Risc OS (of course 32-Bit!) that must be :
------------------------------------------

1. even more stable (perhaps protected memory)
2. stay fast (as it is now)
3. small (well, not as complex as Windoze)

The OS must also supply:

- a lot! of drivers for printers, scanners, dvd, ...
- support for larger hard drives (>1 tera-byte)
(today you can even by ones with > 60 giga-byte and the future won't wait)
- USB

Imago as the hardware:
----------------------

- StrongARM 2 with 600 MHz (which could hopefully simply overclocked to > 800 MHz) as
main processor

- ARM 10 as the additional processor, which delivers the FP-power we will need
in the future


Software:
---------

What the RiscOS scene really needs are developers.
So we need a (optimizing!) C/C++ Compiler, an editor (like Zap) and free manuals.

Of course we need an internet-browser.

by,

xplora/crescent

I also sent RO Ltd an email asking about this, and they couldn't be bothered to reply. Not even a standard reply.
We all have our own ideas about what RiscOS needs, but whether it will get it or not remains to be seen.

RISCOS Ltd is not supposed to make profit as such, just enough to develop and market RISC OS.

If RISCOS Ltd go bust, what I can see happening is RISCOS Ltd disappearing and one of the manufacturers, probably Castle, taking on the development of the OS. There are many people who will not let anything take RISC OS down. I mean, even Acorn's withdrawal from the market and subsequent breakup hasn't even deterred them, so maybe we (myself included) should have more faith.

Perhaps if someone such as Castle took over RISCOS Ltd or took over development of RISC OS it would be beneficial as development could be subsidised by computer and peripheral sales (if needed) and the hardware and software could be developed together.

I can't see Castle lying back and saying "well, never mind" if RISCOS Ltd fails, they'll obtain licensing off Pace to develop themselves!

Perhaps over time we'll end up with one company again...

The OS should be developed by someone with no bias about who uses it. Hardware manufacturers who produce hardware "somewhat off the beaten track" so to speak should, however, pay towards any additional development that their hardware demands.

It's in the interest of Castle, Microdigital, RiscStation and Millipede that RiscOS does not die, but to do something about that may require an extrodinary degree of cooperation between these compeditors. It has been done in a different arena (Linux) maybe the same can happen on our beloved Acorn.

The other alternative is to set up a management committee that oversees the OS development by commercial companies and part time programmers. With the internet the coordination of such an effort may bear fruit. What do you think ?

You say one company is too Microsofty for your liking, well the sotuation now is more Microsofty with one OS company and many manufacturers!

I agree with you about keeping RISC OS alive. What the manufacturers need to do is agree on standards so that all machines are 100% compatible. Also, I maintain that they should decide on a brand (Acorn is my personal choice but that's another argument) and manufacturer under it so you would have, for example, the Acorn RiscPC from castle and the Acorn RiscStation Lite from RiscStation. RISC OS should take on the job of promoting the platform, with everyone contributing to a pool of money for the task. This would give some unity to a fragmented market and provide an image to market to the no-RISC OS market. There would also be more money to spend on one campaign.

Developing like Linux is a good idea, but maybe because we're on propriety hardware (ARM chips etc etc) this would pose a problem when it comes to pushing the RISC OS cause to the non-believers!

It's a shame that RISCOS Ltd aren't interested in replying to customers. I have written, emailed and phoned, each with no reply. If they haven't got/can't afford the staff to deal with the public, they shouldn't plaster their contact details all over adverts in Acorn User. If they are staffed, someone has a bad attitude. Some of these ideas are worth them considering (the subjectof my getting in touch with them), yet they seem closed to communication in all forms.

However, I'm sure the intricacies of business will make that Impossible.

However, I do think (and this wouldn't be impossible, just unlikely) that it would do no harm for a representative of all companies (ROL, RiscStation, Microdigital, Castle etc) to all sit down and plan for the future and market growth, because if it doesn't grow there won't be anything for RS, M and C to compete for. Hmm....

As for the actual OS (Since that was what this thread was about) - a couple more years of //significant// technical development, a few vital additions and some half decent effort to push the stuff and everything could be rosy. Ah. Well.

I agree with Paolo about the way things are heading, but I think 64 bit is going to have a hard time, especially with home users, who have all that 32 bit software. Besides, who needs 64 bit? Mind the hype!

What about home networking, which seems to me as though it is starting to become viable. RISC OS might have the upper hand here, as RISC OS appears in devices such as Bush's Internet TV and Pace products, it must surely be easier to network with RISC OS rather than another OS. Who knows....

I have just got a nice new RiscPC and I have to say, it wipes the floor with my PIII 500Mhz. Speedy, efficient, blah blah blah. I did crash it a few times but that was down to some very old apps and a dodgy CD-ROM. Besides, it only took about 6 seconds to restart, so it was a breath of fresh air compared to Windoze. I think I will sell the PC and build up my RiscPC. You have to take risks and hold principles or what hope is there for the platform? With ARMLinux, it makes the investment safer because I can run lots of apps on that OS too, all on a superior machine. Money well spent.

Free major publicity. Soon. Let's have a scandal or something ;-) Yeah baby.

Who want's to pretend to go out with Geri Halliwell?

Yes, major publicity, and someone who knows how to milk it and give media friendly presentations. Anyone who's ever been to a RO show theatre will know what I mean there.

THIS IS ONE TING THAT CAN PUT RISC OS in the olympus of operating systems!

Now the market must be doing pretty d**n well if Castle can afford to be developing new hardware and consider TV advertising. Excellent!

They are going to be reminding schools just how much better Acorns are for eductation and hopefully when you open PC Pro there will be a huge Castle advert saying "Buy and Acorn, they're great".

Prediction: RISCOS Ltd will go bust and Castle will develop RISC OS alongside the new hardware they are working on. As it should be (just like Acorn). The customer service from RISCOS Ltd is a shambles and Paul Middleton is the worst of the lot. I, for one, will be glad to see someone responsible such as Castle take over the OS - a company who knows how to treat customers.

I would agree that the market is doing well if castle actually did some TV advertising.
Or maybe a billboard campaign
"Sick of your windoze PC? Throw it in the bin and buy something that works/is productive/won't make you want to kill !!!"

I Still think that Castle, RiscStation and Microdigital should at least co-ordinate a strategy on this front.

Good luck to them all, but it still smacks of the same old doomed-to-failure ideas that Acorn kept bringing upon themselves.

Why? Because Castle are just as poor as RiscStation are, and Acorn were, at giving small presentations at small RISC OS shows, so how on earth are they going to manage national marketing????

What about the Geri Halliwell strategy? :)8-<

ROR.

[Edited by 94 at 15:52, 4/8/2000]

I think you will find that Castle and co have learnt from Acorn's mistakes. Well castle certainly have, or else they would just take out a single ad in Acorn User saying "Erm, here's a new computer" and then sit back and make a loss.

Maybe TV advertising would eat up too much money for little gain. Billboards and ad mags and hassling schools is pretty d**n good going, especially for the Acorn world :-)

"Pestering" schools is a good move. Castle realise that schools remain the main Acorn market. Best to keep a hold on your existing core markets and then start the campaign for other markets. Well that's my opinion, anyhow. I'm no marketig expert!!!

I only hope you're right - pleasant surprises I can deal with :)

Maybe Castle have improved on Acorn, but I'm still skeptical (I wish I wasn't) because they still miss a lot of easy points.

Castle (and most RISC related companies) still suffer from a distinct lack of presentation ability. They've got something good, but can't seem to find the words.

They even have difficulty impressing a roomfull of people who are already sold on RISC OS.

An example of an "easy point" that I think they miss would be:

They claim to have a product that can read/write many standard filetypes - such as MS Office - but only because they happen to bundle some software with that capacility - and only then if the user is a) aware of it and b) knows how.

Some documentation about "how to use MS OFfice files" or a custom front end that allows users to drop Word or Excel or whatever files onto it and then loads the appropriate software would be easy for them to achieve.

I can think of plenty more, but this post is getting long already ;-)

ROR!

Yes this is the kernel of problem , i must be able to use all files that i have on PCs in a better way on Risc OS machines and so i will not have fear to lose Windoze... but we did have this program and we did put it in a dust bin... someone remember WinRisc????? this is the right answer to this problems and it will let us ALSO use PCI and AGP cards (for PC of course) on motherboards that have this busses....
:-)

As for winRISC, it was nice, and quite flash :) but I think that particular solution was taking too much away from the whole RISC OS ethos which is why I guess it ultimately failed commercially.

ROR. :)

I do remember Acorn emptying both barrels into both feet in the past on the advertising front.

The big problem is "blinkered" thinking, the approach (a la Acorn) was target education & enthusiasts then it became just enthusiasts and then they couldn't be bothered.

This successfully turned 1.5 million userbase into a few thousand who remained with the platform in spite of rather than because of Acorn.

On the other hand Microsoft never targeted schools and yet they captured the bulk of the business market and are now wiping the floor with Acorn's successors.

Adverts should make potential NEW users aware of your wares. New users will need some convincing (a TV campaign would be best for this as new users WON'T get exposure to the hardware on the high street). A further issue is the location of Acorn shows (do you really think Epsom is hard enough to get to, the Amundsen base in Antartica will be nice in October). What individual who has no experience of Acorns is going to train travel or drive to Epsom just to see an Acorn ?

Put the show somewhere easy to get to (hmm, London). It also gives you a larger audience.

As to Michael Stubbs comments about RISCOS Limited I think moving the O/S operation to Castle would be a mistake (remember RiscStation and Millipede are compeditors of Castles). You'd probably find that those outfits would feel somewhat under commercial threat (now is NOT the time to lose hardware vendors).

As to customer support issues if there are issues then RiscOS limited should take note of them - keeping an existing customer takes a lot less effort than finding a new one !

Adverts must be very well thought out.
I also agree that one manufacturer becoming the OS developer would destabilise things too much.

And whilst Epsom is a really nice venue, it's not ideal (what am I saying, it's useless) for attracting new users.

See "RISC marketing and fashion" thread.

--
ROR!

[Edited by 94 at 14:50, 7/8/2000]

But the fact remains that RISCOS Ltd are all talk and no engineers at the moment! Of course, it has been suggested by RISCOS Ltd that they might commision work (ie sub-contract it out), but have they commisioned work on the 32bit upgrade yet? Have they commisioned work on the hardware independence? They don't seem to be doing much at all, which is why all the rumours and discussions are flying around. I was simply thinking that at least things would be seen to move forwards if Castle or someone else had control. And at the moment, companies are sitting staring at their new hardware with no OS to run on it! For example, the Evolution from RiscStation and Cerilica's Nucleus and who knows what Castle have lurking in their labs!

I have a new RiscPC which is great, but I would feel a lot happier if RISCOS Ltd were boasting about how work on the OS was going extremely well! :-)

Maybe Castle etc will get round this lack of development by fiddling with the hardware. For example, could adjustments be made to bus speeds and the speed of the propriety chips that RISC OS relies on without adjustment of the OS? Couldn't necessary changes to RISC OS be issued as patches by the hardware manufacturers until RISCOS Ltd actually get round to developing the OS?

If so, RISCOS Ltd's lack of action won't matter in the shorter term, but I hope they get moving on the work needed.

[Edited by 160 at 15:10, 8/8/2000]

If RiscOS code is written well it won't be using software timing loops so should scale well irresepective of the speed of the hardware (the current RiscStations run an ARM7500FE at 56MHz rather than 24MHz without excessive re-writes of the OS).

The problem with modifying the old hardware is that the speed improvement will probably not be very dramatic (say at best 2 fold) whereas new hardware (such as Imago) will offer at least 4 fold with oodles to spare in bandwidth for video I/O as well.

The ARM already supports atomic read-modify-write cycles (for semiphore implementation) so (with a little OS giggery pokery) it should be possible to make the Imago multiprocessing.

As for taking the "old" SMP system of the pentium bear in mind that requires specific PCI chipsets from intel (and some of these didn't always work correctly in multiprocessor mode). Another problem would be that this improvement would make intels StrongArm not compatible with ARM Architecture V4 (and ARM10). I believe the best solution is a support chipset that enables the ARM to multiprocess without requiring excessive changes to the ARM itself (other than the addition of bus snooping that is !). Mind you if the chipset implemented a full speed off chip cache it could manage the bus snooping/memory synchronisation issues - what do you think ?

Strange they should choose Cambridge. Could this have anything to do with being near Pace for purposes of development?

I reckon Pace have completed a lot of work on RISC OS and it's advanced enough for RISCOS Ltd to enter the frame and start getting the work Pace has done onto Desktop RISC OS.

Regarding promotion, if you are at university or anywhere else you can put posters up, then email Castle Technology and ask for some free RiscPC posters and stick 'em up where people can see! I have my posters and am ready for action

[Edited by 160 at 00:11, 5/10/2000]