It's time to make computing personal again

Well, I have access in Files to a lot of content from my apps - that’s a decision of the app creator to not use this and keep the content created in the locked area of the app.

For example, the apps from Omni do this, as do obsidian, Linea…

Let’s assign the blame where it should be here.

An iPhone is a very non-typical device. Apple is a non-typical company which builds lock-in to every step of the process.

If you chose to use iAnything then it's a bit late to start complaining about lock in now.

That's not iOS fault. Apps can store their files in a folder visible in the Files app, or can ask the user to open a file or folder from a file provider (also visible in Files app), or to save a file or folder in a file provider (always visible in the Files app).

It's not the 2011 iOS anymore, if an app today hides its video projects from the user, it's entirely the app fault.

I thought most folks had moved on to google docs and ms365?

For the map app, OsmAnd+ is amazing. https://osmand.net/

The reason they don’t grow is much more trivial - they simply have no sign up funnels and are visibly technically complicated. Every step and choice there is arcane and ideologic.

That’s it. Both Open Source and Federated thinks that distribution gateway federation something is something a user must know and be fond of. The user not only couldn’t care less but actively refuses this complexity because they cannot trust their own uneducated decisions. They go for the nearest CorpThing that seemingly just works for everyone and decides everything for them after they tap “Next” a few times.

I would disagree with the implication that everything has to grow, but solely on the grounds that I am not convinced human beings are psychologically mature enough as a species to be that connected with that many other humans and still retain their capacity for acting like a a good human.

The federated networks I am part of are pretty small and we have a lovely time sharing diverse interests, getting to know each other and even disagreeing sometimes without the blind hate, persistent negativity and gotch'a seeking you typically find on places like Facebook, Twitter and Reddit. Too much growth too quickly would destroy that, turning those small federated networks into another cesspool of bad behavior.

However, I am open to hearing why people disagree. My personal experience drives my opinion, so ymmv.

I'm not sure what you mean by lacking a network effect, unless you mean they don't yet have sufficient users to draw in new ones?

I personally switched to Lemmy from Reddit after the API debacle, and I've found it to be an extremely compelling platform exactly -because it was federated. I can curate my feed from hundreds of large and small instances with nary a corporation in sight! It's self-hosting as far as the eye can see, yet it has enough interesting content and discussions to keep me coming back, without any ads or algorithm trying to manipulate me.

It feels like 90's internet full of webrings, and it's glorious.

Well yes and no. There's no existing network (for something new), so certainly there's no network effect making them grow.

On the other hand the pitch to get people to join is weak. I don't pitch it to my friends because (currently) its a pretty poor experience compared to what they are already using.

wouldn't "growth" have to be reinterpreted for such technology? cause part of its appeal is to _not_ grow unchecked. you don't want everybody on there and you want the setup to be a little difficult. about as difficult as it was to hook up with the web in the 90s.

> They lack a network effect that makes them grow. Isn't lack of fast growth a good thing? I swear, I left every social network in the two years after my mom joined.

At some point of a network popularity, it feels like there is an influx of people who want to talk to you but lack reading comprehension to read your answers. Or maybe it's specifically that every "become popular fast" algorithm tries to repeatedly throw you to them.

Curating a corner of web for yourself takes time and effort, and if a social network popularity outpaces you, then you just can't do that.

I feel like the "past" is a shorter timeline for you than it is for me.

For all the examples mentioned in the parent article, PCs were significantly under-powered compared to workstations, much less main frames.

An explosion of hardware development between 2005 and 2020 has lead to an era where hardware outperformed software needs. Now software is catching up.

But there have always been use cases for high end hardware, and always will be.

That’s not true in many domains where doing it on a personal computer would be either too long or too long in asfar as you are are skillful as using faster memory as cache.

video production, climate simulations, pdes, protein folding, etc etc

The part that is especially annoying there is that it's not just about speed, but about AI tools being closely tied to a specific architecture. Lots of them only work on Nvidia cards, but not on AMD. A fallback to CPU is often not provided either. If you don't have enough VRAM a lot of them won't work at all, not just run slower.

It's the least surprising thing in the world! The article is a totally standard bit of left-tech activist writing of the sort that has been widely found online for decades. It used to be a staple of Slashdot, a staple of USENET and it's a staple of HN too. RMS made a living giving talks exactly like this.

What would be actually surprising is to read a full throated defense of modern tech and the companies that build it, and then see an HN thread full of agreement. It's certainly possible, I'd disagree with almost everything in the article. But the sort of people who disagree tend not to waste as much time on HN as me :)

Single player video games are not going to die. And the market seems to punish any push for always online model (which is obviously a scam). I say this because a bulk of market for personal computing is driven by video games.

anybody who really wants to learn to code will just install linux on an old clunker. every uni and highschool student i know who means business does this.

There definitely is neglect and a slippery slope.

The new settings are half-baked and terrible. The OS pesters me constantly to update to the latest revision and I can't turn those messages off, not even close the notification without displaying the settings app. And I don't want the latest revision, because it is buggy, breaks a number of apps, and introduces the "AI" features that I do not want or need.

More and more often good apps get broken by the new OS policies (SuperDuper is a recent example).

The old style of system apps that did wonderful things is gone (think Quicktime player, or Preview), these apps are mostly neglected. The new style is half-baked multi-platform apps like settings, that do little, and what they do, they do poorly.

As a user of mac for 7 years (went back to windows since around 2018), I’m watching my mac buddy from time to time and it’s not what you are describing. It definitely gets stupid-level worse every year and yes, Snow Leopard was peak mac indeed.

never fell into the quagmire Windows did with bloatware, malware and all the viruses (although I think the situation is much better today).

Windows has 10:1 program ratio compared to Mac. You can usually choose from full-bloat gamified experience to a simple tool doing its job. Windows itself is crap by default though, but that’s at least fixable if you know what you want and where to look at.

I don't know what kinds of drugs you are on but it must be very nice.

macOS has objectively getting worse in so many ways it would take a full day to compile a half serious list.

Every redesign has been a pretext for stupid "simplification" making looks airy but much less usuable. This is true both for the system and other Apple apps (iWork redesign is still a lesson on how to ruin things).

They also added a lot of restrictions and all kinds of nonsense in the name of "security" and "privacy".

The updates are slow to the point of making Windows look good, they also always force some sort of programmed obsolescence that you only find out later down the line.

They are unable to keep the development of their OS separate from their Apps (it's actually just a way to force hardware upgrade) and abandoned most local technology to favor their cloud offerings in order to increase their service revenue.

I used OSX from 10.0.0.4 to 10.4 and it was OK then. I recently had to use a Mac for work something-something-tree-whatever and it's slow even on an M1, it's double the weight of an X1 carbon, and the window manager hasn't evolved meaningfully and is junky. I haven't had so many troubles with arranging windows on two screens in almost 20 years.

Maybe people have been slowly boiled? I got my partner on a Mac 10 years ago but would not get her another Mac. Apple's push to make evetything e-waste, foxconn, and the general surveillance in the name of security ensure that. My observation is less that it has aged "like a fine wine" and more that Macs become prisons shaped like a computer.

(Edit: s/has/hasn't/)

> macOS has been very conservative in redesigning the user experience; it's aging slowly like fine wine. There are a few hiccups occasionally but I feel it's a lot more polished and elegant compared to the quirkiness of earlier versions. I don't get this common sentiment that it was better in Snow Leopard etc.

I completely get this sentiment. MacOS since Big Sur has had a quite indeterministic UI, from scrollbars that only appear halfway of the time, to the new app toolbars that truncate filenames and hide the search bar in the overflow menu. The Settings app is still worse than what it replaced, the Mail app keeps randomly appearing, the random confirmation pop-ups are more common than in Windows Vista.

Snow Leopard was (and still is) a bliss, it didn't nag you with "this app is dangerous" bullshit, built-in apps that look and work like an intern's multiplatform UI practice, constant updates... It was an OS primarily designed to let people get work done, not to encourage them to spend all their time on Apple TV/Music/...

I’ve moved away from MacOS for a number of reasons, specificslly the strenghts you highlight I consider weaknesses:

- the oh-so-pretty GUI takes up too much white space, I have little left for actual content. I dont need massive icons with wide spacing since, I need the opposite.

- what is the deal with hiding folders from Finder (like ~/Libraries). Do I honestly need the command line to open this directory?

- every iteration after SnowLeopard I feel their “features” are going backwards and taking away “usability” from me.

- OpenGL stuck on 4.1, Vulkan (Molten) is a 3rd party hack. Seriously?

- its become a case of one step forward, 2 steps backwards.

- I can go on and on, but you get the gist.

I switched to Linux a couple years ago and overall am glad I did, but it's only a partial solution.

As I see it, one way to phrase the problem is that Linux (along with its ecosystem) isn't really user-focused either. It's developer-focused. Tons of weird decisions get made that are grounded in developer desires quite removed from any user concerns. There are a lot of developers out there that do care about users, so often you get something decent, but it's still a bit off-center.

Which distribution can you suggest?

I used KDE (24.04) for a while now. Also used Linux 2000-2008-ish. Have read APUE.

When I win-left/right a window and then resize it, then close it, now win-left/right always resizes it as previous one. There’s no way to reset it to 50:50 (unless logout).

Notification center is obnoxious. Regular “Program stopped working” with zero details. Why do I need this information. Copying a folder in dolphin pops up a progress notification that doesn’t go away when finished. You did something, expect a notification to pop up. You did nothing, still expect it.

Windows either steal focus or fail to steal it when needed, depending on your settings and astrology. Software update nags you to click update, then you click it, then it downloads something for a few minutes (you go doing your things), then a sudo password popup fails to get focus, and it all crashes after a timeout. Or things will pop up right in your face. VNC connection closed? Yes, bring that white empty window to front asap and tell “Connection closed” with an OK button.

Start menu was designed by an idiot, a wrong mousemove and you are in a wrong section. Sections reside exactly on the 80% of bezier paths from start menu to the section content and have zero activation timeout. So you have to maze your mouse around to avoid surprises. Logout, restart, shutdown, sleep buttons in the start menu all show the same fullscreen “dialog” that requires you to choose an action again. What’s the point of separate buttons even.

I could go on about non-working automatic vpn connections, mangled fonts/dpi/geometry if you vnc into a turned off physical display, console that loves printing ~[[A half of the times when you press an arrow. And so on and so forth, the list is so big I just can’t remember it.

Idk how Linux users are using Linux so that they do not meet any issues.

Linux definitely exists.... except that it isn't free from this philosophy either. From the "don't theme my apps" movement, to Wayland's "security above usability" philosophy... I recently even read about some kallsyms functions being unexported from an 5.x release because it could be used to lookup symbols and it shouldn't be that easy to access internal kernel symbols or something.

Not to mention many projects refusing to add configurability and accessibility, citing vague maintainibility concerns or ideological opposition.

Another blatant example is the 6.7 kernel merging anti-user "features" in AMDGPU... previously you could lower your power limits as much as you wanted, now you have to use a patched kernel to lower your PL below -10%...

Everywhere you go, you can find these user- and tinkerer-hostile decisions. Linux isn't much better than Windows for the semi-casual tinkerer either - at least on Windows you don't get told to just fork the project and implement it yourself.

I'm a bit hesitant to call this corporate greed as it's literally happening in the OSS ecosystem too. Sadly I don't have answers why, only more questions. No idea what happened.

I should probably have a pre-defined disclaimer "signature" whenever I write about Mac OS, since I always get this response.

I know Linux exists. In fact, I've been using it as my primary OS roughly from 1994 to 2006, and since then intermittently for some tasks, or as a main development machine for a couple of years. I wrote device drivers for Linux and helped with testing the early suspend/hibernate implementations. I'm all in support of Linux.

But when I need to get work done, I do it on MacOS, because it mostly works and I don't have to spend time on dealing with font size issues, window placement annoyances, GPU driver bugs, and the like. And I get drag&drop that works anywhere. All this makes me more efficient.

But I don't want to turn this discussion into a Linux vs MacOS advocacy thread: that's not what it's about. In fact, if we were to turn back to the main topic (ensh*ttification of everything around us), Linux would be there, too: my Ubuntu already displays ads in apt, as well as pesters me with FUD about security updates that I could get if I only subscribed. This problem is not magically cured by switching to any Linux, it's endemic in the world we live in today.

> It is the legal system that hasn't caught up with how tech scales seemingly small damage.

Most administrations are squishy-soft on corporate crime. If there were regular antitrust prosecutions, violations of Federal Trade Commission regulations were crimes, wage theft was treated as theft, forging safety certifications was prosecuted as forgery, and federal law on warranties was strictly enforced, most of the problems would go away.

In the 1950s and 1960s, all that was normal. The Americans who lived through WWII were not putting up with that sort of thing.

>There are more of us, we are connected globally,

Why is this good?

Also, even if they're not, they're getting squeezed out. It's hard to stay afloat trying to just do a thing without your eye on the "prize" of getting bought out by Google et al.

It's so cool, I could talk about it forever. It's practical enough for the devs to use it as a daily driver (though with linux in VirtualBox or Seoul for some things like running their builds) and theres a few businesses built on it.

But nowhere near as practical as Linux at the moment of course

It's a very similar philosophy to Qubes - one of their open challenges is to port the qubes infrastructure over since qubes is (in theory at least) hypervisor independent. https://genode.org/about/challenges Which would be nice since NOVA hypervisor is dramatically less code then Xen and Nitpicker/Dialog for the management console is dramatically less code than Fedora.

I've looked into it briefly but it seems like too much work for me right now.

The True Genode Way of course is that everything worth having would eventually be ported as a native genode component instead of a qubes style VM. They've put a lot of effort into making that as easy as they can with Goa (a nix-inspired package management and build tool) and adding to their C standard library and ports of popular 3rd party libs like SDL

Also - their defaults are pretty hilarious.

They dont assume you want a RAM-Only filesystem. By default it starts out completely immutable with nothing being able to save anything anywhere.

If you want to save anything to a hard drive you have to enable that driver because they don't assume that you'd need one.

Copy and paste is an optional extra to install

It's wild :p

Not sure! They have a system set up for porting drivers from Linux into userspace components so it bats above its size.

From their description: "It is tested best on laptops of the Lenovo X and T series (X220, X250, X260, T430, T460, T470, T490)", 200 isn't on the list but you'd probably have about as good a time as you can

Oh yeah — who doesn't remember all the "Pamela Anderson" meta tags (thousands) people would put into their HTML files to drive up the page rankings on the various web crawlers.

So easy to game the system before Google. (Now easy again judging by the shitty results I've been getting for years now.)

The reason for Google's massive success has nothing to do with the business acumen or the innovative ideas of its founders - it's because a) in the 1990s, antitrust law in the US was pretty much dead and Google came in at just the right time to take advantage of this fact and make world domination their goal from the outset, and b) Google used their excess reserves of cash to host the worlds most popular website with no ads and no real revenue stream for many years, taking a risky and extremely expensive gamble that their slowly-cultivated vendor lock-in would eventually recoup their losses many, many times over, which it did. In other words, it was dumb luck, being in the right place at the right time. Just like Larry and Sergey, the trust fund babies born to wealthy successful parents with lucrative careers from the 80s tech boom.

In 1992, you more options. Your friends could tell you for free, you could stumble on them yourself, or you could get them with a magazine or book (which you didn't necessarily have to buy, you could just flip through it at the store and memorize the cheats.)

In those days cheating didn't impact other players. This is why pay to win is a bigger problem now.

> you'd call the Sega Hotline on a premium phone number

I remember the ads for that but I've never met a single person who did that. (Or whose parents would be okay with it.) Cheat codes were either shared by word of mouth among friends or in magazines. Or you bought a game genie, but that was more for messing around with a game's mechanics than actual, blatant cheating.

In 2025, the "cheating" has become a business model

> Humans definitely have control over how a corp behaves.

I will counter this with: The argument that the humans that make up a corp are in control over it, and that the corp behaviour simply results from their flawed and greedy characters, is just a convenient way to blame someone because the real problem of understanding what kind of entities corps are, and how to influence and control them is too hard.

Yeah, clearly, thats why we all willingly continue to use Teams.

I agree, I think it would be very reasonable to instate the death penalty for corps. It's difficult though because they've already identified legal structures that make them very difficult to kill and even more difficult to kill in a way where they can't reassemble. Identifying an effective way to truly and permanently kill a corp, so that the death penalty would make a real difference, would be a good start.

So far, your view (that the individuals making up the corp can significantly influence its behaviour) has been the approach the world has taken so far, and yet, corps are not exactly getting more well behaved.

Please reconsider whether continuing to do the same (simple and easy) thing we've always done, that we've already seen does not work.

Take a step back, re-evaluate the problem in a new context, even if you don't end up agreeing with my perspective, attempting to think about the situation in a new light might be helpful.

It might be that something in the way corps grow up, maybe their environment (regulations, lack of same, incentives, consumerism, trade, markets) may influence them to grow into the immoral unethical monsters they often become.

Maybe we should consider them too dangerous and harmful, and simply destroy them, I'm not convinced that's better, but maybe there's a way to understand them at a different level, that allows us to "write articles" they understand well enough to actually adjust their behavior (and not just try and circumvent whatever "obstacle" has been put in their path).

> Corporate creatures are similar, they simply don't share our emotions. That doesn't mean they don't understand or wont cater to them.

I agree, in many ways corps exhibit the same behavior as psychopaths, after all, people is the thing they subside on.

I think their way of cognition is even more alien to a normal human than that of a psychopath is to a normal human.

They may share a fundamental, that we can't make them truly understand why something is wrong, but we may be able to come up with consequences that deter them from undesired behavior.

One difference though, is that we can't make it illegal to be a psychopath, only make some of the actions that only a psychopath would perform illegal.. We could make it illegal to be a corp (I'm not saying this is going to be a better idea.. But we could make it legal to kill corps, if only we can find out how to kill them so they actually die of it [1]).

[1] https://www.youtube.com/watch?v=H73L7M0xsnM

It was funny when the corporatists here blamed consumers for using free (or “free” or whatever) services, falling into the FB trap etc. As if that mattered at all? As if FB and its ilk wouldn’t use the Free strategy every time in order to grow humoungous or fail? Of course they would bet their money on network effects over getting money in the short term—there’s no point in a boutique social network. Those hypothetical consumers who wanted to pay upfront for a 2K user social network (patently irrational but okay) would have remained unserviced.

There never was a choice.

Once upon a time, it was illegal to discount something to gain market share and then charge extra once you've bullied out your competition. Technically it's still illegal, but good luck finding enforcement.

We're seeing the "free" version of that.

True. I hope the pendulum can swing back the other way if services push too hard.

I think it means increase usage of the thing itself, and I think it's a good insight. While there is a natural supply and demand curve, unscrupulous growth-focused businesses optimize their products (unhealthy food) and services (gambling, social media, mobile games) for high levels of consumption (at least for a portion of vulnerable users), irrespective of harmful effects. It's the tobacco industry model reborn.

I think a more generous interpretation of this is simply one that is critiquing planned obsolescence and addicting algorithm. Some things need to increase usage by nature, but how many services have you used that really needed a subscription as a necessary model to work?

How?

Buy-in from the community is indeed the hard part but I have friends running irc and phpbb we hang out in, and matrix is more or less viable to self host for a group chat, it’s just that 100x more people are using Discord and Signal because of network effects, your one account can give you access to a million communities.

I guess activitypub and matrix are meant to be similar in that regard but for whatever reason the learning curve is just a little steeper, so you have to be motivated by ideology to put up with the gaps in usability

Yep! And opting out is either prohibitively difficult or outright impossible for most people

But the hardware availability and affordability gives them the opportunity to learn and experiment if they want to. Even tech-savvy people couldn't do that a decade or 2 ago. Not on a budget.

One of the aspects I've wondered about is the software bundled with either the OS or by the device manufacturer. When PCs (broadly, not just IBM compatibles) were penetrating into the home market they would often come with a suite of tools or demos to show what it could do, or let you create things even if they were the basic editions. Before the internet became part of the furniture, if you'd spent several hundreds on some hot technology there was a good chance you'd buy print magazines for it as well, and they would come with cover discs that exposed people to a lot of what was possible with computing.

Without wanting to sound like a stick in the mud, the focus of computing has definitely changed now. I see it as an interesting thought exercise on how to get someone running around with what is usually a marvel of computing in their pocket to try and imagine that is not the apex of computing, whether to explore what other means of computing offer or what comes next besides a slightly better version of what we have now.

Gimp, LibreOffice, Gog.

What's this about Photoshop? As far as I can tell, using cloud storage is an option in Photoshop, it's not required.

I appreciate your curiosity. The premise is simple: the Web of 2025 is still a primitive sketch that leaves 99% of the medium’s potential unrealized. Why? Because the technical 1% writes all the code, while the rest spend all their time merely navigating that code, generating user data. This rigid division between creator and user is built into every level of the software industry, yet it’s what limits the development of much more complex, massively collaborative, personalized applications based on an abundance of ideas and free exchange of code.

I started the work by recognizing that, in a world of LLM coding assistants, there is no longer a minimum bar for code literacy; it is now a spectrum, along which everyone is capable of unique creation, no matter their sophistication. The winning platforms of this age will be ones built from scratch to accommodate and leverage this new massive creative potential, by dismantling the professional class’ monopoly on software production.

yeah, tough times, taking home 6 figures with good benefits year after year, stock prices at historic highs, taking their picks from the housing bubble, voting for idiots while the world burns... I don't quite get why you're complaining to me about these things, but obviously I'm not talking about people who are struggling to "feed and clothe" their families?

> you’re in for a rude awakening

I've seen how techies spend their time, I'm not the one in for a rude awakening.