1. All machines are Acorn badged. Ideally, RiscStation and Castle etc merge their ventures under the Acorn banner and design and sell as one company. They need to buy the rights to the Aocnr name.

2. RISC OS Ltd comes to some arrangement whith Pace whereby the system is developed for both Pace's needs and the desktop market. After all, it would be to Pace's advantage when it takes off.

3. RISC OS fully 32 bit and work started on 64 bit.

3. Capability, maybe via emulator, to run current RISC OS programs.

4. Hardware improvements such as fast bus (at least 100Mhz), PCI and AGP slots, full USB implementation, all allowing us to use the PC equipment such as graphics cards and USB scanners, not to mention BT's ADSL services. The old podule slots should be retained.

5. High powered advertising in the national press and mags such as PC Pro. Maybe TV and radio. Who's going to buy this future brilliant machine if they don't know about it?

6. Develop the PC card. Make it match current PCs.

7. Do deals with Argo, PC World etc so they are widely available.

8. Get a bloody move on. We're sick of waiting!

Well, there's my opinion. I think we should stick with the Acorn name and I think the market needs to see some mergers and a less fragmented feel. We must succeed!!!

I'm going to make myself unpopular with the following comments, so be warned!

If they could make the OS run on PC's then that would open up RiscOS to the mainstream like never before.
This would solve the problem of our current obsolete hardware. We would not be having to pin our hopes on the SA 2 which even at the stated 600mhz is never going to compete with the 1ghz (and faster) PC chips. And it still doesn't exist yet!
The hardware would be up to date and half the price of current systems.
I like PC's. They are powerfull, Upgradeable and reliable.
Their only drawback is that most of them run Windows.

Also, with so many more people using RiscOS (and they would be), RiscOS software writers would be sure they made there software work with it. It really would be worth their while!

I would really like to see this happen and firmly believe this is the only way out of the deepening hole RiscOS is in.
However, since we know what has happened to all the programmers at RO Ltd, it would take a large miracle (and lots of cash) to make it happen.


[Edited by 78 at 00:38, 25/06/2000]

Once RiscOS has the user base, then it would be worth trying to introduce a new platform.

I downloaded BeOS today (as have 9000 other people from that one site) and gave that a try.
It would be nice to do that with RiscOS

While marketing RISC OS machines under a single brand may prove useful I (and others on the newsgroups) are not convinced at how useful the Acorn brand is. People outside of the RISC OS market remember Acorn for the BBC Micro and branding new machines like this may prove to be a hindrance.

It shouldn't be "After all, it would be to Pace's advantage when it takes off." but rather "..IF it takes off." This might sound negative but RISC OS Ltd will have to convince Pace that it could happen (and try to make it sound that it also has a fair chance of happening) otherwise Pace have no reason to spend resources on developing RISC OS as a desktop OS. Despite what people claim, you don't need a RISC OS box to develop for RISC OS.

32bit I agree with. 64bit is currently only relevant if you port RISC OS to other processors (as suggested by Mad Nurse) which would require a complete rewrite in a portable language. Given ARM's market I rather suspect that 64bit processors may be someway off.

Emulator to run current apps. Well obviously backwards compatibility is nice but if we're going to do this and we're going to port to other processors wouldn't it be easier to just write an emulator and stick a RISC OS style user interface on something like linux?

Hardware improvements yes - old podule slots no. New machines really shouldn't be tied to the old system, unless you can keep them without compromising new them which is practically impossible. I've said on another thread that replacing hardware can be relatively cheap and someone buying a new machine would be aware of the need to do this anyway.

Advertising - yes but we need cash and a decent campaign.

Develop PC card to match top end machines. Tough one. It would be incredibly difficult to keep the card up to date and ultimately it would always be bottlenecked by the bus if placed on, say, a PCI card. My personal belief is that the PC card is aimed at a different market to top end PCs. It's there for people who need the odd bit of Windows but are not power users and don't want two machines. However improving it's spec is always good. Then again if RISC OS is ported to other processors why not run Windows natively?

7 + 8 - yep ok.

Mad Nurse's comments are interesting and they're certainly not unpopular with me. ARM processors will never be cutting edge and given that most people write in high level languages these days the underlying processor is pretty irrelevant. However as I said before - is this not just linux + RISC OS emulator + RISC OS GUI?

Mind you - the x86 must be getting towards the end of it's life otherwise Intel would not be dropping backwards compatibility in the Itanium....

All of this is my opinion, and I maintain the right to be wrong.

As regard hardware\development being too expensive, it would seem Millipede have already gone ahead and developed a motherboard that will solve all the motherboard/peripheral problems in one fell swoop. No doubt Spacetech etc will get down to writing drivers, as always.

On-board rev. T StrongARM with processor upgrade connector
Up to 512 MB SDRAM in SO-DIMMs running at 100 MHz (minimum)
16 MB flash memory

Very high resolution video - 2048 x 1536 @ 72 Hz @ 24 bits
Enhanced SoundBlaster compatible sound processing
Game port with joystick and MIDI I/O
Built-in 100 Base-TX Ethernet

Dual EIDE port to support up to four devices
Dual serial ports (RS232 and RS422/RS485) up to 460 kb/s
Dual USB port
PS/2 mouse and keyboard interface

Audio/video expansion bus with four real-time video streams
Standard podule bus with enhanced data rate
Additional decoded podule connector
Standard processor expansion connector for PC card

As regard processor speeds, ARM and RISC OS are so much more efficient that, like Apples, they can easily beat PCs on speed even though they have lower clock speeds.

I think Imago is going to solve many of our problems.

It's up to RISC OS Ltd to come to an arrangement with Pace regarding the OS and it's development to become fully 32bit and recognise USB and DVD etc. and up to all companies to devise a campaign.

Maybe manufacturers should put some money into RISC OS Ltd to allow them to hire some engineers. It would be a worth-while investment.

:-)

(i). RISC OS emulated on a PC (even at 1GHz) will always run slower than RiscOS on a 233MHz SA Rev T. Do you really want to spend a lot of money so you can run RiscOS at half speed ?

(ii). You would need to load and run Windows first, otherwise the developers would have to code a complete HAL (Hardware Abstraction Layer) for the PC - not a trivial task. This limits you to running RiscOS (a relatively stable OS) on a relatively unstable platform (Windows). Not a desireable state of play I think. People might assume crashes are down to RiscOS and NOT due to Windows (RiscOS might get tarred with the same brush). Also the boot up time would be that of Windows (quite long) PLUS that for RiscOS.

(iii). PCs, contrary to what Mad Nurse said, are NOT always reliable. They can (and do) go wrong. And PC chipsets can cause problems (the i820 for example can have random crashes and cause reboots - as mention in the computer press). I have seen NT systems where you can't install things, Win98's you can't shut down (resulting in scandisks on restarts) and some serious flakeness when dealing with add on hardware.

(iv). An SA-2 at 600MHz will compete with a 1GHz Pentium or Athlon running Integer tasks (which is mostly what processors do). The cache on a 1GHz Athlon is only run at 1/3 the clock speed (less than that of the 600MHz SA-2) and the efficiencies of RiscOS should not be underestimated. My old A3010 (clocked 6-12MHz) manages to scroll through 4000 numbers in the same time as a 166MHz PC (running BASIC on both). When comparing PCs with Acorns you are comparing different architectures so a simple xMHZ/yMHZ type metric is meaningless. Remember the SA-RPC is running with a 16MHz memory bus and only 32K of cache and still manages to be compeditive with 300MHz plus PC which have 100MHz memory and 512K of cache. If the slow memory bus on the RPC is removed and more cache added (or a new approach such as that on the millipede) the Acorn platform will I feel outperform PCs up to and including 1GHz.

(v). Why limit ourselves to PCI, as a standard it has proved problematic (some versions cause data corruptions to disk for example). The 33/66MHz limit is largely one imposed by Intel, the internal bus on a Coppermine Pentium does 133MHz and the AMD Athlon does 400 yet the all get slowed to 33 or 66MHz to suit PCI. I am sure the talented designers at Simtec/Riscstation and Castle could come up with an agreed NON PCI bus that would offer greater speed and reliability.

As to the programmers (or lack thereof) at ROL I would suggest that ROL enlist the help of independant developers and set up a development management structure that allows RiscOS to be development to continue but this development farmed out to independant developers and skilled enthusiasts.

[Edited by 144 at 19:01, 25/06/2000]

[Edited by 144 at 19:02, 25/06/2000]

I think if you want to be an alternative to Windows to which the disaffected can gravitate you should use compeditive NON PC hardware running RiscOS on the processor for which it was written - namely the ARM. Many people dislike (with good cause) PC Hardware and Windows Software, give the computer buying public a real alternative and support it by advertisment and the hardware/software WILL SELL !

The task of re-writing the kernel for a PC would be an extrodinary difficult task as no expertise in PC hardware level software development exists in the RiscOS world (other than maybe Aleph One that is). It would be a very expensive task and I feel the money would be better spent on upgrading the RiscOS hardware spec to support faster memory (DDR RAM), fast wide 128 bit busses (like Millipede's Imago can) and possibly a 64 bit wide asynch bus for add on devices. To that add multithreading to the OS and a true 32 bit model and bob's your uncle (isn't he ?)

The software development for RiscOS I think will ultimately fall to the independant developer community and ROL should act as a "guiding hand" in this - that is assuming they don't wish (or can't) to it themselves.

* Emulation *
First, I don't believe that Emulation of RiscOS is the answer.
Archie already exists. It emulates an Arm2 and requires a Mammouth PC to do it. Perhaps it can be optimised a bit, but it would still need something like a 2ghz PC to emulate a StrongArm chip.
RiscOS must be offered as an Alternative OS to PC users not just an Emulator.
Obviously, RiscOS on PC's would eliminate the need for the PC card.

* Millipede *
Millipede have indeed gone ahead a made their own board, but it does NOT solve our problems.
As I understand it, the SA233 can only utilise memory at a maxium of 66mhz. This in itself is obsolete by PC standards.
I don't see that I help the situation with Peripherals either.
The Imago board is expensive (very) and will always remain so.
It also only fits in that d**n RPC case which is rapidly losing favour with me.

Yes, the case may have been ground breaking with it's advanced all plastic and modular design, but the fact is that I and many other people would rather have a board that fit's in a PC case.
RPC motherboards aren't large, they could easily fit in a MicroATX case, which would give users the choice of a slim desktop case or a Mammouth tower with CD, CD writer, DVD, LS120, Zip drive, 2 hard drives, Old Uncle Tom Cobbly and all!

* Annraoi *
I take your point about PC's not always being reliable.
This will be the case if you have lots of companies making lots of different parts.
However, I maintain that if you choose the parts carefully then there are many configurations of PC that will be just as reliable as an Acorn.
99% of the time, with carefully chosen parts, unreliability is caused by the shoddy OS.

* SA-2 *
A 600mhz SA-2 may compete with a 1ghz Athlon in integer performance. However, when it comes to Games which (like them or not) play a large part in the success of a platform, Floating point is the order of the day.
A 486DX100 will put our SA233 in it's place for this. Your average 300mhz AMD K6 will probably do the same for the SA-2 (I can't back that up, but I think it's a reasonable guess)
600mhz is also the fastest speed currently planned, I believe they will debut at far less.
What happens when in a years time the PC has 2ghz CPU's?
Arm chips aren't designed with the Desktop computer market in mind anymore.

* Factoid! * :-)
The new Thunderbird (which will rapidly replace the old Athlon) does not have the 1/3 speed Cache problem.

* PCI *
Yep, PCI does have it's drawbacks.
It also has it's advantages.
Internal hardware Modems £30
Soundcards
Network cards
SCSI (if you must)
Even the new Voodoo cards will use PCI (as well as AGP)

* Other OS's *
Dealing with Linux and BeOS.
I haven't used Linux, but I have looked into it a bit.
It is not user friendly.
It is bloated. Every bit as much as Windows.
BeOS, has an even smaller market, but is more user friendly (Apparently!)
I downloaded BeOS the other day, and am not impressed. I shall continue to look into it though.
Personally, I think RiscOS is Generally better than either of these.
Yet small though their users bases are, they are large compared to RiscOS.
Imagine how much smaller they would be if they ran on different hardware. Probably smaller than RiscOS if they even survived at all.

* Nearly finished! *
Finally, though it would be difficult and very expensive to produce RiscOS for PC hardware, it would be worth it.
The user base would increase dramatically.

Developing our own hardware will always be expensive to do and sell.
Let the PC hardware manufacturers to that. They have the money to do it.

These are only my thoughts on the matter and please excuse any gramatical errors, my brain seemed to work faster than I could type!
:-)

[Edited by 78 at 21:17, 25/06/2000]

Regarding games, the platform needs promoting as a WORK platform to start with. Games and associated hardware/OS development would follow as the market grew.

I am aware of repeating the point, but I think everyone needs to make a big go of our platform, OS and hardware, and come up with an advance as great as the RiscPC was on release adn the market the d**n thing! Apple didn't give in and port MacOS to the Wintel hardware architecture because it would have defeated the point of their whole system. It is the same for RISC OS.

Intel x86 chip archetecture could be regarded as just a twenty year series of bodges.

Plug and play is a major overhead on system startup. 486 machines start up much quicker than P3s. (If setup right).

Linux has 2 things going for it; it's stable and it's free. On the other hand it is about as user unfriendly as a pc os can get, installation is the worst around (for non programers) and it is absolutely atrotious for graphics and multimedia. Secondly, if you're merely going to whack on a front end GUI what possible connection does it have to the computers you are now using and interested in? You can build your own style of X-Windows if you really wanted to.

Let me just say this before the screaming hoardes decend, I work in a large animation house, and we get to play on some expensive stuff which we push to the limits. I've worked on Linux, Unix flavours in the form of Silicon Graphics and Sun Sparcstations, Macs and NT.
In addition I've been working with BeOS for some time and as far as I'm concerned it is the best OS in terms of multitasking, multithreading, multiprocessor usage (close to the same things really) and power in handling all sorts of graphics and multimedia I've ever seen.

But it's big, and bloated and takes ages to boot up.
Once you start taking things apart you lose your products identity,especially bad in a small community b/c different people like the computers for different things.

Oh I dunno, what do you reckon?

UNIX Airways

Everyone brings one piece of the plane along when they come
to the airport. They all go out on the runway and put the
plane together piece by piece, arguing non-stop about what
kind of plane they are supposed to be building.

Air DOS

Everybody pushes the airplane until it glides, then they jump
on and let the plane coast until it hits the ground again.
Then they push again, jump on again, and so on ...

Mac Airlines

All the stewards, captains, baggage handlers, and ticket agents
look and act exactly the same. Every time you ask questions
about details, you are gently but firmly told that you don't
need to know, don't want to know, and everything will be done
for you without your ever having to know, so just shut up.

Windows Air

The terminal is pretty and colorful, with friendly stewards,
easy baggage check and boarding, and a smooth take-off. After
about 10 minutes in the air, the plane explodes with no warning
whatsoever.

Windows NT Air

Just like Windows Air, but costs more, uses much bigger planes,
and takes out all the other aircraft within a 40-mile radius
when it explodes.

Linux Air

Disgruntled employees of all the other OS airlines decide to
start their own airline. They build the planes, ticket counters,
and pave the runways themselves. They charge a small fee to
cover the cost of printing the ticket, but you can also download
and print the ticket yourself. When you board the plane, you
are given a seat, four bolts, a wrench and a copy of the
seat-HOWTO.html. Once settled, the fully adjustable seat is
very comfortable, the plan leaves and arrives on time without
a single problem, the in-flight meal is wonderful. You try to
tell customers of the other airlines about the great trip, but
all they can say is, "You had to do what with the seat?"

How can I edit one of my replies?

I would just like to point out that although Castle pride themselves on not telling you what great things they are working on until they are ready, it might help if people knew fantastic things were under development. It could well keep people with the platform instead of giving them the impression nothing much was happening.

I am finding a growing number of my friends are sick of Windows. One is about to acquire a nice RiscPC! I still think this is a golden opportunity for the RISC OS platform to push itself forward, what with the 'backlash' against Microsoft.

Remember RiscPCs as they currently stand have lots of hardware limitations (slow I/O bus, slow memory/memory bus and a relatively small processor cache) and yet are compeditive at most desktop tasks when compared to the PC.

When hardware (like) Millipede do come to pass we will have a machine that can compete and may outperform PCs at significantly higher clock rates. I can still remember the benchmarks in PCW that showed a humble ARM2 (no cache, ultra slow RAM) outperform by a factor of 2 an IBM PS-2 Model 8 costing twice the price (which had a processor (386DX) clocked 4 times faster than the ARM and had a substantial off chip cache controller). I know that is ancient history but trust me the ARM is compeditive when it runs RISC OS and when you take off the hardware limitations I think that should silence a lot of the critics.

Give people an alternative both Hardware and Software and they will take it as Michael just showed.

I believe porting RiscOS to the PC is a waste of time and simply puts RiscOS on the same sort of dodgy ground that Windows has to work on and would probably suffer similar problems.

I must agree with Michael on this one there is no point in abandoning the hardware part of RiscOS because the Hardware and OS work together so well that would certainly be lost on a PC port.

We need a super machine and I believe Imago represents the only logical step so far as we know at this moment in time!

When RISCOS programs do crash or throw up errors, it would be useful for the error message to make it clear which program was at fault (this is almost never the case).

Not that we should be concentrating on errors with RISC OS (but my RiscPC has some timing/RAM/PCcard illnesses so I get them more than most :( ... )

My only qualms about the Evolution is the use of PCI which is one of the PC's weaknesses (if you look at the current PCW you'll see the problems ragarding the intel i820 chipset) why put these sort of problems onto our platform when we don't have to ?

Having said that it is nice to see Acorn users having a choice between three alternative hardware models (Evolution/Imago/RiscPC) I would argue that the Acorn user has MORE choice than the PC user in hardware terms - and that can't be a bad thing.

What's the point of hoping about something that the developer has declared as dead? AFAICT the only reason it's still on the RS web page is because it's not been updated for about a year!

With reference to the last para, it /is/ nice to have 3 different RISCOS compatable manufacturers to choose from, and that was all going nicely until ROL hit its current (apparent) wobbles.

So just to focus on something more positive, has anyone heard about either Riscstation's (colour) or Microdigital's (mono) portable projects in the last 2 months?

I'm greatly looking forward to seeing them both hit production... :) I Hope...

Anyone care to dash that hope (or send me into an excited frenzy? ;-)

Or any opinions along the lines of (trying desperately to stick to the original thread here) will these new portables, assuming they make it, have any influence on "RISCOS going mainstream" ?

If the new portables are well specced and well priced, they can only help AS LONG AS THEy ARE MARKETED. I wouldn't mind betting you don't see a single advert for them outside of the RISC OS market.

Anyway, I'll be interested in them!

£2000 !!!!!
This board does also has drawbacks.
Apart from expense, it's still designed to fit in that bloody RiscPC case!
If the Mico Digital wasn't based around that god awfull 7500 processor (MP3's anyone?), I would probably buy one just to have a tower case. And four IDE devices. That work at a decent speed. And an internal modem.

Ah well, no harm in dreaming is there.

Get picking those numbers! :-)