6 ms·

Acorn were indeed way ahead of their time. Especially when it came to flunking software projects. They had the hardware done but the OS was a quick hack job bec

by m0xte 6y ago

Acorn were indeed way ahead of their time. Especially when it came to flunking software projects. They had the hardware done but the OS was a quick hack job because their proper OS (ARX) was late. So they ported MOS from the BBC series and threw a GUI on top. And it initially sucked terribly. I was there for it. I bought a fully stacked A440 when they were released, many thanks to the kindness of a deceased relative and much to the dismay of it mother who said I should have bought history books and art materials instead. Really if they had something RISC OS 2 quality on day one, which would have been a monumental feat like the hardware was they’d have done better. I enjoyed every moment of using the machine after RISC OS 2 came out. Before that I was slightly scared I’d burned a monumental amount of money.

markb139 6y ago
I couldn’t stretch to an A440, but did get an A310. What a great machine. It was blindingly fast for the time (1987). In about 1994 it still out performed a new 486DX100 PC
- mattkevan 6y ago
  I used an A310 and later an A440 as my main machines right up until the late ‘90s when we got one of the original bondi-blue iMacs. Loved RiscOS - I knew it inside out and backwards, and it was so ahead of it’s time, that using one of the terrible RM Nimbus 286 PCs they had at college felt like being suddenly transported back to the Stone Age.
  - cstross 6y ago
    Idle thought: I imagine that as soon as the Apple ARM development boxes drop, someone is going to start evaluating how to port RiscOS to Mac (either on the bare metal or, more likely, via virtualization).
    - m0xte 6y ago
      Never thought of that. I'm getting even more excited now, which is a rarity as I'm a miserable sod.
      - mattkevan 6y ago
        Yeah that would be amazing. As an aside: I'd love to see the alternate timeline where RiscOS continued to be developed, instead of getting bogged down with pointless legal issues for 20 years. What would it be like? Would it have gone through a Mac OS-like transition from RiscOS Classic to RiscOS X? If it was open sourced in the early 2000s, could it have become a viable competitor to Linux on the desktop and Android on mobile? As a lightweight but powerful OS for low-powered ARM devices it should have been perfectly placed.
- m0xte 6y ago
  Yeah the A440 was not cheap. I didn't understand money as much as I do now back then. Equivalent inflation adjusted is about £6500 now (!). Which is why my parents had a heart attack when I just blew it instantly on that. Absolutely would not have bought it if I had some more clue. But to this day I have no regrets as it gave me an excellent foundation, understanding and career in computing. But yes the things lasted a hell of a long time. Mine went until 1998 at which point it was replaced with a Pentium II box running Windows dual booting with Linux. Got 10 years out of that machine which is remarkable considering the progress and change in the 1990s. I'm on a Mac mini at the moment with an iPad Pro as a secondary machine so the moment an ARM Mac comes out I'm back with an ARM desktop again. It's all a big circle. Perhaps ARX was going to be an early macOS X contender... who knows?
  - lproven 6y ago
    I facilitated a talk at the London RISC OS user group by the lead architect of RO, Paul Fellows: http://www.rougol.jellybaby.net/meetings/2012/PaulFellows/ http://www.rougol.jellybaby.net/meetings/2012/PaulFellows/ From what he said about ARX, I suspect not...
    - m0xte 6y ago
      Thanks for this. Will read later. Wasn’t ARX basically Unix reimplemented in Modula2 ?
- guenthert 6y ago
  Not sure whether you're confusing some OS or application performance with the performance of the CPU or are misremembering. A 8MHz ARM2 performed at about 5MIPS. That was impressive in '87, but in '90 my 20MHz 80386 ran at about 6MIPS. The 486DX100 of the mid nineties was much faster.
  - m0xte 6y ago
    Possibly in terms of feel. They really ran a pile of crap on a 486DX100 in comparison. The RISC OS machines of the early 1990s had a similar feel to a modern PC on the performance front with every-day tasks.
mcv 6y ago
I'm not sure if hack-job OSs are 'ahead of their time'. These were also the days of MS-DOS. But it's certainly good to remember that not everything Acorn did was shiny and golden. They deserve to be remembered more than they often are, but they've had their fair share of misses too.
- m0xte 6y ago
  Oh yeah the entire 1990s was a shit show where they burned the entire company to the ground through incompetence.
rahoulb 6y ago
One of the things I loved about RiscOS was a small UI feature. There were no open or save dialogs. Instead, when you saved a file, it minimised to its icon and then you dragged it to the relevant folder. To open a file, you located it and double-clicked.
- regularfry 6y ago
  Yes! This sort of universal drag-and-drop is something I've been hoping other OSes would pick up on, but you only ever get part-way implementations.
- m0xte 6y ago
  I liked that but sometimes it was a pain in the butt. Particularly if you'd forgotten to open the destination directory window already and the save thing was down 3 levels of menus.
- blywi 6y ago
  For me drag-and-drop is not complete without also being able to save by drag-and-drop. On RISC OS you can save your files by dragging the file icon to the place you want to save your file. This can be a directory window, or it could also be another app icon on the taskbar, or an open app window. This way it is possible to create workflows where you have small specialized apps that complement each other. You would open a file by dragging it to the first app, edit it, drag the result to a 2nd app do some more editing, then drag the result to the next app and so on and save only the final result back to disk. This is basically supporting the Unix philosophy of having many small programs that do one thing well, and combining them to achieve the desired result. Instead what we now mostly have on all platforms are huge monolithic programs that try to cram every possible feature into a single package. One reason for this, is in my opinion, that a workflow that is similar to the RISC OS way of dragging from app to app is just to cumbersome on systems without drag and drop between apps, as every step one has to go through the filing system, saving a file by browsing through a directory tree and browse the directory tree again when loading the file in the next app.
  - m0xte 6y ago
boygobbo 6y ago
> the OS was a quick hack job You mean Arthur? I have to agree - the only impressive thing about it was that it was written in BBC Basic. I also spent too much on an A440 (and a multisync monitor that literally made my hair stand on end when I turned it on). Still, I have no regrets. As you say, RISC OS 2 delivered the goods. By day, I was working in monochrome text mode on a clunky old 16-bit IBM XT. By night, I was in the glorious high-res colour 32-bit future. One of the finest computers even made IMHO (except for the mouse!)
- m0xte 6y ago
  Yes Arthur was the hack job I was referring to :) My father was using a 286 by then. It was vastly inferior however he built some software with MS PDS (quickbasic) I think it was and made a lot of cash. Software in the blood.

by m0xte 6y ago

markb139 6y ago

I couldn’t stretch to an A440, but did get an A310. What a great machine. It was blindingly fast for the time (1987). In about 1994 it still out performed a new 486DX100 PC

mattkevan 6y ago
I used an A310 and later an A440 as my main machines right up until the late ‘90s when we got one of the original bondi-blue iMacs. Loved RiscOS - I knew it inside out and backwards, and it was so ahead of it’s time, that using one of the terrible RM Nimbus 286 PCs they had at college felt like being suddenly transported back to the Stone Age.
- cstross 6y ago
  Idle thought: I imagine that as soon as the Apple ARM development boxes drop, someone is going to start evaluating how to port RiscOS to Mac (either on the bare metal or, more likely, via virtualization).
  - m0xte 6y ago
    Never thought of that. I'm getting even more excited now, which is a rarity as I'm a miserable sod.
    - mattkevan 6y ago
      Yeah that would be amazing. As an aside: I'd love to see the alternate timeline where RiscOS continued to be developed, instead of getting bogged down with pointless legal issues for 20 years. What would it be like? Would it have gone through a Mac OS-like transition from RiscOS Classic to RiscOS X? If it was open sourced in the early 2000s, could it have become a viable competitor to Linux on the desktop and Android on mobile? As a lightweight but powerful OS for low-powered ARM devices it should have been perfectly placed.
m0xte 6y ago
Yeah the A440 was not cheap. I didn't understand money as much as I do now back then. Equivalent inflation adjusted is about £6500 now (!). Which is why my parents had a heart attack when I just blew it instantly on that. Absolutely would not have bought it if I had some more clue. But to this day I have no regrets as it gave me an excellent foundation, understanding and career in computing. But yes the things lasted a hell of a long time. Mine went until 1998 at which point it was replaced with a Pentium II box running Windows dual booting with Linux. Got 10 years out of that machine which is remarkable considering the progress and change in the 1990s. I'm on a Mac mini at the moment with an iPad Pro as a secondary machine so the moment an ARM Mac comes out I'm back with an ARM desktop again. It's all a big circle. Perhaps ARX was going to be an early macOS X contender... who knows?
- lproven 6y ago
  I facilitated a talk at the London RISC OS user group by the lead architect of RO, Paul Fellows: http://www.rougol.jellybaby.net/meetings/2012/PaulFellows/ http://www.rougol.jellybaby.net/meetings/2012/PaulFellows/ From what he said about ARX, I suspect not...
  - m0xte 6y ago
    Thanks for this. Will read later. Wasn’t ARX basically Unix reimplemented in Modula2 ?
guenthert 6y ago
Not sure whether you're confusing some OS or application performance with the performance of the CPU or are misremembering. A 8MHz ARM2 performed at about 5MIPS. That was impressive in '87, but in '90 my 20MHz 80386 ran at about 6MIPS. The 486DX100 of the mid nineties was much faster.
- m0xte 6y ago
  Possibly in terms of feel. They really ran a pile of crap on a 486DX100 in comparison. The RISC OS machines of the early 1990s had a similar feel to a modern PC on the performance front with every-day tasks.

mcv 6y ago

I'm not sure if hack-job OSs are 'ahead of their time'. These were also the days of MS-DOS. But it's certainly good to remember that not everything Acorn did was shiny and golden. They deserve to be remembered more than they often are, but they've had their fair share of misses too.

m0xte 6y ago
Oh yeah the entire 1990s was a shit show where they burned the entire company to the ground through incompetence.

rahoulb 6y ago

One of the things I loved about RiscOS was a small UI feature. There were no open or save dialogs. Instead, when you saved a file, it minimised to its icon and then you dragged it to the relevant folder. To open a file, you located it and double-clicked.

regularfry 6y ago
Yes! This sort of universal drag-and-drop is something I've been hoping other OSes would pick up on, but you only ever get part-way implementations.
m0xte 6y ago
I liked that but sometimes it was a pain in the butt. Particularly if you'd forgotten to open the destination directory window already and the save thing was down 3 levels of menus.
blywi 6y ago
For me drag-and-drop is not complete without also being able to save by drag-and-drop. On RISC OS you can save your files by dragging the file icon to the place you want to save your file. This can be a directory window, or it could also be another app icon on the taskbar, or an open app window. This way it is possible to create workflows where you have small specialized apps that complement each other. You would open a file by dragging it to the first app, edit it, drag the result to a 2nd app do some more editing, then drag the result to the next app and so on and save only the final result back to disk. This is basically supporting the Unix philosophy of having many small programs that do one thing well, and combining them to achieve the desired result. Instead what we now mostly have on all platforms are huge monolithic programs that try to cram every possible feature into a single package. One reason for this, is in my opinion, that a workflow that is similar to the RISC OS way of dragging from app to app is just to cumbersome on systems without drag and drop between apps, as every step one has to go through the filing system, saving a file by browsing through a directory tree and browse the directory tree again when loading the file in the next app.
- m0xte 6y ago

boygobbo 6y ago

> the OS was a quick hack job You mean Arthur? I have to agree - the only impressive thing about it was that it was written in BBC Basic. I also spent too much on an A440 (and a multisync monitor that literally made my hair stand on end when I turned it on). Still, I have no regrets. As you say, RISC OS 2 delivered the goods. By day, I was working in monochrome text mode on a clunky old 16-bit IBM XT. By night, I was in the glorious high-res colour 32-bit future. One of the finest computers even made IMHO (except for the mouse!)

m0xte 6y ago
Yes Arthur was the hack job I was referring to :) My father was using a 286 by then. It was vastly inferior however he built some software with MS PDS (quickbasic) I think it was and made a lot of cash. Software in the blood.