It just looks bad. You want to be MacOS? Actually go and copy it, as close to pixel perfect as you fancy.
I've personally experimented with replicating UIs using just a "device" object to a graphical API, so I know very well that it is possible. What prevents this project from just... going out and actually doing what they're trying to achieve? Take high resolution screenshots, carefully understand what's going on, and re-implement it.
The hardest part would probably be compositing (if they do it from scratch), but even before getting there, there's so much wrong already. This really does not have the same finesse of MacOS, has anybody even looked at the screenshots picked by the developers??
This screenshot[1] is a great example of what you're talking about.
* Awful typography (especially bold text)
* Ugly gray background in foreground window
* Windows 95-era buttons in the foreground window
* Weird left padding in the (ugly) dock
* Terminal window has window controls on the right for some reason, and they look like they're from Windows?
All that said, it's still in preview (v0.4pre4), and these things can all be corrected. But I think they'd do well to slap "early preview" all over the place so people don't write it off.
I think there is a critical error in the judgement you make here, we can see that they advertise[0] the finesse of macOS. But what does this really mean? Let's think about it.
But before that, they clearly must be aware that... it's just bad (lightly put), yet it comes off very clearly that they're trying to copy[1] visual elements of Mac. So what does the author think finesse mean?
For my perspective, I'll quote the first definition I get on a Google search, "impressive delicacy and skill." We can argue that Mac does provide this (I believe it does, the experience is very opinionated (which also leads to stuff being intentional) and seamless across their devices, consistent (more than just the potential for consistency, which is what Windows provides), and so on, and so forth.
But, this is nothing like what we see in the current RavynOS. Nothing feels intentional. The advertising is misleading, and I don't think that it's something solvable in the future unless the person responsible for advertising and guaranteeing the presence of this "finesse" update their understanding of the word and clearly defines what that should mean in practice.
Update1: to add to my previous points, if my perception of the word "finesse" is correct, for something to have the "finesse of macOS," it wouldn't even have to look like it, but just achieve the same goal in it's own way (or Apple's way, that's "fine" but leaves a bad taste.)
At the pace they are going, they are years from a usable system. The screenshots you are looking at are from an abandoned approach to doing the GUI tech.
The level of polish that you are advocating seems completely premature to the current level of development.
This is not just a visual design shell. They are proposing a macOS mimic much more deeply at the OS level.
You advertise what you're aiming for, not what you've just achieved. Do you really want some alpha level software with description "an app that draws some windows and sometimes doesn't crash", or do you want to know what the author has in mind for when it's finished?
I would rather they just use KDE and make a nice skin that has a similar scheme to Apples. KDE looks sleek on its own, and then make keybindings as close to Apples as possible. Heck, KDE3 used to have a wizard that asked if you were more accustomed to Mac, Windows or Linux, and set your key bindings that way.
You can get insanely close with KDE, and KDE apps all follow a standard UI approach and the theming will transfer over. I'm pretty much convinced that the best DE is KDE for something full feature. If I had an OS with only the apps made for their respective Desktop Environments, KDE would be the winner for me, everything from games to code editors, browser(s?), and what have you.
The main thing I miss whenever I try yet-another mac-like skin is keybindings that actually work. I have decadeS of muscle memory, so when keyboard binding are different, I instantly bail. Now I understand why that's hard (this is all implemented individually by each app), but it's a non-starter without this.
It depends on how the app implements some event handlers, but yes you're correct, not all apps might honor the keybinding configuration. I think all KDE apps do.
Not to mention that has zero relation wit macOS developer experience, those developers that actually target Apple platforms, instead of using it as fancy UNIX.
Whilst I agree, it was just a development on what I perceive to be their vision. I address this particular point in another comment in this comment thread.
I think this is an effort to create a skin, or graphics layer / window layer to FreeBSD that looks and feels like Apple.
But then it says the project uses the Mach kernel.
Does that mean that they have a version of FreeBSD running on top of
the Mach kernel?
That sounds cool.
Are they creating a clone of Cocoa that will have the same interface
as the original? Or is Cocoa a open sourced from Apple?
Near the top it reads:
""RavynOS aims to be compatible with macOS applications""
Ok well that sounds cool.
Then farther down:
""Trivial macOS and Darwin applications may run directly on ravynO""
I guess it can be a goal that is a long way off still.
I dont know why they would bother with Linux compatibility.
FreeBSD and Mac has to be the most important ones?
It's using FreeBSD's existing Linux compatibility layer, so it's less "building Linux compatibility" and more "making sure they don't break the existing Linux compatibility"
Maybe this is just me but I never considered Apples approach to software installs to be elegant. And I’m not a fan of their unified menu bar (or most aspected of their desktop environment for that matter).
What I’d love to see would be a whole new desktop environment for macOS (homebrew has its warts but is a “good enough” package manager for my day to day needs)
I don't mind the idea of installing applications by dragging the icon to the folder. The problem is that many applications simply don't have the Application folder show up in that window. Seems like Apple relies on developers to make this interaction consistent, which obviously doesn't work.
When a mounted DMG is the foreground window, Finder should slightly adjust the .app handler. The default action will be "Install Application," which will pop up a confirmation window and, if confirmed, copy the application to /Applications.
The "Open" option should still be available for people who, for whatever reason, want to run an application directly from a DMG.
That makes it simple. It doesn’t make it elegant. And I certainly wouldn’t be getting anyone non-technical to run a FreeBSD/macOS hybrid that doesn’t support either platform fully.
Not taking anything away from this project, it’s looks really interesting. But it’s definitely not a mom and pop friendly platform.
Besides, since when have app stores been difficult to use? Package management doesn’t have to be CLI based.
I was forced to use a Mac for work and almost resigned because of it. I happily concede that a lot of it is what you are used to, but to me I need to be able to configure all aspects of my UI / DE ... including things like universal font sizes (which was my biggest issue due to my eyesight and finding that their fonts gave me headaches on my external monitor).
For people who are not "power users" then I suppose I can see the appeal. But the "it's our way or use something else" philosophy doesn't work for me. I will use something else (the problem is when you're forced into a system with that mindset). The computer is a tool, I am the master of the machine (especially as a programmer), and I can't stand feeling like there are barriers between me and how I want the system to behave. "Finesse" to me is the ability to install a completely different desktop environment if I so choose. To type "rm -rf /" as root and not have the machine question my command (very dangerous for the non power users I admit).
It's nice that more people in the FOSS community take note of macOS's design and (historically, at least) attention to detail. However, this project (judging by the screenshots, which is how everyone is going to judge you if you claim to be inspired by Apple's work), they're already making serious (but fixable) mistakes:
- What the actual heck is that window border around the terminal
- Horizontal spacing on the fonts is too dense, the text rendering is absolutely awful - maybe it's just the font but I wouldn't suffer through reading a full paragraph
- The buttons in the installer dialog look like a cheap knockoff of Windows 95
- Different colors for window background, title bar, dock, but somehow the window background and menu bar are the same
- The 3 visible dock icons mix 3 different art styles, the fourth icon is an empty space?...
- Where's the shadows to tell stacked windows apart? Windows did this in the late 90s
- The padding of the window controls is all messed up, with different offsets on all 3 sides; no unified window/tab bar in the browser (even though Firefox under both X11 and Wayland can do it just fine); again different window frame and background colors
- Bright white window borders (absolute worst aspect of the Big Sur redesign), although I'll give them credit for going a notch below #ffffff
- Very poor way to show off the transparent menu bar (the choice of wallpaper + no blur), visually distracts from the contents of the menu bar itself; note that all macOS default wallpapers have a pretty uniform sky color across the top, so the transparent menu bar can be a bit less of a mess
Again, this is all fixable, but the real problem is that half of these things are trivially improved by changing the font, the color, or the padding. But it's this lack of attention to detail (and the fact that despite these glaring issues they've still decided to use these screenshots) that speaks about the project's commitment to details.
It's pretty cool that you've got AppImages or an ObjC compiler, but you're not allowed to use the word "finesse" and make these mistakes!
You don't advertise WIP software as "built with finesse"; and even if you do, you must at the very least put said finesse into the choice of screenshots you present.
Yes, but it only devolved from Mac OS 9 by adding a bunch of unpolished NeXT GUIs, and the development necessitated the temporary removal of features and apps that were already polished.
I‘ve never really worked with Rhapsody, so I couldn’t tell. I came to Mac OS X from NeXTSTEP and FreeBSD, the first version I actually used as a daily driver (with the odd boot to Mac OS 9 for some stuff) was the public beta in 2000 (which was Aqua already of course).
Rhapsody was an unholy mix of NeXTstep implementation with a bastardized version of Apple's Platinum theme, which wasn't designed to support NeXT's UI and display technologies. Steve hated it and continued to use OpenStep on a ThinkPad.
That looks like it's SwitchBoard, not iOS. Users do not have access to the internal Apple test software.
As for Rhapsody, that looked decent, it had a similar vibe to A/UX where it looks like Mac OS Classic but runs Unix.
Perhaps you are referring to Kodiak or Cheetah, but those were pretty GUI-consistent at that time. Granted, up to Panther it was just Aqua and getting it to work consistently and performant, but I don't think it was ever in the state that this RavynOS is. That's mostly because NeXTStep as a foundation already had an existing GUI that was consistent.
RavynOS is having a bad presentation because it looks like some sort of odd partially-done theme based on CDE/Motif.
Yes, but you are straw-manning the argument. This is skankphone UI - a UI used by the technical team building back-end services and features for the original iPhone. The actual UI was being built in parallel, and obviously it never looked like that.
The aspiration to be on par is cool and over time, hopefully they pull it off. But it probably comes down to them having some dedicated graphic designers and Ux designers (and an audio person would be nice too), and foss projects have historically done poorly at recruiting those people.
> I don't code but this looks awesome! Can I still help?
> Absolutely! There will be art, documentation, testing, UX and UI work, [...].
I read this and immediately think the project sees UX as "I can't write code ell oh ell". You need to treat a UX role the same way you treat a kernel hacker role - it's a specialist whose input you must respect, or be left to deal with the consequences.
I think you're reading too much into it. Saying "if you aren't a programmer you might be able to help in various ways, including UX" isn't the same as saying "UX doesn't require any particular skills".
True, but this is literally the only mention of the words "UX", "UI" or "design" in the FAQ/docs; meanwhile they write at length about the various technical challenges and choices. This is inconsistent with their own tagline.
Its worse than recruiting.
The main issue is foss projects work poorly with designers.
Even if you have good designers their powers tend to diminish. Once any disagreement comes up if you cant code, if you cant PR it… wont happen.
Also design is super hard to modularize and split into small pieces. With code random anybody can come and build/fix a feature. Not possible with design. Its mostly big picture work where you need to have deep context and follow it to get desired consistency. For projects like OS gui that requires huge commitment (most likely full time).
But i think copying existing well made software can make this a lot easier. Just need to be very diligent about it and not ashamed (and its a lot less creative work).
Do you think operating systems look polished early in development? Of course they didn't. Most of the work going on isn't visible in a UI. From what I gather in the discord there have been some quite large refactors going on under the hood
What does that mean? It can partially run some trivial macOS apps? The amount of effort just to get basic programs working seems like a ton of effort, for a end project that is just buggy implement of a moving target.
Unlike Windows, MacOS has no problem ignoring backward compatibility. So Wine works be rewriting older APIs and Microsoft always works on backward compatibility, MacOS does not do that. It just works or it doesn't.
I also wonder what that means. What target would this software be compiled for? Apple Silicon or x86?
If x86, then some feasible but difficult approaches exist, such as a translation layer or providing own reverse-engineered APIs. It would be tremendously difficult to achieve practical usability beyond very limited cases. However, by the time it could be done, won't the x86 macOS target be entirely obsolete?
If the target is Apple Silicon, it makes even less sense. The instruction set is ARM + Apple's mods which are proprietary. Moreover, they keep iterating on that instruction set and making new software alongside it. What's to stop Apple from using a Rosetta approach to run their M2 code on M1? If they have a translation layer, they don't need backwards compatibility in the instruction set.
So reverse-engineering would be extraordinarily hard - like, beyond human capability for a small team with no insider leaks out of Apple. Can this project keep reverse-engineering the instructions fast enough as Apple iterates on the arch? Not in reality. x86 is hard but this is that, and a moving target (pun intended).
So... how?
EDIT: Upon looking into it more, it looks like they want to use x86 translation layer Darling + open-source APIs. This seems more like running (maybe) basic GUI applications compiled for Apple's x86 targets that do not rely too heavily on the Apple ecosystem's APIs. So more for fun than for robust day-to-day compatibility akin to Rosetta. So I guess then it doesn't matter that it's a dying target, as it's more of a theoretical thing?
The idea is nice, I don't want to dismiss it. I think attempting it could be a lot of fun. They're kind of brave to call it a project aim though, I don't think it's wise to set such expectations.
What about something like winapps which transparently runs apps in a VM. I know you can't legally run MacOS in a VM on a non mac but it would be hard to enforce against a software project that merely provides the capability so long as it doesn't provide the copyrighted software itself.
I remember 20 years ago Gnome or even KDE were toys compared to the very structured but slow moving GNUStep. GNUStep was also hard to install (no rpms), ugly but very snappy.
But nobody big put any money or devs on GNUStep. Even Sun invested in Gnome (!), what did not made sense at the time because they had invested in OpenStep & the whole thing was Java friendly (I believe Java was a first class citizen in early releases of Mac OS X).
I always wondered if it might have been because there were some legal risks...
> I always wondered if it might have been because there were some legal risks...
What was in it for Sun to invest in GNUStep? Jobs was always resistant towards supporting other platforms. So it’s clear OpenStep was the last attempt at making use of the dying Next line.
I could maybe imagine a sales pitch scenario like “port your Mac apps to Solaris”. But when OSX came around, Cocoa was steered more towards the consumer market, not workstations.
I'm saying that Sun needed an update to the old CDE and invested quite heavily in the development of Gnome to create their Java Desktop System [1]
I believe it was a mistake for many reasons, nobody like JDS, the name itself was a lie and ultimately they just helped their competitor RedHat by turning GNOME into a serious project.
I believe Sun still had all the Lighthouse Design softwares that were made for NeXTSTEP [2]
Jonathan Schwartz founded Lighthouse and actually joined Sun through the acquisition (!)
Anyway at that time the relationship between Sun & Apple had turned very bad & Steve Jobs threatened them when he saw the Looking Glass demo [2]
macOS is more than shiny traffic lights buttons on your windows, global menus, and dock.
It's also a clipboard that works everywhere and not like "oh you cannot do this because X11 this or wayland that", it's also human interface guidelines (the old ones, the new ones are unparseable shit) that make bloody sure that the OK/Cancel buttons appear where you expect them to appear.
It's also some bloody keyboard shortcuts that work everywhere the same and not every bloody app bringing in their own and their own little ways to configure them or none at all. Yeah chrome, I'm looking at you.
It's also that the most used system stuff can be configured pointy-clicky and that stuff that the system can figure out on its own is what it bloody figures out on its own, without me having to drop down to command line (although it is an avenue too, and command line and GUI must be in sync).
It's also that it's accessible out of the box. Try blindfolding yourself, inserting your installer usb stick, and installing your thing and using it successfully. Come on. OS X could do it since Leopard, I think, which was in... 2007?
In other words, macOS could abandon their Aqua looks and go for industrial NeXTStep looks, and still be ahead of almost everything else in terms of consistency, accessibility, and general lack of fuckness.
The ribbon had all the settings for the OS, and was always a click away. Finder worked out of the box. These days, copy-paste is hit or miss, and the window manager can’t be reliably controlled via the keyboard.
Also, even in first party apps, keyboard shortcuts are completely inconsistent. Command-0 no longer reliably restores the main window. In maps, Command-L doesn’t focus the location bar like it does in safari; it repositions the map to your GPS location.
Even the keyboard is a shitshow. Notes aggressively autocorrects technical terms to random words on the latest MacOS, but nothing else does, and I can’t find the second “turn off autocorrect” setting that it must be using instead of the OS-wide one (which is definitely disabled).
Current OS X’s UI and filesystem layout are a cruel parodies of OS 9. It’s like a bunch of Shoggoths or an army of slightly better LLMs tried to recreate OS 9, borg-style.
Apple mostly adds software features by acquiring stuff, storyboarding UIs, demoing a prototype to an exec, and then shipping it, so I guess that’s how it got so weird.
sure, there are inconsistencies, but with some global shortcuts in your muscle memory and a few years of using spotlight under your belt... OSX out of the box is fast and efficient as hell.
But I will relent and acknowledge that your arguments ring very true. How do I get my Messages window back if I accentally close it? Very good point.
The ribbon though? I think your memory is optimistic. It was in my memory ugly, clunky, slow and had some, but few system settings. But I must admit that in my heart OS 7.1 was the high water mark of usability and I resented all subsequent innovations... which it sounds like you might be able to empathize with.
Now imagine that the thing in the OP has even less finesse than what you call a bunch of Shoggoths. More like putting lipstick on a earthworm.
All those projects, hellosystem, this one, I guess I've seen one or two more before — all of them are just kind of missing the whole point.
I would say the current KDE Plasma is somewhat starting to approach any amount of finesse at all. Hope they continue, at least they've learned from the whole KDE4 kerfuffle.
> Try blindfolding yourself, inserting your installer usb stick, and installing your thing and using it successfully.
I don't understand this.
Are you saying its hard to install Linux? Has this ever been true? Usually you smash the Next button. There is a partioning screen, but that is some serious changes you are going to make.
I don't care what OS religion you follow, I'm taking the blindfold off during partioning.
they are saying that you can literally install the OS start to finish it (and configure it for sightless use to boot) without being able to see at all. As in blind. This was part of the point they were making about accesibility features and UI consistency in MacOS.
Yes. You need to be very technical. E.g. there is a keyboard selection thing with multiple ambiguous options while on Mac you have to press like one key to identify it.
Even windows somehow works without requiring a PhD in keyboard layouts.
On modern distros the installer will already have the right one selected if your keyboard layout matches the system language (which is true for 99% of computer users). I'm pretty sure Fedora doesn't show you the list of layouts unless you want to manually change it.
I vaguely recall reading about a project trying to achieve just that on HN a while back. Unfortunately, that is all I remember about it. Back then, it apparently was not quite there, yet, but that must have been at least five years ago, so the situation might be a lot better today.
Unfortunately not. Darling [0] is still at the point that it can only run command line applications. Only the most basic GUI applications are supported. That's still a massive accomplishment that I don't want to diminish, but it's nowhere near the point that WINE was at even quite a long time ago.
I would imagine with Wine the number of people interested to see Windows applications run under Linux/BSD/etc is significantly larger than with macOS/Darling.
that's a pretty wild assertion. I'm immediately thinking of music software: 99% of all plugins, the #1-4 DAWs, drivers for 2/3 of all hardware, at least... not on linux.
So much awful. Bad font rendering. Icons look pixelated. White semi-transparent things look weird. Active and inactive windows look too similar. Instead of just coloring in the window buttons greying the text there should be some more obvious visual change to the actual header. Curve in top of window is pixelate. Very rounded top and share hard rectangle is jarring. Dock is perfectly rectangular with no border. Icons are bad.
On the interface side the global menu is not only a mediocre UI when universally implemented its unlikely to be properly implemented in countless apps making impossible to make consistent.
Yeah, I had the same thought viewing the screenshots. I like that this project exists, but they've got a long way to go with those Windows 95-esque buttons and scrollbar.
I read a comment once that enumerated all the ways even spinning your own linux distro is doomed to fail. It's far more work to properly develop a distro for security and if your distro is just "Ubuntu minus stuff I don't like" then you aren't really maintaining anything that couldn't just be a script you have people run on base Ubuntu.
I suspect most people in this thread haven’t gone deeper than the website.
The project creator is actually very capable - what they’ve done so far is very impressive on a technical level. The project has an effective bus factor of 1, though: I’ve lurked in the Discord for the past year (since it was called Airyx) or so and plenty of people come in but don’t contribute much.
Unless that changes I’m not sure it’ll work out long term.
Well done for runining my weekend! I will give it a try. However, still struggling with NVidia drivers on FreeBSD. For the life of me I can't get them to work.
This is why I switched to AMD video cards. NVIDIA was always flaky under Linux. I finally got a working setup, and then they stopped backporting new OpenGL features to the binary blob drivers for it, breaking all sorts of stuff.
CUDA is a big advantage for NVIDIA, but I don’t use it. Similarly, NVIDIA hardware tends to be a half a generation ahead. Whatever. The AMD drivers are open source, in kernel tree, and reliable over time.
I’m honestly not sure why they would stick to BSD. The first thing many folks do when they start a Mac is install Gnu Tools and other things to make it more like Debian.
Well yeah but FreeBSD is not Linux, and I don't see the connection with your comment. I also use a Linux distro as a daily driver, but I love playing with FreeBSD every now and then.
You can just hear: can I run 'popular macOS application' on it? No.
The average user, for starters, doesn't use a computer (most consumer tasks are done on a phone), and won't know how to install a different OS even if they do.
Run FreeBSD on the server and use the macOS Terminal to connect to it over ssh.
They do mention implementing Cocoa APIs so there is a theoretical future where it can run some macOS apps. But that would be a very heavy lift (not quite ReactOS heavy but still) for little reward. And you’ll probably find everyone is using SwiftUI by the time it’s complete.
It’s a really good idea since apple is abandoning its x86 software. It’s an opening for FOSS to pull in new GUI frameworks/features. Could really help gain userbase.
They use FreeBSD and Darwin open-source code. Darwin is the core of MacOS with the XNU/MACH Kernel. So they are doing nothing wrong that Apple can sue them over it. The trick is writing the code that does what the Non-Open-Source code of MacOS does to run the Mac Apps.
> The trick is writing the code that does what the Non-Open-Source code of MacOS does to run the Mac Apps
And that is where the magic happens.
GNUstep has been trying to provide a useful open source alternative implementation of OPENSTEP for more than 20 years now I believe. I don’t see that happen to be honest.
I realized Brew was released for Linux a while back, and my jaw dropped because… why? Brew is simply awful compared to most Linux package managers. It has gotten better over the years but I still couldn’t ever imagine preferring it over the installed out of the box package manager in quite literally any Linux distro.
Homebrew is my preferred package manager on Linux. It just seems to work more simply than apt or Flatpak for me, and it’s less context switching when I need to use macOS.
For standard apps, I agree. But being able to change the default version of Python, NodeJS, etc. just by running brew install is a lot easier than using apt/apt-get. They just aren't made to support that kind of scenario.
Huhh this is really cool. Never heard of it. And FreeBSD is actually my daily driver.
So apparently it can run macOS apps? More or less. Cool. It'll be hard to find those compiled for intel though going forward. But we'll see.
I'm actually quite happy with KDE on FreeBSD but perhaps their compatibility work can also be applied to vanilla FreeBSD. The only thing I really miss from macOS is Pixelmator.
Subjectively, I find both names terrible. There’s zero chance I’ll vocalise either to someone and they’ll know how to spell it. The current one in particular is bound to have people typing “ravenous”.
But the idea is interesting, and even though the screenshots don’t yet inspire confidence (the typeface is too thin, folder icons are too square, …) I wish the team the best of luck in making their vision come to light.
> The current one in particular is bound to have people typing “ravenous”.
I'd guess it was probably intended, and that makes things worse.
I know - as someone else was commenting - that naming things is hard, but for something like this, finding a name that is easy to pronounce, and easy to spell, should be a high priority.
Off the top of my head, playing with the fact that it is an OS, it runs on a Mac, and is based on FreeBSD (where BSD stands for Berkeley Software Distribution [0]), a couple of quick suggestions:
- BerkeleyOS (let's honor Berkeley)
- FinOS (a play on finess - easy to spell?)
- SteveOS (obvious reference - might be disliked by many - but is it Jobs or Woz? Perhaps WozOS would be better?)
- BillOS or JoyOS (Bill Joy founded BSD in 1976)
- JordanOS (Jordan Hubbard coined the term FreeBSD)
JoyOS is particularly good. It’s short, has a clear cadence that helps to identify its parts (joy-o-s) as has multiple positive reads. At first you’ll just assume “it’s a joy to use” (which could be a slogan) but then has the historical reference that makes the name even better and more meaningful once you learn it.
It's not a joke that one of the hardest problems is naming things. So many promising products and businesses shoot themselves in the foot by choosing a name that hurts adoption.
My personal example of this is the security company “Thycotic”. They could be greatest InfoSec company in the world, but I really hate their name. It looks like the word psychotic said with a lisp.
It’s not just a case of liking names, they should also be functional¹ in terms of spreadability. If this OS eventually does what it is meant to do, I will want to tell people about it and get them to try it.
When making a recommendation in a social setting, there’s a tiny window to get the other person interested in whatever you’re suggesting they watch/play/try. If the name is hard to understand, difficult to remember, or they have to take out a phone to write it down, it’s causing friction that makes them less likely to follow through.
I agree with you in this instance, as early adoption for a project like this relies on people being able to find it and the more people who do and try it, the more testing is done to improve the entire project. I'm reminded of my time on helpdesk when getting people to click on the browser bar was nearly impossible and it basically didn't matter how I asked them to do it. Either they would click into their search bar or they would somehow append the URL to their homepage URL. I learned that my intentions and efforts to do something a certain way don't matter at all if someone else is incapable of doing something I think they can do.
This name is the same way. If someone searches for RavenOS there's another project named that[1]. If they don't see that, they will probably be redirected to RaveOS[2], see "ah cryptocurrency mining" and then back out. Finally, if they miss both of those, there is an Android ROM called Raven OS[3] they might also find. No matter their intention, if people can't find it easily after it being mentioned in conversation it will be harder to grow.
I am not. I speak from vast experience of trying multiple approaches.
Maybe English is the first language wherever you live. That’s not the case for the majority of the world. Even if my friends and I have a good grasp of the English language, switching mid-sentence causes a delay in understanding. It’s not just “raven with a y instead of an e”¹ but “raven, as in <raven in my language>², but with a y³ instead of an e”, which is several transformations to do. I guarantee that would need to be repeated multiple times, with the other person having to take out their phone and asking questions in the midst of typing, and I’d still have to check and they’d have it wrong first time.
¹ And you’re ignoring the part of “then the letters O and S, without spaces”.
² Which itself causes complications, because most people will recognise “crow” as a translation but we don’t make a distinction for “raven” so it’s a word even people with an understanding on English may not recognise.
³ Which is a letter we don‘t use in any native word.
We can agree to disagree. By your logic every small open source project needs to hire a marketing firm to ensure their naming convention is easy to understand and inoffensive in every language on the planet. I think that’s overboard.
The project is clearly being built by someone that speaks English and is targeting English speakers. Notice their website and documentation is all in English?
> By your logic every small open source project needs to hire a marketing firm to ensure their naming convention is easy to understand and inoffensive in every language on the planet.
>I guarantee that would need to be repeated multiple times, with the other person having to take out their phone and asking questions in the midst of typing, and I’d still have to check and they’d have it wrong first time.
Definitely not. I'd just tell the raven os and let them google it. Google would find it anyway.
Which I know you didn't even try before writing this italics laden diatribe because the bigger issue is that the name is already taken (albeit abandoned).
Definitely yes. Again, speaking from experience. Yours may be different, but don’t discount mine.
> I'd just tell the raven os and let them google it.
You missed the point entirely. With each step they have to do, you lose someone. The point is to reduce friction so there is a better chance the other person follows the recommendation. You don’t seem to care if the other person ends up searching for or finding the right thing, in which case you’re irrelevant to the point. The conversation is about people trying to maximise success of a recommendation being acted upon.
> Google would find it anyway.
You seem to be unaware, but Google shows different results to different people. All it shows me for several results is a movie called “ravenous”. To other people it shown multiple unrelated projects. At this point you’ve already lost several people.
Yes, I did. And I didn’t get what I was looking for. I was expecting to get at least the web browser but even that did not show up. Please don’t assume you know what other people do or do not do.
> because the bigger issue is that the name is already taken
No, that is not an issue at all. No one made them choose “raven”, a name already in use by multiple projects. The solution to “this name already exists” isn’t “let’s replace a letter”.
> before writing this italics laden diatribe
You appear to be irrationally angry about this matter. It’s just an internet conversation about naming, it’s not worth raising your blood pressure for. I wish you a calm weekend.
> You appear to be irrationally angry about this matter. It’s just an internet conversation about naming, it’s not worth raising your blood pressure for. I wish you a calm weekend.
Actually my impartial arbiter chatgpt says you're frustrated and annoyed. Always weird when people go straight into projection.
> Yes, I did. And I didn’t get what I was looking for. I was expecting to get at least the web browser but even that did not show up. Please don’t assume you know what other people do or do not do.
Please don't go on the internet and just tell lies.
> No, that is not an issue at all. No one made them choose “raven”, a name already in use by multiple projects. The solution to “this name already exists” isn’t “let’s replace a letter”.
This point was in agreement with you that their name was a poor choice. They should have chosen something else given this was already taken, for the exact same kind of thing no less. If you were less mad you might have realized that.
It is still in pre-release and being worked out. It doesn't run in VirtualBox or VMWare because there is no driver for video on it for the virtual machines.
One moderately frustrating thing about using FreeBSD is that it's always been treated as sort of a second-class citizen by GNOME, and therefore Gtk, and therefore most of the Linux ecosystem which depends on Gtk. So probably the most popular desktop environment is KDE, and it's nice, but it is a certain kind of interface which doesn't appeal to everyone.
Meanwhile, OS X shares a little code with FreeBSD and was partially derived from it around 30 years ago. And the design philosophy of OS X is roughly the opposite of KDE. This situation has engendered at least half a dozen projects over the years, most of which are never really complete, to make a FreeBSD-first DE/UI that works like Mac. And if it can run Mac applications, so much the better!
Unfortunately, this is really hard. Cocoa is a moving target, and contra Wine, Apple is not as pertinaciously obsessed with backwards compatibility as Microsoft. Even much less ambitious (in my estimation) projects like enlightenment 17 and Wayland have struggled to replace major components of the desktop from the ground up.
Not only is Cocoa a moving target, it’s deep and wide. The amount of functionality it offers is staggering and extensive enough that for many types of apps, a dev may not need to import any third party libraries if they don’t want to, with no negative impact on the quality of the final product.
That’s extremely difficult to reimplement, even if you lock compatibility to a particular macOS release.
OS X is derived from NeXTSTEP, first released in 1989 and based on 4.3BSD-Tahoe. Meanwhile, this was updated to 4.3BSD-Reno and ported to the x86 as 386BSD, from which NetBSD was forked in early 1993; it was subsequently reorganized to become FreeBSD in late 1993. NeXTSTEP and OS X later imported a few system components from FreeBSD, while replacing others, particularly replacing parts of the kernel with Mach.
It appears like they’ve aped some of MacOS design language, but it’s clear they’re engineers, not designers. There’s enough “uncanny valley” between this OS and actual modern Apple UI design that it comes off a bit like all those versions of KDE that tried to look and act exactly like Windows XP and always felt a bit “off.” Their screenshots page is a Frankenstein mess of poor font quality and Apple themes mixed together dating back to OS X.1
I think it’s better when a team actually has designers working for them and comes up with their own real UI standards, (see Pop!_OS). IMO Apple’s UI design isn’t particularly good so it’s not like some holy grail we should copy and paste to Linux.
> IMO Apple’s UI design isn’t particularly good so it’s not like some holy grail we should copy and paste to Linux.
It may not be perfect, but it’s pretty darn good - as especially visible when comparing to this effort presented here. There’s a lot of detail and polish in MacOS interface design and UX people don’t realise until it’s missing.
Apple’s UI design certainly has flaws. I really don’t understand why they changed the way the System Settings in Ventura/Sonoma(?) now looks like and behaves more like iOS, as an example.
However, as someone who has used GUIs from SunView/Irix Desktop/DECwindows to iPadOS/MacOS, via Windows 2.0/3.0/95/NT/etc. I am of the opinion that Apple is still the most consistent and easy to learn GUI. Even my 80-year old mother manages quite well with her Mac and iPhone.
To me that is kind of the key strength and why Apple has such a good UI story. Its not necessarily that they perfected the UI. There are certainly parts I dislike. But, its consistent. I have to use Windows 11 at work. It is amazing how some windows have been brought up to date with the new design language of Windows 11, but I open up the Windows Event viewer and still has like the Windows 7 design language.
My interpretation would be that it's about the same level of polish ux/ui-wise as macOS. As with OP, unfortunately I don't see it either - the screenshots show some resemblance with macOS but without any of the finesse.
It's not completed yet. The features are listed under the section "project goals" and the download button has a note calling it a developer preview image.
In my experience with some Linux distros goals alone don't necessarily work out, but the approach with trying to be compatible with actual Mac OS apps is already more notable than an Ubuntu reskin.
In the case of macOS, for me it’s for example things like automation via AppleScript, a coherent UI for native apps, the „polish“ of the interface if that makes any sense
Because its still in major development? Huge refactors are going on still under the hood. It's a project with a small team working in their free time with a big goal and its some way off. I mean, a BSD-based open source OS binary compatible with mach binaries ... it's very cool IMO and I hope more people get involved
Apple‘s secret sauce is not mach binaries but Cocoa. And tons of resources. And clear leadership.
We have decades-old projects like GNOME, KDE and GNUstep and none of them come close to macOS in terms of „finesse“. Despite some of them having corporate backing.
So I just don’t see that happening for a project with a small team working on it in their free time. Far too ambitious, I‘ve seen projects like this fail too often in the last two decades.
My basic opinion is that there are really no great desktop operating systems currently being made anymore. Mac OS used to be it but started losing me after Snow Leopard to be completely honest.
I feel both desktop producing companies that actually matter (can run all the productivity apps I need to do my job) have put this space on the back burner for cloud (Microsoft) or consumer electronics (Apple).
No year has ever been the year of the Linux Desktop and all the Desktop stuff for Linux is too Linux specific to work well portably on FreeBSD.
All I can say is I hope Apple makes iPads able to run full development suites because the desktop appears to be dying a slow death in general.
Or, maybe I’m wrong (would love to be) and things start getting better again.
Indeed seems like post-Snow Leopard was the beginning of the end.
Desktop seems to be converging to tablets. Which is just another way of saying, dumbing down of the UI and restricting what can be done, and offloading to the cloud.
Development suites will all move to containers in clouds just like VSCode does today.
I know some of this is personal taste, but to me macOS really doesn't feel finessed. It feels confusing, distracting and annoying. I suppose a large part of the problem is that I want a useful tool, whereas macOS is designed to be a lifestyle.
It just doesn't look that great either. If you glance at a screenshot then, sure, it looks slick - but you'll routinely run into developer UIs (for example, the settings app prior to Ventura - which has now become pretty to the degree of uselessness).
I think you find macOS confusing, distracting, and annoying because it hasn’t had strong design applied to it in over a decade.
It does get features to sustain people’s interest in Apple hardware and services, which I guess is a lifestyle. But it’s an “Apple lifestyle” not a MacOS one.
A lot of commentators don’t really seem to appreciate how Apple is a consumer electronics manufacturer now, and still evaluate it like people did ca. 2003. Apple could shut down their desktop/notebook computing division entirely and still make a bundle. (They probably would lose phone market share, though.)
I get the impression that MacOS is kind of like Gnome in that they aren't really designed with multitasking in mind. On MacOS, the green button in the corner of the window expands the app to fullscreen. Opening an app in Gnome likewise goes fullscreen.
There are some workflows where that would be OK, but to do my job I have to be in and out of multiple windows constantly. I gave up on Gnome when I was trying to find one of 5 different terminal windows in the exploded window view and it was clear that they'd rather I cycled through them all one at a time instead of what I do on XFCE which is scroll up on the terminal icon in my dock, which quickly cycles through them and brings them to focus.
Personal preference for finesse will always differ so I tend to focus more on the usability aspect of it and accept any finesse that comes along for the ride as a kind of luxury experience lol
It’s designed to multitask, just not in the Win9X-ish style that most people are accustomed to at this point. There are plenty of Mac users who juggle tasks users all day long without issue.
It’s designed around letting windows that are sized to fit content overlap and be at least partially visible if they’re relevant in any way. If there gets to be too many windows to manage (which happens in other desktop environments too; the usability of Windows alt-tab for instance scales incredibly badly as the number of open windows increases), that’s when you start tossing windows into another virtual desktop (“space” in Mac nomenclature), perhaps organizing windows into desktops based on task.
To opt out of fullscreen on the green button, hold down Option/Alt before clicking it. The behavior of the button in that situation is defined by the developer of the program you’re using though so YMMV. Maximizing windows is possible by double-clicking their titlebars, except in Electron apps that have replaced the system titlebar with a fake one that’s missing functionality.
As for GNOME, I’d say it compares to iPadOS much more closely than macOS due to missing or oversimplified functionality. It feels like it’s more oriented towards tablets and small laptops than desktops or even midsize-to-large laptops.
Thanks for the tips :) I believe you pout into words the reason I don't like GNOME - it fell victim to "let's make it touchscreen friendly" at the expense of the way people have been using computers for a long time. Again, it has uses (iPads are perfectly functional and easy to use) but you feel the limitations a lot more when you try to do more advanced things.
My opinion that MacOS wasn't "designed" for multitasking is heavily biased by how I multitask (as you mentioned it's the Win9x style, stemming from using Windows for my formative computing years) so I am not trying to discount MacOS at all - I use it for my music production because Core Audio is rock solid. But then, that's a relatively singular task and as a result I treat it as though its only job is running Ableton at the time.
Multitasking is also different now, since so much of what we do is in the browser anyway.
Same here. I don't see the finesse in macOS. Half the time I have to fight the window manager to get what I want, the other half I have to use 3rd party tools. It doesn't even have very basic stuff like a volume mixer.
I've personally experimented with replicating UIs using just a "device" object to a graphical API, so I know very well that it is possible. What prevents this project from just... going out and actually doing what they're trying to achieve? Take high resolution screenshots, carefully understand what's going on, and re-implement it.
The hardest part would probably be compositing (if they do it from scratch), but even before getting there, there's so much wrong already. This really does not have the same finesse of MacOS, has anybody even looked at the screenshots picked by the developers??