The Icon Bar: RISCOS Ltd.'s new scheme

RISCOS Ltd.'s new scheme

RISCOS Ltd. have announced a new scheme to enable access to more RISC OS 4 updates. However, you won't be able to talk to them about it at the show, because they won't be there - which, having read the full press release, might be a good thing.

The official reason that RISCOS Ltd. decided not to attend this year is that they have "cut back on our presence at exhibitions, since many programmers are happier working in isolation, rather than trying to help users directly with their queries". So no access to the programmers - contractors rather than staffers, which "has been interpreted by some people as 'RISCOS Ltd has NO programmers' which is luckily not the case". Let's hope this does allow ROL to "make the best use of those peoples' skills".

The main thrust of the press release (which I've uploaded to our site, as it contains more information than the version on the ROL site), was the news of RISC OS Select, a subscription-based scheme where for an extra £100 pounds a year you get access to three update CDs - updated RISC OS 4 ROM images plus early access to beta stuff. Access to RISC OS 4 updates - RISC OS 4.5, which will continue to be 26bit - will also entitle you to documents on the website previously only available to registered developers.

And what about developers? The cost of being a registered developer is set to go up - £500 per six months. This entitles you to "direct contact with the RISC OS Development team"; access to documentation will, as previously mentioned, be available to all RISC OS Select members.

RISC OS 5, the 32bit version of the OS required for access to more modern processors, seems to have been put on the back burner. "[T]he amount of work needed to convert major areas of RISC OS such as the kernel and ADFS will take many man months of work to complete and until we get sufficient commitment from our customers, that work will not be undertaken.". So what of the future? "[I]t appears that the availability of 26 bit processors will continue for longer than was expected.". This doesn't, however, address the problem of getting access to processors faster than 200-odd Mhz that we have had to endure for so long.

The quest for continuous revenue streams seem a little Microsoft for my taste, and time will tell if it succeeds or not. But without any commitment to a 32bit RISC OS it looks like we won't be getting access to any new processors any time soon.

RISCOS Ltd.'s new scheme

This is a long thread. Click here to view the threaded list.

Andrew weston

Message #88564, posted at 09:09, 21/5/2001, in reply to message #88563

Unregistered user

So becuase of ROL's decision are we not going to see any speed increases? Are there no improvements being made in 26-bit chips?
The Kinetic was an attempt to overcome the bus restrictions which people had habitually complained about so I think that was a great achievement by Castle and continues the flagship computer's lifespan. It's good to see they have improved it further - once again the only speed improvement that is available it seems.

[ Log in to reply ]

Lee Johnston

Message #88565, posted at 10:45, 21/5/2001, in reply to message #88564

Unregistered user

Andrew - no one is designing chips with 26bit addressing anymore. The only usable chips are the ARM7500 variants and the SA - 26bit mode was dropped way back with ARM8 and even the ARM7T doesn't support it. The 26bit address mode has no place in ARMs and Intels plans and, TBH, quite rightly so.

Castle really are pushing the hardware as far as it will go. Overclocking to 287Mhz was supposed to be "dangerous" - I commented to Nathan on Saturday that Castle must've strapped an ice pack to the back of the 300Mhz models 8)

I seriously doubt we will see any further improvement - if it were possible I suspect Castle would've made a bigger jump this time.

Face it people, without an OS that can run in the 32bit address mode we've reached the end of the line as far as processors go. Without such an OS the only way to keep up is to bolt hardware around the processor (3D hardware, MPEG decoders etc) but this isn't particularly cheap and there will always be a limit at which the processor can feed / remove data from the hardware.

[ Log in to reply ]

Andrew Weston

Message #88566, posted at 12:30, 21/5/2001, in reply to message #88565

Unregistered user

Then WHAT are ROL talking about saying there is no demand for 32-bit chips?! Surely, the hardware manufactures are crying out for them!

[ Log in to reply ]

Nathan

Message #88567, posted at 12:33, 21/5/2001, in reply to message #88566

Unregistered user

I was constantly striving to find some good news regarding the RO market with Lee but we found none although the talk with Roy was enlightening.

Basically we need the impossible. Why do we have ROL? They are a middle-man after all, Pace seem to be doing the hard work. Can we get rid of ROL? I don't see any reason for their existence, a dealer could do much more at a cheaper price. ROL don't seem to consider the market, they just want to squeeze as much money out of you as possible.

OR

We have a more important say in how ROL is run. Maybe open source a few select parts of RO to get development going. The main problem is the lack of support between ROL and dealers, it is horrific.

I have no holdbacks on my feelings now as I am desperate to save the market.

[ Log in to reply ]

paul middleton

Message #88568, posted at 16:04, 21/5/2001, in reply to message #88567

Unregistered user

I can see that the Scheme is provoking some very extreme reactions and I would like to let readers know that further details of whats going to be in RISC OS Select will be included in the next newsletter for Foundation members which will go out later this week.
Details will then follow on the Web Site.

I would love to answer all of the suggestions raised in other comments but commercial restrictions make that impossible.
There are far more factors involved in the future of the RISC OS market than anyone who is not directly involved can ever realise, but suffice to say that some commentators are so wide of the mark with their suggestions that it would be funny, if it wasn't unfortunately so serious that it is usually their defeatist predictions that get believed.

The future development of RISC OS is totally assured, and as I have been pointing out in various presentations since last years Wakefield Show,
RISCOS Ltd is just a small, but essential part of the whole picture.

The enthusiast market, that RISCOS Ltd is primarily concerned with, has a very large part to play in ensuring the future development of desktop versions of RISC OS and I hope that those who are decrying the Select Scheme will read the announcement again and try to appreciate both what we are trying to offer, and what they will be contributing to the RISC OS market if they join the scheme.

As regards 32 bit what many people still fail to see is that even if there was a machine that ran solely in 32 bit mode available now, and there was a 32 bit version of RISC OS to go with it there would be virtually no application software to go with it. We have worked to get new tools to support development 32 bit development released and there will be further announcements about tools later this year. If you consider the current cost of Xscale chips and delivery lead times then you would begin to realise that our announcement is not quite the disaster that some people are predicting.

[ Log in to reply ]

Andrew Weston

Message #88569, posted at 19:10, 21/5/2001, in reply to message #88568

Unregistered user

Please could you clarify 'commitment from customers'. This i what troubles me and I suspect numerous other people.

If you're asking end-users for money before you will continue work then I'm sorry about being indelicate but this seems to me absolutely outrageous.

It is up to the hardware manufacturers surely, to generate the demand and then for them to sell the goods to us. Besides, with 4000 RO4 users and many, many people consistently reading and/or posting to the newsgroups and to web-forums for over 2 years since Acorn's demise, does this not show their commitment and loyalty to this platform.

Please tell us you are not blaming the end-user/enthusiast for restrictng development in this way.

[ Log in to reply ]

DaveF

Message #88570, posted at 19:13, 21/5/2001, in reply to message #88569

Unregistered user

> The future development of RISC OS is totally
> assured

Yes, but the desktop version is dependant on its users which are rapidly depleting (look at peoples comments on how busy Wakefield was). The relationship between RISCOS Ltd and the every day enthusiasts are at a breaking point, and then we're told that development of a 32-bit RISCOS is halted as we don't 'need' it.

> if there was a machine that ran solely in 32 bit
> mode available now, and there was a 32 bit
> version of RISC OS to go with it there would be
> virtually no application software to go with it.

Which is why Microdigital came up with the Omega? Virtually no software? Can you say "32-bit ready Vantage, Ovation Pro and Easi Writer are ready now". If you commited RISCOS Ltd to getting a 32-bit RISCOS then I'm sure nearly all the key developers would get key products ready for it.

> If you consider the current cost of Xscale chips
> and delivery lead times...

I accept that 32-bit processors will be expensive but its not as if we started today RISCOS 32-bit would be ready tommorrow, we need to start now before its too late.

And your vague comments about "There are far more factors involved in the future of the RISC OS market than anyone who is not directly involved can ever realise" well I think that suitablely shows how out of touch you are getting with customers such as myself.

The bottom line is we need these 32-bit processors, we are falling behind in technology so fast because of the limitations we face with 26-bit processors (DVD playback, web cams, digital camcorders, 3d-games, MP3 encoding etc...). With the announcement of Select scheme I understand somebody is going to have to pay to keep the small RISCOS boat - afloat but to abandon the main aim of 32-bit RISCOS, well that is crictical error that leaves the market more like a dead fish than a boat.

[ Log in to reply ]

Gunnlaugur Jonsson

Message #88571, posted at 19:59, 21/5/2001, in reply to message #88570

Unregistered user

Thanks Paul for that remark.

There are some points raised in this little comment you've just made that slightly lessen the blow but one's left to wonder why on earth some of these weren't mentioned in the press release.

First off:
"[T]he amount of work needed to convert major areas of RISC OS such as the kernel and ADFS will take many man months of work to complete and until we get sufficient commitment from our customers, that work will not be undertaken."

How on earth is one supposed to understand that in any other way than that you've dropped all work on converting RISC OS to 32bit? You state it clearly that work on that has stopped. To me that means no XScale or ARM10 in the near future making buying computers like the Omega in order to get hold of new powerful processors out of the question. The Omega will be a huge step forward but now somehow it looks a lot smaller.
And what do you mean by lack of commitment from your customers? Who are you referring to? Hardware developers? Software developers? End users? All of the above? Claims like these require at least some explaining otherwise to the unknowing they sound like the hardware manufacturers don't want the 32 bit OS and that means the end of the road as far as hardware is considered.

Quote:
As regards 32 bit what many people still fail to see is that even if there was a machine that ran solely in 32 bit mode available now, and there was a 32 bit version of RISC OS to go with it there would be virtually no application software to go with it.

This is not entirely so - there are heaps of applications written in BASIC who couldn't care less if the OS was 26bit, 32bit or 2000bit as long as the BASIC interpreter knows how to deal with it.

Quote:
We have worked to get new tools to support development 32 bit development released and there will be further announcements about tools later this year.

Stating this in the original press release would have explained a lot of things and I personally wouldn't have jumped quite so high, although my jump would still have been considerable. RISC OS users have been led to believe that RO 4.5 would be the first step towards 32bit compatibility and RISCOS Ltd. were working on that behind the scenes. Now it comes out that this work has stopped and who knows if it ever started! I'm not saying it didn't but RISCOS Ltd. has lost quite a lot of credibility with Fridays press release so some people might and will understand it like this.

Quote:
If you consider the current cost of Xscale chips and delivery lead times then you would begin to realise that our announcement is not quite the disaster that some people are predicting.

"No more work on 32bit conversion" is hard to interpret as bright and shiny future for a platform that's been waiting five years for any real hardware development no matter the cost of today's hardware. I don't think anyone was under the illusion that RISCOS Ltd. would release a new 32bit ready version this weekend and most realists I think were expecting more like late next year but now that you've told people that no work is being done in that direction at the moment who knows how long we'll have to wait!

Most of us know that upgrading the OS to work on 32bit only processors is time consuming but not working on it doesn't make it take any less time!

Just a final note Paul, before you distribute press releases like the one you did on Friday, please have someone who doesn't know the facts behind the news read it over for you and check out their reaction. It may all make good sense to you but those who don't have all the inside information can very easily read the wrong things from a statement like you just released, specially when it includes angry statements like the one about lack of commitment!

Given that a press release about a development tool is due I'll try and wait. But it had better be good news!

[ Log in to reply ]

Nathan

Message #88572, posted at 21:19, 21/5/2001, in reply to message #88571

Unregistered user

I agree totally with Dave and Gunnlauger. I am *very* surprised that this reaction to the statement was not gauged, call this bad customer service.
I still think the situation is very bad. I assume that we will never get clarification on whether commitment was enthusiasts or dealers but with posts like this we get more people dropping out which puts pressure on the rest to buy more to keep ROL afloat.

The impression you have made (Paul) is that the boat has already sunk and you have given me no more faith than when I first read that post. ROL should be an unbiased company caring for the product it has rather than blackmailing people/companies who are trying to keep the market in existence.

Obviously we don't know everything because so much is hushed up but it seems that something isn't right and that something slightly dodgy would be all to easy to cover up.

I am not happy or pleased and am speaking my true mind about the market for once in my life because the situation can't really get any worse. Is the RISC OS desktop computer market going to be here next year or are we heading for the stagnant pits as a result of bad business practise and bad customer service?

I will await and see but needless to say, unless I see some sort of true commitment to the market from ROL then my alliances will falter.

[ Log in to reply ]

Michael Gerbracht

Message #88573, posted at 16:36, 22/5/2001, in reply to message #88572

Unregistered user

What I don't understand in all those comments is why we ca´n`t make use of 32 Bit prozessors if there is no 32 Bit OS. The Omega is designt to run Applications on XScale processor even when there is no 32 Bit OS (So there Announcement). So we can have DVD decode, MP3 encoding and general faster apps and games. If only they are designed to run 32 Bit. And just the OS is fast enough on existing hardware.

[ Log in to reply ]

David McEwen

Message #88574, posted at 15:24, 23/5/2001, in reply to message #88573

Unregistered user

I think much of the outcry about this statement is from the way it was phrased. It certainly put me in a negative frame of mind. The select scheme itself sounds fine, a method of being bang up to date with the latest OS developments. It also meamns that ROL get a nice bit of cash flow - assuming people sign up for it.

The disappointing bit is the lack of commitment to 32bit, which is to me at least the singular most important step that the OS should be making. The excuse that no software will run on a fully 32bit version is lame. If other hardware companies didn't release hardware developments becuase there was notihng to run on it the industry would be in a very sorry state. Other comapnies deal with transition in various ways...

1. Allow developers to test software on site. Acorn did this with the SA upgrade.

2. Provide developers with pre-production hardware (or a soft-loadable upgrade in this case).

3. Provide an emulation solution. Apple did this when going from 68k to PowerPC.

All the lack of commitment to 32bit is doing is meaning that hardware is having to be more expensive and take longer to develop.

Why can't ROL adopt one of the above strategies ? Sure there is very little that is fully 32bit compliant atm, but we don't have a 32bit OS now. However why can't this be done this year, otherwise we will be afflicted by hardware bodges from each of the different manufacturers trying to circumvent something which should be sorted in software.

[ Log in to reply ]

Nathan

Message #88575, posted at 17:06, 23/5/2001, in reply to message #88574

Unregistered user

Yeah, the main pain is the lack of 32-bit support especially after releasing the 32-bit updates on the ROL site. So how long do we have to wait for another processor, the SA came out in 96 and five years down the line we still don't have the next step.
Stupid me, until we show commitment we can't have it.

Someone, please, tell me why certain parts of RO cannot be open source so that at least something can be done. Or does everyone have to be charged for doing anything bar using RO?

The lack of need for 32-bit is so lame it is like a zero legged dog. We have XScales sitting, staring, wanting a life and we are going to have to wait and see, might not ever come.

The Select scheme is causing rows on the usenet, fair enough but I think that this isn't the essential pain. Perhaps someone to vet press releases may help, I know plenty of people who would do this for free but I doubt this would be accepted.

Seems that noone listens, noone learns. Hardware manufacturers are working their butts off and now stuck with SA110 and a feeble ARM7500. Shameful. I used to like the days where I could argue with people about the good points of RO but I am ashamed of using it now, there are so few good points in my eyes and certainly no foreseeable future....is there....what is definitely coming then?

[ Log in to reply ]

Andrew Weston

Message #88576, posted at 17:48, 23/5/2001, in reply to message #88575

Unregistered user

I agree that the Select scheme isn't objectionable, it's the implication that it's the user's fault that RISC OS is not in the process of being converted for the latest processors.
As others have said, it may be poorly phrased but it's really quite offensive if you dwell on the statement and it's something that RISC OS Ltd have used before and various people said then that it was a kind of moral blackmail.
I don't know from where in the RISC OS community a precedent to this approach has been found.

[ Log in to reply ]

Nathan

Message #88577, posted at 21:12, 23/5/2001, in reply to message #88576

Unregistered user

Put another way, I work on an IT Helpdesk. If I *ever* dared to say, for example, "I am not going to help you in future unless you show commitment to a training course", I would be hung, drawn and quartered.

CUSTOMER SERVICE isn't we blame the users if the market dies. Customer Service is, "Sure, we'll help anywhere possible. We aren't going to let the market die."
It is a great pity that Peter Bondar isn't around because he showed the enthusiasm for the OS, all we have now is an enthusiasm for user cash.

What an extremely sad state of affairs.

[ Log in to reply ]

Michael Stubbs

Message #88578, posted at 23:36, 23/5/2001, in reply to message #88577

Unregistered user

Where is Peter Bondar now? I wasn't up on things back then, just a user. I understand he was very enthusiastic though. Maybe someone should offer him cash to come and rescue us. Chuck out a certain person and replace him with Bondar :)

It's a crying shame Castle didn't get the whole workstation division and not just the hardware :( or that it wasn't spun off as a separate company.

[ Log in to reply ]

David Eaton

Message #88579, posted at 00:01, 24/5/2001, in reply to message #88578

Unregistered user

JOOI - what is the basis for the view that seems to be expressed here that there is a big, immediate market for 32-bit machines?

I know there's a 'large' demand amongst users posting here; but then I also know a bunch of people still using A5000 (or earlier) machines, and others who've only just gone to an ARM710 RiscPC and think it's really cool. It may not be any good for the market in the future, but there are quite a few ordinary, non-power-users who are quite happy with what they've got.

For myself, it was only 6 months ago that I went from ARM710 to Kinetic, and have neither the money nor the need for anything more powerful in either the medium or even reasonably long term. The only things I want are certain bits of software that, AFAIK, /could/ be produced for existing machines but just haven't been. And I answered the ROL survey honestly.

So - and this is a genuine question to which I don't know the answer - is it that those of you posting here about your desire for 32-bit kit are representative of a RISC OS community that RISC OS Ltd are ignoring? Or could it be that you are a minority group of 'power users' that RISC OS Ltd, if strapped for cash to pay programmers for months of work, has to ignore while responding to the majority market from which it might get enough money eventually to pay for 32-bit development later?

I do not mean the above to sound offensive, and as I say the question is genuine: I do not pretend to know what the majority view of the market (as opposed to the relatively few who post to forums such as this) is. So I am asking - are you sure this view is actually widespread enough to constitute a viable market for ROL's efforts?

[ Log in to reply ]

Robert

Message #88580, posted at 00:01, 24/5/2001, in reply to message #88579

Unregistered user

Mmm. Peter Bondar was so damn enthusiastic that he pre-announced Pheobe by a full 2 years. Result, RPC sales dried up and the cash-flow started to look like a dangerous ski-slope. This was definately one of the reasons why Acorn had to be dismantled.
Somehow Peter Bondar at the helm, for all his enthusiasm and thought for the users wouldn't quite inspire me with confidence. Pre-announcing things to that extent, when the spec wasn't even decided was just plain silly.

[ Log in to reply ]

Nathan

Message #88581, posted at 07:38, 24/5/2001, in reply to message #88580

Unregistered user

I think Peter or someone like him could do much better than present management i.e. someone with enthusiasm. I am not saying announcing future plans early, I am saying enthusiasm. We are just run by business people.

Re: 32-bit. If everyone was happy with what they had then new things wouldn't come out. There was a small outcry that with the SA certain things wouldn't run. If technology doesn't advance in an IT sector then its pretty much dead.

You may not need 32-bit as a user, but as a developer it is the only thing that is keeping us going as programs require more and more power. The majority of developers want 32-bit so that they can explore better, more advanced hardware.
While the usenet is bickering about themselves (Select is too expensive, boo-hoo), the real factor is pushing this market to new technology. Without new technology the software will suffer but this is a bit chicken and egg.

[ Log in to reply ]

Lee Johnston

Message #88582, posted at 08:29, 24/5/2001, in reply to message #88581

Unregistered user

I think Nathan sums it up quite nicely. Yes the people who want full 32bit compatibility probably are the minority of power users. However this minority also probably represents the majority of developers left on the platform. You want developers to take up new hardware ASAP because they can begin to get to grips with it / write the software.

I don't know if everyone has seen the post on the newsgroups concerning what the first upgrade of RISC OS 4 will bring (someone posted it from the foundation newsletter). TBH it does seem pretty good. What I can't understand is why this information wasn't released with the original announcement, and then why was it restricted to foundation members only. The information is of interest to everyone. I realise foundation members pay to get information in advance but in this case it has been completely counter productive.

[ Log in to reply ]

Richard Goodwin

Message #88583, posted at 10:12, 24/5/2001, in reply to message #88582

Unregistered user

I've put the list from the Foundation letter up on the site now.

FWIW, some of the new development of RISC OS machines was not based solely on the enthusiast market - the Cerilica machine was being aimed at power users (graphics, especially as used in TV shows etc. where RISC OS still has a foothold) and some RiscStation machines were destined to run Linux not RISC OS but allowed hardware developers to put time and money into a machine that can be used for our favourite OS. Both of these directions will dry up if we don't have access to 32bit processors - not *just* the XScale that Paul seems to be fixing on, but later ARM processors too.

As for the arguement that if we had 32bit RISC OS in 6 months there'd be no software to run on it - rubbish. The reason I upgraded from RISC OS 2 to 3.11 was because in a very short space of time I got fed up of not being able to run new software on my old machine - people took to the new OS quickly and made that a reason for others to buy it too. Would this happen again? Well people are /already/ converting their software to be 32bit aware. Although some old favourites might fall by the wayside, if 32bit RISC OS was released at the end of this year (wasn't it supposed to be ready by this Summer?) I'm certain that there'd be more than enough software to run on it, and some enterprising hackers patching older software in the same vein GameOn and other hacks required to get old programs running when we upgraded to RiscPCs.

Maybe the outcry will cause RISCOS Ltd to change their minds. Probably not.

[ Log in to reply ]

Lee Johnston

Message #88584, posted at 10:27, 24/5/2001, in reply to message #88583

Unregistered user

I don't think it is within RISC OS Ltds resources to change their mind - it depends on Pace.

I've just seen an interesting post on the newsgroups. I won't name names but someone said that they wouldn't post the foundation letter to the newsgroups because if you aren't willing to pay 35quid / yr you should be kept in the dark.

Now given that joining the foundation is voluntary and we're supposed to be trying to expand the market this is an absolutely ludicrous stance to take over something so important. I know from experience that the foundation gets the first look at other information anyway so to restrict information that could reassure all users that RISC OS is being developed would seem extremely dopey marketing.

Then again, as suggested by someone else, not everyone chooses to blindly put money into something they know nothing about.

In the past the newsgroups have been the home to doom-sayers but I think the current discussions are very interesting. I'm under the impression that the "fanatics" are finding it harder to convince the "realists" why they should part with their hard earned cash. That's not good really.

[ Log in to reply ]

David Eaton

Message #88585, posted at 17:09, 24/5/2001, in reply to message #88584

Unregistered user

I quite understand the point about the developers wanting 32-bit, and that without the developers staying the OS will die.

But it doesn't help with the financial side of what I was wondering.

If (and I stress 'if' because I have no knowledge of such things) ROL have very poor cash flow and reserves ATM and simply cannot afford to pay programmers to work on the 32-bit kernal or whatever for 6 months without more income ... or at least a solid indication that there is a large demand for immediate sales at the end of that 6 months ... then the developers wishes, or the fact that it means the OS will eventually die do not come into it: they simply can't do it. Not that you'd expect them to come right out and say that quite so bluntly ... you'd expect, well, something rather like that press release :-/

And speaking as just one RISC OS user with no idea whether I am typical/in the majority/in a minority, arguments about whether this or that will cause the OS to die cut no ice with whether I'm going to buy a 32-bit machine now/in 6 months.

If an appeal went out that the OS would die unless everyone sent in a tenner, I'd send a tenner. But when we're talking about buying a machine that would cost, say, 1000UKP - plus more for software - that's a different matter. Not only do I have no need or desire for such a beast, but I haven't got 1000UKP+ even if I did. Heck, I'd like to upgrade my original x2 CDROM, get a 2nd Mb of VRAM and get some kind of reasonable backup medium for my HD, but can't even afford any one of those ATM.

Nor do I have any desire for any flashy software other than a few internet things that AFAIK could be done with the current machines. DTP-wise, for example, I'm still using a 1996 version of Impression Publisher, which does everything I want to do in a nice friendly way. Downloaded the OvPro demo version, but didn't like it and certainly felt no need for it.

It could well be that if the majority of users are like me, then it is that that is stopping ROL developing 32-bit right now on grounds of pure financial necessity on their part. And it may well be right that the hold-up in development this causes, and the subsequent loss of power-users and other developers from the platform causes the OS to die. Which would be a great shame. But that great shame does not magic me up the money for a new machine, plus all the other money for things I need money for rather more urgently given that I don't have any requirement for a new machine.

Nor does (if this is so) it magic up money for ROL to pay people while the development is going on if they simply haven't got any funds.

Which is all very depressing. But /if/ things are as I have speculated they might be, then it's a bit unfair and no use at all people beating ROL with a stick for making a choice which is the only choice they could make in those financial circumstances. Whatever the consequences of not doing something, if you haven't got the money to pay your staff to do something, then you simply can't do it. Which is why I was asking if those who are doing the beating do actually know that things aren't like that, and that they did have a choice.

[ Log in to reply ]

Lee Johnston

Message #88586, posted at 08:45, 25/5/2001, in reply to message #88585

Unregistered user

David, I think your points are very fair, and also realistic of the majority of RISC OS users who are left. I agree that people shouldn't be forced into upgrading if they don't need to - heck I'm still on an A4000 so I'm more of a culprit than you.

To address some of your points though.

As far as I see it ROL don't have to find the cash to implement a 32bit RISC OS - Pace already have it. ROL have to convince Pace that it's worth their while to let ROL have it. I also understand ROL point of view that XScale processors and the like are going to be expensive and perhaps not worthwhile right now. This doesn't mean they have to say 32bit isn't an issue. Of course what Pace haven't admitted to is having a version of RISC OS that doesn't rely on the IOMD and VIDC. Of these two John Kortnik has proved the VIDC poses the lesser problem with the Viewfinder card.

A new machine does not necessarily have to cost £1000 - the Xscale is just one possible processor. I suspect the ARM810 and ARM9 are somewhat cheaper than the XScale but are also more powerful than the ARM7500. Combined with low end graphics and sound cards it might be possible to produce a more powerful low end machine than is currently possible. Given the cost of the IOMD and VIDC chips it might even be possible to reduce the price of StrongARM based machines which can run 32bit software. (Note that StrongARM machines WILL be able to run the new OS, but it won't offer any benefits other than compatibility)

The availability of a 32bit OS doesn't mean the end for those who cannot afford the upgrade either. Pace are actively stating that applications should be written in such a way as to be mode independent. This is relatively easy if the code is written in C as it only requires a recompile using different compiler options and for the new SharedCLibrary to be loaded. However those running the 26bit OS shouldn't expect to be able to use the software to it's full potential - it's difficult to predict what the difference might be though (if the software doesn't use dynamic areas then 26bit users may be limited on the amount of data they can load to around 64MB as opposed to 4gig on the 32bit OS).

The new hardware will, however, give developers more confidence that writing software is worthwhile. As Nathan points out on the newsgroups, VOTI at least, effectively write software for free. Everynow and then we need an incentive to do so. I'll admit that the prospect of what is achievable with faster hardware is very tempting.

[ Log in to reply ]

David Taylor

Message #88587, posted at 19:41, 31/5/2001, in reply to message #88586

Unregistered user

Should RISC OS be regarded as a National Treasure?
If it were so designated, would it be eligible for
a Lottery Grant? Lobby your PPC.

[ Log in to reply ]

Nathan

Message #88588, posted at 13:12, 1/6/2001, in reply to message #88587

Unregistered user

But Pace would get the money.

[ Log in to reply ]

Marco Frissen

Message #88589, posted at 04:41, 2/6/2001, in reply to message #88588

Unregistered user

I think ROS Ltd made a wise choice. The current ROS lacks quite some features an OS really should need, 32bit compliancy is not one of them. I think Paul is right in saying that there is no immediate need for 32bit support, since the Omega is presumably to work with XScale even without the 32bit OS (makes me wonder how they'll do it though).
I'd rather see a gradual improvement and later maybe conversion of ROS than a quick-and-dirty push to 32bit. Personally, I'd rather see ROSLtd work on a good network stack, with all the bells and whistles one needs. If I'm going to put down £99, together with £25 for foundation membership, I don't know, the price is a bit high for 'minor' patches once in a while... Mind you I just returned from vacation, and haven't read the latest foundation letter... maybe something new is in there...
I'll stay supporting and promoting RiscOS, because even without 32 bit support (which doesn't show from the outside anyway), it still is 100% more userfriendly and functionally better than any other client OS I've worked with. That is what counts, in the end: does the OS do what you want it to do, without any steep learning curves.. PRODUCTIVITY is the word. Amen :-)

[ Log in to reply ]

Nathan

Message #88590, posted at 22:23, 3/6/2001, in reply to message #88589

Unregistered user

Just in case someone tries libel, this is all my personal opinion :) :
I'm afraid I can't agree with Marco. From what I gather MD are making the Omega so that any 32-bit apps run on the XScale and everything else on the SA110. Sounds suspiciously like they have no choice because ROL aren't playing ball.
Gradual improvements slow everything down and we are far behind in the race. 32-bit OS is needed NOW and an emulator/patch program can be written to use 26-bit programs. Evolution didn't occur gradually, it was stops and starts.

There are two perspectives:
1) User
2) Developer

Unfortunately it seems from usenet that users are trapped in their own little worlds, happy to continue as is without a regard as to what will happen and loving ROL every step of the way. Developers, on-the-other-hand, are dropping out of the market. There is no set future, everything we were working for is now part of a blackmail package. I was strongly thinking (and still am) about dropping out of the scene altogether as there doesn't seem to be any future in the desktop RO market. A lot of users unfortunately then say, "Well bog off". Not a common sense attitude.

We need ROL to buck up their ideas and fast. Gain some sort of PR person (I know plenty who will do it for free) and put about some enthusiasm. Some tactics to talk about our market on the outside world would be useful too, you don't need to pay to say the word, "RISC OS". It seems that ROL just sit back while the market tries desperately to spread the word and increase the market size whereupon ROL knock it down again. Acorn was something to look up to, they had the right PR people, ROL don't, don't appear to do anything and are frankly abismal in attitude.
If I was a developer Acorn were happy to help and provided a lot of information, ROL seem to want cash for questions IMHO.

Don't get me wrong, we do have a market, well nearly but if I am to stick around for much longer then *I* want to see commitment from *ROL* and *NOT* the userbase. It is disgusting and irritating and I am highly annoyed at the whole nature of ROL these days hence why I speaking my mind.

[ Log in to reply ]

andrew

Message #88591, posted at 09:47, 4/6/2001, in reply to message #88590

Unregistered user

As I recently said to somebody, people say that Acorn finishing was not necessarily a bad thing but half of their staff are apparently now upgrading RO to run with 32-bit (i.e. FASTER) CPUs. If Acorn had continued this would have been for our purposes.
This is not RISC OS Ltd's fault of course but I'm still mystified why they claim not to see the development of a 32-bit desktop version as important as improving the current version.
A faster OS will increase productivity, very possibly sell more machines and allow for technical advances in software development.
This will at the very least give people who are 'satisfied with what they have' and 'do not need anything better' the option to upgrade when they feel like it.
There may not be a clear-cut need for many things but that is not always a reason to sit back and stagnate.

[ Log in to reply ]

Lee Johnston

Message #88592, posted at 10:04, 4/6/2001, in reply to message #88591

Unregistered user

I tend to agree with Nathan on most of this, although I think the 32bit issue is fast becoming one of whether ROL see it as important or not (regardless of whether it happens now). The press release suggested that it was not deemed important and you can immediately extrapolate that to "there is no future" because there can't be without the ability to run on 32bit processors. Whether it should be done right now is another issue.

As for things like the improved IP stack. Yes I agree completely - I'm just not convinced that these upgrades haven't actually been available for sometime already and we've been waiting for no real reason.

[ Log in to reply ]

Michael Stubbs

Message #88593, posted at 14:50, 14/6/2001, in reply to message #88592

Unregistered user

As a developer, Nathan, why don't you organise with other developers to formally lobby ROL to get on with what the greater majority of us all want and need - 32bit RISC OS, pref. with hardware independence.

Maybe if you and some other well-known developers made your concerns formally heard at ROL, there'd be more chance of them listening to what we want.

[ Log in to reply ]

Pages (3): |< < 2 > >|