And it seems as if another minor, barely-alive operating system will become encumbered by legal bickering between two small companies. The RISC OS scene, which is already a tangled and complicated mess of companies, version number teasing and incompatible versions, might be torn apart even further because RISCOS Ltd might take legal action trying to prevent RISC OS Open Ltd from releasing a RiscPC compatible ROM from the RISC OS 5 shared source project. Should you feel confused, you needn’t worry: so does everyone else.
Let me explain. Wait, let me rephrase that. Let me try to explain. There are two companies working on RISC OS. There’s RISCOS Ltd, who develops RISC OS 4 and 6 for RiscPC and A7000 machines, as well as for various emulators and later on the A9home. These are the guys who recently released the “virtually free” ROM image which you can use in an emulator to run RISC OS 4. RISCOS Ltd acquired the license to work on RISC OS in 1998 from Pace after the demise of Acorn.
The other company is Castle, who develops RISC OS 5 for their Iyonix range of computers. Castle bought the 32bit source code to RISC OS from Pace in 2002, and used that code to build RISC OS 5. I guess someone at Pace thought it was a pretty hilarious joke. Can’t blame him.
In any case, this has obviously disaster written all over it – there were a few back and forths between the two companies, which Drobe neatly sums up:
In 2003 Castle said it bought the OS “lock, stock and barrel” from Pace, however ROL continue to dispute the level of ownership Castle has of the OS, drawing attention to the wording of the buy-up announcement. However Castle maintain to date that it bought the OS from Pace and had instructed its own lawyers to double-check its legal position. A year later Castle and ROL briefly kissed and made up after Castle tried to stop ROL developing its stream of RISC OS soon after Castle announced its purchase of the OS.
Caste is also the company that initiated the shared source scheme, a sort-of open source thing (you may do whatever you want with the code, except sell it “as part of a hardware product”) where the source code to RISC OS 5 would become available to the public, a process guided by a new group called RISC OS Open Ltd, which also takes care of the development of the code. The problem is that RISC OS Open Ltd wants to release a ROM image built for RiscPC machines – exactly, the market RISCOS Ltd caters to. RISCOS Ltd is fine with RISC OS Open Ltd releasing ROMs for the Iyonix range – but it’s not happy with ROMs compatible with the RiscPC.
If you’re already lost, you might as well move on up to another news item, because it will only get worse from here. I’m not even going to attempt to explain what follows – I’ll let the experts over at Drobe do that for me:
ROL’s new strengthened stance on the OS ownership dispute appears to have come about after company director Aaron Timbrell reviewed the paperwork documenting ROL’s licence to develop and distribute RISC OS. This agreement was initially drawn up with E-14, a company formed during the break-up of Acorn that took on the intellectual property rights (IPR) to the technology that we know as RISC OS. This technology was subsequently snapped up by Pace to use in set-top boxes. ROL were, at the time, granted exclusive rights to continue working on RISC OS for the Acorn enthusiast market and from Acorn’s final sources it produced RISC OS 4. ROL say this early desktop-only remit was extended by Pace at a later date – paving the way for RISC OS 4 to be bundled with emulator VirtualRiscPC (which is incidentally published by Aaron Timbrell).Having leafed through the licence agreement and various bits of correspondance and public postings, Aaron now believes RISC OS 5 is a derivative of the source code base that RISC OS 4 was born from. It is understood that some of ROL’s early changes to RISC OS had to be passed back to Pace, and ROL’s position appears, in part, to rest on whether or not these ROL-authored updates made their way into the code that Pace engineers eventually 32bitted and turned into RISC OS 5.
And having fulfilled various obligations set out in the original agreement with E-14 (such as the aforementioned requirement of ROL to share early source code changes with Pace), ROL can now safely declare itself owner of the OS source it picked up in 1998 and all subsequent updates to it, claims Aaron – who believes Castle’s RISC OS 5 also falls under this. Therefore, according to Aaron, ROL own RISC OS 5 and suitable licences must be obtained from ROL before anyone can produce a RISC OS-powered product that falls within ROL’s licence remit. Hence the latest objection to plans to release a free RiscPC-compatible ROM image buit from the RISC OS 5 sources by RISC OS Open using materials provided by Castle.
RISC OS Open Ltd’s response came from its boss, Steve Revill, who tells the community not to lose too much sleep over this one. “The IPR belongs to Castle Technology Ltd and ROOL [RISC OS Open Ltd] have released it under a shared source licence provided by Castle. If ROL [RISCOS Ltd] has any complaint they need to address it to Castle, not to ROOL.” As a second point, he added: “ROOL has no contractual relationship with ROL and are not party to any agreements between ROL and Castle. Accordingly, ROL has no legal basis whatsoever to take action against ROOL as they don’t own the IPR but merely license it from Castle.” Revill states that if RISCOS Ltd has any objections, they should take it to a court of law to settle the matter.
RISCOS Ltd has also stated they have offered a license to RISC OS Open Ltd, allowing the shared source initiative to continue unencumbered with what it is doing. And here you were, thinking the Amiga scene was a mess.
In all seriousness, this is a very sad thing to see. There aren’t that many truly alternative operating systems out there, especially not ones as exotic and eccentric as RISC OS 4/5/6/whatever. Seeing such an already small community fall prey to legal bickering and infighting just makes me cringe – the men and women in charge there need a serious beating with the cluestick: if you guys don’t start working together, harmonising the various different versions, moving towards a unified, single RISC OS, then you guys are done for.
It’s sad and pathetic at the same time.
“can’t we all jsut get along?”
honestly this is just silly. let RiscOS get a little more exposure, god knows it needs it to stay alive.
“””
Seeing such an already small community fall prey to legal bickering and infighting just makes me cringe
“””
In my younger days, I was involved with a number of small volunteer organizations. I learned early that the serious bickering and infighting always started shortly after it was decided, for practical reasons, to open a bank account, and a checkbook appeared on the scene. These organizations were not large enough to have competing organizations to battle, however. So I guess this is a little different. All of our infighting was unambiguously internal.
Has anyone else seen the episode of South Park where Stan’s dad does the worlds biggest crap and Bono from U2 tries to beat him by doing a bigger one? Sort of reminds me of the entire RISC OS scene. I own an A7000, I used RISC OS 2 a *lot* back in the day. I used to write demos and stuff in basic 5. But, RISC OS is probably the most pointless OS in the world. Give me Amiga over Acorn Archies any day
Something the article didn’t seem to explain… RISC OS always ran on 32 bit ARM processors, but before PACE worked their 32bit mojo, it actually ran in 26bit mode, (though earlier versions might have actually ran on 26bit ARM processors.. I forget, and really don’t care enough to find out :-))
This 26bit and 32bit is all about memory addressing and processor status flags.
Old ARM chips (ARM2, ARM3, and their bretheren) only supported 26bit addressing which severely limited how much memory they could support. There was a transition period when ARM chips supported both 26bit and 32bit addressing. Current ARM chips only support 32bit addressing.
The two addressing modes aren’t entirely code compatible, since the original “26bit” ARM chips used bits in the program counter register (R15) as processor status flags. Those status flags were moved out of R15 when 32bit addressing was introduced, although the first chips supporting 32bit addressing would support a 26bit addressing mode which would ensure the status bits were still present in R15.
Some parts of RISC OS (and other code) relied on the presence of the status bits in R15, and thus that code was not 32bit-safe.
Memory addressing!! That’s it!! I knew it was something silly like that.
Mind you, the Acrchies were always billed (to us students who used them by our teachers/professors) as being “32bit”, as I guess the actual processor ran in a 32bit mode.
Remember 26 bits = 64 million addresses. At that time Amiga was limited to 16 million, Atari for some odd hardware reason to even less. Mac were still at 16 million if I remember right.
And Intel machines were using all sorts of incompatible addressing scheme to break 1 million bytes.
Acorn and ARM users must have thought they had it made in the shade.
The 26-bit addressing limitation was only on code; data could be anywhere in a 32-bit address space (4GB), hardware permitting.
So, you’d have to “limit” your code (per process) to the lowest 64MB in memory. That’s still quite a lot of code.
Maybe this is why they released that “virtually free” ROM image. To get some quick cash to be able to declare legal war on Castle.
This whole thing sounds retarded. I don’t know either company (or the OS) that well, but I’m hoping for Castle to win this one (assuming RISC Ltd. continues on). If Castle *bought* the RISC OS code from Pace (who apparently RISC Ltd. did also), how the hell can RISC Ltd. complain about they do with it? They both own it, and both should be free to do whatever the hell they feel like. I don’t think either one has something similar to the Intel/Microsoft duopoly, so what the hell are the bitching about?
RISC Ltd. is sounding like a bunch of arrogant assholes right now, kind of like another company I know all too well.
Don’t believe everything you read!
Both parties have commented on Drobe’s forum that there is no dispute.
From ROOL:
RISC OS Open and RISCOS Ltd are not waving their fists at each other. We’ve enjoyed open and clear communications from the outset and are continuing to talk. Finding the best way for all parties to co-operate for the good of RISC OS has always been one of our aims.
From ROL:
there is no problem, disagreement or conflict between RISCOS Ltd. and ROOL. RISCOS Ltd. are happy with what ROOL are doing and have made offers to them to ensure that their work can continue.
Whilst Drobe generally does a good job of being the most up to date source of information with regards RISC OS news it does like to lapse into sensationalist journalism every now and again- this seems to be one of those times.
So some have wondered what was so special about the OS.
Okay… it booted in 5 secs from ROM.
I has OS filesystems. Not sure of the tech here but the essence is you can write an app and allow the imageFS in the OS to handle the image filetype in the app. EG the word processor that could read and write Word files came on a floppy and was 800k. Compare that to the major OS bloatware.
When I first used it in 1994, you could embed a graphic in a document and resize it, rotate it etc in real time. WOW.
But here is the best thing for me. The writable window/application (where the cursor is) does not have to be the active window/application (wher the mouse clicks are entered. BIG DEAL you may say.
Well lets assume ther is an application eg a word processor. it lacks some function or formatting. We you can use an external app (with that function) to write the output to your wordprocessor. EG !ZAP html, Richard Goodwins web design app etc etc etc. Indeed you could even use the Pluto !speak (speaks your email to you) module to talk other apps content to you.
Ohhh how I miss it.
Cheers
Bob on Mandrive