Drobe :: The archives
About Drobe | Contact | RSS | Twitter | Tech docs | Downloads | BBC Micro

RISC OS 5 GPL breach accusations

By Chris Williams. Published: 7th Feb 2003, 23:47:16 | Permalink | Printable

Linux kernel PCI and I/O code used, ARM Linux developer claims [Updated 8 Feb 2003]

It's not just RISC OS that runs on ARM powered hardware, there's also the port of Linux funnily enough called ARM Linux. Linux, if you haven't heard of this beast, is a free open source UNIX operating system clone developed solely by enthusiasts across the world. Don't panic! This isn't a rant on linux advocacy, we wouldn't do that to you.

Anyone can get the source to it and do whatever they like to it, provided you follow the GPL - the licence the Linux operating system is distributed under. The GPL is fairly complex but simply put, it ensures that software remains free (as in freedom, not necessarily the price). If you modify GPLed software or use GPL sources in other software and then release said software, these changes fall under the GPL too and thereby becomes free. RISC OS on the other hand is very different with respect to the GPL as it's completely closed source and the source code is closely guarded.

Now, to hear of allegations that RISC OS 5 (the 32 bit variant of RISC OS that Castle deploy in their XScale powered Iyonix machines), contains GPLed code is somewhat bewildering. What are the implications of embedding GPLed Linux kernel code in a proprietry system such as RISC OS? It looks like we're about to find out.

Russell King, developer of ARM Linux, has today disclosed his belief that RISC OS 5 features GPLed Linux kernel code. Oh my. After one or two postings to the ARM Linux mailing list, Russell has decided to go public.

"It would appear that Castle Technology Limited, UK, have taken some of the Linux kernel 2.5 code", Russell claimed on the Linux Kernel mailing list, "and incorporated it into their own product, 'RISC OS', which is distributed in binary ROM form built into machines they sell. This code is linked with other proprietary code."

This is pretty illuminating stuff (if it turns out to be proven true mind you, Castle have yet to issue any kind of statement), especially now that Slashdot have thrown the story onto their front page. How many hundreds of thousands of readers do they get? Oh, oh my. For what it's worth, the Linux code Russell claims Castle are using is the source code to the PCI subsystem and I/O resource allocation code.

Naturally, we'll assess the fall out shortly once we're in a position to comment more fully. Thanks to the couple of readers who kindly tipped us off.

Justin Fletcher has informed us that his evidence and further technical material on the Linux PCI code allegedly appearing in RISC OS 5 is now available from his website. The archive steps through his detective work and includes code examples.


Russell publically makes his claim
Slashdot hits out RISC OS 5

Previous: 2003: Five years on from Acorn
Next: Castle and Freeware news round up


Viewing threaded comments | View comments unthreaded, listed by date | Skip to the end

Oh shit! If this is true, Castle are going to have to act on a few things very quickly, although it appears that they have known that they have been discovered for a few weeks now...

Shit shit shit, just when things were starting to look better in RISC OS land... -- Gavin Smith, Carrickfergus

 is a RISC OS UserSparkY on 8/2/03 2:35AM
[ Reply | Permalink | Report ]

It's not all that bad. It just means they'll have to (according to my interpretation and understanding of the GPL) release stuff they've used/changed (ie: the PCI code).

Provided these allegations are proven true, of course. We'll have to wait and see.

Chris, drobe.co.uk

 is a RISC OS Userdiomus on 8/2/03 2:56AM
[ Reply | Permalink | Report ]

Maybe I got carried away reading the several hundred posts on Slashdot...

Still, my (very inadequate) understanding of the GPL was that if a product contained GPL'd source, the product itself needed to be released under the GPL, not just that specific part - i.e. RISC OS as a whole, which Castle couldn't of course do. Erm, maybe I should go and read up on this thing. Still, naughty Castle eh? -- Gavin Smith, Carrickfergus

 is a RISC OS UserSparkY on 8/2/03 3:01AM
[ Reply | Permalink | Report ]

Let's just say I usually believe anything Justin Fletcher says ;-)

I wonder how Castle thought they could get away with it? I can see their thinking; the end was in sight, and they just needed to "bend" the rules of the GPL. They should have known they would have got found out, mind you.

Let's hope diomus's interpretation of the GPL is correct - if it is, AFAICS there is no problem.

I'm torn as to whether I enjoy the excitement of the real world impinging on RISC OS, or whether I'm slightly embarassed for Castle.

 is a RISC OS Usermoss on 8/2/03 9:32AM
[ Reply | Permalink | Report ]

God; reading that Slashdot thread - it's like a group of bullies ganging up on some poor defenceless sod.

 is a RISC OS Usermoss on 8/2/03 9:40AM
[ Reply | Permalink | Report ]

Oh dear, just had a quick look at that there slashdotty page and there do seem some angry young people there! Just how are so many of them aware of applications using code they say is part of the Linux kernel? Surely not reverse engineering copyright material are they? -- Ryan Hitch, Cottingham

 is a RISC OS Userdrjones69 on 8/2/03 11:49AM
[ Reply | Permalink | Report ]

How do we know it's Castle whodunit and not Pace?

 is a RISC OS Userhutchies on 8/2/03 12:13PM
[ Reply | Permalink | Report ]

The comments about the impact as a whole:

I found a GPL on my hard drive, it might be an out of date one, but it says the following:

"If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works."

 is a RISC OS UserSimonC on 8/2/03 1:14PM
[ Reply | Permalink | Report ]

Thas't a good point hutchies.

We all knew that RISC OS is a Pace Product, even if just Castle actually use the here discussed version 5.x

Anyone should that.

For me it sounds not that seriuos. We should pass the whole story once:

- Castle offers and sales the Iyonix and from same dark corners there came statements like prototype hardware, beta-quality software and so on

After the dust has layed :

- Rumours came to light that Castle did not licensed RISC OS from Pace and that they (Castle) are using against any law.

After the dust has layed :

- Castle now should be guilty using GPL code in RISC OS, which itself is not licensed to Castle...

I do not know about the people saying all that and i do not think they really want to harm Castle. But fpr me all that looks like 'hunting witches' (german Hexenjagd).

Of course IF Castle is giulty they should pay. That's for the lawers. But until this has been proofed Castle must be seen as not guilty.

Sincerely Hauke

(I am afraid my bad english made me missing the real point...)

 is a RISC OS UserVLIW on 8/2/03 1:59PM
[ Reply | Permalink | Report ]

The slashdot people do sound furious; but you have to wonder why given that they're pro-MP3 and pro-DVD copying and sharing. "We'll take your copyrighted work but don't you dare screw with the GPL".

As for the Castle vs. Pace thing, I say that it's Castle's product (they have the source after all) so it's their call.

Chris, drobe.co.uk

 is a RISC OS Userdiomus on 8/2/03 2:05PM
[ Reply | Permalink | Report ]

Perhaps someone can help me understand the implications of this. As I understand, the GPL is not a law so it remains to see if Castle or Pace have broken copyright. If they have, it must be proved. Who would own the copyright?

Chris: I agree - I'd say 50% of people that read Slashdot use Linux on a regular basis. Their reaction is completely expected! -- Simon Wilson, Boulder, Colorado

 is a RISC OS Userksattic on 8/2/03 7:23PM
[ Reply | Permalink | Report ]

Chris, why do you assume the folk at Slashdot all have the same beliefs and views?

As for the GPL, you have to release the source to your entire product. It was created that way to ensure that certain people didn't side-step it by putting all the GPL code into some sort of library and just call the functions from there. The LGPL allows you to use code without GPLing your product, you just have to release the changes you made to that library/program.

 is a RISC OS UserEl Capitano on 8/2/03 7:24PM
[ Reply | Permalink | Report ]

SimonC - I read that part of the GPL but unfortunately there's more following that sentence:

"But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License" -- Simon Wilson, Boulder, Colorado

 is a RISC OS Userksattic on 8/2/03 7:29PM
[ Reply | Permalink | Report ]

ksattic, I'm afraid trying to challenge the GPL is a lose-lose situation: 1) The GPL is enforced, Castle end up paying damages/forced to GPL RISC OS 5 (Is that legal given the scope of their license with Pace?). 2) The GPL is struck down. Ooops, Castle don't have permission from the Linux kernel developers to use their code. Thus, they can't distribute RISC OS 5.

 is a RISC OS UserEl Capitano on 8/2/03 7:33PM
[ Reply | Permalink | Report ]

Pace are the only ones legally entitled to released the RISC OS source to the public in any form.

I understand that by having no evidence of Castle/Pace accepting the GPL, it's a copyright matter. Is this a correct assumption? Who exactly owns the copyright? Are the Linux kernel developers a set group of people? (I'm not being sarcastic in any way, I actually have no idea) -- Simon Wilson, Boulder, Colorado

 is a RISC OS Userksattic on 8/2/03 7:44PM
[ Reply | Permalink | Report ]

There is no need at all to agree to the GPL. However, you cannot distribute GPL code without conforming to the terms of the GPL. This is why nobody has dared to take on the GPL in court, since it's less restrictive than normal copyright law.

Either Pace or Castle have taken GPL code and distributed code in violation of the GPL (I have seen the evidence and it's pretty damning). Now, Castle might not have been aware of the violation, but they're in a pickle, since they are still in violation. They do ultimately only have two choices, the ones I listed above. This could kill Castle if they can't replace the code pretty quickly.

 is a RISC OS UserEl Capitano on 8/2/03 7:53PM
[ Reply | Permalink | Report ]

If they've used GPL code then it should be removed.

If they have used GPL code they should request that (FSF/GNU ?) time to replace the offending code.

I am not overly happy at the thought of having ANY Linux code in RISC OS, and if anything the happenings today reinforce it. Think about it if RISC OS is concidered "good" by anyone they'll say "Oh it uses Linux...". The fact the code alleged to be have been taken is only in RO5.XX will be missed (and what about earlier RISC OS versions which CERTAINLY had no Linux code).

By the way WHY was an investigation carried out into RISC OS 5.XX in the first case (the date on the readme there suggests the "research" was started on 28th November (not long after the launch). "Oh I just got up one morning and decided to see if Castle breeched the GPL guidelines" just sounds far too weird....

Additionally the ROM would have to have been disassembled is that not (in itself) a breech of the RISC OS license ?

I am sick and tired of this continual "sniping". If that shower of Linux heads torpidoe RISC OS and smash our platform I for one will switch to Windows (you don't know how much it hurts me to say that).

Forget RPC's, Kinetics, Omegas and the whole shebang and nobody better bring Linux within a hundred miles of me....

-- Annraoi McShane,

 is a RISC OS UserAMS on 8/2/03 8:24PM
[ Reply | Permalink | Report ]

AMS, let me get this straight; either Castle or Pace stole GPL code, and you're pissed off the Linux community are annoyed at this?

Why should the copyright owners give Castle any good-will? Castle have taken from them and given nothing back. In fact, when they were contacted about this, they simply stripped the function signatures from the offending code and carried on as normal.

Castle are firmly in the wrong here. They should be the ones apologising and bending over backwards to make up for their wrong-doing.

I know I certainly wouldn't expect any sympathy if I were pirating RISC OS.

 is a RISC OS UserEl Capitano on 8/2/03 8:53PM
[ Reply | Permalink | Report ]

Why did I start looking at the ROM ? Because I wanted to see what fun and interesting things were present in it. As it was, I found some fun and interesting things and became increasingly surprised as the evidence continued to point to the same thing. That's why I went to such extreme lengths to be really sure of my facts. The text I have provided states with a degree of both waffle and detail how the sections in question were discovered, and the actions taken to be sure that what I was seeing was not just a coincidence.

John Balance emailed me this morning, stating emphatically that there was no linux source within their kernel. My investigation has convinced me that the kernel does indeed contain sections which are identical in form and use to those used within the linux kernel.

I encourage anyone to examine the archive and decide for themselves whether my conclusions are correct.

 is a RISC OS UserGerph on 8/2/03 9:10PM
[ Reply | Permalink | Report ]

So you're accusing Castle of outright lying? That's a very serious allegation, and if they are lying I'm surprised they even answered. It's looking undeniable that they do have some explaining to do, though.

Going back to the various bits of the GPL quoted on here, by me, and then ksattic, they seem to be contradictory, one saying the whole doesn't need to be GPLed and another saying it does.

 is a RISC OS UserSimonC on 8/2/03 9:55PM
[ Reply | Permalink | Report ]

Wow! This is a find. OK, so I'm a bit late. A few things from reading the comments: a)The code MUS have been put in by Castle. Why would pace want a PCI card in a set top box??? b)If Castle is telling the truth and they have nothing to hide, then would it not be in their best interests to compare the source codes in confidence with a witness present. Surely that would finalise this topic.

I hope Castle haven't been playing around with GPL. GPL has to be one of the best things to come to the computer industry and to mess about with it is not a good thing. On the other hand, Iyonix is probably one of the best thinsg to come to the RISC OS industry, and to loose it would be a great shame, and put us back to waiting for the Omega.

On final point: If Castle are 'guilty', do all machines and copies of RISC OS 5.xx need to be recalled?

 is a RISC OS UserSmiler on 8/2/03 10:20PM
[ Reply | Permalink | Report ]

Did I use the word lying ?

If you want to say that, then that's your choice. I said that my investigation indicates the converse of what John has said repeatedly, and I'm waiting (again) to find out what they have to say. If they have something to say which convinces me that their claim is correct and that my investigations were somehow flawed then it will be clear to all what the situation is.

I would not have contacted the copyright holders if I did not believe that something was awry. Russell would not have posted to linux kernel list if he did not feel there was some substance to my beliefs.

I suggest that if you have problems understanding the GPL you consult GPL itself, or at the least the GPL FAQ.

 is a RISC OS UserGerph on 8/2/03 10:23PM
[ Reply | Permalink | Report ]

Try again - my reply seems to have vanished somewhere.

You think something they say isn't true, and publicly state so. It means you believe they are lying. You don't have to use that word to state it.

I would prefer an answer to my question to suggestions of my ignorance on the GPL. If I hadn't read it then I wouldn't have been able to ask the question.

 is a RISC OS UserSimonC on 8/2/03 11:06PM
[ Reply | Permalink | Report ]

PCI - my NetWinder has a PCI bus, and indeed it runs the specific PCI code in question (since it's running ARM Linux).

There's no PCI card expansion, but it's still useful - Video, Ethernet, IDE all sit on it. Whether or not Pace ever produced a STB with a PCI bus, I do not know, but they can be found in STBs by other manufacturers.

-- Chocky, drobe.co.uk

 is a RISC OS Usermrchocky on 8/2/03 11:07PM
[ Reply | Permalink | Report ]

Smiler: Castle have the ability to send out ROM updates to be flashed to the machine, so machines don't have to be recalled. Saying that, I doubt there's any law which would require consumers to give up their machines and it's yet to be proven that Castle have done anything wrong. If they want mine, they're going to have to prise it from my cold dead hands. ;o)

Castle will *probably* release a statement after the weekend. -- Simon Wilson, Boulder, Colorado

 is a RISC OS Userksattic on 8/2/03 11:48PM
[ Reply | Permalink | Report ]

Who wrote the disputed code and was GPL the only license it was ever released under?

That seems to me to be the only possible easy way out for Castle. If Justin raised his concerns with them at the end of November they've had time to investigate and to decide on a response (seems odd they didn't reply to you all the same Gerph...)

The removal of function names from 5.02 that people seem to find so significant may (just possibly) be coincidental - the rom image is getting full and in an attempt to squeeze it back under 4MB such things would probably be the first to go.

Anyway, fingers crossed that this gets sorted out somehow...if Castle are crippled by it it'll end up hurting the entire platform to some extent (ROL and RISCOS.org have already had mud chucked at them by crazed slashdotters...)

And remember that Castle themselves may be innocent (though legally they'd carry the can I guess).

Ick, ick, ick. :-(

 is a RISC OS Usersoutherner on 8/2/03 11:56PM
[ Reply | Permalink | Report ]

A lot of people contribute to the Linux kernel, and the list of contributers for the PCI code is probably quite considerable. While I don't know for sure, I'd say with around 99% certainty that the PCI code is only available under the GPL since the number of people who contributed to it would likely put a license change out of the scope of a reasonable possibility.

Now, the function names are significant, but they're not the only factors. Justin himself has demonstrated that it's not that difficult to alter the Linux PCI code to compile into something that's very similar to the code in that part of RISC OS 5. It's very, very unlikely that the RISC OS developers would choose the same function names and produce almost identical code by pure coincidence.

Now, Castle can save face, but it's going to be very difficult. It would appear they've known about the violation for some time, and yet haven't done anything about it. They look like they've covered it up and now John Balance has denied the existance of GPL code in RISC OS.

If there is GPL code in RISC OS, all it takes is one of the copyright holders to go to court to find out the truth. Even if Castle remove the code, they could still be taken to court for the violation in the first place, especially since they didn't stop shipping as soon as they were informed of the violation.

 is a RISC OS UserEl Capitano on 9/2/03 12:46AM
[ Reply | Permalink | Report ]

IF they are guitly (and I admit the evidence looks like they are, although I want to hear what they have to say about the allegations before making my mind up - just denying them isn't good enough at this stage) then what effect could it have on Castle?

Obviously they would have to stop selling RO5 immediately, but would they end up with crippling penalties? Surely that wouldn't be in anyone's interest.

 is a RISC OS UserSimonC on 9/2/03 1:10AM
[ Reply | Permalink | Report ]

Justin poking around in the RISC OS 5 ROM and then finding something 'suspicious' wouldn't in any way be connected to the fact that Iyonix and RISC OS 5 means RISCOS Ltd are effectively dead and buried, would it? Sour grapes and all that? It's a pretty sad way to spend your time if you have no reason to do it... -- Michael Stubbs, Leeds

 is a RISC OS Userarenaman on 9/2/03 3:53AM
[ Reply | Permalink | Report ]

Oh I'm sorry, I must be mistaken; I could have sworn that ROL are still producing OS 4 updates. As long as ROL keep producing updates for my RiscPC, I'll keep buying them. You can't run OS 5 on a RiscPC afterall.

Chris, drobe.co.uk

 is a RISC OS Userdiomus on 9/2/03 4:13AM
[ Reply | Permalink | Report ]

Would it be possbile that Castle did write the code themselves and it just happens to compile in a similar way to the Linux source, or is that just to coincidental? I would have thought Castle would have stopped as soon as they realsied they were wrong. They a probably still selling RO5 as a protest to those who say it is Linux code when they could have written it themselves.

 is a RISC OS UserSmiler on 9/2/03 8:15AM
[ Reply | Permalink | Report ]

What is all the fuzz about?

How many other closed source products use linux- or BSD code in some modified form (many network stacks etc.) Even the playstation 2 uses a lot of linux inside but nobody complains at Sony.

It's just because it's to easy to look inside a RSIC OS rom that someone discovered this while peeking inside other OSes is usually much more dificult.

Secondly isn't it coincidental that just when there's a new RISCOS machine that "might" pose a commercial threat to other dominant players in the field some forces try (again) to stop these machines from appearing.

M$ has been proven to stole from god-knows-how but just when it's a tiny UK company again building some cool stuff then... easy targets...

I wonder if those slashdotters will have the same attidute at AmigaOS 5 when that appears...

Offcourse nobody is interested in AmigaOS 5 while RISCOS still is quite innovative (I just can't remember which part but...) Geee... perhaps this is a major plot to finally get the sources from RISC OS...

At least we can say is that Castle with its Iyonix PC is BIG news :-)

Isn't this waht we've always wanted ;-)

 is a RISC OS UserEPDM on 9/2/03 9:04AM
[ Reply | Permalink | Report ]

Smiler, it's very unlikely such an event would occur. Think about writing a calculator program from scratch. How likely is it to be identical to the calculator that's bundled with RISC OS?

There's no point in being API compatible with Linux, seeing as RISC OS and Linux are very different systems. Thus, there's no good reason to keep the same function names. Choosing the same function names and using the same naming convention is very unlikely, but not outside the realm of possibility.

The real evidence comes from Justin's changes to the Linux code. Within a very short space of time, he'd made some small changes to the Linux code to make it almost identical to the RISC OS code. The odds of that being coincidence are very small indeed.

 is a RISC OS UserEl Capitano on 9/2/03 12:29PM
[ Reply | Permalink | Report ]

EPDM, you are wrong on so many counts, I'm almost inclined to believe you're just trolling.

First, let's tackle the BSD issue. People are free to use BSD in closed source products, no problem. The BSD license pretty much gives you the code, but puts the author out of any responsiblity for it.

As for looking inside other OSes, people discovered Be were violating the GPL, along with a number of non-OS things.

Sony are not violating the GPL. The PS2 does not ship with any Linux code. You can get Linux for the PS2, but Sony will also give you the source to any GPL components, including their kernel changes.

I won't go into how competitive the new RISC OS machines, but I cannot see many serious developers reverting to an OS that has no memory protection, a poor library collection and co-operative multi-tasking.

MS haven't been proven to have stolen any GPL code. They'd be nuts to even contemplate it considering how many people are watching them. They stole from people like Stac, and they were made to pay for it. Not enough, mind you.

As for AmigaOS 5, can you show it has any GPL code in it?

And what part of RISC OS is still innovative? Name one major change that hasn't been done before in the past 6 or 7 years. I highly doubt anyone outside the RISC OS community would want the source. Most of it is ARM assembler, and most has been done better before.

I just can't believe people are defending Castle over this. Would you think it was OK if I now started giving out copies of RISC OS to anyone who asked?

 is a RISC OS UserEl Capitano on 9/2/03 12:38PM
[ Reply | Permalink | Report ]

El Capitano : "Would you think it was OK if I now started giving out copies of RISC OS to anyone who asked?"

Only if it was OS 5 ;-p

 is a RISC OS UserNodoid on 9/2/03 1:33PM
[ Reply | Permalink | Report ]

As Justin pointed out earlier, JB of Castle has confirmed by email that RISC OS 5 does not contain GPL code.

This means one of several possibilities:

(a). Justin misinterpreted what was seen in the OS to show that it contained Linux driver code when it didn't (e.g., it was a mere co-incidence).

(b). That Castle's developers used the function calls DESCRIBED by the Linux documentation - but used their OWN code to create the functionality (so the headers looked the same, but the code within the functions weren't).

And if Linux heads want to complain about that bear in mind that that is what WINE does in Linux when implementing Windows ;)

(c). Both Linux PCI GPL code and RO5 GPL code share a common ancestor which was available to both.

Things look not so bad now, time to think about ordering that Iyonix and formatting the PC drive to remove Linux I think....

-- Annraoi McShane,

 is a RISC OS UserAMS on 9/2/03 3:02PM
[ Reply | Permalink | Report ]

Annraoi, the problem with that is that Justin has provided a demonstration of how to completely re-create that code using 2.5.x PCI code. That would be pretty freaky to be due to chance, no? The headers were where he found indicators that something may be amiss; the similarities in compiled code appears to be spooky... -- Andrew Hill,

 is a RISC OS Usermd0u80c9 on 9/2/03 3:09PM
[ Reply | Permalink | Report ]

AMS: " Things look not so bad now, time to think about ordering that Iyonix and formatting the PC drive to remove Linux I think.... "

Dear me, you'd don't happen to post on /. a lot do you? What a completely naff reaction. Just because there is a (strong, having read Justin's report) suspucion that Castle have used GPL code, you're going to remove linux? Who are you aiming this action at? Who are you trying to get at? If there if there is a violation then the copyright holders are within their right to persue it. Even if there is not, they should look into it, and Castle should provide evidence that everything is cool, and people can go back to business as usual.

Kneejerk reactions like that are the kind of crap you get from a /. troll. Angry kiddies without a grip on reality.

-- Dougal

 is a RISC OS UserDougal on 9/2/03 3:22PM
[ Reply | Permalink | Report ]

Too late my copy of RedHat 7.1 is now history. And it's not petulance on my part, I wouldn't want to be ever accused of using any GPL code I would have thought. If it ain't on my machine I can't be influenced by it in any code I write can I ?

As to Castle, yes if they've used ANY Linux code it MUST be removed and due appologies made - with an undertaking NOT to use any GPL code in future (or if so only within the terms of the GPL license). I personally would recommend they keep well clear of the stuff.

That raises the question of will GNU seek punative damages (I figure any amount of serious money would put Castle out of business wouldn't you ?).

Now what of the reverse case, what if Castle show they've NOT used any GPL code ? Will it be just ah well that's ok no harm done (?). Is that really fair do you think ?

Sincerest Regards

-- Annraoi McShane,

 is a RISC OS UserAMS on 9/2/03 3:40PM
[ Reply | Permalink | Report ]

Hi Andrew,

I am not saying that there is no reason to investigate this. If for no other reason that to "clear the air".

If Castle HAVE purloined/stolen/adapted GPL code contrary to GPL they should remove it and replace it by their own code.

They should swiftly appologise to the GNU community, and promise NOT to do this again.

On the other hand if (inspite of the similarities) it is provable that Castle DIDN'T copy the GPL code then what ? What would YOU suggest ?

-- Annraoi McShane,

 is a RISC OS UserAMS on 9/2/03 3:44PM
[ Reply | Permalink | Report ]

Enough. This is usenet "quality" slagging off. Take it there, Drobe doesn't need it, nor the ill thought out commentary.

Do something useful, instead of using this as an excuse to waste loads of time complaing as RISC OS users are so often keen to do.

-- Chocky, drobe.co.uk

 is a RISC OS Usermrchocky on 9/2/03 4:05PM
[ Reply | Permalink | Report ]

I take it that's directed at me.

If that's the case please accept my sincerest appologies.

-- Annraoi McShane,

 is a RISC OS UserAMS on 9/2/03 4:12PM
[ Reply | Permalink | Report ]

Enough? Hmm. Stifle a good debate cos you wuv CTL so much and have too many sets of rose-tinted goggles? Poor.

 is a RISC OS Usertwattle on 9/2/03 4:21PM
[ Reply | Permalink | Report ]

It's directed at everyone, including Ian.

-- Chocky, drobe.co.uk

 is a RISC OS Usermrchocky on 9/2/03 4:39PM
[ Reply | Permalink | Report ]

I agree with Mr. Chocky - there is far too much uninformed debate taking place here. We haven't even heard what Castle has to say on the subject! I cannot see how this wild speculation can possibly present the community im a positive light - it will only serve to frighten away any new users who might have been enticed by the generally positive press coverage surrounding the launch of Iyonix.


 is a RISC OS UserNeilWB on 9/2/03 8:27PM
[ Reply | Permalink | Report ]

The amount of speculation on Slashdot was shocking. People were posting email addresses (riscos.com, riscos.org, castle.uk.co) and blame was being thrown about without any basis. -- Simon Wilson, Boulder, Colorado

 is a RISC OS Userksattic on 9/2/03 8:41PM
[ Reply | Permalink | Report ]

With an open mind to freeware software I can understand that stealing someone's code is a bit cheeky, but it's not like Castle have nicked the whole damned OS.

Personally, I think the likes of Justin Fletcher and Russel King are simply being pathetic. ARM Linux just comes across as a joke - has the thing been updated (except for the Kernel :)) in the past two years? I definately think that these guys have got far too much time on their hands - how about writing something useful instead of trying to take down a GOOD THING lads?

Castle has finally brought RISC OS into the 21st century and it's sad that anyone (especially a major contributor to the RISC OS scene) should make such a fuss over such a trivial thing. It's not like practically every other OS has PCI and I/O capabilities anyway. Do

I used to support Linux, but having suffered with various distributions since SuSE 7.2 and RedHat 8.0 (I think) I can gladly state that the whole thing is just a big joke. The GUI's available are rubbish, as is the majority of the software on offer (which most people would agree to if they've installed the stuff that comes with the distro discs)

To be frank RISC OS is a far superior product, if not technically at least in terms of its usability, supporting software (especially the amount of good quality freeware) and open-minded development community.

I don't think this issue will evolve into anything more than a slaggin match between the Lunatix nerds and the decent minded folk at Castle. It certainly won't stop me buying an Iyonix at any rate (even if it costs me 10,000 to get a second hand one) :)

Signing off - oops, dropped me pen :)

 is a RISC OS Useriamnotamused on 9/2/03 9:09PM
[ Reply | Permalink | Report ]

Just a quick question:

Has anyone bothered to look at the sources for the BSD PCI? Could it be that these are the same as used in Linux?

If that was the case - Castle would be using the same code as Linux but legally based on what people have been saying about BSD. That would also mean that Linux people have been "stealing" BSD code and putting it under a GPL!

Of course this is pure speculation, I haven't bothered checking this out myself and I may be completely off the mark here.

-- Gunnlaugur Jonsson,

 is a RISC OS UserGulli on 9/2/03 9:18PM
[ Reply | Permalink | Report ]

[quote]Has anyone bothered to look at the sources for the BSD PCI? Could it be that these are the same as used in Linux?[/quote]. Yes. It's not. Has anybody actually managed to come up with any refutation of the evidence other than "coincidence"?

 is a RISC OS UserEl Capitano on 9/2/03 9:35PM
[ Reply | Permalink | Report ]

Two points raised by iamnotamused:

"I can understand that stealing someone's code is a bit cheeky, but it's not like Castle have nicked the whole damned OS."

A small amount of theft is still theft. *If* it is the case the Castle took GPLd code and broke the license they're in the wrong.

"I definately think that these guys have got far too much time on their hands - how about writing something useful instead of trying to take down a GOOD THING lads?"

Burying your head in the sand is not a good defence.

From Gulli:

"Has anyone bothered to look at the sources for the BSD PCI?"

In Justin's article he points outhe checked the BSD PCI code and it's completely different.

-- Dougal

 is a RISC OS UserDougal on 9/2/03 9:37PM
[ Reply | Permalink | Report ]

[quote]but it's not like Castle have nicked the whole damned OS.[/quote] Whether it's the whole OS or just sections, it's wrong. If Castle have done it, I can't see how anyone could possibly condone their behaviour and then claim to be against piracy. Of course, if Castle can show their innocence, then I apologise for my accusations. However, so far nobody has managed to refute the evidence.

 is a RISC OS UserEl Capitano on 9/2/03 9:37PM
[ Reply | Permalink | Report ]

Argh, sorry about the funny quote tags, I thought I was using a different forum engine :)

 is a RISC OS UserEl Capitano on 9/2/03 9:38PM
[ Reply | Permalink | Report ]

El Capitano: snap ;)

-- Dougal

 is a RISC OS UserDougal on 9/2/03 9:39PM
[ Reply | Permalink | Report ]

I'd like to know how RISC OS would be affected by the GPL license. Would it:

a) Mean the whole OD would need to be GPLed or the code be removed.

b) Only the code for the PCI/IO interface be GPLed.

If (a) surely Castle only need examine and re-write the necessary parts of the code (which is already half done as interfacing with RISC OS must already have been specifically written)

If (b) can't they just release the code as a module? Even if it's free, no-one without OS5 is going to download it unless they want to reverse engineer it - or they're a loon.

Sorry for being so dumb, but it's late and I'm not a techie bloke - just a plain old BASIC/ARM programmer and I have no interest in developing software for Linux (seeing as Linux developers can't be arsed to write decent GUIs and software to develop decent software with - sorry, just being inflamitory for a change)

 is a RISC OS Useriamnotamused on 9/2/03 10:34PM
[ Reply | Permalink | Report ]

iamnotamused: You have no idea what I've been doing for the past two years and are in no position to comment on whether it is worthwhile or not.

 is a RISC OS UserGerph on 9/2/03 10:35PM
[ Reply | Permalink | Report ]

Sorry about the comments earlier guys. I recognise the hard work you have contributed to the RISC OS/Linux community of the years, but I do not fully understand your reasons as to your decision to publish your finding at such an important possible turning point in the future of RISC OS's history.

I just can't understand why you have to kick up such a fuss over such a trivial misdemeanor. I love RISC OS as much as anyone (I've been using BBC/Acorn computers for nearly 20 years) and the advent of the Iyonix, offering capabilities that finally bring the OS into the 21st century can only strenghten support for one of the few remaining examples of British home computer technology.

I guess the last couple of days have been enough stress for everyone, especially you two and Castle. I guess we can only wait and see what happens now - hope everything works out in the end.

 is a RISC OS Useriamnotamused on 9/2/03 11:39PM
[ Reply | Permalink | Report ]

Gerph still got toooo much time on his hands though :P (and just in case this gets taken seriously... :P )

However, much as we love you Gerph, we love RISC OS more - so this damn well better work out nicely!

 is a RISC OS User[mentat] on 10/2/03 12:17AM
[ Reply | Permalink | Report ]

"I just can't understand why you have to kick up such a fuss over such a trivial misdemeanor." But the people who wrote the code in the first place don't see it as trivial. These people worked hard on that code and placed it under the GPL because they wanted to keep it free (in the GNU sense).

 is a RISC OS UserEl Capitano on 10/2/03 1:21AM
[ Reply | Permalink | Report ]

iamnotamused: No stress for me. I went through that when trying to decide what to do with the information I believed I had uncovered. I don't feel that (if it is the case) stealing the code from others is a small misdemeanour. Having been at the receiving end of plagiarism in the past, I have been exceptionally careful to attribute those who require it, and to not copy things when am prohibited from doing so. I was so utterly astounded with what I had found that I had a number of individuals examine what I had seen to check that I wasn't actually completely nuts. I, too, could not believe that they would do such a thing. This is why I have been - what I would say was - very thorough in corroborating what I believed I was seeing and going through the recommended process. Castle had ample opportunity to provide an explanation both before and after the copyright holders were informed and the events of the weekend might have been avoided.

If you think that you should not report something that is wrong, then... well, that's where I fall down on understanding. If you want to blame me for reporting what I feel is a serious issue, then you're quite welcome to; it's not like I mind. Just lay off the hate mail, peoeple :-)

In any case, it may all be water under the bridge tomorrow (well, later today), and it's up to Russell and the other copyright holders what they wish to do about the whole issue in any case.

 is a RISC OS UserGerph on 10/2/03 3:38AM
[ Reply | Permalink | Report ]

Did Castle actually say they were going to release a statement today, or was that just rumour? I really think they need to say something. -- Gavin Smith, Carrickfergus

 is a RISC OS UserSparkY on 10/2/03 4:01PM
[ Reply | Permalink | Report ]

Iyonix users were told to expect something within a couple of days after the weekend. I think it's best if we sit tight until then.

Chris, drobe.co.uk

 is a RISC OS Userdiomus on 10/2/03 4:12PM
[ Reply | Permalink | Report ]

Good idea. Even I've managed to post some silly things about this, despite not knowing much about it, like a fool.

 is a RISC OS Usermoss on 10/2/03 4:32PM
[ Reply | Permalink | Report ]

I asked: "Has anyone bothered to look at the sources for the BSD PCI?"

And Dougal (and others) replied: "In Justin's article he points outhe checked the BSD PCI code and it's completely different."

Thanks for sharing that - haven't had time to read Justin's article and wanted to make sure someone was checking. Let's hope someone somewhere gets off his behind and tells people what's going on with RISC OS.

 is a RISC OS UserGulli on 10/2/03 4:36PM
[ Reply | Permalink | Report ]

There's a mini-FAQ that sums up the present situation in the article just above this one.

Chris, drobe.co.uk

 is a RISC OS Userdiomus on 10/2/03 5:06PM
[ Reply | Permalink | Report ]

Oops, Iconbar beat Drobe to Castle's response! ;) -- Gavin Smith, Carrickfergus

 is a RISC OS UserSparkY on 10/2/03 7:32PM
[ Reply | Permalink | Report ]

Sigh. So discard.

Chris, drobe.co.uk

 is a RISC OS Userdiomus on 11/2/03 1:26AM
[ Reply | Permalink | Report ]

Please login before posting a comment. Use the form on the right to do so or create a free account.

Search the archives

Today's featured article

  • RISCDomain magazine reviewed
    A media watch special
     9 comments, latest by druck on 30/10/07 8:55AM. Published: 20 Oct 2007

  • Random article

  • !Stego update: Hide confidential files in spritefiles

     1 comment, latest by mavhc on 16/2/03 9:38PM. Published: 22 Sep 2000

  • Useful links

    News and media:

    Top developers:
    RISCOS LtdRISC OS OpenMW SoftwareR-CompAdvantage SixVirtualAcorn

    CJE MicrosAPDLCastlea4X-AmpleLiquid SiliconWebmonster


    RISCOS.org.ukRISCOS.orgRISCOS.infoFilebaseChris Why's Acorn/RISC OS collectionNetSurf

    Non-RISC OS:
    The RegisterThe InquirerApple InsiderBBC NewsSky NewsGoogle Newsxkcddiodesign

    © 1999-2009 The Drobe Team. Some rights reserved, click here for more information
    Powered by MiniDrobeCMS, based on J4U | Statistics
    "We accept Drobe likes to be [controversial], no problem there - but a sinister pattern has appeared over the past year or so"
    Page generated in 0.5221 seconds.