Comment by hbn | Hacker News Reader

Comment by hbn | Hacker News Reader

Liquid Glass on macOS is such a joke. Most of the redesign was just turning buttons into Fisher Price-looking circles and ovals. I'm typing this from Safari which looks so stupid in Tahoe. The tab bar is a giant oblong oval with a bunch of tab titles and icons floating on a solid background, only separated by a short, faint vertical bar that doesn't go to the top/bottom to truly separate them. The current active tab is a small oblong oval within the giant oval. The perfect visual metaphor for tabs which Safari set the trend for in macOS is gone.

And then just above is a bunch more ovals and circles. The sidebar button is an oval, the back/forward buttons are in an oval, the Wipr extension icon is in an oval, the URL bar is an oblong over, etc. And (at least in light mode) this is all white ovals on a white background. It all looks so amateurish.

I'm so glad that Hack Alan Dye is gone and I pray to God that Stephen Lamay can get us back to reason. I doubt they'll do an overnight Cmd+Z update in macOS 28 or whatever, but perhaps he can direct Liquid Glass in a direction that isn't just rounding things for the sake of it.

Liquid Glass is Apple’s Windows Vista. They had a ton of fun with Vista in their “switch” ads, if the Windows team were in better shape they could have a field day just screenshotting Tahoe on Social Media. Lucky they’re distracted with their own challenges.

Liquid Glass does have some good points, but it feels like someone turned in C- level work.

I see the Vista comparison a lot but I'm not sure I agree with it. I never thought Vista was that ugly, I thought it was more most of the computer hardware people were buying at the time just wasn't capable of running those visual effects (and I recall it was pretty buggy too)

It had a glassy aesthetic but the similarity doesn't go much further than that description. They didn't make all the buttons into glass blobs floating on top of the content with distracting warping effects; the window chrome was still generally separated from the content.

It's the year of the Linux desktop. Break free of the walled gardens, there's no good excuse to throw your money away anymore. ElementaryOS and a few other projects have superb Apple flavored UI and UX. Apple just wants your money; they don't give a flying rat's ass about you or your needs.

Let liquid glass be your red pill - come join us in the real.

Not everyone uses Electron apps exclusively.

The cross platform scene is much different these days. Electron apps suck, but at least they suck equally across all platforms. And there are many Electron apps.

But a lot of people rely on Adobe, Microsoft or Windows-only, Mac-only apps. I don’t see that changing anytime soon, unfortunately.

Breaking free is easier than ever. You don't need walled gardens.

AI is making handling the edge cases that kept people locked in almost trivial. Any workflow, custom spreadsheet, specific OS-only app can be worked around, easily. Staying stuck on Apple or Microsoft is a choice - they're no longer returning value concurrent with the money they charge.

You're free to continue giving them money, but the reasons to do so make less and less sense each day that goes by.

"Adobe Creative Suite not running on Linux can be worked around easily" is something that people have been getting wrong for decades, but injecting AI into the premise is a new frontier of funny.

What's the AI workaround for Illustrator/After Effects/etc.? You're not suggesting generating vector art or video assets via LLM replaces these, surely?

I'm very curious what their workaround plan for something like U&I MetaSynth would be.

Given than Adobe has their own GenAI trained on work licensed for that purpose in Photoshop, I think they would disagree with you on that.

Offering to add it to the workflow doesn't mean they think it can replace the whole product for all users - if they stop shipping the rest of the features, then that'd be Adobe "disagreeing with me on that".

I use Adobe Illustrator daily at a very high level and have about 25y of source files in its private format, as well as a bunch of plugins I rely on. How well can Linux deal with running a version of it written in this decade?

Inkscape is not an option, nor is anything involving importing PDF/SVG, those have to expand a huge ton of stuff that's represented much more compactly in an .AI file. It's about as large a difference as that between an executable file and its source code.

I don't think there is an answer. The best you can do is probably running Windows in a VM and limiting its use to applications that you really cannot replace. It's been a while since I used a VM on Linux, but VMware had a thing called Unity Mode where you can have application windows from the VM on your Linux desktop:

https://techdocs.broadcom.com/us/en/vmware-cis/desktop-hyper...

It seems like they removed it from VMware 17.6, but maybe another VMM still has this functionality?

I haven't tried it but I've seen this project promoted in some tweets and maybe even here: https://github.com/winapps-org/winapps

VirtualBox (with Guest Additions) has Seamless Mode if you're using X. I'm really not sure about the status of it using Wayland.

https://www.vectorpea.com/ and https://www.photopea.com/ are the lowest barrier to useable alternatives. You can even save them offline and convert to PWA, with very little friction. vectorpea and photopea should handle your .ai files admirably.

Inkscape, Affinity, other open source alternatives exist, but have a remarkably different UI and don't capitalize on your muscle memory.

The feature overlap is bordering on complete, but there are some Adobe Illustrator only perks, for sure. Most of it you can make up for with any of the frontier image AI models.

There are plugins - if you're well versed in how they work, converting between AI and vectorpea should also be a piece of cake with AI.

They didn't ask you for alternatives, they asked you how well Illustrator works.

Hahaha no, I spent several minutes waiting for the 184m .ai file I was working on today to upload to Photopea. It displayed something that looked like my art for a half a second before crashing the browser tab so hard that its address bar went blank.

There are plugins - if you're well versed in how they work, converting between AI and vectorpea should also be a piece of cake with AI.

I'm an artist, not a prompt jockey. My interest in spending even a minute of my life trying to convince a plagarism engine to spit out something half as refined as Astute's plugins for Illustrator is absolutely zero.

> Breaking free is easier than ever. You don't need walled gardens.

There’s nothing that comes even close to Photoshop. Same for a lot of similar professional tools.

> AI is making handling the edge cases that kept people locked in almost trivial

Not for anything remotely complex. Let’s see how that looks in 5 years, but I’m skeptical.

Photopea.com exceeds CS6 but keeps the UI. All sorts of plugins, great project. You can also use various hacks like photoshop UI for Gimp and things like that, but I found that route to be unusable.

I've converted over a dozen weird edge cases of spreadsheets and access apps and ancient scripts used by departments into standalone little apps or browser apps, ranging from budget and finance related bookkeeping to tracking sales to licensing management. The only advantage Excel has over this is ease of maintenance - it's a lot easier for someone to guide themselves through updating things on a spreadsheet, or to break an idea down into multiple pages, etc, if spreadsheets are what they're familiar with.

If you're an engineering or finance firm dependent on an obscure, unique Excel feature, I could at least see the argument that your use case is too hard to migrate off of Windows.

Photopea is a really impressive project, don't mean to diminish it in the slightest, but it's a toy next to Photoshop.

> exceeds CS6

I wish. It's not even on par with Photoshop 4. No LAB mode, handles only 4 profiles. I could be here all day listing missing features. Also, have you tried to open a 6GB PSB file with it?

I've used Photoshop almost daily since 1994. I really wish there was an open source competitor. There isn't.

This year I switched from the Linux desktop to MacOS. I finally got tired of how unprofessional Linux operating systems were being run.

I think Liquid Glass looks good.

We aren't given an option at my work, but if we were I'd still choose the Mac anyway. I love the Mac and that's why I care so much about this design regression. I like that it unlocks with my watch or fingerprint from a wireless keyboard, I like that I can push files and browser tabs between my Mac and phone just by sharing, I like that if I can push my mouse off the side of the screen and control my iPad with my keyboard and mouse with zero setup, or if I want one more monitor I can turn my iPad into that with 2 clicks. I could go on.

They just need to get back into the mindset that design is how it works. Not forcing some aesthetic into everything with the superficial idea of "focusing on content" as a backwards justification for making everything transparent cause someone thought it was prettier.

> They just need to

Linux is for people who want to get rid of "they". If "they" start screwing things up, you switch to a different "they". Alternatively, you become "they" by forking the project.

> Alternatively, you become "they" by forking the project.

This doesn't make sense for the vast majority of people.

Linux desktop doesn't have the vast majority of the niceties that living in the Apple ecosystem gives you. If I was going to rebuild any one of them for Linux, it would easily become a major project that would suck up all my free time.

> This doesn't make sense for the vast majority of people.

That's fine.

> Linux desktop doesn't have the vast majority of the niceties that living in the Apple ecosystem gives you.

And it never will should nobody actually step up and put in the work to make it a reality. Linux needs users willing to do such things.

The original free software business model is that people would pay programmers to work on the features they needed and the results would go back into the commons in the form of upstream patches. I've actually made some money this way. It was nice.

About the UI I become "they" but installing the GNOME extensions that I need to make my desktop look like 99% of what I would it to look and behave. It takes a few minutes to get to 80%, a few hours to get to 95% and days (a few minutes here and there) to 99%. Those huge menus and tabs on GNOME terminal eventually became skinny with a good deal of CSS and AI.

Do most people want to get through that research? Absolutely no, I don't expect many people to follow me into that rabbit hole. They can get the default or Windows or a Mac, no problem with that.

But with AI, "go through with that research" is tell Claude code how you want things to behave, and it'll make it happen for you.

You are correct but only in part.

Claude did tell me where to fiddle with CSS, but its suggestions where not always on spot. It cut the time I spent on it and maybe it made it possible, because I wouldn't have dissected the source code or inspected the UI in the right way (GTK inspector). All the process still cost me a few weeks of five minute attempts now and then.

I think that there is no way an agent (and I was using the chat UI) can take control of my desktop, patch CSS, restart it, take advice and give me what I want. Not yet.

By the way, I have an autohiding Windows 95 like app bar at the bottom. The text inside the app items is still a bit too small and some icons in the terminal top bar are still too narrow. I think I have more CSS to fix but I'm in no hurry.

> I think that there is no way an agent (and I was using the chat UI) can take control of my desktop, patch CSS, restart it, take advice and give me what I want. Not yet.

Claude Cowork can do such things if you set it all up.

You're kinda skipping over the 75% of my comment where I said I like everything else, including things Linux can't replicate.

I skipped over those arguments because they miss the point. You're not going to get Apple tier caretaking from Linux distributions.

Nothing inherently prevents Linux from replicating those features. It's just that somebody's gotta put the work in to make it happen. That somebody could very well be you if nobody else cared to do it.

My laptop has fancy RGB keyboard LEDs. Manufacturer shipped a shitty Windows app to control them, as well as the internal fans. My choices were: either give up on those features or implement them myself. So I reverse engineered what I could and made a Linux program to control the LEDs. Threw it out there on GitHub just because and woke up one day to discover that not only did I have users but somebody else had independently built a GUI on top of it.

The idea is not to change Linux distributions into Apple tier experiences. The idea is to convert you into a contributing user who is capable of solving his own problems and making his own features. The idea is to elevate users from consumers to contributors who take ownership of their systems.

That's the only way you can ever be free of "they" and whatever "they" decide to impose on you. Being at their mercy is a dangerous position to be in. Forcing low contrast glass UI on users is a nonissue compared to that one time where they threatened to start automatically scanning everyone's devices for CSAM. There's really no limit to what they can do to you should you relinquish ownership of your machine to "they". That's what I want people to understand.

> The idea is to convert you into a contributing user

Until you want to contribute but is stonewalled and gatekept by overzealous devs to the point that you lose all interest in contributing and just give up. Which means you are back to using a “they” computer—not a big corp “they”, but a “devs somewhere” “they”. Pretty much exchanging six for half a dozen.

I've run into this problem myself. I've tried a few times to contribute to GNU software but for whatever reason my code just never made it in. Definitely sucks to spend time learning a project and sending patches only to end up with nothing.

Still preferable to being completely at someone else's mercy though. It is always better to have the option to maintain a fork if needed. I actually maintained my own custom fork of bash for quite some time. About a year later upstream added the feature I wanted so the fork was no longer necessary. It was a pretty good exercise though. Maintaining the fork turned out to not be that hard. All I had to do was pull upstream, rebase my branch on top of upstream, recompile and repackage. It is always better to have the power to do this than to not have it.

This is the Linux equivalent of “corpo speak”—as in, it’s complete bs.

The absolute VAST majority of Linux users are not out there forking their own distros or creating their own WMs. They, more often than not, are just fine with using whatever is being cooked up by the big names in the Linux space, people aren’t giving up the “they” just because the “they” is now KDE or GNOME.

Hell, even Torvalds himself has gone on record to say that he just wants his computer to work and is happy using Fedora and GNOME (the very definition of a “they” Linux).

It's not "corpo speak", it's a principle. It's fine if most users use popular stuff instead of creating their own. The point is to empower them to make different choices if they want to. Whether or not they actually exercise this power is irrelevant. The important part is the fact they have this power at all.

I built a freestanding lisp interpreter that runs directly on top of the Linux kernel just to prove this. Zero dependencies, native system call support. I know that everyone is going to want stuff like glibc instead. But it was possible, so I did it.

And this is more of that same type of “corpo speak but for freedom” bs I was referring to—the principle doesn’t matter if the results are the same, in the grand scheme of things it quite literally does not matter if you have the ability to customise the entirety of your system IF you will NEVER actually customize it, and for the absolute vast majority of users that is very much the case.

For those that such options do matter, it is absolutely essential (I’m in this category). But for the common user, it’s just another thing that their system does that they don’t understand and have no desire to spend time learning. Most think like Torvalds himself: they just want a computer that works and gets out of their way.

> But for the common user

I'm arguing against the "common user" migrating to Linux though. "Linux is for people who..." I just didn't want to say it out loud. If this stuff doesn't matter to you, then obviously you shouldn't use it. I'm just trying to convince others that it does matter.

I want Linux to be the programmer's system, and I want a world with more programmers in it. I have no interest in "common users" other than the fact they might one day become programmers themselves.

I think people should stop replicating "Apple-flavored" user interfaces on Linux. That just leads to constant disappointment.

I'd rather Linux developed an identity of its own. I feel like keyboard driven tiled windows are the closest it has to that.

What's a good invoicing app for small businesses / freelancers on Linux?

What's a good calendar app on Linux?

What's a good e-mail client?

What's a good photo and image editor?

I really want something between Sequoia and Tahoe. (Probably mostly Sequoia, but with targeted applications of Liquid Glass.) I don't like how Tahoe treats everything as floating on top, as if properly dividing windows into sidebars and panels is wrong... There's so much extra padding and rounding now, I hate it. Everything's lost the depth, detail and cleanliness it used to have, replaced by this bubbly mess. Like, sheets don't even slide out anymore, they overlay like on iOS. The charm, expressivity, and, well, Mac-ness is gone.

I love Liquid Glass - the blur and refractive effects are so pretty and technically impressive - but it should be used tastefully instead of this nonsense. I feel like Tahoe in general is straying way, way too far from the battle-tested Cocoa foundation and into this total top-down crap. Liquid Glass feels like some sort of shareholder-enforced enshittification.

macOS is supposed to be defined from the bottom up; it always has been. There has always been importance in having a solid base; a robust foundation for developers to build on. HIG, Cocoa, CoreGraphics, all of that is in service of this. The user experience and vertical integration is a result of this and couldn't exist without it.

There's so much wrong with Tahoe that goes against everything Mac has ever been. We don't want to dumb down the interface; that has never been the goal. The goal has always been to make the interface intuitive enough that anyone can learn it. macOS and iOS are fundamentally different platforms with fundamentally different design constraints and considerations.

Icons being able to escape the squircle was supposed to be a reflection of the fact that apps on Mac are less contained than apps on iOS. They have more expressive power and more advanced capabilities. You're working closer to the metal and in a less controlled environment. Because of that, you can do more and you're not constrained to the flows of the system.

iOS always hasn't been this. The constraints of touch are different than the constraints of the desktop. Steve Jobs spoke about this a lot back in his day, about why iOS is so much more locked-down than Mac.

But Mac has always been a platform for freedom and control. And Tahoe strips the soul of that.

"shareholder-enforced enshittification" what on earth could this possibly mean?

I assume the subtext is something like: "Customers are being abused to create the short-term illusion of improvement, to satisfy myopic investors in the financial markets and the personal compensation incentives of executives."

Can't speak for GP but I got the feeling that after Apple embarrassed itself shipping almost none of the Apple Intelligence features announced at WWDC 2024, they scrambled to get something drastic out the door to show they're still "innovating" and "doing big things"

Shareholders want to maximize stock price, therefore they choose psychopathic CEOs willing to do literally anything to achieve that. People who view reputation and goodwill as just capital to be spent. Giving out free service to get people hooked then turning the screws on them is a proven strategy.

None of the siblings got it right. By 'shareholder-enforced enshittification' I meant when shareholders (or, generally, anyone from the top) enforce a direction that doesn't align with what's natural of the foundation. So the system ends up being stretched to afford it, corners get cut / shortcuts get taken, and then that becomes the final shipping version.

"None of the siblings got it right."

Perfectly proving my point. It's a meaningless argument built on vibes that every reader interprets differently.

> There's so much extra padding and rounding now

I don’t like it either, but I wonder if that’s to support the touch-enabled Macs that the rumor mill is reporting about right now.

In any case, Tahoe has many other issues beyond padding.

There are definitely other ways to do it than making everything look like this.

> But Mac has always been a platform for freedom and control.

My impress has always been the opposite: MacOS is "opinionated", and the user can either accept the Apple way of doing UI or can take a hike.

MacOS has offered token customization, such as allowing the user to change the color of menu bar highlights, but any substantive change required 3rd party intervention, which would inevitably cease to function at the next upgrade.

These days the OS is even more locked down, making it all but impossible to modify OS files.

Catalyst was already sort of a death knell, since it's an admission that it's ok to port over iPhone/iPad HIG to mac. Maybe swiftUI too, since it's replacing appkit and all its various affordances.

Reply on news.ycombinator.com