-03-sh-invader...

I put a lot of work into this. It decidedly was not one-shot, and it would definitely be way more bloated and buggier without my work - but my work was obviously way less than it would’ve been a couple of years ago, and most of it was prompting (“Please time ‘x’ and try ‘y’ method for improving efficiency”; “Could we use ‘z’ data structure here?” etc). At the same I probably wouldn’t have even done it a couple of years ago, and if I had it certainly wouldn’t have been as well tested or arcade-accurate.

How much is it ‘mine’? How worthwhile is it? Do I understand it as much as I think I do? The blog post is really about that - not the project.

> I think the effort of understanding a project to the extent that I can write about it well is almost equivalent to just doing it myself in the first place.

I’m not so sure this is true. At the same time, I have learned things from projects I’ve used LLMs to work on - and not just about how the LLMs work or how to use them (though I’ve learned a lot about that too).

I enjoyed reading about this project, FWIW. It’s ‘cool’ to see the hardware being used, and I liked reading the write-up and the code to learn how it worked.


I genuinely don't understand why you would spend time complaining about how someone else uses their free time. What are you getting out of this soulless process?

I think it's a good thing, because using a LLM like this is the first time most people have used a computer as a computer. It's the first time people have actually made something via an intermediary that executes on their computer as opposed to installing an application built by a developer.

In so many ways, LLMs are the computer revolution realized. A lot of people just don't know how to write a driver, what it is, or how to investigate this. I certainly wouldn't. Nor would a lot of trained mathematicians and scientists. It's not immediately obvious how you'd tackle a problem like this to them and it creates this feeling of learned helplessness.

Now, thanks to the age of magical machines. This person wanted a piece of their chidlhood back and voila, they did and now can share it with the world if they so choose.

I do think that people should get better about attributing the machines though. I wouldn't be comfortable claiming ownership or authorship over code written by a machine entity so I do my best to mark machine-authored code as machine authored code.

An example, I know about snell's law and a little bit about optics. Nothing beyond the standard undergrad stuff (right now). I've tried Apple's liquid glass and I didn't like it. I think a part of it is that it's not translating the physics well and most implementations fall flat. It sort of either exaggerates or doesn't. Two nights ago as a passing fancy, I decided to work with codex and gpt to make it into a library that I can use for a larger research project. https://github.com/areoformed/daphnis-glass https://as-we-may-think.halfdrop.workers.dev/examples/as-we-...

It made me happy to watch a machine go off and do checks + validate the equations I gave it and to watch it write,

    > The numerical checks agree with the central magnification limit and your rainbow equation. An important consequence for the UI is already clear: the distance between the drop and the page can change both magnification and orientation. “Refraction strength” alone won’t describe the effect.
But as I did nothing more than write up a bit of insight, prod the machine, and make a detail request, I made sure that the machine entity got attribution for it, "Developed by Codex in collaboration with areoform. MIT licensed."

I wouldn't have been capable of turning that insight / thought process into code before, but thanks to the machines now I can, and I can even share it under a permissive license so that other people can find some small joy in it?

Is it soulless? Perhaps. But consider this, I found joy in turning a passing thought about physics into something that tangible that I can use for myself. That wouldn't have been a possibility for me in November 2020.


> I genuinely don't understand why someone would want to work like this. There's obviously no practical use in having a 26-year-old novelty webcam work at 5fps.

So? Why does everything have to be practical? It can be fun to get something working like this, even if you use an LLM to do it.


>What is anyone getting out of this completely soulless process?

A blog post, which is content. Most bloggers are wannabe "influencers" and "thought leaders."


> It runs at about 5 frames per second. That's the limit of USB 1.1 and the sensor clock, not the code.

I agree. For basically two or three short paragraphs too...

When I scroll through someone's personal blog and every single post is AI written I get that AI nausea/fatigue feeling.


Yes, you're entirely correct. But you're also underestimating the people's potential to squander great opportunities like these. Look at AMD for example, they had a good head start over nvidia yet they haven't done anything remotely as good.

This is kind of the wrong approach when linux namespaces exist and rootless namespaces are approach full spectrum.

The Linux kernel is a gigantic attack surface with a ton of legacy stuff most modern applications will never need. Namespaces do not really help with this, but userspace kernels can. (Haven't looked into Ginux enough to know if it does.)

That is also why I believe the namespace approach, and by extension Docker-like containerization, is a dead end when it comes to isolation. I hope something like gVisor will win.

This is even more important since agents have turned isolation from a datacenter topic into everyone's concern.


> The Linux kernel is a gigantic attack surface with a ton of legacy stuff most modern applications will never need. Namespaces do not really help with this, but userspace kernels can. (Haven't looked into Ginux enough to know if it does.)

This is incorrect. If you can't trust Linux because it is gigantic, you can't trust a whole another new kernel. Userspace is not blackmagic.


Another new kernel might be 1/6000 the size of Linux and therefore have 1/600 times the exploitable bugs—or none, using formal methods. The question is how much a new user-mode kernel can shield Linux from attacks. Usually the new kernel depends on Linux to separate it from the malicious code!

> Another new kernel might be 1/6000 the size of Linux and therefore have 1/600 times the exploitable bugs—or none, using formal methods. The question is how much a new user-mode kernel can shield Linux from attacks. Usually the new kernel depends on Linux to separate it from the malicious code!

Is there any new language close to rust which can help being verified using formal methods or would some other language be more suitable for this job?

Maybe using verus + rust, it seems that atmosphere[0] is an interesting project in this space

[0]: https://mars-research.github.io/projects/atmo/


Of established languages, Rust, OCaml, F#, and Lean seem to be promising for this kind of thing, but people seem to be vibecoding a lot of new languages right now with this use in mind.

You writing that in response to a vibecoded kernel.

No, I writing that in response to "If you can't trust Linux because it is gigantic, you can't trust a whole another new kernel. Userspace is not blackmagic." The nature of Ginux is irrelevant to that proposition, and, as I think I have clearly shown, it is not well thought out.

The context is not another kernel as an abstract idea here, it is specifically a linux userland kernel. You can't take stuff out of context and make pointless general assertions.

Whether your general assertion was pointless or not, it was false, evidently because you didn't think about which facts about the situation at hand were relevant enough to mention. I'm not convinced that you would have made it to a true statement by doing so, and your efforts in the comments above don't seem to be headed in the direction of a correct thesis, but if you bother to explain what you mean in a coherent way, I'll be happy to engage with that instead.

But I'm not going to try to guess what you wanted to say and then discuss that, because I'll probably guess wrong.

Instead of complaining when people point out that what you said was false, you should thank them for taking the time to correct you and take the opportunity to learn from it. Then, be more careful about what you say next time, so that you say things that are true instead of false.


No, your assertion that you can bolt a userspace kernel on top of an insecure and bloated kernel to make it secure is false. A userspace kernel is just a vm like any other, in fact, a less secure vm at best.

So you should carefully read and consider the conversation before you write, least you come across as arrogant and foolish.


I haven't previously made an "assertion that you can bolt a userspace kernel on top of an insecure and bloated kernel to make it secure", but, as it happens, you can; that's how web browsers execute JavaScript, how Cloudflare Workers run Wasm, and how Firecracker runs virtual machines on Linux, for example.

Whether this works or not—or, more defensibly, how great the risk is—depends a great deal on the details.

I'm not really interested in how I appear to you, so I recommend you don't waste your time telling me about it.


But those are not Linux kernels and have very different memory model, hardware access, and so on and so on.

But I don't think this is a technical conversation anymore, you have some other problems to address and it is best you do that.


Actually, Linux kernels are overwhelmingly the most common thing to run in Firecracker VMs, and at any rate that isn't relevant to your assertion that you cannot "bolt a userspace kernel on top of an insecure and bloated kernel to make it secure"; that would still be mistaken even if the userspace kernel did have to have a very different memory model, hardware access, and so on and so on, from the underlying insecure and bloated kernel.

As for whether this is a technical conversation or not, I think there's an interesting split: all of my comments in this conversation have been technical, and informative to boot, while the technical content of your comments has been limited to trivial statements that are easily shown to be incorrect.

Instead, your last two comments have contained personal attacks, which I have so far refrained from. (You may have noticed in the past that, when you repeatedly launch personal attacks on people, they usually respond by personally attacking you too, but I have not done that.) Your latest comment is over 50% personal attack by byte count.

Why is that? Have you considered thinking long enough about one of your comments to post something technically correct, original, or both?


Firecracker is not a userland kernal, it is a type 1 vm since it is basically a lick of paint over kvm, running linux inside firecracker has nothing to do with userland kernels. So I am going to stop at that. Like I said, it is not a technical discussion for you it seems.

In that case, wouldn't it make more sense to build the Linux kernel without those features, or to replace them with sufficient mocks?

You might want to expose more to trusted apps and less to more threatened ones. With kernel shim like gVisor you can have this with little effort.

Ah, good point, thanks!

I’m still fascinated by the approach, though. It seems to me that it sits somewhere between cgroups and virtual machines.

I do wonder what problem it solves, as I don’t think that it provides the same security guarantees that micro-VMs like firecracker do?


It’s the same approach as gVisor. Fairly decent isolation if your application can work on it, but your SOL if a scenario comes up that doesn’t work. VMs give you pretty much 100% Linux compatibility at the expense of density, optimal resource allocation and startup time. gVisor gives you excellent startup time, resource allocation and density at the expense of some app compat and performance at the extremes

Edit: at least with gvisor you get close to a decade of work on compat. gVisor can support some rootless containers scenarios and systemd these days.


I’d be very interested in it if it started to approach VM style isolation without needing KVM access (or qemu).

But I’m not sure if that is really possible.

The use case I’m thinking about is a sandbox exec isolation on an HPC cluster without Docker. I know singularity/apptainer exist, but I’m always looking for something that could make that environment more flexible. It is a tight security environment, so we are always limited in what we can do.


It all becomes a religion to some people. hypervisor escapes are a thing. People always point to Vt-x or AMD-V as having stood the test of time, and that might be true, but the software layer on top certainly hasn’t. Yet you will have “security professionals” tell you that you can’t trust anything beyond a hypervisor. Or they might tell you that a user space kernel is just as secure. They might tell you KVM is a no-go because of bad Type1/Type2 hybrid architecture compared to Xen or HyperV Type1. All have had their fair share of vulnerabilities over the years. All less than Linux process isolation, but still.

I really like Virtualization to keep things separate, but I never thought of it as an impenetrable wall. For example, I'd not trust an AI agent even inside a VM in my own system.

OTOH, running a messy environment with rollback support is much easier with a VM than a container.

Containers are fun for packing messy executables into their own environment and run them now and then, though.


I don’t find VMs to be that much easier to manage (in terms of rollback) than containers. The opposite in fact. VMs are these giant blobs that can be stateful or stateless or a mix of both. While containers have clear separation between the state (stored on volumes or in an external store like a db) and the “application” stored in immutable tags or digests.

Rolling back is easy as long as you understand where the state, if any, is which is actually the hard part of any application. At least with containers you have all of them stored in a well defined registry, while with VMs you would be rolling your own VM snapshot storage system. Not super complicated to do in the grand scheme of things, but it’s another thing you need to get right. But VMs/applications that can rollback would be just as easy (if not easier) to rollback with containers.


>I really like Virtualization to keep things separate, but I never thought of it as an impenetrable wall. For example, I'd not trust an AI agent even inside a VM in my own system.

Remember the submission about how Auto Review is safe? Their standard of safe is that it will take a playwright dependency of another project in your home directory and add playwright based tests to your current project that hasn't declared playwright as a dependency. Amazing.

The issue literally isn't even about a hypothetical security argument anymore. Your sandbox is now a policy that tells the agent what it is allowed to do or not.


> All less than Linux process isolation, but still.

.... Right. Exactly. Everything has some vulnerabilities, but if an attacker has to use a relatively rare VM escape, that's a huge win.


I wonder if this can grow into a capability-based security tool.

You can combine hypervisors, user space kernels and linux namespaces all into one solution. The attacker will have to escape all of them.

Rootless namespaces are getting better, but they don't eliminate the host kernel attack surface

If it has to go through another layer , hopefully without ring0 privs, which more than likely may filter out a number of exploit vectors, how is this the wrong approach ?

Breaking this layer doesn;t automatically give ring0 privs across the hardware.


Isn't this also adding a capability sort of like OpenBSD's pledge?

> Telling someone's research and exploration is wrong...

Never change, HN, never change :D


The project says:

> That research phase is complete.

So not sure what part of it now constitutes "research and exploration".


Please stop using AI to author your articles. It was really hard to read it despite how interesting the topic was. I dropped halfway because it was just too annoying.


I was bothered not so much by the slopful article as by the AI-written text in the dashboard that iKnow produces. I ran the program on my Mac and it did display charts and graphs with names of people and organizations I’ve interacted with over the years, but I couldn’t figure out what it all was supposed to mean.

One example: "323 people, 122 events and other records, 200 employers and 1 stored group — as the store on this Mac links them. Read live just now. Colour is what each node is to you: the stored group in red, family and closeness by iKnow’s rules on the kinship scorecard. Size is how many rows in the store name it."

Claude etc. can be prompted to produce clearer explanations than that.


Yep I cannot fathom why people feel OK shipping anything like that publicly. The headers are also all completely stupid.

> Every node as a list — the graph is not the only way in

> Family, as the scorecard has it

> Reachable, exactly as stored

Oh good I was worried it was going to show me something altered

> WHY RESOLVABLE MATTERS

Oh fuck off. This isn't Axios! I hate these pathetic brainrot stupid phrases Why it matters What Happened What Changed, how patronizing. Instaclosed, instakilled python, what a fucking waste of my time and a waste of an article too


> Six people, six relationships, and a subtitle saying nobody typed any of it.

> a closeness score above the cut · nobody typed it

> A dashed line is a conclusion nobody typed.

> Nobody stored a sleep schedule.

> The hours the log goes quiet - a sleep schedule nobody wrote down

> published by metadata that contains no words.

This is Claude-ism that is particularly infuriating, it never makes any sense.


Yeah seriously. Article becomes unreadable immediately after the intro. I believe this author probably could actually write two years ago, it's a shame people are just giving up on that and choose to turn out stuff like this instead.


I'm having more trouble spotting AI text these days. It used to jump out at me. I'm not sure if it's that the LLMs are better or it's just becoming so normal to see that I stop recognizing it. This one didn't scream "slop" to me other than being quite long-winded for a blog post that I admittedly did not finish on first read. Going back and scrolling deeper into the post, it does become much sloppier.


> quite long-winded for a blog post

That's actually exactly what does it for me. After the intro, it gives you way too much specific detail. No one cares about the exact file path or whether that path is TCC Full Disk Access protected. What does that even mean? Why is this important enough to be in the first sentence of the section? I don't care that there's 51 database files and 181 other things and 278 doodads. Who says stuff like "triple store?" It uses bizarre language with far too much detail, and like you said, your eyes start to glaze over because there's so much irrelevant information. No human would've chosen to write all that tedious detail in their post about something much more interesting. It's bad writing. It's slop.

A human would've said something like "It's a database in <Library/IntelligencePlatform> with three columns that contain..." and left all the rest of it out or maybe in some extreme details footnote.


>this one didn't scream "slop" to me other than being quite long-winded for a blog post that I admittedly did not finish on first read

Maybe you're just looking for GPT 3.5 era tells?

Because this one sure screams AI slop loudly and clearly:

"The identity group is the tidy story: same surname, same home address, an iCloud Family flag — and it is the small half. "

"The graph is the overview, and an overview is easy to shrug at. The pages behind it are not."

"Six people, six relationships, and a subtitle saying nobody typed any of it."

"Below that is the part I find hardest to argue with."

"This is the page's own honest point: the graph points at the face clusters and can name almost none of them (three of the twenty-nine, in the fixture). It does not need to."

"None of this is content. It is all metadata, and it is enough."


As I said, below the fold the tells are more obvious. I didn't get that far on first reading, because I was already getting bored.

I think on the top part he had his self-written summary which gave to the AI to expand, and the second part just had some bullet points or shorter prompt.

My bad. In my defense, only the technical parts are AI. But, thanks for the feedback; I thought I had unsloped it enough...


It's not possible to un-AI text, only replace it wholesale.

Suggestion for next time: "Review my prose for accuracy, grammar, appropriate strength of claims, etc. Flag any issues for my attention. Do not provide text; this work should not contain any AI-written content." That'll turn it into a very good check that doesn't take your voice away.


[flagged]


It's already technically possible for LLMs to generate all kinds of styles of text. In fact older models are better at it. It's the newer models that are nearly impossible to steer. This makes me less hopeful for the future. There is apparently some incentive for Claude and ChatGPT to lock in these very recognizable patterns.


AI writing is fine in a constrained space.

If you create a strict template, like a PowerPoint deck or tightly formatted report, the LM does a really, really good job. at populating that template with good data.

If you give it an open license to just create stuff, it’s always terrible for the general audience — but it may be fine for you or a small group that is familiar with the subject.

I compare it to fast food. An uneducated teenager is quite capable of putting together a McDonald’s lunch within the constraints of the system. If you let the kid loose in a kitchen designed for trained chefs, the result would not be a good lunch.


> but it may be fine for you or a small group that is familiar with the subject.

One of the problems with LLM writing is that the "small group" is very often only itself, maybe including the prompter. Details of the prompts or implementation or internal chain-of-thought nits leak out into prose, divorced from context.

Many READMEs are now littered with proud declarations of the removal of Foo, with any explanation purged with it. It exists in this project, too. A header of one graph is "Where each fact came from, decoded out of the bitmask." Why should the audience care about this implementation detail at this location? Just label it "Sources" and move on.


Yeah, the LLMs “thought” process does not distinguish between internal monologue and presentation well.

That’s why you funnel it into a template.


>Complaints of AI are like a form of friction that eventually goes away.

This. Already there is evolution in AI-related comments here.

For example, a year ago they often featured remarks like "I know AI when I see it: here are unmistakeable 'tells.'"

Then there would be a lively back-and-forth about the validity of the stated giveaways.

Such "How I know it's AI"-type comments no longer appear.


"It's not X, it's Y" is still the most obvious one. Abuse of passive voice is another. We stopped talking about it because it gets boring to say the same things over and over.


This has not been my experience. People talk about LLM-isms (Claude-isms especially) all the time. "The absolute shape", "load bearing", "seams", "smoking gun", genuinely empty preambles, constant sloganesque phrasing. If you're seeing it less in conversation, it's not because it's improved.


Genius-level response. And honestly, that's valid.


[flagged]


I’m sorry you feel defeated, but such attitudes won’t help.


Help what?


Help improve or restore the quality of content.


Been doing that for decades. This is just another tool.

Check out my work. I don’t hide it.


It is funny to me how slop enthusiasts argue from the past and the future to deny the present state.

"Slop is here to stay so you better get used to it because slop is inevitable and you will be left behind."

It is both already a cemented permanent presence and destined to arrive by an unbreakable prophecy at a date TBD.


If you came away thinking I was enthusing about AI slop, it's probably because I was too subtle in comparing it to a food I abhor while saying there are tidbits, non-slop tidbits, that I could appreciate. Apologies for that human generated mischance!


I'm actually rather bemused to see many young folks, arguing against LLMs and ML, and seeing more experienced folks, embracing it.

I'm pretty experienced, and have found LLMs to be quite useful.

Like any new tool, there are deficits and peculiarities, but these are becoming less of a concern, and the utility of these tools is advancing rapidly.


We get it. You are saying that in the future there will only be mushrooms to eat and as a result everyone will enjoy mushrooms.

Now think about how much you hate mushrooms.

Do you see the problem with your argument?


> In my defense, only the technical parts are AI.

A defense would be "I just used it for spellchecking" not "I just used it for the meat and the important parts people care about, the rest is mine!"


Yes, that was a weird defense; essentially, he used it for the actual interesting parts of the article.


So most of it?


So only the parts that matter, you ai slopped. Great work.


You can’t unslop A.I. prose. Once it’s written, the content is poisoned.

“Sure there was a turd in the punch bowl but I did my best to remove it, so you should be able to drink it.”

Nope.

And the technical parts are what matter, saying you did that with A.I. tells me “I don’t care about this part” so what the heck are we even doing here?


Interesting for me it's the opposite. I care about this so I let AI do it since LLM are better writers than the majority of people. I'm not so arrogant to think I'm a writer of that caliber.


You are. I promise you. Almost every single person is a better writer than an LLM. You're wilfully sabotaging yourself.


Maybe LLMs are better writers than the majority of people, but that's only because the majority of people just don't write. If you have any level of competency, you are a better writer than an LLM, because good writing requires empathy. It required thinking about what the audience knows, what they want to know, what kind of writing they are comfortable with, and so on.

LLMs fail at those things.


LLM are terrible writers though. Worse than most humans unless they have some mental disability.


Agreed, I haven't read most of it because it's poorly written: too verbose, points are buried, it's three times longer than it should be. I still don't know what this database is used for and don't want to go through that writing to find out.


I second this. Unfortunately you just can’t just “unslop” it, much like you can’t just “shake the dirt away” from food that fell into a bucket of… you get my point, it won’t be good to eat anyway. Bonus additional ironic parallel could magically manifest if you can guess the bucket content.

Anyway, it’s a shame, this could have been excellent work. Instead, I find reading it unpleasant and part of my attention rots away wondering if you even checked what the LLM wrote for you. You probably did, and I could also do it myself, but still…


I disagree. Hand-authored articles can be slop too. This research is great, and your subjective feelings about word cadence don't affect the information. I guarantee you read AI output every day and don't know it.


Hand-authored engagement or clout (LinkedIn) farming articles and corporate press releases are usually slop, most others aren't. I'd be surprised if corporate press releases today weren't largely AI-written - no one would know the difference.


The criticism can be restated, then, as “you used AI to write it, and unlike others who use AI to write things, you did a bad job directing it and editing the work.”


This comment is not directed at you, but this is the best place I could find to insert my thoughts on this "nO LLm slOP" distraction:

----

The research is jaw-dropping enough for the average reader to comprehend (e.g: I am not a techie, but a former IBEW datacenter wiretwister).

It could be covered in pube-sprinkled-turds and would still be worth reading.

To me the "slopLOL" crew is akin to people who'll take their time to correct others' their/they're/there usage:

Just read the fucking sentence (LLMs can and do and aren't so poopy-you-suck-for-being-human).


We know there’s interesting information in there. That’s not the point.

How news is delivered matters. If you don’t believe that, why would news outlets compete with each other? Even if the facts are the same, would you prefer to get them from a good or bad writer?


>would you prefer to get them from a good or bad writer?

I am smart enough to not need perfection. I just want the data. Write however tf you want, good'll'be read.

And so will LLMs.

I'll suggest adaptation.

>why would news outlets compete with each other?

Competition for the masses' con$umption$?

It seems to generally work; shouldn't have to obfuscate the `man` pages for this always-running daemon... how many people even know what `man`pages are (obviously not here)? how to access them? why they would even want to?


> always-running daemon

This is completely tangential to your point, but I'd bet that this daemon isn't always running. Apple is fond of putting as much system functionality as possible into daemons that (a) can be spun up quickly when needed in response to Mach IPC messages, (b) can be killed safely at any time to free up memory, including with SIGKILL if necessary, and most importantly (c) operate in the tightest, most locked-down sandbox they could devise.

For background computational tasks like this one, it can also be helpful for getting thread prioritization right. If some computation doesn't need to happen quickly, a daemon can easily make sure that it gets assigned to efficiency-optimized CPU cores, tell the OS scheduler to give priority to more important stuff, reduce precision of timers to coalesce wakeups, and make sure that any thermal throttling affects the background stuff before it slows down anything the user will notice. It's a pretty slick system.


Well, good for you for not caring, I suppose. The vast majority of society does.

> I am smart enough

Knock yourself down a peg, kid.


I will, sure, but I've been writing about this for almost a decade; and I know more than most people about this topic, despite being a bluecollar electrician.

Also, my IQ scares even me; I've chosen bluecollar so I don't have to be evil as much with it.

----

Companies do not give two shits about most users, with (e.g:) AMD saying the quiet part out loud, recently, with their recent gamers-last announcement.

Also see: Apple's invasive attempts to categorize all local data, by default; if you want to opt-out (some service, like OP's, you cannot!), it takes dozens of clicks for each new software – intentional dark-patterns to make opting-out cat-and-mouse.

----

>Well, good for you for not caring, I suppose.

I care too much, is the problem (at least from my POV). Companies are soulless, and only care for shareholder value (legal requirement).

Data is the new_oil is the new_steel is the new_tin is the...


> I've been writing about this for almost a decade; and I know more than most people about this topic

> Also, my IQ scares even me

Oh, brother.

I mean this with the best of intentions and care: have you been psychiatrically evaluated lately? The language you’re using is consistent with friends I know who are off their stabilizing medications.


[flagged]


Why? "There is gravel in the food" is a perfectly valid contribution. You don't need to write a compelling story about it.


Because the article itself is way more interesting than the question of with what style it was written.


It's like the saying beauty is only skin deep, but ugly goes clean to the bone. Bad AI writing makes an article "unreadable" - you'd have to force yourself to read it, and people don't get paid for reading stuff on HN. The HN value of an unpleasant to read article is basically zero.


I really don't want to waste more words on it here.


I think the difference might be that you find bad style easy to ignore and I don't.


If the applied style is painful, can't one just take a lightweight model and have it transcribe the content in a style that one personally finds pleasurable? Even phones should be powerfull enough to chew through such a task locally these days (even across languages), and once you define the criteria for what you enjoy you'll never have to be pained again (unless you enjoy it offcourse)


The criteria I enjoy is "written by a human".


You can pay someone to rewrite it. There's loads of college students who wouldn't mind extra spending money.


You don't have to leave a comment if you are too lazy to read it. It's not even if the words were written mirrored. Anyone who can read English including children can read the output of AI without needing to force anything. Maybe you need to see a doctor if you need to force yourself to read it.


I wouldn't worry about the down votes. Human voice will eventually go the way of the pops, and hiss of dusty records. Eventually those raised on AI generated prose will come to appreciate it in the same way we have nostalgia for 96kbps mp3s.


"How dare you have a preference that is not my preference?"


Having a preference of writing style and being unable to read a writing style are not the same.

The comment upthread was:

> Bad AI writing makes an article "unreadable" - you'd have to force yourself to read it

That doesn't literally mean "unable to read", it means it's unpleasant. I don't think anyone in this thread has literally meant "I could not possibly read it", just "it's painfully unpleasant".


Its interestingness is zero if it’s so aggravating to read that you just give up.


> Can we add a rule that if you blame someone for using AI, you _also_ have to bring something interesting to the table?

No.


I agree. One comment constructive is enough. It feels like some guys pushing the topic away.

It annoys me as it takes my attention with them so with all respect: user aharmetz (and others) please keep it in one post that you’d prefer non-AI text. Point.

Thank you.


I’d rather this over the very long tradition in this site if people bellyaching about site design and layout on otherwise u related comment threads


> Can we add a rule that if you blame someone for using AI, you _also_ have to bring something interesting to the table?

Criticism is a valid 'something interesting'. And growing numbers of us also don't want to feed the US mega-companies LLM training sets.

More and more people are keeping human-generated stuff to themselves, cause of that.


Who does PasswordAuthentication=yes in this day and age anyway.


We have clients (regulated banks and insurers of various sizes) that send us feeds via SFTP. You might be surprised what proportion of them ignore our “we strongly recommend key based authentication” and ask for password auth instead.


When I set up a client sftp drop I had the luxury of saying "key based access only"

Then it was a surprisingly customer service intensive process to get the keys from the customers. I never did manage to automate it fully. Weird enterprise systems I have never heard of with strange key formats. ssh1 keys, email mangling them. Their internal processes requiring a password.

And way to many "I am sorry but that was the private key you sent me, That key is now compromised and you will need to remove it and generate new keys anywhere it was used"

I almost regret not allowing passwords, almost, despite the headaches key based is still much better.


About to say, users just drop the key on their desktop and then download random exe's from the internet and run them.

Still safter in general, but not bulletproof.


At least in that case, if the account is compromised we definitely didn't leak the credentials. That we don't have them to leak is one of the advantages of key-based auth.

[of course not having the credentials to leak does not protect us from this sort of bug, it is just one factor in a multipart security stance]


If you refused, would you get overruled internally?


Previous company: probably. It could be a dealbreaker for the client, and we bent over quite readily on a number of issues considered minor (keys are best practise, but passwords are so common it is difficult to argue that they are so bad they should be refused).

Current state (now we have been bought and are effectively more than an order of magnitude bigger): I definitely don't have opportunity to make a unilateral decision like that. Even though the larger company has a more “fair enough, feel free to go elsewhere” attitude when a small client says “if you can support × that is a dealbreaker for us” then we could previously afford to take as our stance, some of the bigger clients are definitely paying us enough, and do use passwords, that someone above me would overrule a keys-only decision.


it's the default for most systems. I can't remember a time I had to not turn it off.


Don't most cloud providers default to that? I know I had to force some to deploy with a key and even then PasswordAuth was often on and had to be disabled.


If you run a cockpit service on your server you will need password authentication enabled, unfortunately.


I went and checked this. I can admin my server with Cockpit using either the web service (using my username/password) or the desktop client (via private key) without issue even with SSH password authentication disabled. I do usually disable the web service so had to reconfigure to check.


As this is apparently affecting openwrt, probably a fair number of openwrt users, if I had to guess


I thought OpenWrt used Dropbear by default?


People who use AuthenticationMethods "publickey,password"

You must present a key first after which you must enter a password.


Just do no password at all.


Putting the one in 1fa? No thanks.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: