My hypothesis is today's "modern" OS user interfaces are objectively worse from a usability perspective, obfuscating key functionality behind layers of confusing menus.
It reminds me of these "OS popularity since the 70s" time lapse views:
As a kid, the OS's supported me in learning. They were simple, intuitive and rewarding. I'd click around and explore, and discover cool things like a Wheezer music video, or engaging puzzle games.
There was no one who could help me when I got stuck, beyond maybe an instruction manual. I just had to figure it out, mostly by trial and error. I learned so much, eventually being able to replace hardware, install and upgrade drivers, re-install the entire OS and partition the hard drive, figure out networking and filesystems. It built confidence.
Now my kid sits infront of an OS (Windows, Mac, it doesn't really matter) and there's so much noise. Things popping up, demanding attention. Scary looking warnings. So much choice. There's so many ways to do simple things. Actions buried deep within menus. They have no hope of building up a mental model or understanding how the OS connects them to the foundations of computing.
Even I'm mostly lost now if there's a problem. I need to search the internet, find a useful source, filter out the things that are similar to my problem but not the same. It isn't rewarding any more, it's frustrating. How is a young child meant to navigate that by themselves?
This looks like a step in the right direction. I look forward to testing it out.
This is one of my biggest frustrations with modern GUI computing. It's especially bad with Windows and Office, but it happens on iOS and macOS too to an extent. Even though I've had Office installed for weeks I still get a "look over here at this new button!" pop-up while I'm in the middle of some Excel task. Pop-up here, pop-up there. It's insane the number of little bubbles and pop-ups and noise we experience in modern computing.
Even on gnome, I regularly have applications stealing focus when they decide they're the most important thing. As well as being really annoying, it's a security risk if an application steals focus while you're typing your password or otp key
> if an application steals focus while you're typing your password
Definitely. Bitwarden does this ironically. It will pop-up "UPDATE AVAILABLE!" half way through typing a passphrase. Why not suppress the pop-up if the user is typing, or make it non-modal? Every few days I am interrupted just trying to unlock a vault.
Did you install an extension for that ? By default Gnome prevents this and shows a notification instead. It's extensions like Steal my focus that allows focus stealing
I'm using PopOS, so there's a chance that it's related to that, but no I haven't. And I've tried the gsetting that supposedly helps prevent it. I've only ever seen Zoom and VSCode do this; I get the feeling that they don't quite follow the same conventions that 'native' Linux applications use, or use some trickery (as they see themselves as the most important thing on your computer).
Apple has kind of made things worse in the recent macOS, where my phone's notifications show up on the desktop now. Like, man, I was already drowning in them before anyway, I don't want them on two screens now.
You probably already know this if it was bothering you, but just in case: System Settings -> Notifications -> toggle off "Allow notifications from iPhone".
I would also argue, regardless of what mobile OS you're on, quieting, delaying, or disabling notifications on a regular basis (and taking stock of what you let through) is prudent.
I'm always shocked whenever I see someone having hundreds or even thousands of unread notifications. It pains me to see that instead of controlling their feed to only what interests them, they've just let everything pile up forever, completely unread.
For example, on Discord, I sometimes see people with unreads for every server, and dozens to hundreds of completely unread DMs, just because they don't know that you can turn notifications off for the stuff you don't care about. Instead of doing that they just learned to ignore everything, leading to a disorganized mess.
I'm somewhat familiar with what typically leads to this (usually something like ADHD), but when you let it go for so many years it's such a big task to fix it that the fixing never happens and you're just kind of screwed for eternity.
In Discord, I have no unread servers and no unread DMs, despite being at the server limit. This is because all my servers are completely silenced and all my DMs are read immediately. My only unread email is one I marked as such because I still plan to reply to it soon. I have the attention for every single notification because I aggressively optimize the notifications I receive to the point where they all are typically things I care about. Back in the day I would instantly report every email to SpamCop, typically in under one minute, but I eventually stopped doing that because there's no point.
I simultaneously do and don't understand people who just submit to a flood of irrelevant garbage. Control it!!
> Instead of doing that they just learned to ignore everything
Leading to notifications being ignored, and not mattering at all.
A well curated set of notifications that only gives you the things you actually need is superb, but incredibly difficult to get right.
Learning how to ignore all of the noise is probably a more valuable skill for a future where control is slowly wrestled away from the user. A certain Black Mirror episode (Fifteen Million Merits) comes to mind.
> Leading to notifications being ignored, and not mattering at all.
You found my point! Some notifications may be important, but if one has learned to ignore all notifications, then they won't catch the important ones anymore.
When I get important notifications, I can act on them immediately, because I have not learned to ignore all my notifications, because I don't need to. I have taken care to block any notifications I don't care about, leaving only the ones that I do.
> A well curated set of notifications that only gives you the things you actually need is superb, but incredibly difficult to get right.
Ehh... maybe it's incredibly difficult to fix once you are already drowning in them, but since I immediately nuke anything I don't like from orbit, and have done so my whole life, there are a lot fewer things I don't like than if I hadn't to do that.
> Learning how to ignore all of the noise is probably a more valuable skill for a future where control is slowly wrestled away from the user. A certain Black Mirror episode (Fifteen Million Merits) comes to mind.
I have a really hard time ignoring noise. I think that is because of my autism. Ignoring noise would be a nice skill, I guess, but I feel significantly better when the noise is simply not there in the first place. I'd imagine most would, but for some reason I seem to break a lot more easily when uncomfortable.
> > Learning how to ignore all of the noise is probably a more valuable skill for a future where control is slowly wrestled away from the user. A certain Black Mirror episode (Fifteen Million Merits) comes to mind.
> I have a really hard time ignoring noise. I think that is because of my autism. Ignoring noise would be a nice skill, I guess, but I feel significantly better when the noise is simply not there in the first place. I'd imagine most would, but for some reason I seem to break a lot more easily when uncomfortable.
I share this discomfort, but only in physical space. I struggle with visual and audible noise in the real world, but my screen I’ve become very adept at ignoring. Giving up on inbox-zero in my personal life and ending up with 2000+ unread emails in my personal mailbox is where that started.
I unfortunately don't share this ease. As an example, I have to use custom CSS to completely remove blocked messages from Discord, because once I know a blocked message exists I can not seem to ignore it. The only way to properly protect myself from blocked users is to ensure that I can never even know they're there.
I do a similar thing for muted DMs. Once I know that someone has sent a message, I cannot seem to avoid checking the message. The only way for me to properly take a break is to ensure that there's no way for me to accidentally discover the existence of new messages. So I have some JavaScript that completely removes muted DMs from the list so that I won't even see them jumping back up to the top every new message.
I don't know what causes this to happen or how to fix it, but I do know that ignoring things has always been nearly impossible for me. Whenever there is anything I need to protect myself from, I need to also protect myself from ever noticing any related activity.
Maybe this is some sort of OCD, I'm not sure. Pretty sure it would meet the criteria, I guess.
That sounds like a lot of work. Alternatively, I can just ignore them all. I don’t care if there’s a red circle with a number in it. I don’t notice it and it doesn’t bother me.
Again, it is a lot of work if you have to do it at once. If you've done it naturally over the course of your life, you would never have had a big pile of things to take care of at once.
> Alternatively, I can just ignore them all. I don’t care if there’s a red circle with a number in it. I don’t notice it and it doesn’t bother me.
Suit yourself. Personally, I care about responding promptly to certain things, like instant messages or emails, but I don't appreciate unwelcome distractions.
> Personally, I care about responding promptly to certain things, like instant messages or emails, but I don't appreciate unwelcome distractions.
Sure, if I’m being paid to respond right away then I will. The “instant” in “instant message” describes the delivery time, not the required response time. Email, IM are things I all treat asynchronous communications. Anything urgent should be a phone call, ideally. If there’s a particular email address I consider important I’ll make a separate inbox for it.
Otherwise, I find sort of fun to see how big the number can get.
On my work email I just mark everything as read at the end of the day so the number starts over each day.
Discord is really awful about this, and that's one of the many reasons I dislike it. Its defaults are bad. I should be able to set, at the account level, the default that servers don't send me notifications. I believe by default, I only get @everyone and @role notifications, which is only a few taps per server. But every time I join a new server, I have to remember to do it. If I didn't actively care about notifications from one particular server, I'd just block Discord notifications entirely and stop managing them in their app at all.
It's not really a matter of not knowing it can be done, it's that the mental effort of curating it is not really worth it (short term, anyway), because then I have to decide what's a worthwhile notification and what isn't.
The worst example of things popping up I've seen is Youtube's sponsored content warnings. These immediately appear above the active click target (video thumbnail) when you move your mouse over it, hijacking the expected action there. If you click things as a single action like every skilled mouse user does (rather than pointing and then clicking as two separate actions), it's physically impossible to react in time to avoid clicking them. And because only a part of the click target gets hijacked it's too inconsistent to learn to avoid it by intuition. I've clicked them accidentally several times and every time it's been disturbing because it feels like some serious and unexpected error.
Omfg I do not need national political news shoved into my face in on the left side of my taskbar while I'm trying to focus on work, thank you Microsoft, k thx bye!
Google Android phones do so freaking much of this too. Open the Google app (or swipe left from the home screen on the Pixel Launcher), gone are the days where it's a copy of their original homepage and is just a search bar, now it has news... Go to the search bar, and it shows trending searches. Die in a tucking fire!
Humans hate being bored but only dick-around and learn things like this when they're bored. Speaking personally, I guess.
Since the 90s we've found "better" ways at "curing" our boredom. Put this UI on a modern OS in front of a kid today and they would just download Steam, Chrome and Discord. And be assured, they're very proficient at the in-and-outs of those platforms.
Just some random thoughts I had, not sure any of it tracks...
I used to think I watched TV or scrolled Reddit because I didn’t have the energy to pursue more interesting things. I blocked Reddit and TV. Turns out I have plenty of energy, those were just stealing it from me.
Yeah I agree. The trial and error mentioned in GP needed a good amount of focused time commitment, the sort of thing at a premium in the modern attention economy of tech.
> I learned so much, eventually being able to replace hardware
As a young teenager in the early-mid 2000s, I learned the hard way what the little standoffs are for by killing a motherboard by screwing it directly into the steel case :')
Never made that mistake again, that's for sure. And I share all the same experiences as yourself
There should be a club for people like us: we learnt the hard way to double-check and never ever fully trust ourselves, especially with hardware connections.
The first Mobo I ever purchased with my own money was insta-fried exactly like that, it still hurts a little to think about that.
I did the exact same thing - mounted a shiny new board straight onto the case, powered it on, and… nothing. Spent hours troubleshooting before realizing I'd basically shorted the whole thing
I share your nostalgia. I did learn a lot by exploring Amiga OS and later Windows 98 without anyone to help or an internet connection. It was fun, but we had time to spend back then. Today time feels much more scarce and I no longer appreciate if I have to learn by trial and error. Now it feels like you have to keep up with the tech progress, which is crazy fast. And youth has too many things to do that are more appealing, like social media, youtubes, loads of games etc. For them the exploration of OS or some software is not attractive anymore.
There're also performance issues. Building muscle memory (which means offloading tasks from working memory, leaving it open for learning) can't happen if you're constantly trying to figure out when the system is going to actually respond to your input.
We largely abandoned an unbelievably efficient form of human input in favor of big fat dumb slow touchscreens. Can you imagine where we'd be as a species if we got our shit together 25 years ago and standardized a few of the most important keyboard shortcuts and layouts and that was the default everywhere? I won't advocate terminal-only, but even classical GUIs with windows and icons and all could have been much more efficient if keyboard input and navigation was given priority instead of the comedy of using a pixel-precise indirect pointer to fart around a virtual screen select some button when...there are buttons under my fingers.
We did standardize a lot of keyboard shortcuts and layouts. And it was fairly widespread, even, with both Windows and macOS retaining some bits of it (sadly, fewer and fewer with each new update).
This is so true. The most frustrating thing in the world to me is waiting for the UI to catch up to my actions… that should just never happen in 2025. Not only is it frustrating to wait, but as you elegantly stated it forces the menial task to enter working memory.
Win32 would buffer keystrokes so that a sequence of commands wouldn't be lost even if the UI took long to respond (e.g. if a dialog took long to open), but that has mostly been lost in the era of web apps and other similar bullshit.
Modern web (and web-like) apps all too often don't even bother supporting keyboard for anything other than text field input. What I find especially infuriating are the dialog boxes where the only thing you have is a textbox and a button, and yet you cannot press Enter to submit the dialog once you finish typing - nope, you have to reach out for the mouse and click that button.
This is why I commonly have a trackpad in addition to a mouse when I use Windows. Literally, mouse on right, trackpad on left. It's much faster to use the trackpad for annoyances like these. I also find that trackpads are ten times better for scrolling.
These kinds of things almost always give me an uncanny-valley feeling. Here I'm looking at the screenshot and can’t help noticing that the taskbar buttons are too close to the taskbar’s edge, the window titles are too narrow, the folders are too yellow, and so on and so forth. (To its credit, Wine is the one exception that is not susceptible to this, even when configured to use a higher DPI value so the proportions aren’t actually the ones I’m used to.) I’m not so much criticizing the theme’s authors as wondering why this is so universal across the many replicas.
Computing is largely a cargo cult thing these days.
The problem is that the interfaces these bootleg skins draw "inspiration" from were designed on the back of millions of pre-inflationary dollars' R&D from only the best at Golden-Age IBM, Microsoft, Apple, etc.. BeOS, OS/2, Windows 95-2000 do not look the way they do because it looks good, they look the way they do because it works good, countless man hours went into ensuring that. Simply designing an interface that looks similar is not going to bring back the engineering prowess of those Old Masters.
I’m less inclined to attribute it to “these days”, as I remember the contemporary copycat themes in e.g. KDE and Tk looking off as well. Even Swing with the native look-and-feel didn’t quite look or feel right, IIRC.
As a (weak) counterpoint to supplicating ourselves to the old UI masters, I submit Raymond Chen’s observations from 2004[1] that the flat/3D/flat cycle is largely fashion, e.g. how the toolbars in Office 97 (and subsequent “coolbars”) had buttons that did not look like buttons until you hovered over them, in defiance of the Windows 95 UI standard. (Despite Chen’s characteristic confident tone, he doesn’t at all acknowledge the influence of the limited palettes of baseline graphics adapters on the pre-Win95 “flat” origins of that cycle.)
Also worth noting are the scathing critiques of some Windows 95 designs[2,3] in the Interface Hall of Shame (2000). I don’t necessarily agree with all of them (having spent the earlier part of my childhood with Norton Commander, the separate folder/file selectors in Windows 3.x felt contrived to me even at the time) but it helps clear up some of the fog of “it has always been this way” and remember some things that fit badly at first and never felt quite right (e.g. the faux clipboard in file management). And yes, it didn’t fail to mention the Office 97 UI, either[4,5]. (Did you realize Access, VB, Word, and IE used something like three or four different forks of the same UI toolkit, “Forms3”, among them—a toolkit that looked mostly native but was in fact unavailable outside of Microsoft?..)
None of that is meant to disagree with the point that submitting to the idea of UI as branding is where it all went wrong. (I’ll never get tired of mentioning that the futuristic UI of the in-game computers of the original Deus Ex, from 2000, supported not only Tab to go between controls and Enter and Esc to submit and dismiss, but also Alt accelerators, complete with underlined letters in the labels.)
> Despite Chen’s characteristic confident tone, he doesn’t at all acknowledge the influence of the limited palettes of baseline graphics adapters on the pre-Win95 “flat” origins of that cycle.
It's right in the second sentence: "...Windows 1.0, which looked very flat because... color depth was practically non-existent."
> I’ll never get tired of mentioning that the futuristic UI of the in-game computers of the original Deus Ex, from 2000, supported not only Tab to go between controls and Enter and Esc to submit and dismiss, but also Alt accelerators, complete with underlined letters in the labels
I think that's because they used the stock UI toolkit of the original Unreal Engine, which also had all these things. If you recall, UT'99 actually had a UI more like a desktop app at the time, complete with a menu bar and tabbed dialogs:
In modern times telemetry can show how well new designs work. The industry never forgot how to measure and do user research for ui changes. We've only gotten better at it.
I've had an alternate theory for a while. Prior to verbose metrics, UIs could only be designed by experts and via small samples of feedback sessions. And UIs used to be much, much better. I suspect two things have happened:
- With a full set of metrics, we're now designing toward the bottom half of the bell curve, ie, towards the users who struggle the most. Rather than building UIs which are very good, but must be learned, we're now building UIs which must suit the weakest users. This might seem like a good thing, but it's really not. It's a race to the bottom, and robs those novice users from ever having the chance of becoming experts.
- Worse, because UIs must always serve the interests of the bottom of the bell curve, this actually is why we have constant UI churn. What's worse than a bad UI? 1,000 bad UIs which each change every 1-6 months. No one can really learn the UIs if they're always churning, and the metrics and the novice users falsely encourage teams to constantly churn their UIs.
I strongly believe that you'd see better UIs either with far fewer metrics, or with products that have smaller, expert-level user bases.
I don’t believe either is the primary driver of modern UI design. Cynical as it may be, I think the only things that get any level of thought are:
1. Which design is most effective at steering the most users to the most lucrative actions
2. What looks good in screenshots, presentations, and marketing
The rest is tertiary or an afterthought at best. Lots of modern UI is actually pretty awful for those mentioned bottom of the bell curve users and not much better for anybody else in terms of being easy to use or serving the user’s needs.
Proper use of analytics might be of assistance here, but those are also primarily used to figure out the most profitable usage patterns, not what makes a program more pleasant or to easy to use. They’re also often twisted or misused to justify whatever course of action the PM in question wants to take, which is often to degrade the user experience in some way.
There's a much simpler explanation. At some point, the UI becomes about as good as it can be. It can't really be improved any further without changing the whole paradigm, and just needs to be maintained.
But product managers inside the large corporations can't get promoted for merely maintaining the status quo. So they push for "reimagining" projects, like Google's "Material Screw You" UI.
And we get a constant treadmill of UI updates that don't really make anything better.
Just because they're measuring doesn't mean they're measuring the same things as before.
The goal in 1995 might be "The user can launch the text editor, add three lines to a file, and save it from a fresh booted desktop within 2 minutes".
The goal in 2015 might be "we can get them from a bare desktop to signing up for a value-add service within 2 minutes"
I'd actually be interested if there's a lot of "regression testing" for usability-- if they re-run old tests on new user cohorts or if they assume "we solved XYZ UI problem in 1999" and don't revisit it in spite of changes around the problem.
Telemetry may tell you the "what" but, at best, it will only allow you to infer the "why". It may provide insights into how people do things, yet it will say nothing about how they feel about it. Most of all, telemetry will only answer the questions it is designed to answer. The only surprises will be in the answers (sometimes). There is no opportunity to be surprised by how the end user responds.
It can look better. This is basically a distro with Chicago95 out of the box and not well configured. If you take the time it can look more like 95. The Chicago95 screenshots IMO look better:
Ouch. That screenshot is uncomfortable to look at. The window title bars are painfully narrow, the frame borders have inconsistent thicknesses, the Start menu overlaps the taskbar, the vertical centering of text is wrong.
The answer to your question is that these replicas are of low quality. This one looks like the whole thing was made by someone (or a committee of people) lacking attention to detail.
> The whole UI as branding thing has utterly killed usability.
This is caused by a change in who is hired as UI/UX developers. In days past it was HCI experts and engineers, now it's graphic designers. "Pretty" is the order of the day, not "useful". "There are too many menu items" is now answered with "So let's hide them" when it used to be "How can we organize them in the UI us a simple, discoverable manner?" But then that "overflow" menu (really? Needed menu commands are now OVERFLOW?) gets crowded so they start just removing features so the UI is nice.
Having worked with amazing HCI experts over the years, you've hit the nail on the head. It's wild how much design is done for designs sake at my work, with nary a nod to HCI given. The a11y team try to patch over it as best as possible, but we end up with a mess, and I'm treated like a pariah for pushing back on some of it
I'm glad to hear my suspicions and impressions are accurate. I have a number of friends who have gone into UI over the past 20 years and while they're all very smart people, In my opinion exactly 1 of them has what it takes to work in UI/UX. The rest are mediocre graphic designers with average to no training in graphic design, and zero experience in anything technical. Only one knew what Fitts law was, or what HCI meant, or what GUI meant
>This is caused by a change in who is hired as UI/UX developers.
„UX/UI developers“ is a strange name for it.
In 2000s the web enabled more sophisticated presentation designs and there was a push from client-server to web-based applications using incredibly strange technologies for building UIs — HTML, CSS and JavaScript, which gave the rise to UX design as a interdisciplinary job (HCI+digital graphics design). By 2010 the internet of applications kicked off and in mid-2010s moved to mobile, dramatically increasing the demand for UX designers. By then it actually mattered more who is hiring designers, not who is hired. Since only relatively small fraction of hiring managers does understand the scope of this job even now, they even started calling it „UX/UI designers“ or „Product designers“ as if that name change could help, still judging design work by often-fake screenshots on Behance rather than by case studies in the portfolio. Even HCI professionals are often reduced to mere graphic designers by those managers who skip research and apply „taste“ to a science-based discipline. At the same time, since UX design is one of the most well-paid and less stressful creative jobs, a lot of people switched to it without proper education or experience, having no idea what is statistical significance or how to design a survey. And voila, we are here.
I agree. It's a UI Engineer. User Experience is just the fluffification of the title to something that sounds expansive and nebulous when it's actually pretty focused and critical.
Then you don’t understand it too. Software solves user problems by offering experience, not UI. Not every solution requires UI, but every solution creates user experience.
It's interesting especially because it seems like companies today pour tens of millions into "accessibility", but I never see a thing's usability in terms of simple and easy-to-do-what-I-want UX fall in to the same category.
Companies are outsourcing testing. I'm not surprised that they get rid of UI testing. Back in the day companies used to invite people to sit down and use their software. Nowadays they just push out whatever they have and then start collecting bug tickets. Then they let the community to vote on the tickets. It's basically a huge "pay for being a beta tester" scheme.
It’s a completely predictable result if you think about it.
Old style UI was developed with the findings of countless man-hours of UX research performed by field experts, while branded UI is typically whipped together purely based on trends and vibes in an evening by a visual designer who’s probably never performed an ounce of serious research or user trials. It’s natural that the latter is only going to be good at the most superficial of purposes. UI as branding is the McMansion of UX.
I think it’s worse from a time wasting standpoint, really - a lot of modern UX does have thousands of hours of UX research dumped into it, but with faulty metrics driven goal seeking and internal politics bolted on. I agree that Vibe Branding killed UX in the way you describe in the 2000s (remember when every company had some abominable Flash site?!), but now, we’ve come full circle: from the ashes we’ve allowed warring factions of UX researchers to return to create hundreds of carefully constructed disparate systems with no consistency.
I don’t think we’re quite back to where we were, because branded UI widgets are almost always devoid of functionality compared to their traditional UI toolkit counterparts. If a feature is even slightly “power user”, branded UI widgets probably don’t implement it, even in tools made for technical users.
One of my favorite examples is tree-style lists (“outline views” in AppKit nomenclature). On macOS these have a very convenient functionality where holding down option while expanding/collapsing a section performs that action on all children as well, and it’s practically never implemented in custom-built tree widgets even in cases where the primary audience skews Mac-heavy.
It's also repeating what the hellscape of inconsistent skinned UIs did in the late 90s and early 2000s. People are looking back at those times with a rather selective memory.
The themed UIs of that era were very superficial -- if they applied to serious software at all, they were just a cosmetic layer on top of an otherwise well-engineered interface, and could be easily disabled. Most people I knew, for example, disabled the theming engine that shipped with Windows XP. Most applications that supported UI skinning still had a default or fallback UI that adhered well enough to modern conventions.
Not so much anymore. The abandonment of any coherent organizing principle to UI layout in favor of pure aesthetics has been a massive regression. Reasonably complex software often doesn't even include menu bars anymore, for example.
When I used XFCE as my daily driver, I once tried installing Chicago95 just for nostalgia and it stick as my daily driver for almost a year! The UI is less distracting than modern UIs and there's something to it that makes it easier to just know which window is open over which window that's lacking in modern UIs (I think it's the over-reliance on soft shadows and the borderless windows).
Eventually, I stopped using it because: 1- it was always annoying to send an screenshot to someone and have to explain that no, I wasn't using Windows 95, and why; 2- the grey-ish look of everything started to bother me over time; 3- I wanted a more integrated desktop experience and moved to KDE Plasma. Still, I configured my Plasma to work like old Windows: window titles on taskbar, zero to none animations, etc.
I also dailied it on XFCE. The UI was very utilitarian and purposeful. I suppose aesthetically it is unimpressive and not streamlined, but it serves the purpose of being a good interface to do a task.
Same as you say, people have asked me a lot about it and even asked me if I could set it up for them. The theme is evangelizing Linux a little bit, and that is interesting. In the right hands, these UI principles could convert many people to some product.
>and there's something to it that makes it easier to just know which window is open over which window that's lacking in modern UIs (I think it's the over-reliance on soft shadows and the borderless windows).
I think this started with Vista, I remember watching a video criticizing the new love of glass effects on UI chrome as it got rid of or minimized the color/shading difference between focused/unfocused windows. The example the video used was 6 notepad windows and pick which one was focused, and the main cue you'd need to look for is that the window with focus had a close button colored red.
Thin borders and minimalist/hiding scrollbars is another one that annoys me, give me something graphical for my gaze to grasp.
This entire comment section has me repeatedly thinking to myself "you don't run into that problem with i3-alikes" over and over. My choice not to put up with modern UX bullshit is feeling strongly reaffirmed right now. Not that it needed to be.
> it was always annoying to send an screenshot to someone and have to explain that no, I wasn't using Windows 95
That's not a negative, that's a fringe benefit as an endless source of entertainment.
Microsoft Windows programs hid functionality under layers of menus and the registry. MacOS, at least, surfaces much less functionality, because it offers sensible defaults. I never had to do anything akin to fiddling with the Windows Registry.
I did like some Windows things, though, like the ribbon, and reconfigurable UIs. Today's UIs are more immutable, for the worse.
I'd agree macOS surfaces much less functionality but I feel like it's more "because they don't want you to feel like there is a choice to make in the first place" rather than "because the defaults are ideal for everyone". Over time it feels like "layers of menus" have definitely made their way into Apple's software anyways.
The replacement to the registry seems to half be "magic CLI incantations for settings which can't be found in the GUI for some reason" and half "here's a $4.99 app to 3 finger tap to close tabs".
And what difference to end user it makes where exactly the key/value data is stored? No real difference whether the data is HKEY_CURRENT_USER\Software\MyAppName or com.my.app when you're trying to coerce some internals whose configuration is not exposed because you're not worthy of it
It was common in the Windows 9x days for the two Registry Hives (SYSTEM.DAT and USER.DAT) to get corrupted leading to an unbootable system or to get fragmented and/or full of disused values from poorly-uninstalled software leading to increased memory usage.
Here are some KB articles to check out for context:
There is - removing a wonk preference namespace is as easy as `rm ~/Library/Preferences/com.cheapskatesoftware.wonko.plist`. Whereas the Windows Registry is a monolithic piece of gunk you need a Microsoft editor for to zap something
I’d say that a quick defaults command is probably on the whole more friendly than trawling around in the arcane mess that is the Windows registry. It’s not as friendly as it could be, but at least it’s a somewhat human readable one liner.
It’s also reasonable to back up plists and/or sync them between machines like some users do with their dotfiles, because they’re just files.
It’s not a 1:1 mapping, but much power user functionality in macOS is designed to progressively reveal itself as the user becomes more technically capable, a type of design known as progressive disclosure. This allows newbies to not feel overwhelmed while also allowing power users to feel at home.
The problem is that way too many people approach macOS with the Windows way of doing things firmly planted in their minds as “correct”, which interferes with this process. For example, over the years I’ve encountered numerous posters complaining about how macOS can’t do X thing, after which I point out that X thing is right there as an easy to find top level menu item, but the poster in question never bothered to take a look around and just assumed the functionality didn’t exist since it wasn’t surfaced the same way as under Windows or KDE or whatever they were coming from.
Of course there are things macOS just doesn’t do, but there’s plenty that it does if users are willing to set their preconceptions aside for a moment.
If you approach macOS the Linux or BSD way, it feels like Windows Powershell.
Of course you can use brew and stuff, setup you dev enviroments etc.
But when it comes to system settings, its bad, very bad.
Also stuff like docker, k8s suffer performance and usability.
Docker, etc are going to suffer on anything that’s not Linux due to how coupled they are to Linux. Even WSL isn’t as good as bare metal Linux in that regard. To me it speaks to a need for return to platform agnosticism in dev tooling more than anything.
One thing that I always hated about macOS is the menu bar placement.
Ironically, in the long run, it has proven to be an asset for the simple reason that any macOS app has to have a main menu with commands in it, if it doesn't want to look silly. So this whole modern trend of replacing everything with "hamburger" menus that don't have the functionality isn't killing UX quite so bad there.
Although some apps - Electron ones, especially - stick a few token options there, and then the rest still has to be pixel-hunted in the window. Some don't even put "Settings" where it's supposed to be (under the application menu). Ugh.
On the last paragraph, something is better than nothing, though. It’s always bugged me that Electron doesn’t offer Windows and Linux users a way to enable the menus that’ve been provided for the Mac version.
As somebody who recently had to switch to Mac for work, my experience has been the exact opposite of this. Every other OS I've used since Windows 95 I've been able to get to grips with the same way: start off using the mouse to find my way around the UI, and introduce keyboard shortcuts as and when I find them useful. Eventually I get to the point of being able to use either exclusively keyboard or exclusively mouse for most tasks.
MacOS seems to _require_ some unergonomic combination of both from the get go. Some basic things are easy with the keyboard but hard/impossible with the mouse and vice versa. The Finder app doesn't even have a button to go 'up' a directory for god's sake.
It helps to keep in mind where the OS and its core user base is coming from.
For the case of the up button for example, prior to OS X the Finder was a spacial file manager where each folder had a single corresponding window that remembered its position on screen, allowing users to rely on spacial memory to quickly navigate filesystems. Its windows didn’t even have toolbars, because they weren’t navigator windows — every time you opened a folder you got a new window (unless that folder’s window was already open, in which case it was foregrounded).
So when OS X rolls around in ~2000 and switches the Finder to navigator windows, they’re looking at what existing users will find familiar. Back/forward is easy since most had used a web browser by that point and map cleanly to most people’s mental models (“I want to go back”), but up? That’s a lot more rare. A handful of folks who’d used an FTP client might’ve been familiar with the concept, but few outside of that few would’ve, and how “up” relates to a filesystem is not in any way obvious. And so, the Finder never got an up button, just a key shortcut because anybody advanced enough to be hunting down shortcuts is going to understand the notion of “up” in a filesystem.
> I never had to do anything akin to fiddling with the Windows Registry
If I recall correctly, when I got my first Macbook, I had to edit plist files or something similar in order to do basic things like permanently showing hidden files, showing the full path in Finder, show file extensions for all file types, increase the animation speed so the computer didn't feel slow as molasses, etc, etc.
Maybe these things are now easier to configure via GUI on macOS?
One thing that is weird is that you’re expected to look around the menu bar holding the option key as the menu contents change when that is pressed (also applies to tray icon menus, e.g. WiFi icon shows a lot of stuff when option-clicked.) IIRC some of what you say can be toggled with option menu items.
The last time I was setting up a system, it's still very difficult to find in the menus. If it's not discoverable and I have to know the incantation/shortcut to do it, then that's bad UI.
MacOS is pretty cursed. The equivalent to registry fiddling is doing anything in ~/Library/Application Support
It still has "Services" as a hold over from Next that is completely broken and unused (but still present in every app for some reason). Now you also have the joy of diving deep into the Settings every time an app needs some sort of permission.
I'd say something about .DS_Store files, but that's not really UI.
I'd much rather work with plain-text human readable property list files with straightforward `defaults` commands than the hive of Hell called the registry.
> I never had to do anything akin to fiddling with the Windows Registry.
I don't believe you. You have never, EVER, NOT ONCE run a terminal command to change an option on MacOS? I just refuse to believe anyone on HN hasn't altered preferences in the terminal on MacOS.
Not with anything near the frequency of Windows. The last such thing I remember doing is restarting the locate indexing service with launchctl. I do lots of things in the command line, of course, but not so much to configure MacOS itself.
There is an entire ecosystem of free and paid Mac apps meant to augment the Mac experience because MacOS does not provide functionality or configuration needed for a sensible computing experience out of the box.
I think, the main difference of MacOS and Windows is, that Windows allow drivers from 3rd-party. MacOS does not. Drivers means also hardware. So you can build your own PC. Same as with Linux.
This is the Apple secret of success IMO. No 3rd-party drivers and hardware, means, it will just work and no one will blame you for stuff 3rd-parties messed up.
But its also like: There is only a red and blue t-shirt. Choose. No gray, no white, no yellow, no printings.
macOS allows third-party drivers too, Apple just wants vendors to write them in userspace rather than kernelspace. That’s probably not the worst thing, because proprietary driver code is notoriously shoddy and should be run somewhere that limits the blast radius.
That’s mainly restricted to graphics cards. Audio cards like used for production as well as I/O (USB, etc) and networking cards have drivers and work fine given you have a PCI-E slot to plug them into, and of course almost anything external connected via USB or Thunderbolt works fine. For GPUs, it’s only a specific subset of users that needs a discrete GPU, especially as the GPU built into M-series SoCs has become powerful enough for most uses outside of high-end gaming.
Selecting the appropriate tool for the task at hand is crucial, in my opinion. However, I believe the choice is often influenced by companies mandating the use of Microsoft and Mac systems due to cost and maintenance considerations, rather than allowing employees to choose between Mac, Windows, or Linux based on their preferences. Proprietary software that only runs on Mac or Windows, never was an argument, because you can just RDP stream remote desktop apps or use the browser.
> My hypothesis is today's "modern" OS user interfaces are objectively worse from a usability perspective, obfuscating key functionality behind layers of confusing menus.
I think if you went back and actually tried to use these old UIs you would realize that one of the reasons stuff isn't hidden behind layers of menus is that in a lot of cases the 'hidden' features just didn't exist back then.
Mobile too, I'm sick of those OSs updating every 18 months because some product person along marketing decides the wheel has to be reinvented otherwise there will be no buzz.
I have a harder and harder time navigating both iOS and Android as time goes, should be the opposite.
> The dominance of Windows is crazy, even today, Mac desktops and laptops are comparatively niche
I was actually surprised that macOS/Linux/ChromeOS together are >20% of all desktop/laptops. I would have expected Microsoft machines to be closer to 90% than 80%.
I'm surprised by the breakdown as well. At least according to these two citations in Wikipedia, the breakdown is:
---
For desktop computers and laptops, Microsoft Windows has 71%, followed by Apple's macOS at 16%, unknown operating systems at 8%, desktop Linux at 4%, then Google's ChromeOS at 2%.[3][4]
---
[3] "Now more than ever, ChromeOS is Linux with Google's desktop environment". About Chromebooks. 1 August 2023. Retrieved 25 September 2024.
[4]"Desktop Operating System Market Share Worldwide". StatCounter Global Stats. Retrieved 9 March 2025.
Honestly, I don't think anyone has done real user research on basic interfaces since Microsoft did it for Windows 95. I'm pretty sure that was the last publication I've seen.
It's a lot of time, effort, and expense to run user research, but the potential benefits to the users are big.
For modern Windows operating system, i.e. >= Windows 10, getting rid of ads, weather, auto-update and search bar can improve UI usability significantly.
Windows 95 was not stable enough back then. I believe Windows 2000 was the first OS that was easy to use and relatively stable. XP and 7 are both solid options too.
I have used both MacOS X and modern MacOS (15) and the UI of X is definitely way better than the UI of 15. It is more clean cut.
Yes. I'd love to see someone take the basic design of Windows 95 or even early OS X and reimplement it not so much visually, but tactilely. Make something that works as well, is as simple but isn't nostalgia.
In any case though, this particular attempt at giving a complete Windows 95 experience is quite cool.
The Windows dominance really puts things in perspective... Even with the rise of mobile and Macs making some headway, it's amazing how entrenched it still is on the desktop.
Give today's XFCE4 a try, on debian stable or testing. It's a remarkably no bullshit GUI for use with xorg. You can of course still install all the gnome and kde libraries and run all the gnome and kde applications, though things will look a little bit mismatched.
> The dominance of Windows is crazy, even today, Mac desktops and laptops are comparatively niche
But for whom and to do what? People around me, like my wife and mother in law, are happy with an Android phone. If they use a computer, the only thing they need is a browser.
I wouldn't be surprised if 90%+ of all Windows users would use Windows for one thing: click on a browser icon.
That's why after one malware too many I confiscated my mother-in-law's Windows laptop and got her a Chromebook.
I'd say that's why the shit UI/UX doesn't even matter anymore for most users: the only use a browser anyway.
What do you mean exactly? Like the Menu-Issues in Windows 10?
Because from a UX perspective, basically nothing has change. UI of course, but UX is the same like in the 90's following "The Design of Everyday Things" by Donald A.
I think its more about the change management, expectations. For example in Win XP you had the option to use the NT theme. As a user: "I can decide when to move on to the new design."
Usually around 50% of your users are conservative about change. You have to keep this in mind when u change design.
On the other hand, if you sell a product with subscription, you have to introduce new feature, else user will move to another product. But, when you introduce new feature, UI gets more complicated and user will blame you for that.
Like making window borders 1px wide, even as screen pixel density increases. It's darn near impossible to resize a window anymore.
Like making buttons auto-hide unless you mouse-over them. I don't remember when this came in, but the default PDF viewer in something did this, and I spent _weeks_ being baffled that some jerk made a PDF viewer that couldn't zoom in on the page, until I randomly waggled the mouse for some reason and the missing buttons magically appeared. I have no words for how upsetting this was.
Like having icons-only for many functions, with no text-and-icons or text-only option to replace them. I'm sure some people are fine with that, but other people can scan a screen for a desired word MUCH faster than they can scan for a desired icon, and removal of text labels is just an insult to that segment of the userbase.
Like no longer highlighting, or even having, hotkeys for many menus. I can alt-space or alt-menukey my way through a late-90s menu tree _way_ faster than I can mouse through it, even with today's better mice, but that simply doesn't work anymore in a great many programs.
It's one thing for people who've never known a different UI to just be slow in this one and that's all they've known, and that's fine for them I guess, if it's pretty and they prefer that, or if keyboards frighten them.
But for people who have DECADES of reflexes invested in these shortcuts to suddenly find that they don't work anymore, and we're forced to SLOW DOWN and be less productive than in the past, that's a high insult.
Microsoft spilled tankers of ink in the 90s talking about how their new GUI patterns would make people more productive by unifying these things across programs (which was true; in the DOS era every program made up its own shortcuts and ways to access them), and folks who learned them are now being punished for trusting MS with our loyalty.
Thank you for the recommendation, I'll look at installing that.
But the problem with add-ons is that every machine you use will have a different combination of them installed. Maybe you're at work and don't have privileges to install them. Maybe you forgot to install one on your desktop even if it's on your laptop, etc.
And building it into "a lot" of Ubuntu-based distros lacks discoverability. I might have that now on the machine I'm typing this on, but it does me no good if I don't know it's there. (Everything in Linux has worse-than-terrible discoverability, but that's another rant entirely.)
MS's dominant position meant their defaults Just Worked everywhere, and when those defaults were good, they were really, really good, by virtue of their ubiquity. Then they fucked us by using their dominant position to just... I don't know... completely lose the plot? Aside from HiDPI fractional scaling and support for large monitor "maximize to a quadrant" and stuff, I can't point to a single MS UI improvement since the XP days. Everything else has just gotten worse, fragmented, and for no good reason.
I agree, thats bad. And for example the "icon only" thing follows a bad but hip UI pattern where designers assume the knowledge of the icon meaning of the user. They should not in my opinion.
I mean, in the end, you can decide.
1. To learn all this new patterns in windows.
or
2. Move on to another, more stable window environment like gnome or KDE or whatever.
In the end, its all about the effort for now and on the long run. And you get forced to calculate that because of the introduced change.
Have you used Windows 11? When right clicking, there's now a context menu within the right click context menu. To see what you could see in Win10, you have to right click, then select "See more options" or something. Which just opens up the "old" Win10 context menu with a totally different visual appearance than the Win11 one. Talk about jank and bloat.
I'm sorry, but absolutely no. Fuck no. Nothing from Microsoft has even been in the same building as a copy of "The Design of Everyday Things", or as a copy of any old-school UX book from before UX meant "Electron". UX is just as much about the "how" as it is about the "what", and Microsoft has been failing everyone lately on this count.
Thats not true. For example simple stuff like Toggles to the hole Windows management is derived from that. IMO the huge change in Windows 11 is how the Menu, App Starter and so on works (if you use the mouse).
Lately I've been strongly considering helping migrate my parents to Linux. Their needs are primarily web-based with some basic productivity tools mixed in, and Windows has just been getting more and more hostile. On top of this, they're at an age where they're now more susceptible than ever to various scams/attacks, and shutting down an entire category of problems by removing Windows from the picture is increasingly attractive.
I had forgotten that Chicago95 exists, but this might be exactly the right thing. They'd immediately find it familiar, and while the theme isn't the whole story, this would go a long way in easing the transition I think.
A chromebox mounted behind the monitor did the trick for me. Haven't had an emergency wipe/reinstall in years. Also, a tablet with keyboard takes some of the pressure off having a "computer" and you can go with iOS or Android depending on what phone they use.
I've been evaluating a Lenovo ThinkCentre m920q tiny I picked up for not very much money on eBay (the m720q models are even cheaper) and they seem like perfect machines for the task.
My parents use some tools and hardware that require a full OS so the tablet route isn't an option, but I'm starting to really like the idea of deploying a couple of these micro PCs.
Out of the frying pan, into the fire. Exposing your parents to total surveillance (from one corp to another) is not what I'd characterize as safe or friendly. Linux is fine these days if the hardware is supported, and you can use an immutable distro if extra reliability is warranted.
Strong disagree. I mean tracking is what it is, but it's happening regardless and if they are using Chrome to browse and Google services they are being tracked.
ChromeOS seems to work really well though and is dead simple and intuitive. It used to be incredibly awesome with crouton but that's mostly dead. Crostini is acceptable though. I would absolutely recommend ageing people getting a chrome device for security and simplicity.
Plus running any android apps on desktop gives even more software options than any other desktop for most consumers.
Reliability is a commodity these days, and we don't use Google services.
Apparently, ubiquitous surveillance is acceptable to you. Not an uncommon stance, but I'd never recommend it to others. It's a disservice when freedom and privacy respecting choices, that are just as reliable, already exist.
You also can't properly back up the system unless you install a custom ROM and take on the associated maintenance burden. (At least Android. I'm not sure to what extent these things are left up to the whims of the developer on iOS.)
I installed ubuntu for my mother, she just needs to download pdfs and read pdfs, look at images, use gmail. Sometimes she opens a document with LibreOffice, but no power usage.
Seems to work, the maintenance is also now super easy, ssh, update. Something wrong and she needs support? I ssh, open up a tunnel and connect via remina to her desktop to explain.
I had a situation once when Ubuntu did literally not go into the Desktop Environment anymore, but all I did was update and upgrade packages and it started working again.
IMHO, Fedora's Atomic Desktops[^1] are the way to go for that. Automatic upgrades you can roll back if something breaks? Yes, please.
Universal Blue[^2] has some spins that got a glow up, but their dev team gives a bit of the "everything old is bad" vibe.
OpenSUSE's MicroOS[^3] desktops aren't ready for nontechnical people, but their atomic upgrade strategy is much faster and simpler (btrfs snapshots). I'm keeping an eye on it.
My daily driver is NixOS and part of me really wants that level of predictability and rollback for them. For a brief period, I had started thinking through what it might look like to remotely manage this for them. But my ultimately goal is to help them achieve autonomy, and only step in when necessary.
I may have understated their needs somewhat. Most of what they do is browsing and document editing, but a few key use cases make a real computer necessary (or at least highly desirable):
- Document scanning
- Label printing (my mom buys/sells stuff on eBay)
- My dad still works and writes proposals/manages invoices/does complex taxes
At a minimum, they need a full desktop environment. Most of these things have decent 1:1 Linux alternatives, but one or two might necessitate a single-purpose Windows VM when all else fails.
Two pretty decent used micro PCs will also cost less than a single iPad.
I like themes like this. The only thing that hampers the authenticity for me, and this isn't the fault of the author really, is the super high resolution fonts compared to what was available back then. There's just something charming about low resolution fonts that are readable enough on screen, probably nostalgia.
I think any type of pixel font authentic to a couple decades ago won't look good on a 4K monitor, unfortunately. It got to the point where I ordered a 1024x768 monitor just to play old games with a period system.
Pixel fonts don't accurately represent the 90's UIs because we don't use CRTs anymore. The poor souls buying the very first terrible flat screen monitors may have used computers like that, but most of that era was experienced using smudgy, edge blurring CRTs.
You could probably create a CRT-filter-based font for high resolution screens (though you'd probably still need to optimise for subpixel layout for accuracy, even on 4k monitors).
Most people vastly overstate the effect that CRT displays had on the appearance of legacy software.
Yes, very early on, when people used TVs or cheap composite monitors as the display devices for their computers, there were blurry pixel edges, bloom effects, dot crawl, color artifacting, and all the rest.
But by the '90s, we had high-quality monitors designed for high-resolution graphics with fast refresh rates, with crisp pixel boundaries and minimal artifacting. CRT filters overcompensate for this a lot, and end up making SVGA-era graphics anachronistically look like they're being displayed on composite monitors.
CRT monitors did not have "crisp pixel boundaries". A CRT pixel is a Gaussian-blurred dot, not a "crisp" square as it is on modern displays. What "high-quality" CRT monitors did have was higher resolutions, even as high as 1600x1200, where individual pixels are basically not distinguishable.
By the early '90s, high-quality CRT displays had low dot pitches or very precise aperture grilles in addition to supporting a wider range of refresh rates, and better clarity of display was a major selling point.
People were typically using 640x480 or 800x600 in GUI enviroments, and most DOS games were at 320x200. 1600x1200 was incredibly uncommon, even where the video hardware and monitors supported it -- people were usually using 14" or 15" 4:3 displays, and that resolution was way too high to be usable on displays that size, and the necessarily lower refresh rates made flicker unbearable at higher resolutions.
At the common resolutions and with purpose-built CRT monitors, pixel boundaries were quite clear and distinguishable.
> At the common resolutions and with purpose-built CRT monitors, pixel boundaries were quite clear and distinguishable.
Being able to clearly resolve individual pixels (which I agree was a thing at resolutions like 640x480 or 800x600. 1024x768 is pushing it already though) is not the same as seeing "crisp" boundaries between them. The latter is what I was objecting to. 320x200 (sometimes also 320x240 or the like) is a special case since it was pixel-doubled on more modern VGA/SVGA display hardware, so that's the one case where a single pixel was genuinely seen as a small square with rather crisp boundaries, as opposed to a blurry dot.
Another issue with modern recreations of old UIs is that the dimensions are usually subtly wrong, which for me ruins the feeling. Some of that is related to the fonts having different height, but in many cases it is just that something is one-pixel off and just looks wrong. For the 95-style UI the common issue are control borders (especially the high-light side of "3d" controls), of which there is a huge amount of examples on the screenshot.
I actually would think less of this as a look back into the past but hopefully as a real alternative to the current DEs, which obviously then needs to have high res fonts. That would be nice.
If we're talking Windows 9x, the "original fonts" could also be TrueType, hence arbitrarily resized. Yes, the original Windows 95 included a pixel font for the UI but then TrueType fonts like Verdana and Tahoma were added soon after that and were commonly used.
You're right in that there's nothing stopping one from doing so (I even use an integer scaler for old games on my main computer), it's just a tradeoff between "doing what's possible" and "having the most authentic experience one can".
If we're talking about the subjective experience of recreating "a child's bedroom computer" from the mid 90s-early 00s, a widescreen aspect ratio alone would be jarring, since my conception of a monitor for such a system is a 4:3 CRT. So for me, little else would reach that level except a system with the same aspect ratio and a similar DPI.
Not only that, but UI design itself has undergone many shifts since that era to account for the types of monitors those UIs are being designed for. There's not as much of a need for pixel-perfect design when vector-based web UIs dominate the desktop application space nowadays, relegating those that go back to older UI paradigms to enthusiasts who still remember earlier times. Or maybe people who develop for fantasy consoles.
I should mention while I'm at it that those sort of faux-pixel art shaders used in some games come off as quite jarring to me since I expect UIs to be meticulously laid out for the original screen size, not just for blowing up arbitrary content 2x or 4x on a huge widescreen monitor. I sometimes feel those are meant to represent a nostalgic feeling of some kind, being pixelated and all, but really it just makes me wish there were some alternate reality in which people still designed games and desktop applications for 800x600 or 1024x768 monitors again.
It's interesting at present how there's stuff like 4K and then there's the playdate with a relatively tiny handheld resolution, but relatively little interest for new content for those types of resolutions in-between.
> If we're talking about the subjective experience of recreating "a child's bedroom computer" from the mid 90s-early 00s
Is that what this project is going for? I understood it to be attempting to apply design elements from that era to create a superior UI for a modern "child's bedroom computer".
- Windows 10/11. Especially in 11, it's easiest just to type the start of an app's name into the search box. As opposed to the two clicks it takes to get to the "traditional" menu where you still have to scroll to find it.
- Gnome (only on fresh Linux installs, usually replaced with Mate pretty soon). Has a smartphone-style app grid, but here, too, its quickest just to type the start of the app's name.
- Mate: Modern, but still has the Windows 95 paradigm (easy enough to collapse the two toolbars into just one bottom one). Still my favourite desktop environment.
Not all fancy graphic stuff is good. And don't even get me started on how hard it is to drag an app window to another screen these days - on Windows. You really have to find the 2% or so of the top bar that's still draggable and not cluttered up by other stuff.
I used winaero tweaker [1] to disable web search, the search is infinitely better now.
You can do the same tweak by editing the registry [2] if you don't want to download an app for it (though the app includes a lot of additional useful tweaks).
You can download openshell and it will replace the start menu from the start menu from whatever era you want (XP, 7, 8, I think 10 too). It's open source as well
Dunno. On the relatively clean Windows 11 on my laptop, on which I'd never run Teams before, I typed "tea" into the search box (no ENTER) and while it also offered to look up on the web about "Tea" it did show "Microsoft Teams (personal) (app) at the top right with a large, clickable icon.
I think they drank the Macintosh kool-aid and expect you to have all your commonly used apps pinned. In Win11 this even looks sort of like a Mac dock. Still made no sense to ruin the usability of the start menu which they invented.
I am 99% sure that this is the result of academic GUI design and focus testing.
I read all the Windows 8 development blogs, and everything they wrote about seemed absolutely justified. Then you actually use the thing and it was a nightmare.
Same with their approach to hardware, the Duke Xbox controller tested really well, but then someone with daintier hands went to use it and uh actually its great for like 15% of the user base.
I think the desktop operating systems of that era were at a sweet spot. They were technically advanced enough to render good-looking, crisp color user interfaces. However, most people were still novices at using computers, so OS designers consciously designed their operating systems to be as clear as possible. Applications tended to be written for each individual platform and to follow its UI guidelines.
Windows 95, NT, System 7 and System 8, BeOS, and NextSTEP all had really clear UX. You always knew where to drag a window, what was clickable, where to find settings, etc.
An aspect of System 7/Mac OS 8/9 that I find criminally underrated is how flexible it is.
For those versions, a good bulk of the “system” isn’t part of the system proper but instead implemented by way of extensions and control panels loaded at startup. The OS itself is extremely minimal, basically just enough to provide a barebones desktop and act as a substrate for apps to run on. Everything else, including “core” functionality like audio and networking, was implemented in an extension.
This meant that you could pare back the OS to be extremely lean and only have the exact functionality you personally needed at that precise moment and nothing else, and doing so didn’t require you to recompile a kernel or anything like that — just create an extension set that only loaded what you needed. This was excellent for use cases like games and emulators where you wanted every last ounce of resources to go to those, and nice for single purpose machines too (no point in loading game controller support on a box that only ever runs Photoshop and Illustrator).
Of course the way it was implemented is awful by modern standards, but the concept is golden and I think there should be OS projects trying to replicate it.
I remember creating different extension sets using the built-in Extension Manager and the third-party tool Conflict Catcher. I had sets for gaming, video editing, and normal usage. It was a simple matter of selecting the correct set and rebooting. Or you could hit shift on startup and start into a minimal system without any extensions.
There's a good reason the third-party extension manager was called "Conflict Catcher," but the power and flexibility such a system grants users is unmatched.
"I've come up with a set of rules that describe our reactions to technologies:
1. Anything that is in the world when you’re born is normal and ordinary and is just a natural part of the way the world works.
2. Anything that's invented between when you’re fifteen and thirty-five is new and exciting and revolutionary and you can probably get a career in it.
3. Anything invented after you're thirty-five is against the natural order of things."
-- Douglas Adams
It's ok for people to waste time building stuff they think is cool. Did it need to be a distro? No but it also didn't need to exist. I'm glad it exists though, I think it really whips the llamas ass!
This attitude is why Linux based operating systems have such poor market share on the desktop. Opportunity costs are real. Friction is real. You don't see Windows creating a new OS for a single theme. You don't see macOS do it either.
I see it as one of the consequences of freedom, but perhaps also a gap in packaging where they can't bundle up their changes in a form to be applied onto another base.
That 'base' is one issue I've been thinking about with linux, I have similar concerns about the cost of everyone being able to make their own distro for their own slight variation on something else. It's not that I think it's a bad thing to pathfind in new areas, but the replication in building/supporting it all, getting users to pick between 4 similar variants of the same thing, and accounting for "you're using KustomLinux, which is 2 steps removed from CommonLinux" and all the little differences between them. It's an interesting contrast against standardization, but I can't help wondering how it would change the approachability of linux if the starting point was limited to one of the big distros and then variants are layered on top of that.
I agree. I've gone down the "make Linux look like Windows 95" path a couple times and while it's a fun yak shave, it's tough to get all the little changes and details right on one system, let alone your friend's system who want to play with the same DE for a day.
The reason Linux has poor desktop market share has nothing to do with a fun themed distro someone created as a side project.
> You don't see Windows creating a new OS for a single theme. You don't see macOS do it either.
I'd also consider the behavior of both Windows and OSX to be a warning to avoid, and not an example to emulate.
But the line doesn't always have to go up with every breath everyone takes. It's ok to do stuff just because it's fun. Not every single action needs to increase market share.
Why would the Linux ecosystem -- a diverse community of lots of different individuals and organizations all pursuing their own particular goals -- be singularly concerned with increasing desktop market share of Linux as a whole, and all pursue that in the exact same way?
I wouldn't call it an entirely new distro. It's just an Fedora image bundled with the necessary changes to create the UX. It doesn't provide its own software repositories. It's more like an unofficial Fedora Spin.
I see, it still seems like the kind of project that would be much better suited to a DE packge-group style release. I think very few people will want to reinstall their OS just to try it.
would be nice if it would have wine installed so it can run most windows apps for where there is no good linux alternative.
xyplorer, sumatra, irfanview
I love the niche of enthusiasm that exists for the Windows 95 UI. It's not an original point that aside from nostalgia it's a really clear and usable design; but that leads me to wonder, are there any modern UIs/themes/etc that are inspired by (rather than necessarily directly mimicking) Windows 95?
Would be interesting to see what a modern version of Windows 95 would look like, or what general design lessons can be learned from it's niceties.
And mine just said
]▩
as it waited for you to type an Applesoft command. It is always weird when people say something is from "your childhood" as opposed to theirs. I remember the 1990s, sure, but I was already an adult.
There are some pretty good desktop environments for Linux which emulate the Windows desktop, so that old Windows users would feel at home immediately.
But I've never seen them emulate the filesystem, which is what took most old Windows users the biggest effort to understand. And the Linux filesystem raises it to a new level of complexity, which makes every old Windows users want to go back to Windows immediately.
With "old" users I don't mean experienced users.
Is there some kind of overlay which does all this `C:\User\afidel\Desktop` mapping for those users?
My Windows 11 C:\ no longer has a `Documents and Settings`. It still works, though. Windows 8 has it as a hidden junction targeting C:\Users, so it was already deprecated there.
That's just the chicago theme I'va had for years now on my XFCE, I think it's a bit lame to steal that theme (not from Microsoft but the maker of that xfce theme), put it in a repository, give it a different name, and pose it as your own work?
The repo literally adds nothing, just a name change.
I think the main difference is that it’s a fully set up installable ISO of Fedora Atomic XFCE with the theme and other tweaks preloaded which is a convenient thing to have, I guess. Also they do credit Chicago95 pretty much at the top of the README.
I've done something like this with the windows 2000 look from time to time, but something I found frustrating with theming xfwm4 window decorations is the non existent ability to create a horizontal gradient across the top of the window like in Windows 2000.
I believe this is the reason you cannot find a proper windows 2000 theme for xfce.
anyone remember XPDE? I'm not sure it was ever finished / packaged in any major repo but came across someone doing a walkthrough of it the other week and it looked pretty complete.
It is hard for me to distinguish between the functional simplicity of desktop computing in that era, with the overall excitement that the explosion of connectivity brought to the world. The internet was a lot of fun and had so many surprising corners. Practically all of it was personal, niche, or experimental content for awhile.
I wonder if Windows 9X was really all that exceptional, or if it was just what people remember driving with as they navigated the new world.
The best modern equivalent to that desktop paradigm I've found is LXQt, although when I use it I find I kind of miss some of the accouterments of the modern desktops.
This looks great with some apps that have matching themes, but I wonder if it quickly falls apart once you rely on apps with very non-consistent UIs (audio/video software, Discord, Spotify, Slack, and basically all other Electron-based apps). Although I guess there might be some matching CSS injection hacks available for the Electron ones?
Neat! It's probably just nostalgia, but I still don't think any modern desktop has been as good as Windows 2000 was. Perfect blend of minimal without hiding things (well, excluding the Office 2000 menu hiding disaster which we will conveniently ignore)
This is awesome. IMHO the Windows 2000 UI was the pinnacle of UX, it's still unmatched to this day. I'd love to see the SerenityOS UI ported on Linux+Wayland someday.
What are the benefits/drawbacks of using this vs. actually running Windows 95 or XP?
I'm assuming the PC will be mostly used for "educational software" (games), which you would want to run on XP. What benefit is there to running Fedora?
Fedora is a modern Linux desktop distribution with active development and support for modern softwares. While using an old OS such as Windows XP is a huge security risk, with very minimal possibilities of running modern softwares. Even major browsers like Firefox and Chrome won't run on XP anymore.
95 and XP are different in terms of software selection.
Linux Benefits: Security patches (safety), software updates (convenience), hardware support, freedom, and privacy. Not to mention a modern browser, terminal, TLS, and filesystems like exfat.
Fedora runs Wine as well. Maybe not every 95/XP program will run, but I'd guess a lot of them do. You could also run the others in Virtual Box when needed.
Ok, but you don't need any of that if you're just running 1990s/2000s Windows educational software on an air gapped OS with no Internet access, which is what you would want for this application.
Why would you want to mess around trying to get programs to half work on Wine when you can just have the real thing?
If you want your kid to have a web browser for educational purposes, I feel like you had probably might as well just hand them the iPad.
I can testify that Xfce+Chicago95 is great as a daily driver, and one of the most consistent GUIs across different distros (some differences in defaults still exist)
Yeah as far as GUIs go, Win95 isn't in the "nostalgia" category for me, I was already well into adulthood.
I kind of get the appeal, but it's also unnecessarily skeumorphic/fake-3d and there were some UX things that made little sense especially lumping all the window controls all together (including the destructive "close" X) where MacOS smartly separated them.
The fake 3d is actually very useful in communicating what is a button or another interactive piece of the interface and what's not. The modern clean uis where everything is a thin rectangle or just text that you are supposed to click are a nightmare.
Old screen caps of UIs with depth feel so relaxing to look at, and I don’t think it’s just a nostalgia effect.
It’s like there’s always just a little extra brain power and attention being used by modern flat UIs, and you get to shut that off when you look at a depth-enhanced UI.
The most important part is that controls are consistent across applications. In that regard, tools and libraries that implement some look and feel rather than deferring it to the underlying environment are a disservice to users.
Windows, for instance, has dozens of ways to do that, and you can find parts of Windows that use an archeological version of the controls. Nobody, it seems, bothered to reimplement the older APIs on top of the new ones.
I agree. Early Macs had to give buttons a 2D distinctive look. A good thing was that the look and feel were part of the OS and not the application, so everything would be consistent.
Interesting that modern macOS now have them next to each other, like Windows.
You'd be hard pressed to call the Window 95 UI pretty, but it is really functional. I'm still a firm believe that the majority of the work we do with computers today could be done within the Windows 95 shell. We need 64bit, more memory, faster CPUs, GPUs all that, but the modern UI aren't really "better", if anything many of them are more confusing. I think a lot of office works would be happy to just have kept the Window 95 era UI for Windows and Office.
Note that NeXTstep actually did the right thing with window controls -- close is on opposite side of the window from the others, so you can't accidentally hit it.
Windows3 and Motif hid this stuff under a menu, so wasn't a huge concern.
But then Windows95, and then (oddly) MacOS through this away in favour of throwing them all together.
Awareness of spatial patterns / frequency of use seems to have been higher among early UX/UI designers than after. I guess maybe because mice became more accurate?
Yeah. My first computer I soldered the chips/resistors/capacitors/etc to the PCB... It had 1K of RAM, and the screen memory had to come out of that too...
Someone wrote chess for it.
[aside]
It was a Sinclair ZX-81, and I was 11 at the time. My parents bought the kit and a second-hand black & white TV with a dial-tuner (no pushbuttons to change the channel) as an Xmas present ...
I loved the TV, it was my TV when we only had one other in the house. I watched everything on that TV (even snooker and swore I could tell which ball was which)... After a couple of months, my dad started to get annoyed I'd not bothered to build the computer, so I was dispatched to the shed to build it.
A few days later (hey, I was in school), the thing worked and I was working my way through the (rather excellent) manual that came with it, getting to know it. One of the logic chapters had an example:
[P]RINT 1+1=2
(It was tokenised input, so you just pressed P and PRINT would come up in the built-in BASIC). Anyone here can see that the answer would be logical-true because 1+1 does equal 2, and indeed the computer printed "1" on the next line.
Anyway, flush with this futuristic knowledge, I set it all up using the family TV in the lounge, and we went through the same thing, just to prove to everyone that it worked...
[P]RINT 1+1=2
1
"I knew it. You've buggered it", said my dad in disgust as he got up and walked out the room. I tried to explain the (new to me) concept of logical truth to him and how the computer represented it, but I don't think he ever really believed me...
[/aside]
Anyway, that Sinclair ZX81 fundamentally changed my life. Computers and computing opened up a whole new world. Some 45 years later I'm about to retire from Apple as one of their most senior engineers, having been here for the last 20 years. Anyone with any Apple device is running some of the software I've written over the years which is kind of cool, but it's time to bow out.
> I get significantly more work done when I unplug my computer from the Internet.
I tried this, but documentation is often a huge problem. Increasing amounts of it are primarily online and not particularly straightforward to mirror locally.
you could try Zealdocs - it was a reason to go reading more of the projects own documentation pages. There are converters for whatever-doc/ssg-framework a project uses to Docset, the packaged format it relies on. Though ignoring that, how approachable the docs are to different levels of experience is another thing.
I learned heaps from treating a REPL as an (offline) escape room, in terms of how to get inline help, variable introspection and debugging tricks. Not every language offers a convenient one though.
A Windows 3.x UI theme would get the childhood nostalgia going much stronger for me than a 95/NT4/98/2000/Me theme does.
Occasionally I boot up Windows 3.1 in a VM and play a game of Solitaire, for old time’s sake - I can run Windows 95 in a VM too, but it just doesn’t have the same pull.
Win3.1 works great in DOSBox, and you can get an even more authentic '80s-'90s IBM-compatible experience using PCEm or 86box, which emulate a wide variety of PC motherboards, graphics cards, and sound cards.
I remember a couple kids with Windows 95 PCs at home. They seemed like such richie rich's. We'd all play Wolfenstein when we'd sleep over at their houses. My childhood computer was a WebTV, hacked to get dialup internet for free, on a 100 lb CRT TV from Goodwill. I finally scraped up the money for an actual PC some time in highschool.
Usually the observation that “a widely used thing is objectively bad” is strong market signal for entrepreneurial opportunity in a big tam.
I for one would welcome a set of deeply integrated ui improvements in a Mac that included a better file manager, better window management, better desktop search, a contact manager just that worked, a messaging client that just worked, audio and camera controls that just worked, a calculator that didn’t suck, etc.
That Git GUI got my attention. Does anyone have an idea what it might be? The titlebar says aurora but when I searched for it all I got was an commercial product with AI nonsense stuck into it.
well this is simply adorable, down to the neocities hosted website
this already is my stack, tho I need to get over to the immutable variety. i'm on fedora 41 i3-spin, using chicago95. been using it for years, with the plymouth bit that gives you a proper win95 startup background during boot
That surprises me. The major DEs (KDE and Gnome) have felt reasonably polished to me for quite a few years now, they haven't felt buggy to me for even longer, and they always felt like they catered to power users unlike Windows and Mac that actively get in your way when trying to customize things.
Several times they've been ahead of the curve and I've been surprised to learn that Windows or Mac were only just getting a feature that I'd had access to for years or more.
Of course at present I only use i3-alikes. I'd characterize those as a return to the sanity of the past.
This looks wonderful to use. Is there video playback?
I’ve been thinking lots about the YouTube algorithm, how it’s based on engagement, not educational value, and how it’s such a huge missed opportunity for our kids.
So I’ve been building my own YouTube exploration app, for my own kids.
It needs the BSOD screensaver from XScreensaver, for sure. And, maybe, DOSBox-X or DOSEmu2.
Also:
- Pan
- Sylpheed
- Audacious with the WinAMP theme
- Hexchat kinda has a MIRC vibe
- Parole looks like WMP from < v9 releases
- You can't simulate a dialer, but with trickle you can mimic a 56k/ISDN connection pretty well
- SyncTERM for BBS's
- ScummVM, with just a bilinear filter, because I played tons of adventures
- There's an SDL2 reimplementation of Space Cadet Pinball at github.
- Trigger Rally would look like a great shareware game
- Pidgin, hands down. Either you were an AOL user in America, or a MSN user in Europe. It has emoticons, not emojis. Add that annoying notification theme with a sound and that would be the very late 90's/early 00's (my early teen years)
This looks neat, but the problem with all these nostalgia fueled projects is that if you use it seriously you’re basically LARPing being in a kinder past. Eventually it just gives an empty feeling, and you long for days when interfaces didn’t just look like this for fun, it’s because that’s what they simply were, and this was the state of the art. But you can never go back to those days.
/data/UNREALTOURNAMENT_settings.png){kind=link}
