Linked by Thom Holwerda on Thu 23rd Oct 2008 14:19 UTC
Thread beginning with comment 334806
To view parent comment, click here.
To read all comments associated with this story, please click here.
To view parent comment, click here.
To read all comments associated with this story, please click here.
Guess what ARM originally stood for? That's right, Acorn RISC Machine. No points for guessing who created the RISC OS...
http://en.wikipedia.org/wiki/ARM_architecture
You'd think so; (I read somewhere that the ARM architecture was designed for, or in parallel with, RiscOS) but it's been brought up in the past on OSNews that RiscOS has no memory protection, preemptive multitasking, or multi-user support...
RISC OS is slightly later than ARM. Acorn originally had plans for a much more ambitious OS for ARM, but they realised that they couldn't finish this in time, so they made a much simpler OS called "Arthur". This was later extended to become RISC OS.
RISC OS certainly as memory protection, and has had this since Arthur. No pre-emptive multi-tasking, but that is not nearly as large a problem as you would think. And for netbooks, I really think multi-user support is redundant.
But I agree that RISC OS no longer offers many features that you can't find in other systems, and it lacks many of the browser-related technologies (Flash etc.). What it does have is a font system that gives readable text at low resolutions and an advanced GUI that runs with little resources (both CPU and memory).
Ten years ago, it would have been an obvious choice, but now it is just one of many plausible choices.
RISC OS is slightly later than ARM. Acorn originally had plans for a much more ambitious OS for ARM, but they realised that they couldn't finish this in time, so they made a much simpler OS called "Arthur". This was later extended to become RISC OS.
Yes, the original OS project was called ARX. It was being developed at Acorn's Palo Alto Research Centre, and mostly written in Modula-2. It was to have a unix-like design with a Mac-like GUI. Sadly whilst the project was being written by a bunch of very bright academics, they had little real-world experience, and it would also seem no competent project managers. With no sign of delivery, ever, and the launch of the Archimedes looming it got axed. Arthur (short for Acorn Risc by THURsday) was essentially a quick hack based on Acorn's old 8-bit OS for the BBC Micro, spruced up by a bunch of game developers, with hardly an OS developer to be found. Arthur got renamed RISC OS for version 2.
RISC OS certainly as memory protection, and has had this since Arthur. No pre-emptive multi-tasking, but that is not nearly as large a problem as you would think. And for netbooks, I really think multi-user support is redundant.
RISC OS 2 and 3 most definitely did not have memory protection. A buggy application could crap all over any part of memory, bringing the machine crashing down in a heap.
One of my hobbies 15 years back was cracking RISC OS games. The lack of memory protection was very handy.
The co-operative multi-tasking system was lousy. If an app misbehaved your GUI would become useless; your mouse pointer would still move but that was it. Any app taking up too much CPU, and not ceding control back to the OS often enough, made for a sluggish and unresponsive system. A buggy app getting stuck in an infinite loop could demand you reboot to regain control.
15 years back I'd drunk the kool-aid and was sold on RISC OS's co-operative model. In retrospect, I put up with a load of crap and didn't know better.
(NB I think that post-Acorn these flaws got partially addressed, but I'd stopped developing for RISC OS before Acorn died.)
Unfortunately RISC OS is inherently an insecure and unstable OS.
You're right tho, multi-user is absolutely pointless on a netbook. :-)
Ten years ago, it would have been an obvious choice, but now it is just one of many plausible choices.
Ten years ago these major flaws in RISC OS were there, so I wouldn't say it was an obvious choice. Acorn knew this and in Feb 1997 they announced a replacement OS called Galileo, which was not expected to be RISC OS compatible.
Which incentive to rewrite yet another kernel?
Linux is there with support for all kinds of hardware, it can be made to boot in 5s (on a SSD, 10s with an HDD), it has real time support, nice thread performance, etc.
So the kernel is a "done deal", what remains to be done is to build a good OS with it (I consider current Linux's user space to be just average: lack of responsiveness is capital sin in my book).
Unfortunately many prefers to reinvent the wheel..






Member since:
2005-08-28
You'd think so; (I read somewhere that the ARM architecture was designed for, or in parallel with, RiscOS) but it's been brought up in the past on OSNews that RiscOS has no memory protection, preemptive multitasking, or multi-user support... It's basically another AmigaOS 3 or MacOS 7; amazingly capable for its time or even after, but lacking internal features people expect nowadays.
As for size, I've mostly resigned myself that we're not going to see those tiny OSes with megabyte footprints any more, unless we're ok with giving up a lot of features...
Edited 2008-10-23 22:00 UTC