This is great, thanks for releasing your work. Very impressive.
You may get some interest from others in the retrocomputing/permacomputing sphere if you implement an Uxn emulator; it is extremely simple and can run on very limited hardware. https://100r.co/site/uxn.html
Vintage hardware would be a great host for Uxn programs, so I suspect this would generate some excitement.
Ok, I built the floppy image now. dd'ed it on a floppy and powered my IBM PS/1 up. Despite some nasty sounds of the HDD bearings that went away after 30 seconds, the floppy does not boot on this machine. Just a black screen. 386SX-25 2MB RAM. Maybe 2MB RAM too less, but I thought at least something will happen. :-)
For PS/1 you'll need the 16-bit version from https://github.com/luke8086/gentleos. A floppy image is provided in releases. Note you only need to copy the first 64KB, the rest is just padding for emulators.
Good catch, the yellow and blue colors are totally inspired by BeOS :D I'm even adjusting the default VGA palette to get the right tints in 16-color mode.
I think it's that yellow bar what it makes it look like BeOS. And maybe the right hand menu bar. But once you check a B/W version, it doesn't look like BeOS that much.
Nice! The project also has a 16-bit variant https://github.com/luke8086/gentleos, not clear if it works on 8086 IBM PC, but I'll give it a go. Been looking for a reason to power up my IBM PC again.
This reminds me of the era when operating systems felt more approachable and visually distinct. Modern UIs are often cleaner, but many of them have lost some of the personality that older systems had.
> A hobby operating system for vintage 32-bit PCs.
I am all in favour of great projects, but why a differentiation between 32-bits or 64-bits? I don't understand that. Is a computer that is 32 bit or 64 bit, either way which, not worthy?
Edit: I understand a motivation if it is on simplicity choosing one or the other, but other than that I don't see why that should ever be a goal worthy to be pursued. Software should really "just work" no matter the number of bits and bytes.
In this context, 32-bit means the minimal requirement. You can absolutely run even the 16-bit version on a 64-bit PC, provided it has BIOS/legacy-boot mode.
It only won't work on modern pure-UEFI systems because that would require writing full stack of USB drivers for keyboard and mouse, and that would be a huge task.
> Edit: I understand a motivation if it is on simplicity choosing one or the other, but other than that I don't see why that should ever be a goal worthy to be pursued. Software should really "just work" no matter the number of bits and bytes.
"Vintage" usually refers to actually old stuff, while "retro" refers to new stuff that looks/sounds/feels like old stuff. So GentleOS is a retro OS designed to run on vintage hardware.
(That distinction wasn't clear to me either, so I had to look it up - TIL).
x86 boots in 16-bit real mode. Then you need to specifically transition into 32-Bit, and from 32-Bit it can be transitioned to 64-Bit Architecture...
The last step (32-bit to 64-bit) can a bit of a can of worms especially on older platforms where 64-bit implementations can differ greatly and 32-bit "just works tm". 32-bit is quite well supported and has enough resources to make some interesting programs work without much hassle.
I think the author has made the decision not to support 64-bit mode due to needing to balance the complexity and usability of the project. It is a hobby project after all.
Since the author maintains a 16-bit and 32-bit for this project I suppose if you wanted you can always fork and maintain a 64-bit version if you wanted to.
I think there's a couple of different forces at play in the convergence of GUI design that we're seeing in the past 20 years. First, there's been a huge amount of widely accepted research that shows what the most accessible way to design an interface is. Things like Google's Material Design and Apples Human Interface Guidelines come to mind. Second, the widespread availability of those two specific design guides make it increasingly common for developers to just create to those. It ensures that things just work and are increasingly portable. Third, we're in a landscape where API stability and design is prized. That's partially because of the number of times that design has been broken by updates and version changes. It takes many years for developers to update their applications when a new back-end is developed, and the time in between gives broken applications, and ugly looking fallback. You can look at running GTK1 apps on modern GNOME, or X11 apps rendering on Wayland over the past decade for an example of this.
All that said, I truly miss the days when we had interface skinning. There was a skin for OS X called UNO that was absolute perfection in my eyes, and it was ported to an old version of Android back when skinning was a thing. There's nothing like it available now. Even GNOME is highly against theming and skinning now, apparently because they like breaking with every single release rather than maintaining an API/ABI and skinning support. The themes that were available for Windows XP were so much fun, even if you had to swap out DLLs to get them working.
I have a theory that corporations make new UIs to entertain people through them. First, to create the feeling that something is happening, and second, to increase screen time.
Old interfaces were far more practical for getting work done, and therefore obviously boring.
For me, as someone who is supposed to use technology as a tool and not as a source of amusement, the new interfaces of the major OSes feel unacceptable. But the other billion people chatting and scrolling are the real consumers, not me — and as a result, we now have the interfaces we have.
The link is to a designer talking about how technology has led us to a design world that is mostly driven by nostalgia. I personally don’t see it as being applicable here as it deals with big design houses not hobbies, but I can see why someone might think it is.
What a designer might call nostalgia an actual user of an OS might call standard, or maybe even intuitive. The point of an operating system is to be used. If it’s pretty, that’s a bonus. Usability by the target audience is the primary concern.
Copy/pasting my comment from the other post:
This is great, thanks for releasing your work. Very impressive.
You may get some interest from others in the retrocomputing/permacomputing sphere if you implement an Uxn emulator; it is extremely simple and can run on very limited hardware. https://100r.co/site/uxn.html
Vintage hardware would be a great host for Uxn programs, so I suspect this would generate some excitement.
> The only future plans are bugfixes, optimizations, and adding more apps.
Perfect. Nice to see a platform target stability instead of constantly reinventing itself and its APIs. Definitely want to give it a go!
I think this is fantastic! I love that the code is so clean my dumb ass can understand it despite not using C much.
The code is cleaner that what I was expecting from a C repo. Also, it's quite a feat to fit this into 4 MB on a 386.
Part of why these images look so nice is because these systems were not so locked down.
A pre-build floppy disk image would be great, so I could run it on my IBM PS/1 from a floppy.
Ok, I built the floppy image now. dd'ed it on a floppy and powered my IBM PS/1 up. Despite some nasty sounds of the HDD bearings that went away after 30 seconds, the floppy does not boot on this machine. Just a black screen. 386SX-25 2MB RAM. Maybe 2MB RAM too less, but I thought at least something will happen. :-)
There seems to be an 8 meg image here: https://github.com/luke8086/gentleos32/releases/download/202...
Someone prepare a set of floppy disk images so you can get the proper installation experience.
Including the mandatory corruption on the last one.
This is a hdd image, I guess.
For PS/1 you'll need the 16-bit version from https://github.com/luke8086/gentleos. A floppy image is provided in releases. Note you only need to copy the first 64KB, the rest is just padding for emulators.
GUI looks a but BeOS inspired, but somehow even cleaner.
Good catch, the yellow and blue colors are totally inspired by BeOS :D I'm even adjusting the default VGA palette to get the right tints in 16-color mode.
I think it's that yellow bar what it makes it look like BeOS. And maybe the right hand menu bar. But once you check a B/W version, it doesn't look like BeOS that much.
More like a Win 3.1 theme with BeOS colors and a NeXT desktop.
Made me think of Breadbox Ensemble, which is GEOS, and was really lovely.
Nice! The project also has a 16-bit variant https://github.com/luke8086/gentleos, not clear if it works on 8086 IBM PC, but I'll give it a go. Been looking for a reason to power up my IBM PC again.
What a lovely-looking OS! Also great to hear that the project isn't aiming for infinite changes!
Will be digging out some old hardware to test it out very soon, this is exciting!
<3<3<3
[dead]
This reminds me of the era when operating systems felt more approachable and visually distinct. Modern UIs are often cleaner, but many of them have lost some of the personality that older systems had.
> A hobby operating system for vintage 32-bit PCs.
I am all in favour of great projects, but why a differentiation between 32-bits or 64-bits? I don't understand that. Is a computer that is 32 bit or 64 bit, either way which, not worthy?
Edit: I understand a motivation if it is on simplicity choosing one or the other, but other than that I don't see why that should ever be a goal worthy to be pursued. Software should really "just work" no matter the number of bits and bytes.
In this context, 32-bit means the minimal requirement. You can absolutely run even the 16-bit version on a 64-bit PC, provided it has BIOS/legacy-boot mode.
It only won't work on modern pure-UEFI systems because that would require writing full stack of USB drivers for keyboard and mouse, and that would be a huge task.
“Vintage” 64 bit PC’s aren’t a thing.
> Edit: I understand a motivation if it is on simplicity choosing one or the other, but other than that I don't see why that should ever be a goal worthy to be pursued. Software should really "just work" no matter the number of bits and bytes.
Not really how software works.
> “Vintage” 64 bit PC’s aren’t a thing.
Just sold my SGI Indigo 2 for 900 $ ! Vintage 64 bit is absolutely a thing. :-)
they said PCs
The DEC 3000 would like to have a word with you.
It even has 64 bit "word" size!
Itanium was released 25 years ago now...
Whilst that’s definitely old in computer terms, even “retro”, is it old enough to be “vintage”?
Personally I’d have said it isn’t. But these terms are subjective.
"Vintage" usually refers to actually old stuff, while "retro" refers to new stuff that looks/sounds/feels like old stuff. So GentleOS is a retro OS designed to run on vintage hardware.
(That distinction wasn't clear to me either, so I had to look it up - TIL).
Don't worry, this is portable to both vintage word sizes: 32 and 16.
x86 boots in 16-bit real mode. Then you need to specifically transition into 32-Bit, and from 32-Bit it can be transitioned to 64-Bit Architecture...
The last step (32-bit to 64-bit) can a bit of a can of worms especially on older platforms where 64-bit implementations can differ greatly and 32-bit "just works tm". 32-bit is quite well supported and has enough resources to make some interesting programs work without much hassle.
I think the author has made the decision not to support 64-bit mode due to needing to balance the complexity and usability of the project. It is a hobby project after all.
Since the author maintains a 16-bit and 32-bit for this project I suppose if you wanted you can always fork and maintain a 64-bit version if you wanted to.
32 is vintager vintage
32 bit vinteger ;-)
You can do 32 bit voolean too, great for that vintage bitmasks to store application flags. =]
I’m just going to leave this here… https://www.instagram.com/reel/DY_UkH8Ry0V/
That video's bullshit. People prefer old UIs because modern ui is shit, not because they're not creative.
Computer programs are tools. It doesnt do anyone any good if they're unusable in the name of chasing moronic trends.
I think there's a couple of different forces at play in the convergence of GUI design that we're seeing in the past 20 years. First, there's been a huge amount of widely accepted research that shows what the most accessible way to design an interface is. Things like Google's Material Design and Apples Human Interface Guidelines come to mind. Second, the widespread availability of those two specific design guides make it increasingly common for developers to just create to those. It ensures that things just work and are increasingly portable. Third, we're in a landscape where API stability and design is prized. That's partially because of the number of times that design has been broken by updates and version changes. It takes many years for developers to update their applications when a new back-end is developed, and the time in between gives broken applications, and ugly looking fallback. You can look at running GTK1 apps on modern GNOME, or X11 apps rendering on Wayland over the past decade for an example of this.
All that said, I truly miss the days when we had interface skinning. There was a skin for OS X called UNO that was absolute perfection in my eyes, and it was ported to an old version of Android back when skinning was a thing. There's nothing like it available now. Even GNOME is highly against theming and skinning now, apparently because they like breaking with every single release rather than maintaining an API/ABI and skinning support. The themes that were available for Windows XP were so much fun, even if you had to swap out DLLs to get them working.
> First, there's been a huge amount of widely accepted research that shows what the most accessible way to design an interface is.
Focus-group based and UX research was a lot more intense in the 1990s compared to today, and late 1990s UIs are still among the best available.
I have a theory that corporations make new UIs to entertain people through them. First, to create the feeling that something is happening, and second, to increase screen time.
Old interfaces were far more practical for getting work done, and therefore obviously boring.
For me, as someone who is supposed to use technology as a tool and not as a source of amusement, the new interfaces of the major OSes feel unacceptable. But the other billion people chatting and scrolling are the real consumers, not me — and as a result, we now have the interfaces we have.
Not really a theory.
Software makers treat UIs the way auto makers treat paint and body styling.
Like almost all bullshit that is believable by a multitude, it is couched in some very small truths.
So, you poop everywhere you go. Interesting.
I don't really understand GP's message. User's comment history seems pretty normal, why would they drop a random IG link here? Wrong article?
The link is to a designer talking about how technology has led us to a design world that is mostly driven by nostalgia. I personally don’t see it as being applicable here as it deals with big design houses not hobbies, but I can see why someone might think it is.
What a designer might call nostalgia an actual user of an OS might call standard, or maybe even intuitive. The point of an operating system is to be used. If it’s pretty, that’s a bonus. Usability by the target audience is the primary concern.
To anyone interested in the video, but without an Instagram account, gramsnap sometimes works. (Imginn.com sometimes works for viewing IG profiles.)
https://gramsnap.com/en/instagram-reels-viewer/
The video is a word salad of self-aggrandizement and judgement. Don't bother. The dude holds up fashion as if it is some paragon of virtue.
The first sentence of your comment seems to describe 99% of social media.
But sir, what about memory safety? Rust would have been the better choice given the insecure nature of vintage hardware.