> When a program needs to allow the user to choose a file, drag and drop is again used, with the window providing a drop area to collect the file
The neat part was you could "save" from one app into another, without having decided to actually save the file yourself yet at all.
I have to echo the comments about the mouse button "Adjust". Being able to move windows about while they preserve depth position without some obscure shortcut was very useful.
Over the years I've grown to appreciate the extent to which whatever vision there may have been behind RISC OS originally the lack of a proper GUI toolkit and serious OS internals held them back such that by Win95 Windows really was better. At exhibitions in 94/95 Acorn devs themselves were conspicuously more interested in running NetBSD than RISC OS, and it always seemed a shame they didn't make a more serious effort to get some descendant of the RISC OS desktop ported over to a UNIX like kernel, rather like a more serious shot at the ROX desktop, but in truth Win95 won the late 90s desktop paradigm war convincingly.
ajb 5 days ago [-]
Acorn just plain didn't have the manpower. They failed to effectively enter the international market and so other companies with a wider customer base out-invested them.
They had a UNIX clone in 1988. The guy that did the kernel, Mark Taunton, used it for his daily driver until some time after 2000 but they never ported their GUI to it.
The internals of riscos were creaking by the end; it didn't have a proper library system and only had cooperative multitasking. There was an internal project ('galileo') to replace it, but it suffered from second system effect and NIH and never saw the light of day.
fidotron 5 days ago [-]
> They had a UNIX clone in 1988. The guy that did the kernel, Mark Taunton, used it for his daily driver until some time after 2000 but they never ported their GUI to it.
Are you referring to RISC iX? I had no idea there were serious users of that to be honest.
A few years ago some Acorn strategy documents leaked from the early 90s and it showed they basically knew the game was up long pre-Risc PC. I don't think any number of people would have helped them by 1992. There was this odd void where everyone (such as Xara) kind of knew the PC was going to take over everything, but it just wasn't quite there, and then suddenly it was.
ajb 5 days ago [-]
Yes, RISC iX. I've no idea how many users it had to be honest .
They kept going for a period by virtue of owning ARM. When the shareholders persuaded them to list it directly, the game was over. Before that they bet on set-top-boxes and the supposed PC replacement, the NC (thin client).
gjvc 5 days ago [-]
>the NC (thin client).
I was thinking about this the other day. It seems to have come to pass with smart TVs and consoles like the PS5.
ajb 5 days ago [-]
In a sense. NCs were really for business use though
Probably the closest thing we have is a Chromebook.
Lammy 5 days ago [-]
> rather like a more serious shot at the ROX desktop
ROX-Filer is still to this day my file manager of choice for any X11 system.
fiddlerwoaroof 5 days ago [-]
Yeah, ROX-Filer is an amazingly good bit of software, once you figure out its UX quirks
sillywalk 5 days ago [-]
> serious OS internals
There is a "moonshot" effort to redo the internals of RiscOS from RiscOS assembly.
Yeah, I'd seen this. I've not been a participant in the RISC OS community since the 90s, and clearly do not share the same enthusiasm some have for aspects of it.
I'm honestly of the view the wisest thing to do would be focus on making it play well with being emulated, right down to enabling enhancements when doing so (such as graphics acceleration), but then not being in that world it really is not my call at all.
trebligdivad 5 days ago [-]
Yeh that was pretty neat - as I remember the OS/libraries had a concept of whether the place you had dragged to was file storage or another application, so that an application could know not to mark itself as 'saved' if you just dragged to another application.
lizknope 6 days ago [-]
I searched for "ARM" and didn't see it in the article. The Acorn ran on ARM (that's what the A in ARM stood for anyway, Acorn RISC Machine)
You can run RISC OS on a Raspberry Pi which is also ARM based.
If they want to guarantee its future, maybe they should consider funding a RISC-V port as well, or instead.
lproven 5 days ago [-]
I believe that at one point there was a term in a licence agreement from one of the companies involved that basically prohibited it ever being ported to anything else... but that's long ago. I wrote about the Moonshot effort recently:
... And I'm still talking with ROOL about the directions they're examining. The restriction may no longer hold true.
msephton 5 days ago [-]
Thanks to the author for writing about their exploration and learning. I'm also of the mind that user interface designers should dig deep into the past: System 7, RISC OS, BeOS, OS/2, Amiga OS, FM Towns, X68000, GEM, etc. So much good stuff.
lproven 5 days ago [-]
Agreed.
Virtually every FOSS desktop environment out there feels very firmly built on the Windows and OS/2 tradition, and there is very little indeed that comes from any other tradition.
The ideas of boldly experimenting with alternatives seems to be confined to "hey, what if we moved the default button into the title bar?"
5 days ago [-]
Sophira 5 days ago [-]
I grew up with the RISC OS desktop, and it's one of the biggest reasons why I'm so comfortable with the multi-window interface of the GIMP, because it's very reminiscent of how things worked on RISC OS.
jlarcombe 5 days ago [-]
To this day I miss the way RISC OS handled focus, window/app activation, file loading and saving, menus, and the three-button mouse.
RetroTechie 5 days ago [-]
To those here familiar with RISC OS (like, having used it as daily driver for extended period, past or present), what about it would you regard as most worthy of preservation?
Things about its low-level internals? Specific aspects of how its parts inter-operate? Particular nifty features in the UI? General look & feel / user experience?
sillywalk 6 days ago [-]
I do like the "Adjust" mouse button.
EDIT: especially that you can use it to do multiple menu selections without actually closing the menu.
rjmunro 5 days ago [-]
> The Save dialog is normally presented as a menu, into which you enter a filename
> This menu however is deeply confusing.
No, this is the greatest piece of the RISC OS UI. It makes all other systems save dialogs seem like total nonsense. You just put files where you want them. The only thing is that the OK button shouldn't exist.
ranger_danger 5 days ago [-]
One of the biggest problems I've noticed with younger computer users is that they have no idea where they saved a file. Having to drag it to a specific folder seems like it would be harder to forget in that case.
fidotron 5 days ago [-]
The end of the Mac spatial Finder with the OS X transition was not entirely welcomed by some people on that basis, but we live in a world that prioritizes thinking in words and looks down on visual/spatial reasoning.
lproven 5 days ago [-]
TBH, stuff like Microsoft's "fluent" interface ribbons, and GNOME banishing menu bars and using buttons and hamburger menus, feels like the exact reverse of this to me.
fidotron 5 days ago [-]
The whole point of a spatial finder is it leaves things physically in the same place, which is very definitely not true of the ribbon menus (or their precursor in Office 2000 where menu item visibility was reliant on how often you used them), or anything that tries to hide/reveal things based on assumed intention.
Burger menus are just rotated tab bars that can adapt equally badly to mobile and desktop.
lproven 3 days ago [-]
Good points.
(I hated that O2K feature and so memorised where to go to turn it off.)
The neat part was you could "save" from one app into another, without having decided to actually save the file yourself yet at all.
I have to echo the comments about the mouse button "Adjust". Being able to move windows about while they preserve depth position without some obscure shortcut was very useful.
Over the years I've grown to appreciate the extent to which whatever vision there may have been behind RISC OS originally the lack of a proper GUI toolkit and serious OS internals held them back such that by Win95 Windows really was better. At exhibitions in 94/95 Acorn devs themselves were conspicuously more interested in running NetBSD than RISC OS, and it always seemed a shame they didn't make a more serious effort to get some descendant of the RISC OS desktop ported over to a UNIX like kernel, rather like a more serious shot at the ROX desktop, but in truth Win95 won the late 90s desktop paradigm war convincingly.
They had a UNIX clone in 1988. The guy that did the kernel, Mark Taunton, used it for his daily driver until some time after 2000 but they never ported their GUI to it.
The internals of riscos were creaking by the end; it didn't have a proper library system and only had cooperative multitasking. There was an internal project ('galileo') to replace it, but it suffered from second system effect and NIH and never saw the light of day.
Are you referring to RISC iX? I had no idea there were serious users of that to be honest.
A few years ago some Acorn strategy documents leaked from the early 90s and it showed they basically knew the game was up long pre-Risc PC. I don't think any number of people would have helped them by 1992. There was this odd void where everyone (such as Xara) kind of knew the PC was going to take over everything, but it just wasn't quite there, and then suddenly it was.
They kept going for a period by virtue of owning ARM. When the shareholders persuaded them to list it directly, the game was over. Before that they bet on set-top-boxes and the supposed PC replacement, the NC (thin client).
I was thinking about this the other day. It seems to have come to pass with smart TVs and consoles like the PS5.
Probably the closest thing we have is a Chromebook.
ROX-Filer is still to this day my file manager of choice for any X11 system.
There is a "moonshot" effort to redo the internals of RiscOS from RiscOS assembly.
https://www.riscosopen.org/news/articles/2025/03/28/moonshot...
https://www.riscosopen.org/content/documents/risc-os-moonsho...
I'm honestly of the view the wisest thing to do would be focus on making it play well with being emulated, right down to enabling enhancements when doing so (such as graphics acceleration), but then not being in that world it really is not my call at all.
You can run RISC OS on a Raspberry Pi which is also ARM based.
https://www.riscosopen.org/content/sales/risc-os-pi
https://www.theregister.com/2025/04/01/risc_os_open_moonshot...
... And I'm still talking with ROOL about the directions they're examining. The restriction may no longer hold true.
Virtually every FOSS desktop environment out there feels very firmly built on the Windows and OS/2 tradition, and there is very little indeed that comes from any other tradition.
The ideas of boldly experimenting with alternatives seems to be confined to "hey, what if we moved the default button into the title bar?"
Things about its low-level internals? Specific aspects of how its parts inter-operate? Particular nifty features in the UI? General look & feel / user experience?
EDIT: especially that you can use it to do multiple menu selections without actually closing the menu.
> This menu however is deeply confusing.
No, this is the greatest piece of the RISC OS UI. It makes all other systems save dialogs seem like total nonsense. You just put files where you want them. The only thing is that the OK button shouldn't exist.
Burger menus are just rotated tab bars that can adapt equally badly to mobile and desktop.
(I hated that O2K feature and so memorised where to go to turn it off.)
• <http://toastytech.com/guis/indexriscos.html>
• <https://guidebookgallery.org/guis/riscos>