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

Wakefield 2006 live news

By Chris Williams. Published: 13th May 2006, 06:29:35 | Permalink | Printable

Just like news on t'telly

Today's the day of the Wakefield show, and you can follow all the news over the course of the day from here. Show report to follow soon; photographers, please get in touch.

Live Wakefield show news

Previous: ArcEm ported to Amiga OS 4
Next: Acorn is a legendary brand says new Acorn

Discussion

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

common pics please!

 is a RISC OS Userhighlandcattle on 13/5/06 1:48PM
[ Reply | Permalink | Report ]

Thanks for the live news. "Qercus to consider dropping to 6 to 9 issues a year" ... AFAIK that is not a drop but an increase ;-)

 is a RISC OS Userhzn on 13/5/06 1:59PM
[ Reply | Permalink | Report ]

Nice to see castle are "evaluating new arm chips", its just a shame that arm chips are still relatively slow. I read back in 2002 plans for a 1Ghz Arm so not sure what happend to it? Hopefully we'll see some faster processors as demand for more powerful mobile phones increases.

 is a RISC OS Usermark1282 on 13/5/06 2:08PM
[ Reply | Permalink | Report ]

Why a faster one? The other option is to put several of them on the mainboard and tell RISC OS how to make use of them...

 is a RISC OS Userhzn on 13/5/06 3:24PM
[ Reply | Permalink | Report ]

Multiple processors won't make the Iyonix faster at processing images in PhotoDesk or encoding MP3s. Agreed, the RISC OS desktop is easily nippy enough on a 600MHz XScale, but some things simply require more oomph than even a 1GHz ARM (or several) would provide.

That said, it would be a very nice and potentially useful project. Anyone remember Hydra though? It's been tried, and the development required could be huge.

 is a RISC OS Usertamias on 13/5/06 4:00PM
[ Reply | Permalink | Report ]

"hzn"

"Why a faster one? The other option is to put several of them on the mainboard and tell RISC OS how to make use of them..."

I THINK NOT. Risc OS is cooperative multitasking so dual core's and Multi processors are a bit of a no no.

 is a RISC OS UserJwoody on 13/5/06 4:42PM
[ Reply | Permalink | Report ]

Have RISC OS look after the GUI and have Linux drive the processors ;-)

 is a RISC OS Usertweety on 13/5/06 5:06PM
[ Reply | Permalink | Report ]

In reply to Jwoody: I know that RISC OS currently is not into using more than one processor but that is something that can be changed. For starters you could put the odd driver onto a separate processor like the network card, USB, screen, harddisc etc. freeing the one for apps from all those background and interrupt driven things. OK that would then be an intelligent IO processor type setup but it would offer quite some processing power to the apps.

 is a RISC OS Userhzn on 13/5/06 5:28PM
[ Reply | Permalink | Report ]

Good idea - one step at a time :)

 is a RISC OS UserAW on 13/5/06 6:02PM
[ Reply | Permalink | Report ]

Woo, posting this from the train on the way back to London! Isn't tchnology wonderful?!? Using an ARM powered PDA of course. Apparently CJE sold out of all the A9s they had with them, which is good news. MartinW said that currently using Artworks' PDF export is the only way to get CMYK output on RISC OS but his next project, with Jon Tytgat, will be a Postscript level 3 printer driver. Then any app can will be able to do it. Martin will also be looking at porting the new version of GIMPprint when it comes out, although its release has been imminent for several years!

And big thanks to the Wakefield group for organising another excellent show :-)

 is a RISC OS Userhelpful on 13/5/06 6:32PM
[ Reply | Permalink | Report ]

"hzn"

"I know that RISC OS currently is not into using more than one processor but that is something that can be changed. For starters you could put the odd driver onto a separate processor like the network card, USB, screen, harddisc etc. freeing the one for apps from all those background and interrupt driven things. OK that would then be an intelligent IO processor type setup but it would offer quite some processing power to the apps."

Sorry NO CAN DO. you cannot have USB, harddisk using a separate processor under coopertive multi tasking. When Risc OS does an I/O it has to stop and cannot do anything in the background. Especially true when an application renders a font. The font may miss the cache and have to go off to do I/O.

 is a RISC OS UserJwoody on 13/5/06 6:35PM
[ Reply | Permalink | Report ]

I don't know if this would be possible but how about some kinda daughter board that could accept multiple processors but risc OS would simply see it as one processor. Kinda like a distributed computing system. If that is a crazy idea. How hard would it be to re-write portions of the Risc Os code to support multiple processors. If Microsoft can manage it surely RISC OS LTD can.

 is a RISC OS Usermark1282 on 13/5/06 7:48PM
[ Reply | Permalink | Report ]

Can i nominate hzn for funniest post of the week? Stultifying.

 is a RISC OS UserChimpy on 13/5/06 7:52PM
[ Reply | Permalink | Report ]

"mark1282"

I don't know if this would be possible but how about some kinda daughter board that could accept multiple processors but RISC OS would simply see it as one processor. Kinda like a distributed computing system. If that is a crazy idea. How hard would it be to re-write portions of the RISC OS code to support multiple processors. If Microsoft can manage it surely RISCOS Ltd can."

I think you have a case of DREAM ON STANLEY"

 is a RISC OS UserJwoody on 13/5/06 8:03PM
[ Reply | Permalink | Report ]

mark1282: I'm sure RISC OS Ltd. could do a very nice job of it, if they had anywhere near the same resources as Microsoft.

 is a RISC OS UserSimonC on 13/5/06 10:03PM
[ Reply | Permalink | Report ]

Yeah I suppose microsoft have a slight advantage with over 60,000 employees and billions of dollars. Shame they still can't manage to put together an operating system as reliable as Risc OS though. Speaking of risc os, anyone know what the new features in the select 4 are?

 is a RISC OS Usermark1282 on 13/5/06 10:24PM
[ Reply | Permalink | Report ]

The Iconbar have a photo of a motherboard in what looks like a portable case. It has a name "Mainstone handheld division" on it. Anyone know anything about it?

 is a RISC OS Userknutson on 13/5/06 11:12PM
[ Reply | Permalink | Report ]

knutson: it's a PDA/mobile phone development board with a little PXA XScale running NetBSD. It's designed to help engineers to produce drivers for the chipset; Gavan on the NetBSD stand has experience with this sort of thing and produced the NetBSDiyonix port. The Mainstone kit is about as portable as a park bench.

 is a RISC OS Userdiomus on 14/5/06 12:51AM
[ Reply | Permalink | Report ]

I noticed the photos of the Phoebe at the Show and am quite curious. I know that she was nearly completed at the time that Acorn was broken up - but what happened to her after that ? What were the barriers that prevented the buying of the rights to the design and then completing and releasing the finished product ? I can make some guesses of my own but it would be nice to hear from someone in the know ^_^

 is a RISC OS UserJohnB on 14/5/06 7:48AM
[ Reply | Permalink | Report ]

"mark1282"

"Nice to see castle are "evaluating new arm chips", its just a shame that arm chips are still relatively slow. I read back in 2002 plans for a 1Ghz Arm so not sure what happend to it? Hopefully we'll see some faster processors as demand for more powerful mobile phones increases."

What did Castle say? and what new chips ?

When PC's went through the 1Ghz mark, they had to add larger caches and speed up the front side bus, otherwise the extra Mhz was just a waste of time. Not sure mobile phones would want to jump through that loop unless they have to.

Also what are these demands for faster mobile processors? I thought mobile phones already did video and after all they are unlikely to increase screen resolution much so why do they need faster processors ? More memory I can see but faster processors I am not so sure.

Personally I would not hold my breath for faster ARM's. I am not convinced the mobile phone industry needs them and I think Intel will start pushing their own processors for PDA's

 is a RISC OS UserJwoody on 14/5/06 10:46AM
[ Reply | Permalink | Report ]

In reply to JWoody: "Sorry NO CAN DO. you cannot have USB, harddisk using a separate processor under coopertive multi tasking. When RISC OS does an I/O it has to stop and cannot do anything in the background. Especially true when an application renders a font. The font may miss the cache and have to go off to do I/O." This has absoultely nothing to do with cooperative multitasking! When an app reads data from a disc it asks the OS to fetch and deliver the data. How this is fetched, that is from some buffer in RAM, from the cache of the harddisc, from the harddisc, from some LAN disc doesn't matter to that app. While the system waits for that data to arrive in its buffers there is absolutely no reason to not do other things like handle network requests, run other apps, play music in the background etc. Or, simply think of the second processor which does the IO like as some intelligent, fast controller for the odd device. Do you assume that while some code inside RISC OS is waiting for some data to arrive from some external device that it just sits back and does nothing (except for handling interrupts). I dearly hope it does not work that way, though formatting a floppy disc does suggest that it does waste resources that way. Then that is a chance to speed things up.

In reply to Chimpy "Can I nominate hzn for funniest post of the week? Stultifying." Sure, if that makes you happy! But just think of modern PC systems with their chipsets which are used for special operations, or the idea of co-processors, or the idea to let the chip in the graphics card do some work, or to take an older more simple device, that is the DMA controller. All that can run in the background and when done deliver the results to the main processor. And that using DMA, which pretty much means to let some extra hardware do the odd word, Geminus does reach quite some speed increases.

 is a RISC OS Userhzn on 14/5/06 11:09AM
[ Reply | Permalink | Report ]

"hzn"

"Do you assume that while some code inside RISC OS is waiting for some data to arrive from some external device that it just sits back and does nothing (except for handling interrupts). I dearly hope it does not work that way, though formatting a floppy disc does suggest that it does waste resources that way. Then that is a chance to speed things up."

YES THATS EXACTLY WHAT IT DOES

When a system like Windows/Linux/Mac OS does some I/O it hands the processor over to some other task. It called proper Multi Tasking.

When Risc OS does I/O it just has to wait.

When Risc OS does copying a file from one location to another. It does offer to task switch but not while its actually doing any I/O i.e. It only offers to switch when the application does a wimp_poll. So at best what looks like "multitasking" is Risc OS is just a conjuring trick

As to changing things it would take a total rewrite from the ground up and I don't think Risc OS Ltd have even the resources to dream about such things.

SORRY TO SHATTER YOU ILLUSIONS

 is a RISC OS UserJwoody on 14/5/06 11:26AM
[ Reply | Permalink | Report ]

jwoody: What you describe is called pre-emptive multi-tasking, not proper multi-tasking. What RISC OS does is no more of a conjouring trick than that is, after all, neither of them are really running more than one thing at once (putting dual core processors aside for a minute). Whilst pre-emptive mult-tasking has its advantages over co-operative (and some disadvantages, although they are less apparent these days), it's wrong to say that it is the only proper way of doing things.

 is a RISC OS UserSimonC on 14/5/06 1:48PM
[ Reply | Permalink | Report ]

Best thing about the show (IMHO) was the generally upbeat note of many people there.

As John Cartmell said, "You don't need to feel you need to apologise for using RISC OS".

There was new hardware, new softwarew releaseases, and it was really good to see development restarting on several applications....

 is a RISC OS Usermarkee174 on 14/5/06 1:55PM
[ Reply | Permalink | Report ]

RISC OS does allow some disc operations to go on in the background using DMA. My TWAIN drivers for example given you have a SCSI card that supports DMA, that it is stuck in a DMA slot, that it works with your scanner in DMA mode and that you have enabled it in the scanner driver SCSIbits setting, reads one block of data whilst processing the previous one. Genuine parallel processing. Long ago Acorn issued a desktop floppy formatting program which I have a feeling used background disc operations. I think this was a reaction. It worked but didn't look pretty. Maybe it would do better on more modern machines. I have another feeling that someone set out to use DMA support for the printer port. Windows XP would you belive grinds to a halt if you print to the parallel port. That's because MS have not used interrupts (never mind DMA) to handle the parallel port. Instead XP sits there polling the printer port at 3GHz or whatever. Buy a USB printer is what Bill tells you. Multiple processors are the way to go, we have been here before though with the Simtec Hydra card - must be 10 years ago now. The RISC PC of course would let you insert two processors. Again more dreams, did someone work out how to install two ARM processors, like your old ARM 6 card as well as your new Strong ARM card.

 is a RISC OS UserDavidPilling on 14/5/06 2:21PM
[ Reply | Permalink | Report ]

"SimonC"

"What you describe is called pre-emptive multi-tasking, not proper multi-tasking. What RISC OS does is no more of a conjouring trick than that is"

NOT TRUE pre-emptive multi-tasking is when a process gets interruped at a time slice interval.

Straight multitasking is when a process switches tasks when it does I/O.

Cooperative multitasking where the machine waits ( apart from interupts ) is NOT VERY efficient. It takes millisecs to service an I/O and you can do a lot of processing in that time. Risc OS only gets away with it because users spend most of the time doing only one thing. Go to something like video conferencing or web serving and you want to run more than one thing at once and you are severely dissadvantged. For example if you tried to run a database server on Risc OS and did nore than about 70 I/O per second you would hit a brick wall. Not true on other systems.

It also means you cannot exploit dual cores or multiple processors unless you write the applications to do this themselves.

Sorry I disagree when Risc OS does a copy of a file from one place to another its nothing better than a conjouring trick. i.e. You can claim its multitasking but its not using the time when I/O are taking place to do useful work.

 is a RISC OS UserJwoody on 14/5/06 3:25PM
[ Reply | Permalink | Report ]

"David Pilling"

"Long ago Acorn issued a desktop floppy formatting program which I have a feeling used background disc operations"

Don't think so, my understanding was that it just did a wimp_poll between each track or I/O operation. so again gave the impression it was multi-tasking but was not doing anything useful while the I/O was taking place.

 is a RISC OS UserJwoody on 14/5/06 3:30PM
[ Reply | Permalink | Report ]

Jwoody: I wasn't making a point about whether one was better than the other. I know the basics of how each system works, even if I'm not good on the details. I just fail to see why you refer to co-operative multi-tasking as a conjouring trick, but not pre-emptive. They are both tricking the user into thinking the computer is really doing two or more things at once, instead of rapidly switching between them. Whether or not one method works better in a given situation doesn't make the other one any more or less of a trick.

 is a RISC OS UserSimonC on 14/5/06 4:13PM
[ Reply | Permalink | Report ]

JohnB wrote>"I know that she [Pheobe] was nearly completed at the time that Acorn was broken up - but what happened to her after that ? What were the barriers that prevented the buying of the rights to the design and then completing and releasing the finished product"

The workstation division, responsible for Phoebe, the Risc PC, STB's etc., was sold to Pace - AFAIK for 200K GBP by Element-14 (the ghost of Acorn's past). Much of phoebe was lost - in the sense that multiple VIDC's and the like never made it into a production machine. Some other elements (e.g., PCI) did. To my knowledge some prototypes were made (I even seen one on TV once - the flourescent yellow case is quite distinctive). I suspect as Pace had the rights to the design some elements of it made it into their STB's and some may have been reused in the Iyonix - which was developed by ex-Pace engineers (under the Tematic company name) for Castle.

Sadly a case of "what might have been..."

 is a RISC OS UserAMS on 14/5/06 4:32PM
[ Reply | Permalink | Report ]

"SimonC"

"I just fail to see why you refer to co-operative multi-tasking as a conjouring trick"

Because its not multitasking as a computer science person would understand things. Sure it enables people to do several things at the same time, but is wasteful and just makes it seem like its doing the right thing

If you went back to mainframes or servers then you want the processor to be doing things whilst I/O which takes an age in CPU terms is going on. Processors are the expensive part of the system, it pays to get them busy.

99.5% of the computer world use pre-emptive and mutitasking to do work whilst I/O is happening, because from a computer science point of view its the correct thing to do. Not waste expensive CPU cycles waiting for I/O.

If I run my PC it runs UD "Looking for a cure for cancer" Bittorrent, Skype. whilst I run some application. All of which work far superior on a true multiprocessing operating system. I could not contemplate running the same apps, assuming they were available under Risc OS.

 is a RISC OS UserJwoody on 14/5/06 4:45PM
[ Reply | Permalink | Report ]

What killed the pheobe?

Back in the day I used to make Acorns for a living.

The info that we got was that the IOMD2 chip didn't work properly.

& then Acorns workstation division got wound up :-(

I remeber a meeting with Acorn about pheobe - they pulled the case out of a box & it was eww what colour is that (done by the same people that did the 1st zip drive apparently...) & then Uh Oh BGAs on the board, not good - only test we had for them at the time was to bang the board on its side at the end of the line and see if the chip fell off!

 is a RISC OS UserROHC on 14/05/06 5:08PM
[ Reply | Permalink | Report ]

To Jwoody: "99.5% of the computer world use pre-emptive and mutitasking to do work whilst I/O is happening, because from a computer science point of view its the correct thing to do. Not waste expensive CPU cycles waiting for I/O."

Pre-Emptive Multitasking vs. Cooperative Multitasking has *nothing* to do with the question of the system having to wait for I/O, or not - and vice-versa!

Let us take the harddisc reading of several blocks for example. The app asks the OS to read the data; the os initiates that harddisc I/O and then gives the CPU to some other app since the harddisc needs some time. When the OS gets control back - be it due to taking control in pre-emptive multitasking or due to the current app passing control to the OS - then if the harddisc data is there it can pass that to the requesting app and thus give control to that one. That is something which can be done in RISC OS too.

 is a RISC OS Userhzn on 14/05/06 6:49PM
[ Reply | Permalink | Report ]

"hnz"

"Let us take the harddisc reading of several blocks for example. The app asks the OS to read the data; the os initiates that harddisc I/O and then gives the CPU to some other app since the harddisc needs some time. When the OS gets control back - be it due to taking control in pre-emptive multitasking or due to the current app passing control to the OS - then if the harddisc data is there it can pass that to the requesting app and thus give control to that one. That is something which can be done in RISC OS too."

hnz you need to invest in a Risc OS Progammer Reference Manual" and read it. You would then hopefully not make so many outlandish and incorrect statements about Risc OS.

With Cooperative Multitasking the only time the operating system passes control to another task is when a task reaches a statement in the program that is a SWI_Wimp_Poll. That is if an application does I/O it has to wait for that I/O to complete, the only think that happens in the back ground is interrupts. If it is doing multiple I/O's as David Pilling has stated it can overlap I/O's via DMA. If it does 10 I/O's then it is still going to take 9 x 17 msec per I/O. All time where the CPU is idle apart from Interrupts from timers etc.

When multitasking systems such as Windows/Linux/Mac OS do an I/O they task switch to some other application/process so that while the I/O is happening the CPU can be busy.

I AM SORRY TO SAY THAT YOU ARE WRONG AGAIN.

MULTITASKING versus CO_OPERATIVE HAS EVERY THING TO DO WITH WAITING FOR I/O OR NOT !!!!!!!!!!!!!!!!!!!!!

 is a RISC OS UserJwoody on 14/05/06 7:44PM
[ Reply | Permalink | Report ]

All those exclamation marks, you notice? Five? A sure sign of someone who wears his underpants on his head. Opera(ting systems) can do that to a man. ;-)

(Terry Pratchett, Maskerade)

 is a RISC OS UserStewy on 14/05/06 8:00PM
[ Reply | Permalink | Report ]

In reply to hzn:

Re. Quercus:

I agree that 6 to 9 issues published regularily are much better than a publishing lottery of few more issues irregularily. I hate to pay for promises, I like to pay for reliability and quality.

Re. Multi-processor RISC OS:

Sorry, Herbert. Currently I don't agree. First, it is essential that RISC OS running on 1-chip computers is developed further and that this RISC OS is bug-free and stable. So, in addition, it is important that there will be faster and better CPUs available. For example, you like to see VIDEO DVDs ? Faster IONYXes or faster PCs, that doesn 't matter. In my opinion the 1 GHz which was mentioned is very slow. We need very-low-voltage CPUs with 3 or more GHz, with much bigger caches, with FPA, better power management and so on.

RISC OS for multi-chip designs is currently a dream only. I don't believe that RISCOS Ltd is willing and financially able to develop such kind of RISC OS. But perhaps future versions of VirtualAdjustSA can use multi-core INTEL CPUs, running i.e. 2 RISC OS simultaneously on a Pentium Dual Core, or 4 RISC OS on planned Quadro or Tetra Core Pentiums. Or it may be possible to run 2 RISC OS on a Single Core Pentium with hyperthreading.

 is a RISC OS UserGollum on 14/05/06 8:55PM
[ Reply | Permalink | Report ]

RE: Qercus

John sells a number of editions so its in his interests to produce as many as possible. if something has to give, he should be congratulated at choosing to focus on fewer high quality copies rather than just churn them out monthly regardless.

 is a RISC OS Usermarkee174 on 14/05/06 9:17PM
[ Reply | Permalink | Report ]

Re JWoody: You simplify to much. In Risc OS, the main application gives control to an other application when it calls Wimp_Poll, correct. But, like you wrote, interupts can happen and guess what they are used for: I/O. You don't write to a serial port, you write to a serial port buffer. You don't read from a serial port, you read from a serial port buffer. If there's nothing there what do you do. You wait, but I give control to Wimp_Poll and check again when it returns.

The difference is, you write your programs in an other way. You don't use blocking calls.

Peter V.

 is a RISC OS UserPeter on 14/05/06 10:09PM
[ Reply | Permalink | Report ]

"Peter"

Serial port being handled by interrupts is one thing, but Disk I/O is not the same.

Take for example rendering fonts. The character may not be in the cache hence the need for disk I/o. The system cannot hand control over to some other task because the rules of the Wimp do not allow that so FileCore never issues a Wimp_Poll on your behalf. Nor can you do disk I/O by just reading or writing to a buffer

 is a RISC OS UserJwoody on 14/05/06 10:20PM
[ Reply | Permalink | Report ]

Re JWoody: What you see is that slow I/O (serial port/MIDI/DMX, Keyboard, Sound, network etc) uses buffers/interrupts. Faster processes (Disk/Screen) don't.

So what happens when a new font has to be cached, millions of cycles are waisted when the head of the harddisk seeks the right track (a small delay when printing to the screen). Under interrupt, the slow processes keep on running.

But who will finish the printing to the screen first? The pre-emptive system has to wait until the other interrupts are finished. The co_op system continues at ones (OK, I know, you can not see the difference it is late half a milion clock ticks but for time critical systems it can make a difference).

Peter V

 is a RISC OS UserPeter on 14/05/06 10:54PM
[ Reply | Permalink | Report ]

Can we split off all the stuff about CMT/PMT to somewhere else as its never ending and has already degraded in to the usual ill-informed slanging match, and get this topic back to what happened at Wakefield, as there is precious little of relevance.

 is a RISC OS Userdruck on 15/05/06 09:32AM
[ Reply | Permalink | Report ]

I was tempted to reply to jwoody's ill informed rantings before now but managed to hold back.

However, WHEN YOU RESORT TO THIS !!!!!!!!!!!! If you're correct in what you say, shouting about it won't make it any more valid. The problem is you're wrong. No let me re-phrase that. You're lack of imagination and programming expertise prevents you from understanding what is/is not possible under either CMT or PMT.

A multi processor CMT OS is perfectly possible and our favourite OS would require a re-write.

 is a RISC OS Usermripley on 15/05/06 1:50PM
[ Reply | Permalink | Report ]

In reply to JohnB:

AFAIK I believe that with the breakup of Acorn a number of Pheobe cases were sold/given away and some users have used them as new homes for their RISC PC.

 is a RISC OS UserCol1 on 15/05/06 1:52PM
[ Reply | Permalink | Report ]

Col1:

Roy Heslop found the first production batch and was selling them.

What was most interesting about that phoebe is that it's case hadn't been modified for the podule slots

 is a RISC OS UserROHC on 15/05/06 2:43PM
[ Reply | Permalink | Report ]

In reply to Woody

Hmmm, very good point. Your first couple of exclamation points seemed kind of derivative and orthodox, but your next three clearly made up for that, in a stunning tour de force. And just as the argument seemed to be dragging on for the next five or so exclamation points, you really hit home with your last four. And I don't think any reasonable person could argue after that compelling last one.

 is a RISC OS Userjeffd on 15/05/06 3:45PM
[ Reply | Permalink | Report ]

Just thinking back on the subject of multi proccessors in risc os, as i beleive this is the way forward. Technically the RISC PC could support multiprocessors, even though only one was a StrongARM. I run a distributed computing project called distributed.net (kinda like seti at home, using spare cpu cycles) and it can utilise the x86 and the ARM at the same time. Also I was reading on Wikipedia that the phoebe would support multiple ARM proccessors. If this is true doesn't it mean that mutli strongArm support would be quite possible today?

 is a RISC OS Usermark1282 on 17/05/06 02:20AM
[ Reply | Permalink | Report ]

"Mark"

"on the subject of multi proccessors in risc os, as I beleive this is the way forward."

Only works if applications are written specifically for Multiprocessing, current applications are not.

Risc OS cannot exploit multiprocessors without a rewrite of the operating system and or applications. The former would require a total rewrite and is unlikely to happen given the size of the Risc OS market

 is a RISC OS UserJwoody on 17/05/06 09:22AM
[ Reply | Permalink | Report ]

//Only works if applications are written specifically for Multiprocessing, current applications are not. //

Not a problem. Most apps don't need a lot of power anyway.

RISC OS could be rewritten over time to make use of extra processors itself, once the hardware was available (and, hopefully, a standard). Hang on... deja vu.

 is a RISC OS UserLoris on 17/05/06 1:30PM
[ Reply | Permalink | Report ]

Loris: Yes, it sounded a bit familiar to me as well - didn't we cover most of this about a month or two back? :)

 is a RISC OS UserGulli on 17/05/06 5:57PM
[ Reply | Permalink | Report ]

A lot of what goes on here seems to be re-treading the same discussion, be it the state of RISC OS, multprocessors, or whether or not emulators are bad.

If an application wants to make use of multiprocessors then it needs to be written to do so. If you've got two single-processor applications on a dual processor system then one can run on each, simultaneously (memory access etc. not being considered here).

This should not be taken to mean that I think it's easy / hard / impossible to rewrite RISC OS to do this, or whether or not existing RISC OS apps would work if there was such a version of the OS available.

 is a RISC OS UserSimonC on 17/05/06 7:27PM
[ 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

  • Software hosted by Drobe: Your guide
    A round-up of users' mini-websites on drobe.co.uk
     12 comments, latest by neilwhite on 17/11/07 9:57AM. Published: 5 Nov 2007

  • Random article

  • R-Comp offers RISC OS emulation solution
    Look, it's a PC in a small box
     64 comments, latest by dgs on 17/10/03 3:28PM. Published: 10 Oct 2003

  • Useful links

    News and media:
    IconbarMyRISCOSArcSiteRISCOScodeANSC.S.A.AnnounceArchiveQercusRiscWorldDrag'n'DropGAG-News

    Top developers:
    RISCOS LtdRISC OS OpenMW SoftwareR-CompAdvantage SixVirtualAcorn

    Dealers:
    CJE MicrosAPDLCastlea4X-AmpleLiquid SiliconWebmonster

    Usergroups:
    WROCCRONENKACCIRUGSASAUGROUGOLRONWUGMUGWAUGGAGRISCOS.be

    Useful:
    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
    "Drobe won't credit us for breaking stories. Oh well, that's journalism for you."
    Page generated in 0.3801 seconds.