Could open source RISC OS bring back users? | RISC OS news and information from Drobe


Beta! \| About us \| Contact \| Submit news \| RSS \| Twitter	Webspace \| Tech docs \| Downloads \| BBC Micro \| Gallery \| Wallpaper

Could open source RISC OS bring back users?

Published: 10th Dec 2006, 12:17:19.

The jury deliberates

RISC OS should become open source to lure back ex-users put off by the cost of Iyonix and A9home hardware, according to an opinion piece published last week. Ex-RISC OS user Mike Reed said in an article titled 'What would it take for me to consider RISC OS again' that the platform's workstation hardware was "very expensive." He backed efforts by RISC OS Open to reveal the RISC OS 5 source code, which he argued will encourage people to port the operating system to more widely available and therefore cheaper ARM-powered hardware.

The article, published on the mainstream IT news site OSNews, adds to the growing belief that RISC OS can only move forward by opening up its source code, so that more people can work on the operating system.

Mike wrote: "My hope would be that at some point in the near future, in becomes possible to build a complete version of RISC OS from the sources.

"If this were to happen, it would open up the possibility of either community-based or commercial groups moving the OS to other hardware platforms. ARM powered motherboards, PDAs, set top boxes and even phones are ubiquitous and many of these could, in theory, serve as a platform for running RISC OS.

"Other groups having access to the source code of RISC OS means that further development of the platform can be focused around software development - in this case, porting the OS - rather than, usually more expensive, hardware development."

The article also charts the OS development split between RISCOS Ltd and Castle, the state of third party software development and other issues. Some of the more technical points are arguable - for example, RISC OS does have pre-emptive multitasking thanks to the Taskwindow and Wimp2, although the operating system layer and kernel are strictly single-threaded. Also, RISC OS does have memory protection for applications, and one application cannot trash the workspace of another directly. It is possible for an app to elevate its privileges using SWI OS_EnterOS and find other mapped-out programs, or abuse SWI Wimp_TransferBlock to ruin another application.

Mike told us: "The article itself is bound to be a bit controversial as I am not going to pull any punches, and I intend to be quite critical about the current state of the platform."

Links
What would it take for me to consider RISC OS again from OSNews
Mike's website - with posts about how he wrote his OSNews articles

Discussion

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

Very intresting article, though isnt RISC OS being used on settop boxes already?

MikeCarter on 10/12/06 12:35PM [ Reply | Permalink | Report ]

Excellent thought, to promote discussion and awareness in the wider "potential new people" community.

I agree with the open-sourcing sentiment in the main, but am aware that "TiVO-isation" and related system vendor lock-in of their own IP is an important part of the market which RISC OS cannot ignore.

If there is one thing missing from the article, it is the thought that VirtualAcorn and other emulators provide a route to commodity hardware for the developer desktop, and form an important piece of the jigsaw.

steelpillow on 10/12/06 12:35PM [ Reply | Permalink | Report ]

I've no idea whether open source (ROOL) or a fully commercial OS (ROL) is the way to go. I'll support whichever one looks like it's going forward and progressing the platform: it strikes me that the OS model is far less important than the competence and vision of the people behind it. At the moment I'm feeling very gloomy about ROL's ability to deliver, and a little frustrated at the delay to the Castle/ROOL initiative. But I don't have a personal beef against either: if they can produce the goods (and fairly soon), I'll support them. Both should take a leaf out of the Peter Naulls/Martin Wuerthner/Colin Granville book of software development: if you keep your promises, people will pay; if you don't, they won't.

lym on 10/12/06 12:51PM [ Reply | Permalink | Report ]

I am glad that you highlighted some of the erroneous statements in the article. Although it is often trumpeted as one of RISC OS's deficiencies, it does have memory protection, and this situation is improving all the time (especially in RISC OS 6).

I believe that many users see the supposed lack of memory protection as something upon which all other ills can be blamed. Many other things can go wrong, such as single-tasking programs that have claimed vectors being terminated abruptly and failing to release them. Addressing some of these other problems would be more productive than carping on about memory protection all the time.

thesnark on 10/12/06 1:14PM [ Reply | Permalink | Report ]

What would bring me back is not the open-sourcing, but some actual signs of worthwhile development. (Which of course, the shared-sourcing of it may bring.) From a user's point of view, there isn't a lot in RISC OS Select that's all that useful for me, nor in RISC OS 5, which has even less new. It still looks as dated as it did in 1995, and still has problems with randomly crashing from time to time. It's a shame :-/

The hideous price of most RISC OS kit is pretty off-putting, too. Especially when designs are 4 years old.

rjek on 10/12/06 2:27PM [ Reply | Permalink | Report ]

In reply to lym > I've no idea whether open source (ROOL) or a fully commercial OS (ROL) is the way to go.

I'd say, both can work: - A commercial OS is ok, if development takes place at a resonable speed enhancing the OS - really enhancing it... and for a price to match that (pity this didn't happen in the recent past for one of the OS forks) - An open source OS is ok, if development takes place at a resonable speed enhancing the OS - really enhancing it... now time has to show if there are enough users out there who will work on the open source RISC OS

hzn on 10/12/06 7:06PM [ Reply | Permalink | Report ]

I bought a BBCB way way back in the early 80's and have stuck with Acorn/Risc OS every since. However, on the 20th December my pay packet is huge thanks to my multinational company becoming private and thus all my stock being purchased for double what it was 6 months ago (never mind the hundreds of free shares every employee was given when we split off from Motorola!). East Kilbride shops won't know what's hit them this christmas.

So in the new year what do I do:

A. Purchase a new RISC OS machine? If so which one the Iyonix/RO5 variant or the A9/RO6 variant. Neither of which do the basics that even a mobile phone costing a tenth does these days! Two OS's....dumb.

B. A PC......Ok no way, let's forget that one!!!!!!!!!!!!!!!!!

C. An apple. A company whose OS is clearly moving on and whose applications work with todays internet content.

RISC OS doesn't even do the basics and it costs a fortune for that lack of functionality. The idiotic bitching that goes on now is a sure sign of a dying platform. It's all a real shame since RISC OS has such huge potential and could knock the socks of M$ and Apple if the resources were allocated correctly with a clearly defined SINGLE development path.

Any "open source" work on RO5 will inevitably duplicate some of what has already been done on RO6. Likewise any work on RO6 will inevitably duplicate some of what has already been done on RO5. Any new work on either will have to be duplicated on the other for apps to continue to function or be stuck back where the OS split. Dumb dumb and dumber. This split has existed for years with no sign of been fixed meanwhile the world moves on with phones, PDa's, PC's, TV's and Games consoles all being multimedia rich and RISC OS incapable of showing a simple 10 second internet video clip!

mripley on 10/12/06 8:54PM [ Reply | Permalink | Report ]

To be fair, all new functionality of RO5 could be incorporated into RO6, when RO5 is shared sourced.

JGZimmerle on 10/12/06 9:53PM [ Reply | Permalink | Report ]

JGZimmerle: "To be fair, all new functionality of RO5 could be incorporated into RO6, when RO5 is shared sourced."

Are you sure the licences are even compatible?

guestx on 10/12/06 10:11PM [ Reply | Permalink | Report ]

"Could be". This won't happen. It is clear from the comments in these forums that RISC OS is split into two camps. Those who think the future is Castle/RO5 and those who think the future is RO6. The vitriol expressed is, as I've said before, indicative of a dying platform. These two camps WILL NOT merge unless the language (and thus attitude) expressed NOW differs from that expressed over the last 4 years.

It's time to wake up. There are too many suicidal self interested very vocal groups at work and I have all but given up. I will keep an eye on proceedings until the new year. But I do not hold much hope which is so sad.

I have tried to avoid pointing my opinionated finger. But at this stage I will. Castle: who clearly attempted to corner the RISC OS market and failed miserably having failed to understand that it is entirely different from the PC market.

mripley on 10/12/06 11:38PM [ Reply | Permalink | Report ]

I am clearly an idiot. I waste thousands of pounds on computers which the rest of the world would never touch for the price, but I don't care. Why do I bother: 1) I love them 2) They're unique. Sure, I'm really hanging out for a proper web browser and media player, and yes there isn't enough pro audio stuff or video editing stuff to keep me solely working on RO, but then that's what the Wintel/Linux box is for.

To get things going, bitching won't help, putting your money where your mouth is will. It is all well and good to point to where problems have occured but in the end this is not a democracy, nor is it a FOSS community. In the end it is just another capitalist market - the companies are here to make money. Unless you can program (which I can't), or have enough money to buy the rights to the OS (which I don't) then the only way to make sure things don't fall over is to keep buying as you can afford to keep the platform afloat.

I enjoy computing on RISC OS. And for that it is worth my money.

timephoenix on 11/12/06 9:24AM [ Reply | Permalink | Report ]

mripley>"I have tried to avoid pointing my opinionated finger. But at this stage I will. Castle: who clearly attempted to corner the RISC OS market and failed miserably having failed to understand that it is entirely different from the PC market."

Rubbish, and rubbish at a number of levels.

First Castle bought the IP of RISC OS from it's owners (Pace), you know the crowd whose (c) symbol appeared on things like RISC OS Ltd's RO4.0 ROMs that I have in my RPC at home. That surely gives them some rights surely?

Second the reason RISC OS (IMHO) is in difficulties is because the OS and Hardware were *not* owned by one vendor early enough after Acorn's departure. This meant ROL couldn't do things unless hardware vendors "did" something first (e.g., USB). By all rights ROL should have "defined" the interface - they didn't and left it to hardware vendors - hence the *two* USB implementations. If (for example) Castle had had both Hardware and OS at the time we'd have ONE OS and ONE USB stack. The "split" you complained about would *never* have happened.

And what's wrong with cornering the market - it certainly would have avoided your "bete noir" an OS split. More Iyonixs would have sold - we might even have had more OS development of RO5.

Thirdly there is *no* difference between the RISC OS market and the PC one. You sell, make money and pay for developments; if you *don't* there ARE no developments and the thing stagnates.

To do all the things you decried earlier (like run browsers and streaming video and the like) you usually have to *license* expensive IP from big players. You want Java you pay Sun (AFAIK Acorn paid 500K for the earlier version), you want WMA you *pay* Microsoft, you want to display DVD you pay the CSS licensing outfit (30K USD per annum), and MPEG LA some dosh for MP3 and MPEG-2. All that costs money. Additionally you have to develope code to implement these features (that also costs money). As ROL don't have any income other than from selling software to a few people - and Castle sell to other outside customers *as well as* hardware to *us* then they'd be a better bet.

If they (Castle) are now Open Sourcing the OS that's also a good pragmatic move. If *we* want the platform to survive the *we* have to get behind this - and not point fingers or engage in other equally unproductive pursuits.

AMS on 11/12/06 1:55PM [ Reply | Permalink | Report ]

I think it's unreasonable to have expected much better things from a strengthened (fortified?) Castle. Acorn were a much bigger company and were still totally unable to keep up through the further development of their own hardware platform and their own operating system, especially with a reliance on what are now considered to be embedded CPU technologies. There were approaches which could have lightened the burden: standardise on commodity hardware (which happened to an extent with the introduction of PCI slots); adopt other software technologies (even if ARM Linux hadn't been considered acceptable, there was always NetBSD and various other operating systems). The only benefit that I can see for Castle and company as they slog along on the old Acorn path is the proliferation of ARM-based systems, mostly for the embedded market, which means that they don't have to bankroll the whole hardware design process themselves if they don't want to.

guestx on 11/12/06 3:13PM [ Reply | Permalink | Report ]

This article is suggesting that the real problem with RISC OS (as far as keeping users is concerned) isn't the OS itself, but the hardware it runs on, so you've got to ask would open sourcing change that? Would hoards of programmers suddenly appear to take on such a mammoth task of porting to a different system, and presumably also include some sort of transparent emulation to keep the old stuff running? I rather doubt it.

SimonC on 11/12/06 6:19PM [ Reply | Permalink | Report ]

In reply to Simon C: Yes, I think open sourcing would change that.

It would reduce the cost of VirtualAcorn, and/or ROM images for the free emulators, down to a low enough level to attract users of cheap PC hardware into the RISC OS world.

The other RISC OS market, embedded systems, is not famous for high-priced hardware.

steelpillow on 11/12/06 7:05PM [ Reply | Permalink | Report ]

Sorry to follow one post with another, but I just found a relevant thought on the Xara forums (remember Xara?). By jedfrechette, about halfway down the page at: [link] Xara are also undergoing the opening-up process. The support they receive from the FOSS community is also underwhelming - the reasons why are interesting.

steelpillow on 11/12/06 7:36PM [ Reply | Permalink | Report ]

SimonC:

Open-sourcing wouldn't by itself bring the extra development we need, and I doubt it's the panacea some seem to think. But as well as the hardware issue, there are dozens of improvements to the OS which could be accelerated by using eager volunteers rather than relying on ROL or Castle to do the work. Unicode printing, improvements to the Window Manager, such things would be useful and well within the capabilities of some of the key third-party developers still around. But the key point seems to be this: Castle are no longer developing RISC OS and ROL seem incapable of releasing a product at all. So open-source is the only option left, whether or not it proves sufficient to get the platform back on track.

lym on 11/12/06 7:55PM [ Reply | Permalink | Report ]

Interesting Xara-related reading, there: do a half-hearted "opening" of your code, fail to get any interest because of licensing issues, see no extra developers wanting to do your legwork. If RISC OS "Open" fails to get interest by replicating this strategy, I'm sure we'll hear the complaints about indifferent open source developers. This is the "I told you so", posted up front, that I hope we won't need to revisit.

guestx on 11/12/06 9:46PM [ Reply | Permalink | Report ]

GuestX>"I think it's unreasonable to have expected much better things from a strengthened (fortified?) Castle."

I think it would have avoided issues like 2 USB stacks and the current "two" OSes. Having the OS and Hardware development separate meant that ROL had its hands tied behind it's back. Castle (and other hardware developers) had to "slot" their requirements into what ROL had scheduled at any given moment in time. In short as a mechanism for developing the RISC OS platform it was limited and restricting. Let's say "fleetness of foot" was not one of the descriptions I'd have attached to it.

GuestX>"Acorn were a much bigger company and were still totally unable to keep up through the further development of their own hardware platform and their own operating system, especially with a reliance on what are now considered to be embedded CPU technologies"

The chose not to - and that's a different issue. Take MicroDigital, Castle or Ad6 *all* have developed new hardware *after* Acorn. All have less resources than Acorn had. Acorn didn't complete a new machine after the StrongARM upgrade - how come? Again they chose not to.

lym>"Open-sourcing wouldn't by itself bring the extra development we need, and I doubt it's the panacea some seem to think."

I don't think anyone is saying that it is. But *something* has to be done. If it means more minds are applied to updating/upgrading or even just *documenting* what RISC OS does and how it does it then that is a good result. My only concern with open sourcing is that some people can get a tad "religious" about it - and indeed you'll see that in animated comments about porting RISC OS to non-native platforms - their interest is *open source* and NOT RISC OS. Then there are others who are keen open source enthusiasts who port software *to* RISC OS - the contrast is quite marked. Which, do you think, benefits the platform more ?

AMS on 12/12/06 8:31PM [ Reply | Permalink | Report ]

AMS:

Did you read the rest of my post after the first sentence? I support the open-source initiative, much as you seem to.

lym on 12/12/06 8:52PM [ Reply | Permalink | Report ]

I did read it - and was agreeing with you and citing other considerations

AMS on 12/12/06 9:04PM [ Reply | Permalink | Report ]

In reply to AMS:

"Rubbish, and rubbish at a number of levels."

You are entitled to your opinion which is all your reply is.

You are entitled to interpret events incorrectly.

You write like an agitated Iyonix owner.

mripley on 15/12/06 12:57AM [ Reply | Permalink | Report ]

mripley:

Your memory of past events appears to be very suspect.

I remember Paul Middleton stating that he was not going to produce a 32bit OS, as nobody wanted one, when at the time a requirement for 32bit was one of the main topics of discussion.

If Castle had not done their own thing, we would probably still be 26bit only and dying very fast, if not already dead.

Stagnation was not an option, when 26bit addressed cpu production was expected to be finishing.

ajb on 15/12/06 6:22PM [ Reply | Permalink | Report ]

mripley> You started one paragraph in one of your earlier contributions with the words "I have tried to avoid pointing my opinionated finger. But at this stage I will..." so who exactly is just gushing forth with opinion me or you?

The ROMS in my RO4 RPC *do* contain a copyright message composed of parts (c) Risc Os Ltd and (c) Pace. This very site carried the news that Castle had *bought* the head license from Pace ergo whatever *was* (c) Pace became (c) Castle.

They [Castle] also (as ajb points out) brought to completion the 32bitting of RISC OS and the Iyonix and *now* the Open Sourcing of RISC OS which one can say is a positive thing.

Open sourcing means the source (with some strings attached) is available to *all*. We can all both contribute to and benefit from these developments.

Please treat it as a positive development rather than hurling brick bats at it.

mripley wrote>"You are entitled to your opinion which is all your reply is. "

Nope, I stated "First Castle bought the IP of RISC OS from it's owners (Pace)". This is *fact* and even was covered on *this* esteemed site". Unless you dispute that Pace owned the RISC OS IP (in which case what did Acorn/e-14 sell them, what did they sell Castle).

I stated "Second the reason RISC OS (IMHO) is in difficulties is because the OS and Hardware were not owned by one vendor early enough after Acorn's departure.". While that *is* an opinion I believe it to be true. Logically dividing RISC OS and Native Hardware in the way that was done did seem to stymie development. If you wish to argue otherwise feel free to do so and I'll consider such comments on their merits.

I stated "Thirdly there is no difference between the RISC OS market and the PC one. You sell, make money and pay for developments; if you *don't* there ARE no developments and the thing stagnates. "

Surely that's true. Do you think MS or Apple would be in as good shape if their sales were at RISC OS levels?

mripley wrote>"You write like an agitated Iyonix owner."

I think you left a "NA NA" off the end of that ;-)

AMS on 15/12/06 6:44PM [ Reply | Permalink | Report ]

ajb: "I remember Paul Middleton stating that he was not going to produce a 32bit OS, as nobody wanted one, when at the time a requirement for 32bit was one of the main topics of discussion."

Nobody wanted one? IIRC nobody wanted (or was able to) to pay for it.

scl4c0rn on 15/12/06 7:05PM [ Reply | Permalink | Report ]

scl4c0rn:

Could it be that Paul Middleton later priced himself out of the market, re. 32bit OS, or the terms offered to Castle were too onerous for them?

Castle must of had a reason to go to Pace.

ajb on 15/12/06 7:22PM [ Reply | Permalink | Report ]

ajb:

That's probably why there is no RO Select for the Iyonix (yet).

scl4c0rn on 15/12/06 9:48PM [ Reply | Permalink | Report ]

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

Create a new account
Forgot your password?

Search this website

This week's poll

Recent developments have left me feeling...

Assured ROS will appear on new hardware		55%
Pleased OS desktop features are being developed		10%
ROL and ROOL will eventually kiss and make up		5%
App developments are critical		10%
Dave Holden sleeps easy at night		19%

Discuss this | Archives