Castle opens RISC OS future wishlist

By Chris Williams. Published: 18th May 2004, 16:06:17 | Permalink | Printable

CTL compares notes with users

Castle have published a meek road map of their view on the future of RISC OS development, which is about time too, given that they own it.

The RISC OS 5 developers have pledged to keep the Norcroft C/C++ compiler toolset up to date, but they say updating the RISC OS core is "commercially rewarding work" and "has the highest priority". Such development has, according to Castle, led to further support for ARM processor cores.

Future desktop enhancements, under the codename Merlin, will include: filer enhancements, extended graphics capabilities and network services, USB performance and connectivity, audio input, window manager enhancements, ROM application (Paint, Edit, etc.) enhancements and last but not least, bug fixes. These updates will be gradually released to end users, and Castle may introduce a pay structure for some features.

Castle have also opened a features wishlist, billed as a "6 week period of open consultation with the RISC OS community to assist Castle's management in allocating development resources through its internal Green Paper."

The Merlin website adds: "At the end of the 6 week consultation, Castle will begin processing the Wishlist database and compare it to their 'Green Paper'. Once processed, the Wishlist database entries will be made available along with Castle's comments and preliminary projections of future development plans."

The parallels between today's announcement and the early days of RISC OS Select are uncanny: RISCOS Ltd. also talked of C/C++ compiler updates, also opened a wishlist, also published a road map of desktop enhancements and also launched a scheme of periodic updates. Castle have also all of a sudden begun, uncharacteristically, pre-announcing activities and products: firstly, beta USB2 support is demonstrated at the Wakefield show, and then talk of un-named projects involving 50,000 units. As the divide between RISC OS 5 and 4 ever widens, Castle are either feeling the heat from the combined popularity of VirtualRPC and Select, or they have much to talk about.

Links

Project Merlin website Press release

Previous: Omega ethernet ships
Next: Omega XScale support in 2004?

Discussion

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

How about a SINGLE version of RISC OS instead of the forking around 3.8?

It's such a duplication of effort with two ADFSs, three, or four USB stacks.....

It's funny how Castle seem to be the one in the corner though now as I'd have thought between Select for RiscPC, Omega, A75 and VA they must be selling more 4.x than 5.0x

If Castle want to continue to develop RISC OS rather than just bug-fixing the Iyonix, then they need to buy ROL out, jees it's only a one-man-band!

simo on 18/5/04 4:54PM
[ Reply | Permalink | Report ]

Sounds exciting. I agree with simo that it's odd to have two (or more) seriously different paths being taken.

harmsy on 18/5/04 4:57PM
[ Reply | Permalink | Report ]

Firstly- I applaud this initiative from the great people who brought us the Iyonix and, recently, the great looking Panther. Asking your target market what it wants is a good step forward. In the RISC OS field progress need not only to be made, but be /seen/ to be made.

I feel that it is about time Castle learnt from the example of Select. Hopefuly this 'consultation' can lead to some of the features of Select appearing on 'Merlin' though duplication of effort within the platform does seem crazy. Maybe liaison with ROL can be one of our suggestions on the wish list (I know this requires a two way dialogue).

That said I think that the community should get behind this initiative and provide constructive suggestions. Focusing on the negative can only be harmful for the platform as a whole.

Herm on 18/5/04 5:00PM
[ Reply | Permalink | Report ]

I vote for them buying out ROL too!

Ah, if only it was as simple as writing a comment on a drobe forum ;-)

Adam

adamr on 18/5/04 5:10PM
[ Reply | Permalink | Report ]

Sounds like a good idea, and sounds a lot more than Castle "just bug-fixing the Iyonix". If they are actively seeking users opinions, then it sounds like they are really developing the OS.

It is odd to have the OS going in different directions, and different companies developing each side. Of course the people to make the comments about the split to are Castle, and they are now asking for consultation with their users. Perhaps a chance to raise these concerns.

j5m1th on 18/5/04 5:14PM
[ Reply | Permalink | Report ]

Marvellous! Let's provide a whole load of enhancements that have already been written. This argument is going to keep going round and round and round.

jonix on 18/5/04 5:25PM
[ Reply | Permalink | Report ]

Please please get talking with RISCOS Ltd and Microdigital. Let's have some dialogue going! We just need some putting together of RISC OS 4 and 5 that will suit all the hardware that we have in development.

But then again perhaps this is just a consultation about the future of RISC OS *5*, not RISC OS itself??!! Aargh!!! If Castle own RISC OS then take control of all of it dammit! I'm so tired of the messing around...

rod on 18/5/04 5:27PM
[ Reply | Permalink | Report ]

RO4.xx = emulators RO5 = new ARM based hardware

Seems a simple split to me ;o)

ROL won't have to write a HAL or 32 bit compatibility, Castle won't have to provide Windowsy GUI 'enhancements' ;o)

I feel that we actually need is software new applications (Cino is a good example) to *increase* the number of things that we can do using RISC OS. Maybe changes to the OS will enable/enhance this process.

blahsnr on 18/5/04 5:54PM
[ Reply | Permalink | Report ]

Good news then, nice to see Castle being proactive about engaging their customers (and potential customers).

If I were them I'd advise *against* buying out ROL for a number of reasons:

(1). What ROL have can't readily be moved to RO5 - so why pay out a wodge of cash for it and then pay again to have it adapted to RO5 (2). If Castle had to continue to support existing RO products under VA it would be a waste of resources and if they *didn't* it would probably P*ss off people who bought into that "hybrid" computer nonsense. (3). How p*ssed off would Omega users be if Castle *didn't* support Omega as ROL had ? (4). All the above sounds expensive and that money and effort would be better spent on the hardware platform and the OS.

The one unified hardware and software platform is the only sensible way for Castle to go (IMHO) that means Iyonix (or its successor) and RO5.

AMS on 18/5/04 6:44PM
[ Reply | Permalink | Report ]

The biggest mistake one could make here is to start putting "talk to RISCOS Ltd" on the wishlist. That is not an operating system enhancement. Here is a chance to put forth sensible and constructive suggestions as to what to add into the OS. It would be childish in the extreme to clog up the system with political and commercial comments.

Castle are taking RISC OS forward. This further commitment to the desktop market is brilliant news.

It's time to except that the future of RISC OS lies with this company. It has the money and engineers that RISCOS Ltd doesn't and to boot it has infinitely superior management and strategies, as has been demonstrated time and again.

I can't see why anyone holds any sort of emotional or ideological attachment to RISCOS Ltd, who have, with the exception of releasing RISC OS 4, been mostly useless. RISC OS's glory days were when Acorn was developing the OS and the hardware together, which is precisely what you have now with Castle. With RISCOS Ltd, we ended up with some 56Mhz machines and a RiscPC with faster RAM (Kinetic). With Castle in control, you now have the option of buying a 600MHz XScale machine. It's a no brainer!!

blahsnr:

Agreed. Software development is key. But with Castle talking of further support of ARM cores and this open dedication to the desktop market, it seems a pretty stable commercial environment in which to release software.

arenaman on 18/5/04 6:51PM
[ Reply | Permalink | Report ]

AMS: "What ROL have can't readily be adapted to RISC OS 5"

Oddly enough, Jack used almost exactly those words when he gave his presentation at ROUGOL last night.

Castle and RISCOS Ltd *have* talked about ways and means of co-operation, but seem to have reached the mutually acceptable conclusion that neither will gain much by trying to re-transplant one company's development plans over the top of the other.

Surely the userbase is better off accepting that reality, and pushing forward further development of RISC OS functionality, which is what Castle are offering - rather than diving into the politics of which company armchair critics believe should own which company.

As blahsnr quite rightly said, ROL are developing RISC OS 4 and are likely to make most of their sales through PCs with emulators (plus the A75 that isn't really aimed at the ordinary user), and Castle are developing RISC OS 5 that is aimed at the end user who wants to use real RISC hardware. It's a neat split, and it all makes sense.

What may start to upset people (perhaps has already?) is Castle increasingly stepping into Acorn's shoes. Selling tens of thousands of systems with RISC OS on real ARM hardware means Castle end up with vastly more resources and influence than anyone else in the RISC OS market, and the small players may quite understandably feel threatened.

dgs

dgs on 18/5/04 7:06PM
[ Reply | Permalink | Report ]

Does this ever mean RISC OS Ltd and Castle will co-operate?

sebgate20 on 18/5/04 7:24PM
[ Reply | Permalink | Report ]

arenaman said "RISC OS's glory days were when Acorn was developing the OS and the hardware together"

Glory days my pants! "It was like Romeo and Juliet, except it ended in tragedy..."

adamr on 18/5/04 7:35PM
[ Reply | Permalink | Report ]

Well OK, if it doesn't make sense to try to port existing code between RO4 and 5, then at least make any NEW features version-independent - like one ToolBox, CLib etc.

Having two ADFS's being re-developed for the same reasons at the same time is crazy. Just make one that will work on all machine types - if RPC, VA and Omega can use completely different hardware and all use RO4, then surely the Iyonix's much-heralded HAL can cope with using the same ADFS?!

simo on 18/5/04 7:36PM
[ Reply | Permalink | Report ]

Simo>

I am not sure that the two ADFS developments are aimed at solving the same problem. Castle's ADFS will extend support for UDMA to CD/DVD devices (UDMA already works for the HDD). UDMA doesn't exist on ROL's implementation (it's not present on the VA implementation as the access is ultimately done through windows and Omega (to my knowledge) uses it's own IDEFS).

If Castle had to take ROL's ADFS it would be a step backwards - CTL's ADFS *already* does UDMA - it would have to give that up to use ROL's ADFS wouldn't it ? And the HAL works under RO5 it wouldn't help port CTL's ADFS to the RPC either (quite apart from which the RPC doesn't do UDMA

AMS on 18/5/04 7:47PM
[ Reply | Permalink | Report ]

Sorry 'bout the last sentence there - I'll try English the next time

AMS on 18/5/04 7:48PM
[ Reply | Permalink | Report ]

Well, UDMA is used when you use VA, of course, because Windows's IDE drivers do it. That, along with modern standards of block cacheing, is the reason VA blows everything else away in disc speed terms.

nunfetishist on 18/5/04 8:01PM
[ Reply | Permalink | Report ]

dgs: "Surely the userbase is better off accepting that reality, and pushing forward further development of RISC OS functionality, which is what Castle are offering"

That's what people said when ROL came along. Better support this company (ROL) which is pushing forward development of RISC OS in the wake of Acorn's demise.

I bought Acorn, I bought Psion and now look where those companies are. I have this horrible fear that if I buy an Iyonix that (a) the £x00s I've spent on Select are well and truly wasted and (b) Castle could end up making STBs or embedded systems with no interest in the desktop market in a couple of years.

I don't want to see my investment in RISC OS (the OS, not ROL) wasted. I don't want to buy a computer from a company that *may* in a few years have moved on to better, more profitable things.

jms on 18/5/04 8:04PM
[ Reply | Permalink | Report ]

nunfetishist>

"UDMA is used when you use VA",

yes I know, but re-reading what I wrote you'll see that I meant that the ADFS implementation provided by ROL is the bog standard one - it doesn't know how to do UDMA and that it is actually Windows (through VA) that does the clever stuff.

Moving ROL's ADFS onto RO5 would be a nonsense as it is less advanced (relatively) than the one in RO5 already (never mind any further improvements CTL will make to it).

AMS on 18/5/04 8:11PM
[ Reply | Permalink | Report ]

adamr:

A3000, A3010, A5000, A7000, RiscPC to mention a few of the machines that Acorn produced, along with necessary changes to RISC OS.

Acorn machines were at one point far ahead of rivals in terms of speed and capability.

Compare to what happened (or mor accurately, didn't happen) whilst RISCOS Ltd 'developed' the OS, and clearly the Acorn days were the glory days. Just because Acorn stopped making desktop computers and then disappeared doesn't mean the time before this was bad, does it?

arenaman on 18/5/04 8:12PM
[ Reply | Permalink | Report ]

jms:

Considering Castle are asking what desktop users want in the OS and are clearly commited to the desktop market, I don't see what you are worrying about. Especially since if you bought a Pee Cee for example, the same risk applies. Considering Castle are the ones developing new hardware and the OS to work on such hardware (such as IyonixPC), Castle own RISC OS and RISCOS Ltd clearly has no intention in looking to the future in this fashion (unless someone else pays the cost) then in my mind buying an IyonixPC is a better investment in the OS.

arenaman on 18/5/04 8:19PM
[ Reply | Permalink | Report ]

jms: "Castle could end up making STBs or embedded systems with no interest in the desktop market"

You need to listen more carefully to Castle presentations. That's not Castle's aim, and they certainly don't have the naive approach that Acorn had regarding big external customers hopefully descending on them with big chequebooks. Yes Castle *do* want to take advantage of investment from larger companies, but the reason they're putting huge efforts into the enthusiast market is that thousands of enthusiasts spending a thousand or two pounds each is still a very worthwhile amount of money - that's not going to change fast.

Do you know of any companies who got rich making set top boxes? I don't.

Watch this space for a more detailed examination of some of what Castle have been doing and saying recently - it may meet some of your concerns.

Incidentally, I've been a Select subscriber all along, I've bought an Iyonix and am considering buying a second Iyonix. I don't consider my Select subscription wasted, I could always keep a Risc PC and put RISC OS 4.39 on it for fun.

dgs

dgs on 18/5/04 8:22PM
[ Reply | Permalink | Report ]

As long as the APIs are the same so you won't need two versions of programs. But of course there needs to be the same bug fixes and features so that programmers can use them over the whole platform. If Select features were standard for Iyonix, VPC and Omega users and an option for RiscPC users more programs would use them.

We're basically going to have 3 groups of users, ROS4, ROS4.5 and ROS5, so everyone will still have to write for ROS4 users.

mavhc on 18/5/04 8:28PM
[ Reply | Permalink | Report ]

> That, along with modern standards of block cacheing, is the reason VA blows everything else away in disc speed terms.

Well, that and the fact that that you're probably referring to the RISCOSmark benchmark which doesn't test the disc access speed at all under VA because it just repeatedly loads a 1MB file which has been cached by Windows and is thus little more than a memory copying test.

If that's what your RO app happens to be doing then fine, but it's unlikely to be representative of most apps.

adrianl on 18/5/04 8:29PM
[ Reply | Permalink | Report ]

dgs: "I don't consider my Select subscription wasted"

If you no longer use a RPC (or only rarely) and Iyonix is the way forward, then what benefit has years of spending money on Select provided? None!

As a 'normal' user I find the situation of two different versions of RO seeming to pull further apart utterly depressing. Anyway, I look forward to reading the upcoming Castle article.

jms on 18/5/04 8:40PM
[ Reply | Permalink | Report ]

jms:

That's like saying what's the point of buying a new car now when in a few years you will replace it. It doesn't mean the car was a waste of money. You buy what you need or want. If you used it until it was time to move on, then it wasn't a waste of money. Apply same logic to RISC OS.

arenaman on 18/5/04 9:02PM
[ Reply | Permalink | Report ]

I think I agree with jms, an easy analogy would be that if you were a Mac user, you used Mac OS X 10.0, then 10.1, 10.2, and then 10.3 with your Power Mac G4, but when you upgraded to a G5 you had to go back to 10.0, 10.3 was not compatible.

Having now used an Iyonix, I can see what the Select fans are talking about, whereas before, I quote frankly did not see what the fuss was about. Select has some stuff (like alpha-blended PNGs) which allow you to achieve effects which are just not possible on an Iyonix. Not until (if) Vantage appears on it, anyway. That's not to slight the Iyonix in any way, it's cool, it would just be cooler with Select.

thegman on 18/5/04 9:16PM
[ Reply | Permalink | Report ]

arenaman:

I'm not quite sure I follow the car analogy ... but if I use it, I feel it's like buying a new car that doesn't have the aircon, electric windows and GPS system of my current car, but it'll have a bigger engine and uprated suspension. I might then have to wait another couple of years for this model to get aircon, etc. Personally, I prefer a car with aircon to one with a big engine.

jms on 18/5/04 9:31PM
[ Reply | Permalink | Report ]

thegman: Comparisons with other operating systems are not always appropriate. Other operating systems have a tendency to get slower with every iteration. (It's one of these sad facts, even despite developer efforts to avoid it). But most RISC OS developments, including RISC OS 5, tend to make things faster with new releases, not slower.

Long may that continue.

If the G5 ran at twice the clock speed of the G4, with the cache, disk access and so forth to back it up, and similar features, then I doubt anyone would be using a G4 Mac even if they *were* the only ones to have Mac OS X 10.3 or whatever. (Does this comparison look silly now? It's your comparison).

dgs

dgs on 18/5/04 9:34PM
[ Reply | Permalink | Report ]

jms: It's more the case that the Iyonix *does* have the air-con, and electric (front) windows, and cruise control and remote central locking. These are standard features (USB, networking, DHCP?) that some other cars are only just getting.

The Iyonix also has the large engine, obviously.

The Iyonix arguably *doesn't* have the shiny R18 alloy wheels, the racing stripes, and the ten-CD autochanger as standard. Maybe Castle's PANTHER pc is the racing stripes one.

If your current car has a GPS navigation system, and you use it all the time, then you *definitely* don't want to do without something similar in your next car. Is there anything that is similarly vital in your Select software, that isn't provided with the Iyonix? i.e., can you be specific?

Are we fed up with car analogies yet?

dgs

dgs on 18/5/04 9:46PM
[ Reply | Permalink | Report ]

The tale of two stories? M$ and DOS Iyonixs and RISC OS How does the story of M$ and it's DOS system go? As I understand over the years, the DOS idea was not dumped just to please past M$ users. Is this correct to say that M$ may have had a more appealing system if it did dump DOS ages ago? So are we recreating M$ history by hanging on to RISC OS? - Just to please past owners and users of the that system? I personally have a RISC OS 4 and 3.7 machines and would like "something" to leap forward into development whichever system it may be RO4 or RO5. It appears in my views that "some" (but could be "many") Acorn and mid 90's RISC OS successors don't like to spend much money on computers. The problem you get (like me) when these darn machines last too long!!! If Castle makes this "Quantum Leap" in development, I'm speeding up my savings for an Iyonixs too. All the same, most of the conflicting opinions I read are still of interest and should be evaluated. I registered with the Merlin Site to get my 2 New Zealand dollars worth of say (Is NZ$2 worth much in £pounds??) Cheers, Steve.

Sawadee on 19/05/04 01:31AM
[ Reply | Permalink | Report ]

IMO Acorn went down because they effectively stopped developing their own unique hardware-technology after the RiscPC was released. By the time they noticed this mistake and restarted development of their chipset for the Phoebe, it was too late. They also did not use the little influence they had on ARM properly. The ARM3 had multi processor capabilities. Why was this dumped? IMHO if you want a brigt future for RISC OS, you need to support the companies, wich develop both the OS- and the hardware-technology. We need our own chipset, because otherwise we will always be dependant on low-end chipsets for the embedded markets. Of course any company doing such a development has a lot of catching-up to do. MD for example first had to duplicate the efforts Acorn did betweeen the first Axx machines and the Phoebe. Now that this has been done, they will continue development to catch up with the rest of the industry, starting with features like FPAs (floating point accelerators), graphics acceleration (tailored to the WIMP), etc, wich can be implemented in the Omega and then speeding these up with the newer FPGAs from Xilinx (already available) in new machines, when the Omega's FPGAs have been used to their full potential.

JGZimmerle on 19/05/04 08:06AM
[ Reply | Permalink | Report ]

And because the OS developers do not have to put in lots of effort to adapt to new chipsets wich are incompatible with the previous ones, they can concentrate on keeping up with Windows and MacOS on user-recognisable features.

JGZimmerle on 19/05/04 08:21AM
[ Reply | Permalink | Report ]

In reply to Blashnr:

O 4.xxx does NOT only= emulators!

Theres also the RPC, A7000+, RiscStation and Omega who could all, presumably, benefit from any developements ROL do to RO4.xxx.

Col1 on 19/05/04 09:05AM
[ Reply | Permalink | Report ]

Col1: The RiscPC isn't manufactured any more. The A7000+ may still be manufactured, but it's hardly likely to be sold in any quantity. Likewise the RiscStation. Omega sales have been, shall we say, fairly limited.

I assume blahsnr is talking about future development and sales of new machines, perhaps to new users, not small numbers of sales for obsolete machines like the RiscPC.

This also leads us back to the question of why people who have bought new machines should continue funding RISC OS 4.xx developments that will be of no benefit to them.

dgs

dgs on 19/05/04 09:24AM
[ Reply | Permalink | Report ]

JGZ Acorn went down because they unfortunately were sitting on a very large pile of valuable ARM shares.

I suspect that vast majority of Acorn shareholders did very well out of the breakup deal and hence from those quarters I can recall relatively few complaints.

Undoubtedly if Acorn had carried on with Pheobe they would have solved the technical problems they had encountered. However at the then proposed price of around ukp2000 it makes even the Omega look like a bargain.

Acorn would have carried on making losses even with Pheobe and would have carried on having to deal with the 'Acorn community' constantly whingeing on about what Acorn were doing wrong.

Castle have taken an enourmous risk in "putting Acorn back together". They have their own customers for RISC OS outside the desktop market, and therefore have a vested interest in protecting their brand reputation.

Of course there is the danger that a one horse race could 'damage competition'. However bearing in mind the huge range of software, hardware prices, economies of scale, and raw power advantages of the PC and Mac world, don't you think that Castle would have competition aplenty if they were the only player in the RISC OS market?

blahsnr on 19/05/04 09:29AM
[ Reply | Permalink | Report ]

Colin dgs has got it right.

If the market for RO4+ for the older Acorn and ARM7500FE computers was so large, why then did ROL have to launch Select to raise funds?

Ignoring that for a moment tell me what sort of developments could these older machines (or emulators Omega or Windows style) benefit from, that would advance the RO4+ branch?

HAL? 32bit compatibility? UDMA enabled ADFS? PCI graphics card support? DVDFS?

blahsnr on 19/05/04 09:47AM
[ Reply | Permalink | Report ]

JGZ said "We need our own chipset, because otherwise we will always be dependant on low-end chipsets for the embedded markets."

I don't get this argument. Surely we want to be able to use the cheapest, most readily avaiilable components? What's the point of redesigning, for example, a graphics chipset, when other companies who specialise in this have already done the hard work? What happens if the people making "our own chipset" suffer problems, go out of business etc? We'll be back to where we were when Acorn closed down.

jonix on 19/05/04 09:52AM
[ Reply | Permalink | Report ]

In reply to JMS: "I'm not quite sure I follow the car analogy ... but if I use it, I feel it's like buying a new car that doesn't have the aircon, electric windows and GPS system of my current car, but it'll have a bigger engine and uprated suspension. I might then have to wait another couple of years for this model to get aircon, etc. Personally, I prefer a car with aircon to one with a big engine." Well yes, you would wouldn't you. Presumably that's why Castle now want to do some work on the aircon and get it installed in their 'car' (RISC OS 5). The (silly) arguement around here, however, seems to be that since other car manufacturers have put aircon into their own cars, Castle have to use their aircon otherwise it will be a Terrible Thing. As a customer buying the car, however, surely if you get the aircon you're happy, regardless of where it came from?

senduran on 19/05/04 09:55AM
[ Reply | Permalink | Report ]

The improvements that Castle state are not specific and you could say that every single OS release has met these criteria from Arthur onwards ! I strongly suspect that Castle realise that they be losing the initiative of OS improvements and are trying to deflect attention from the advances in Select. The bottom line is that the Riscos market is small and duplication of effort is utter madness. I'll say it again can somebody physically close to Jack and Paul knock their stubborn heads together. The excuse that features in Select can't be ported to RO5 that's bull.... It's a smokescreen for "Jack and Paul can't agree a way forward but don't want to admit it".

Reality check to all those singing the praises of Castle as the "only" way forward. If you want more applications then you MUST have a single OS otherwise developers will have horrendous problems getting code to work with both versions. These problems will increase as the differences increase. I did a quick search for some rough numbers and from what I can glean if an application developer concentrated on RO5 only, they would lose 80% of the userbase !!!!!!!!! frigging hell come on people. The survey on Iconbar states that 20% of users are RO5, 71% are RO3.7,4,select, the remainder are pre 3.7. So the economics of the current riscos marketplace would dictate that an application would make more money ignoring RO5 not embracing it to the exclusion of all others.

Now I totally agree that the future is ARM hardware not emulation but there is still more none xscale ARM hardware in use. It would be a no brainer to provide an upgrade path through RO3.7,4,select to RO? where RO? is the branch re-joining. People would use applications and an OS with super duper features which go slowly on a Strongarm but fly on an Xscale. They would then be desperate to upgrade to the latest hardware are you listening Jack ?. Otherwise people are amazing at living in ignorance with old applications. You bribe people with software features to purchase better hardware ! (Microsoft do it all the time with unbelievable success). You also solve the problem with people having invested in Select (more users than RO5 don't forget) who don't want to see their investment as a waste of money.

mripley on 19/05/04 10:16AM
[ Reply | Permalink | Report ]

JGZ re own vs generic chipsets Developing drivers for generic chipsets saves all of that time and energy re-inventing the (hardware) wheel. Companies developing chipsets, graphics cards, etc generally know what they are doing and spend a lot of time optimising their products (video cards for instance).

If your OS supports many chipsets then as a hardware manufacturer you can tailor your OEM hardware product to the wishes of the customer. Neither you nor your OEM customer are dependent on just one chipset or expansion card supplier.

If you buy a generic chipset and the supplier accidently includes/reinvents something patented by someone else, they get sued, and the worst that happens is you have to find a compatible chipset.

If you develop your own chipset and you accidently include/reinvent something patented by someone else, you can be sued and there is no redress.

blahsnr on 19/05/04 10:16AM
[ Reply | Permalink | Report ]

JGZimmerle:

Acorn went down for a simple reason - they sucked at business.

Designing their own hardware was actually part their downfall. A company with such a small market was never going to survive by designing everything themselves when far cheaper and more powerful components were available. What Acorn designed was often superiour to what was available but unfortunately at a heavy cost.

If Acorn had decided sooner to start implementing stuff like PCI, even AGP (if technically possible) maybe the Phoebe would have been a more realistic goal. Acorn would never have had a chance designing graphic chips up against the likes of ATI and nVidia today. Better to use them than to chase with them on chip design.

Far to much off-the-shelf stuff was being redesigned by Acorn instead of just using what others had already created and were selling for maybe a quarter of the price Acorn could ever dream of getting their stuff for. Sometimes the off-the-shelf stuff was even far superiour to what Acorn created, some simply just more readily available.

At a time it seemed that Acorn wouldn't have been able to sell a winter jacket on the North Pole - even if it was full of castaways from Ibiza because they didn't believe it was their target group!

That's why they failed - IMO

Gulli on 19/05/04 10:27AM
[ Reply | Permalink | Report ]

@jonix:

Maybe you want to use the cheapest, most readily available components, but I don't. I want high-quality, powerful components. I now they have their price, and I am prepared to pay it, just like most Omega owners.

The points in redesigning the graphics chip were compatability with the VIDC and to get graphics accelaration tailored to the needs of RISC OS's GUI, rather than Windows' GUI. But my point was actually more about the mainboard-chipset, specifically the northbridge (IOMD in the RiscPC). There simply is no high-end northbride designed for ARM desktop computers available from any of the big ARM manufacturers. What we want are things like FPAs, additional intelligent caches with instruction reordering to make optimal use of the slow ARM chips, acceleration functions for things the ARM is terribly slow with, like iDCD/DCD, the ability to use new CPUs without having to wait for a northbridge that could be shoe-horned into a desktop computer, etc.

JGZimmerle on 19/05/04 10:28AM
[ Reply | Permalink | Report ]

Acorn went down because they did not have anything to offer other than a dated OS and old hardware-technology. Because of that they had become unable to make enough money themselves to offer an additional value to the huge pile of ARM shares the owned to their own shareholders.

JGZimmerle on 19/05/04 10:33AM
[ Reply | Permalink | Report ]

Mr Ripley ASSUMING the poll on the Iconbar could be considered accurate in any way and looking at the 'latest' incarnations of RISC OS 27% use Select most regularly vs 20% using RO5.

Considering RO5 is less than two years old, and requires you to buy a new machine, whereas Select is more than three years old and requires no new hardware, I would have thought that many more would be using Select.

Also interesting that RO4 = 29% (a fiver year old OS) and still 15% using RO3.5-7 (10-8 year old OS). I'd say that the Iyonix/RO 5 has done remarkably well.

On this poll you would most likely only lose the interest of 29% and not 80% as you suggested.

As many Select like features are available for RO5 (PhotoFiler and CDROMFS now 32bit share and freeware utilities provide other features) it would be relatively easy to ship a version of RO5 with most of the visible Select features implemented.

blahsnr on 19/05/04 10:40AM
[ Reply | Permalink | Report ]

@mripley:

100% agreed. And I may add, that Select is already following exactly that strategy with things like the thumbnail view in the filer. It only makes sense if you either only have very small pictures or a very fast machine.

JGZimmerle on 19/05/04 10:42AM
[ Reply | Permalink | Report ]

JGZ&mriple Re bloat read here the advice for A440 users; [link]

IS that an example of bloated code or simply a limit of the hardware, in this case an A440? ;o)

We are no longer running 8MHz processors with 4Mb of RAM. A StrongARM RiscPC is now 8 years old and PC's run processors with clock speeds of around 15x an SARPC (and lets not start talking about RPC hard disc and RAM speeds).

What current MS OS or current Windows application is going to run happily on an 8 year old PC?

blahsnr on 19/05/04 10:55AM
[ Reply | Permalink | Report ]

JGZ> cheap and readily available does not necessarily mean low quality.

jonix on 19/05/04 11:21AM
[ Reply | Permalink | Report ]

Julian.

If you want high end graphics capabilities (and I odnt nessesarily mean screen redraw) like deconvolution, 3D rendering of isosurfaces and native xyzdatasets then you use high quality components. IE ditck riscos and get an SGI.

I dont think high end components are required for the apps that still exist in RISC-OS-land.

AND cheap is not nessesarily good or bad. EG some graphics cards I've been looking at are custom designs and cost 2000pounds, however I'm getting a 600pound 512Mb card for my entry level machine.

All the best Bob

anon/130.209.49.128 via 130.209.6.40 on 19/05/04 1:17PM
[ Reply | Permalink | Report ]

S. Williams - my 71% is based on all the users of RO3.7,4 and Select running the OS on a RiscPC (a reasonable assumption). It is these owners that are likely to upgrade their hardware if the applications have developed enough. Now here's an if I posted on CSAMisc today. "If" the Omega is released with a fully functional Xscale then 71% of the current userbase has a hardware and OS upgarde path that is fully compatible with their existing hardware/OS/applications. The Iyonix/RO5 path would be a dead end since a developer has a choice either 71% or 20%, a no brainer. The Select side also has the added advantage that the current emulation users can upgrade to an Xscale with no software issues.

If the OS paths are merged such that Omega and Iyonix are running the same OS then neither path is a dead end. However this has to happen asap before the differences become too great. I think most people want this.

If Select is being developed to the point where RiscPC hardware is an issue (re Julians point about thumbnails) then , to me anyway, it suggests that the developers know something.....like a functional Xscale card for the Omega ?

regards,

Malcolm

mripley on 19/05/04 3:48PM
[ Reply | Permalink | Report ]

mripley:

"a developer has a choice either 71% or 20%, a no brainer"

Which developer would that be, then? I've yet to see a major app written by anyone outside of ROL that makes full use of features present in Select and /requires/ Select for full functionality. Given that many RISC OS users are still using RO 4.0x and below, it makes a lot of your argument pointless speculation. The RISC OS market is such that it's a brave move to support /one/ variant of the OS alone. As a case in point, NetSurf uses its own internal functionality to render images. It doesn't use the IFR system present in some Select versions as doing so would result in either extra code complexity (increasing the likelihood of bugs) or lack of image rendering on non-Select OS versions.

It's all very well having new APIs and functionality in the OS (as Select does) but it's all a bit pointless until developers no longer have to (or need to) support RISC OS 4.0x and below.

Equally, noone appears to have considered the possibility that enhancements made to RO5 by CTL would be API-compatible with those made by ROL to Select. There's no denying that the duplication of effort in implementation seems futile from the outside, but if APIs between the two branches are identical or sufficiently similar, a lot of the "71% or 20%" choice you say a developer would have is largely removed. You're then limited to deciding whether to embrace the future (Select and RO5) or continue to wallow in the past (4.0x and 3).

"The Select side also has the added advantage that the current emulation users can upgrade to an Xscale with no software issues"

What evidence do you have that those using emulation would return to native hardware given the opportunity? The presence of VirtualAmp alone worries me a great deal. Not because it isn't an excellent piece of software (which undoubtedly it is) but because of the dangerous precedent it sets. Next we'll have a thin wrapper around mshtml.dll (IE's HTML rendering engine) and so on. Why is this a dangerous precedent? It effectively means the end of RO as we know it. Why would software developers bother to put time and effort into products for native RO when they can make a quick buck throwing a wrapper around some Windows API? (This argument holds for a hypothetical VA running under Linux too before someone accuses me of anti-MS bias). One thing I certainly don't want to see is RISC OS as a glorified skin for another OS, which is the way I see things going. If anything, VA has become too good, which can only be a credit to the VA team though it poses serious questions for native hardware manufacturers and software developers.

John.

jmb on 19/05/04 4:39PM
[ Reply | Permalink | Report ]

mripley How so a no brainer? The Omega WITHOUT X-Scale is significantly more expensive compared to an Iyonix of a similar spec (this is ignoring the lower performance of the Omega).

Go and check the figures I posted here; [link]

An Iyonix WITH Aemulor Pro is at least ukp270 cheaper than an Omega of similar spec. How much money are you going to need for 32bit software upgrades bearing in mind that the Iyonix ships with a full suite of internet software including Oregano2?

Don't forget also that RO4 has some compatibility issues for older software and these are not going to go away because you are running it on an Omega!

As for RiscPC hardware being an issue it has been an issue for a long long time. It was the fastest thing we had but long past its sell by date, now at least we have Iyonix.

blahsnr on 19/05/04 4:48PM
[ Reply | Permalink | Report ]

Let's face it. Divergence now is going to kill RISC OS stone. cold. dead.

You're going to get apps developers having to make a choice, or support both, as jbm rightly points out. If the API differs enough that various RO features start to differ on a user level, you're very quickly going to end up in a situation similar to a Linux kernel running varying systems on top (a la Gnome/KDE/etc.). Linux has a big enough userbase to support this. RO simply doesn't.

No matter how you look at it, this is going to cause one side or another to fall over unless they have a meaningful dialogue.

heds on 19/05/04 5:34PM
[ Reply | Permalink | Report ]

heds: I disagree. The work Castle is doing is far from killing RISC OS stone dead - its keeping it alive.

If RO Ltd have shipped 1000 copies of Select for example (Im guessing at the figure btw) and Castle have shipped 100,000 copies of RISC OS 5 on some "ARM powered device", I know as a developer which market I'd be looking to support, even if the APIs were different (which at present they arent)

Cheers,

/Neil/

spellinn on 19/05/04 5:54PM
[ Reply | Permalink | Report ]

Castle should have consulted and hired RISC OS Ltd to do RISC OS 5 in the first place instead of sidetracking them. They could have had an ally and a knowledgeable company ready to advance RISC OS 5. Now they have an enemy.

jerryf on 19/05/04 11:45PM
[ Reply | Permalink | Report ]

jerryf: "Castle should have consulted and hired RISC OS Ltd to do RISC OS 5 in the first place"

Castle never needed to "consult and hire" RISCOS Ltd to produce a 32-bit version of RISC OS, that was part of RISCOS Ltd's original mission statement. Castle only went it alone when nothing had appeared several years later.

Without wishing to be rude, it seems that you're commenting on some quite involved "political" matters behind RISC OS, without actually paying any attention to the reality. Please don't try to mislead people like this!

"Now they have an enemy"

Do they? Who? Why so?

dgs

dgs on 20/05/04 00:17AM
[ Reply | Permalink | Report ]

jerryf:

What claptrap! Who are you to say what Castle should have done? Whether you like it or not, Castle's recent moves have been a resounding success - it's given them ownership of the OS that is central to their future, as well as the future of the RISC OS desktop market. It's also allowed the release of the IyonixPC. RISCOS Ltd is hardly any more knowledgeable than Castle - stop and think what Castle has achieved and how much longer they've been developing, compared to RISCOS Ltd. Also think where the engineers currently working on RISC OS at Castle came from.

RISCOS Ltd failed, so Castle moved to protect their business and the market. Tough luck for RISCOS Ltd. That's business. Why do you think Castle is so much more successful in the market than the other businesses? They act like a real business. Like Computer Concepts used to.

If Castle hadn't moved, we'd quite likely still be waiting for a 32bit RISC OS and the hardware to go with it.

anon/80.6.181.115 on 20/05/04 01:49AM
[ Reply | Permalink | Report ]

Having followed this discussion, only one thing is clear - support Castle. They are professional in their business outlook, and are actively seeking and taking on customers. What are ROL doing? Supporting a few thousand Select subscribers.

If I was a newcomer to this scene, what do you think I would do? I wouldnt give a toss about the politics, I'd go for something which I feel has some future - Iyonix. Castle have the resources, and they have RISC OS 5 to develop further to meet the needs of their current/future customers.

If I was a corporate customer who approached Castle with a few to developing a project, what do you think I would do if Castle turned around and said that OS development is dependent on ROL and their handful of engineers, and that turnaround time is counted in several months, if not years? I'd quite rightly go elsewhere.

<flameproof> This all seems like a very silly discussion. Why must it go on? </flameproof>

sascott on 20/05/04 03:29AM
[ Reply | Permalink | Report ]

John,

Your first paragraph is exactly what I'm saying about development problems and the branch is currently stifling development. There are two paths nobody knows which way things will go but if the Omega/Xscale becomes reality then its choice between 20% Iyonix/RO5 or 71%/Omega/Select. You could develop both for a short while but the paths will diverge as time goes on and it will not be cost effective to support the 20% side. If the Omega/Xscale becomes a reality then it is not "wallowing in the past" as you put it but would actually be more advanced than Iyonix/RO5. It also allows for a cleaner upgrade path for those at the fork and a cost effective upgrade path for those already treading down it.

What evidence do I have that emulation users would return to native hardware ? Well I don't it's based on the discussions and defense of emulation that have been bandied about the various newsgroups. There are more voices saying it will promote a migration to RISCOS with the inevitable uptake of hardware than those that say it won't. Just for the record I was always one of those who argued the latter, like yourself, but am willing to concede to the majority opinion. Again this is one of those wait and see moments. However it would not matter if the RISCOS fork was rejoined.

Just for record I'm at the RO4 branch and waiting for :

A) Select on Iyonix in which case I'd be an Iyonix owner by Xmas at the latest. B) Omega is complete with an Xscale in which case I'd be an Omega owner by Xmas. C) Select/Omega branch dies a death in which case I'd be an Iyonix owner whenever. D) Plans to merge the OS's in which case I'd purchase Select for my RiscPC and then whatever Xscale hardware has the rejoined OS when it appears.

The reason for choice B over Iyonix/RO5 is that the market potential is 71% versus 20%. Now if people migrate over to Iyonix such that it becomes 50:50 then change choice B to an Iyonix since as has been said before Castle seem to be more professional. The problem is there are an awful lot of "personalities" and prejudices flying around to really know who is professional and who isn't.

Stephen : if you want to comment about it being a silly discussion then say so without preceding your statement by contributing to the "silly discussion". That's just hypocritical. Unless you are saying that one side of the discussion is silly in which case that's just plain nasty. Everyone is entitled to their opinion aren't they ?

regards,

Malcolm

mripley on 20/05/04 09:21AM
[ Reply | Permalink | Report ]

Castle should have made those they need to side with them. Instead they stepped on just about everybody's toes and did everything themselves. Now they reap the benefits. Hardly any development for Iyonix. RISC OS split. They should crash and burn.

jerryf on 20/05/04 09:46AM
[ Reply | Permalink | Report ]

mripley You can have virually all the user visible features offered by the current release of Select on an Iyonix today. What are you waiting for?

ROL could have released a 'Select for Iyonix' set of modules, many months ago. I would have been happy to pay money for some of the Select features but now my money has gone to other companies offering Iyonix happy alternatives.

More lost sales for ROL must make their shareholders happy!

blahsnr on 20/05/04 09:47AM
[ Reply | Permalink | Report ]

mripley:

You appear to have missed my point entirely. The fact remains that it makes no difference what fancy new features ROL or CTL add to their versions of the OS. Developers will generally try to (or have to in order to make a project viable) cater for the lowest common denominator. Alas, this is RO3 or RO4.0x. (hence my point about the lack of developments relying on new APIs in Select) . Those using Select are not "wallowing in the past" as I was careful to point out. I would contend, however, that it is those users running RO3 or 4.0x who are. If this platform is to move forward, a concerted effort is needed to move those users onto newer OS versions - be that RO5 or Select (I really couldn't care less which option they choose). How this is to be achieved is anybody's guess, however. The alternative, of course, is for the extra features to be made available for older OS versions though I feel that this works around the issue rather than solving the root cause.

As I stated before, the fork in itself is not necessarily a bad thing. API divergence is. What I don't really understand is the assumption by many that APIs will be different. If CTL have any sense (and it appears they do) any functionality they add to OS5 which duplicates that in Select would have the same API. If that doesn't happen, then as heds has already said, we can wave goodbye to any sensible future for the OS in the desktop arena.

John.

jmb on 20/05/04 10:00AM
[ Reply | Permalink | Report ]

mripley: A) Select is a scheme to keep the older equipment as upto date as possible. Castle via the Merlin project are the future. B) Sooner or later Microdigital are going to have to "wake up and smell the coffee". The 26bit processor that the Omega is based on are no longer manufactured, ArmTwister was a nice idea, but it they cannot get it to work. They are going to have to make the Omega an XScale based RISC OS 5 machine. So, there really is no point in furthering Select for the Omega. C/D) Select will survive in the short term, in fact it serves a vital function because there are so many 26bit machines out there. Eventually it will wither when these machines are replaced.

JWCR on 20/05/04 10:03AM
[ Reply | Permalink | Report ]

"You can have virually all the user visible features offered by the current release of Select on an Iyonix today"

No, you can't. Which is why so many people aren't buying Iyonix until it gets Select.

"More lost sales for ROL must make their shareholders happy!"

Er, they have nothing to sell to you. That's not a 'lost sale'. Castle is the one actually losing sales.

jerryf on 20/05/04 10:04AM
[ Reply | Permalink | Report ]

jerryf: > > "You can have virually all the user visible features offered by the current release of Select on an Iyonix today" > "No, you can't. Which is why so many people aren't buying Iyonix until it gets Select.

Apart from a few cosmetic things like thumbnailing and alphanumeric sorting in the Filer, I don't really feel that there is a great loss from moving from Select to RISC OS5. The main thing holding back sales of the Iyonix is the price, and that is unavoidably high because the RISC OS market is so small.

JWCR on 20/05/04 10:13AM
[ Reply | Permalink | Report ]

What is this 71% vs 20% notion all about? There are too few people to market products to as it is, so developers simply use the RISC OS 3.7 feature set. With RISC OS Ltd holding onto everything they've developed rather than making some of it backwardly available (like the nested WIMP) they've basically locked developers out. As an example, look at NetSurf. Amongst other things, it has to provide support for PNGs that have alpha transparency. This is the perfect use for a new Select feature - but it can't be used as it alienates a huge percentage of users (and almost all of the developers). As such, a alpha transparency module was written, valuable development time is thrown out the window and another feature goes to waste. As much as I respect the fact that RISC OS Ltd have to keep themselves afloat, there are some components that really should be made available otherwise developers will simply find their own solutions and RISC OS Ltd's efforts will be wasted. In my eyes this includes Toolbox fixes (as they have been - the versioning situation between OS5 and Select/Adjust here is a joke though), alpha channel sprites, and basically the stuff that affects developers and not users.

not_ginger_matt on 20/05/04 10:17AM
[ Reply | Permalink | Report ]

But surely if RISCOS Ltd were to 'give away' their improvements there would be less of a reason for people to upgrade. When I was at Uni I needed DHCP to connect to the network. because of that, I upgraded to select. At some point things have to stop working on old versions of RO to encourage users to upgrade. I don't need a faster machine at the moment, so for me speeding up my productivity with cut+paste icons was worth the upgrade. If for example Netsurf required Select to work correctly, then perhaps more people would upgrade to use it. Yes you would loose some users. But you always will. (Losing the developers is more of a problem I conceed).

Rimmer on 20/05/04 10:31AM
[ Reply | Permalink | Report ]

jerryf Most user visible Select features ARE available for Iyonix.

I wrote to ROL explaining why I was letting my Select subscription lapse (after 2 years of subscribing). It was mainly because I could find precious few advances over RO4 that made it worthwhile me contributing to the scheme anymore especially for OS upgrades targetted at obsolete computers.

I also said that if certain Select modules such as the improved CDFS and thumbnail filer were to be made available as stand alone products for the Iyonix then I'd be interested in buying them.

I have had no reaction to either my email or letter I sent. Since then WSS have released CDROMFS and PhotoFiler for the Iyonix. So accordingly as ROL are not interested in my business anymore WSS get my money. Thus very much a lost sale for ROL.

JWCR PhotoFiler from WSS is now Iyonix happy so thumbnailing (including Artworks files) can be had on the Iyonix. Or you could use something like Thump.

blahsnr on 20/05/04 10:35AM
[ Reply | Permalink | Report ]

I think I must have missed something.Where is the Omega 32bit OS coming from...it is needed for an XScale processor ins't it?

jlavallin on 20/05/04 11:10AM
[ Reply | Permalink | Report ]

jlavallin: I think it's coming from the same place as the Lightning upgrade for the Mico.

dgs

anon/195.217.253.5 on 20/05/04 11:13AM
[ Reply | Permalink | Report ]

Rimmer: I'm not suggesting that RISC OS Ltd give away /all/ their improvements - if you read my post I specifically listed the ones that I believe would actually start the OS advancing and new features being used. Alpha blended sprites are an example of this, DHCP certainly isn't. Your argument for NetSurf and Select is unfortunately flawed as if NetSurf required Select then the developers who are on an Iyonix would be unable to contribute any further even if they were willing to waste their money. I've just made a retrograde step of going from OS4 to OS3.7 (long story - if anyone has some s/h OS4 ROMs for sale, let me know!) and yet NetSurf still runs fine due to Acorn's release of the nested WindowManager. Realistically, no-one is going to buy Select simply because the OS supports alpha-blended sprites. Bearing in mind that no developer is going to use them because their audience would be reduced, they'll write their own module and bypass Select totally. So, we end up with a million and one independant versions of a function that has already been written by RISC OS Ltd because no-one can use their version (heck, the documentation to most of their things is only available if you have Select!)

not_ginger_matt on 20/05/04 11:18AM
[ Reply | Permalink | Report ]

I understand all the arguments for and against RISCOS ltd releaisng components for other systems, but it just seems like a catch 22 situation. RISCOS can't afford to make changes if they give them away for free, but no software is going to support those changes unless they do.

Rimmer on 20/05/04 12:06AM
[ Reply | Permalink | Report ]

And how are you going to actually create alpha blended sprites without the Select version of !Paint that lets you? It's a very wise decision to encourage developers to follow a single format rather than to independantly create their own - hasn't the fact that there are two different USB stacks tought anyone anything?

not_ginger_matt on 20/05/04 12:44AM
[ Reply | Permalink | Report ]

Richard Wilson I wouldn't know what an alpha blended sprite was if it bit me on the bum.

However I do know that from a 'technical' point of view the usual reasons why a piece of code cannot be used elsewhere in an otherwise similar system are a) it has been incompetently implemented b) it has been deliberately implemented in an incompatible way (for instance to try to get more business for your company) c) it has been implemented with no thought to similar target platforms. d) it was the only solution available at the time given resource and funding limits. e) your client has insisted it is done in a particular way for their own curious reasons. f) it is in fact very simple to re-use it on the similar system but for political reasons you don't want people to know this.

blahsnr on 20/05/04 1:11PM
[ Reply | Permalink | Report ]

S Williams An alpha blended sprite is where instead of having a choice of whether a pixel is transparent or not, you have a choice of various levels of transparency. This allows you to do lots of clever things ranging from having sprites of ghosts that show proportion of the background through them to making the smoothing around the edge of icons work correctly (generally RISC OS sprites are blended to grey so that they look nicer in the Filer, but this makes them look horrible on fancy pinboards - now you can have the edges correctly semi-transparent so it looks better on any background [are you listening RISC OS Ltd?]). Anyway, the (apparent) reason why many Select features cannot be ported across to RISC OS 5 without much work is due to 'low level kernel changes'. I can appreciate this for many of their changes, but that line simply doesn't apply to the alpha sprites unless they've done it in an insanely perverse manner. I'm sure I'm sounding like a broken record with this issue, but it's directly affected me and many hours of my evenings because of this idiotic policy of theirs. As is probably becoming very apparent, I've absolutely no time for RISC OS Ltd because of their constant failure to decide upon a development direction, see where the market is going, or even do simple things like provide documentation for the enhancements/changes they have done. Having said all this, I'm not especially happy with Castle developing RISC OS. It's nothing to do with any of the same reasons though - they've proven themselves to be everything the market could want them to be - my issue is that we're going back to the days of Acorn with the OS developer also making the hardware. And please can everyone not now hijack this forum and start moaning about the competition being MicroDigital and an Omega that they saw (heck, possibly even touched!) at a show once that seemed to go wrong when they did something that they can't remember but they've also heard that everyone else has a buggy one so it must be true and any user who has one that's been nigh on flawless is either insanely lucky or a liar, as I'm talking about the principal of having more than one hardware developer and not specifics.

not_ginger_matt on 20/05/04 1:39PM
[ Reply | Permalink | Report ]

Richard Wilson I can understand your point about about competition. However with such a small market and such enormous external competition (MS Apple Linux to name but a few) how on earth is the internal competition at this time going to help the RISC OS market? Especially when most of the internal 'competition' seems to be in the form of 'buy a Windows box and an emulator it's the future of RISC OS'.

As for Castle developing RISC OS well what is the alternative? ROL were supposed to bring us RISC OS 5 but after 5 years they still don't seem to have managed it. Castle have their own non-desktop customers and have a much greater vested interest in promoting (and protecting) the RISC OS brand than ROL seem to.

blahsnr on 20/05/04 2:08PM
[ Reply | Permalink | Report ]

S Williams I'm talking ideal world here, not reality - given the current climate of the RISC OS world, I can think of no better OS maintainer than Castle/Tematic. They seem to go out of their way to help developers (look at the documentation available on their website) whereas RISC OS Ltd seem to simply close everything off (lack of publically availble APIs, nonsense regarding locking out the 32bit SharedCLibrary from Select.) In reality, the only future I see with RISC OS Ltd is peddling their wares for emulation. As much as I dislike the principle of this, there is a real hole in the market with respect to laptops that it nicely fills. Given time, this need will hopefully vanish and a native laptop will be available - an excellent opportunity for a second hardware company that wouldn't be competing with Castle but simply complementing their offerings.

not_ginger_matt on 20/05/04 2:40PM
[ Reply | Permalink | Report ]

A long list of interesting comments here. Many comments are still the merry-go-round topics of ROL4 Select Vs Castle Iyonix with the odd Ancient History fact about the Acorn etc. etc. All our wish list comments should include though the comments of the above as it is important no matter what the outcome of our computer platforms future is, I am keeping my fingers crossed for the Iyonixs/RISCOS industry and all it's users that our "Road Map" as stated in the article does NOT end up like President George Bush's "Road Map" to Middle East Peace!!!!! I wonder if M$ Windoz company really listen to their users? Cheers Steve.

Sawadee on 20/05/04 10:02PM
[ Reply | Permalink | Report ]

What is all the fuzz about?

I bought my first "Acorn" because I wanted some tasks to be accomplished. In my case I wanted an affordable DTP system which did the things that I needed it to do. The A5000 back then deliverd it. The system was fast, more or less stable and had the applications/tools that I needed to perfom DTPwork.

Now I understand the 2 paths that are being walked by ROS and CTL with regards to the OS. But basic functionality is the same on both versions. So most application software runs on both versions. Now I think that we currently need more and better application software (at affordable prices).

So please stop all this fuzz about which RISCOS from which company. Let both CTL and ROS do their respective job (especially since they both appear to be doing good things albeit independantly from each other). And lets do something on the end-user application front!

Because no matter how good (or bad) your OS is, if you got no apps the whole things just fails.

Also read [link] ...especially the last lines ;-)

quite interesting stuff about OS's

anon/213.193.182.110 on 20/05/04 11:02PM
[ Reply | Permalink | Report ]

spellinn: "If RO Ltd have shipped 1000 copies of Select for example (Im guessing at the figure btw) and Castle have shipped 100,000 copies of RISC OS 5 on some "ARM powered device", I know as a developer which market I'd be looking to support, even if the APIs were different (which at present they arent) "

Would those (hypothetical) 100,000 units all be the full version of RO5 as found in the Iyonix or a cut-down embedded version? (I'm just curious to know).

jwcr: "Select is a scheme to keep the older equipment as upto date as possible. Castle via the Merlin project are the future."

I'd like to say a rude word followed by an exclamation mark to that. That's not how Select was sold to prospective buyers and it's not a phrase I've heard said until, well, this last week or so when Castle's proposals were made public.

If Select is simply to keep older equipment up to date, then surely it and RO5 need to be as compatible as possible so they both have something to be up to date with?

jms on 20/05/04 11:14PM
[ Reply | Permalink | Report ]

Epdm> Read what, exactly? Can't see any OS related stuff on there.

jonix on 20/05/04 11:20PM
[ Reply | Permalink | Report ]

jms: Could you name a key piece of software that works on RISC OS 5, that doesn't work under RISC OS 4.37 ?

dgs

dgs on 20/05/04 11:34PM
[ Reply | Permalink | Report ]

@jms: "Select is a scheme to keep the older equipment as upto date as possible. Castle via the Merlin project are the future." This is just marketing-nonsense, just ignore it.

JGZimmerle on 20/05/04 11:37PM
[ Reply | Permalink | Report ]

JGZimmerle: "This is just marketing-nonsense, just ignore it".

It's a pity to see "marketing-nonsense" on these forums, Julian.

Could you explain to me how comments about the Omega having "the ability to use new CPUs without having to wait", are not "marketing-nonsense" ?

dgs

dgs on 21/05/04 00:19AM
[ Reply | Permalink | Report ]

> units all be the full version of RO5 as found in the Iyonix? No. The core stuff would undoubtedly be largely unmodified and most RO5 apps could presumably be used with little or no modification. Since the I/O capabilities would differ from those of the Iyonix, different driver modules would be present and those which aren't required would be omitted, just as they are in the Iyonix RO5 which has no PS2Keyboard, Econet, MIDI or Joystick modules, for example.

adrianl on 21/05/04 01:40AM
[ Reply | Permalink | Report ]

dgs:

I can name two - Aemulor and Cino :-)

/Neil/

spellinn on 21/05/04 08:52AM
[ Reply | Permalink | Report ]

"Could you name a key piece of software that works on RISC OS 5, that doesn't work under RISC OS 4.37 ?"

Aemulor?

mrtd on 21/05/04 08:57AM
[ Reply | Permalink | Report ]

spellinn: Thank you, but I doubt those would be of much use on the Risc PC anyway :-)

dgs

dgs on 21/05/04 09:13AM
[ Reply | Permalink | Report ]

Richard Wilson: And how are you going to actually create alpha blended sprites without the Select version of !Paint that lets you?

With the latest versions of Composition or Variations on anything running RISC OS 3.5, 3.6, 3.7, 4.0x, Select or (for that matter) 5.0x. :-p

(ok, its years since I used 3.5 - but it should work!)

Rob.

southerner on 21/05/04 10:30AM
[ Reply | Permalink | Report ]

southerner: "and how are you going to actually create alpha blended sprites without the Select version of !Paint that lets you? ... With the latest versions of Composition"

What do you do with it once created, apart from looking at it in Compo on a non-Select machine?

anon/138.3.180.155 via 193.32.3.82 on 21/05/04 2:09PM
[ Reply | Permalink | Report ]

Cool nick. ;-)

Edit it? Combine it with other graphics and save it as a JPEG? Convert it into a PNG? Send it to a Select user?

southerner on 21/05/04 10:24PM
[ Reply | Permalink | Report ]

jms: "I bought Psion and look where those companies are". Psion is off Edgeware Road in London, and are still making portable machines, and doing quite well for themselves. Was there a point you cared to make?

mrchocky on 22/05/04 11:42PM
[ Reply | Permalink | Report ]

Language support (Like what you find under "Encoding" in MS). I use sites nearly every day that require Thai language. I use our schools MS system computers, but cannot "yet" read Thai language on my RISC OS because I have not successfully learned how to get type 1 fonts into my computers font manager (I have a Thai/English keyboard though) and have unsuccessfully tried to convert and install True Type One Thai fonts into my Oregano2 browser I bought only a few weeks ago.

I have an Acorn/RISC OS Type One to Font converter software, but something was missing when I got it to the computer's font manager so it didn't work. All I need is a basic bundle of True Type One Thai fonts to put in my computer's font manager and in the Oregano2 programme. It would be nice to have a few basic common languages support features as "standard" in our computer ;-)

Cheers, Steve.

Sawadee on 23/05/04 09:54AM
[ Reply | Permalink | Report ]

mrchocky: I did say Acorn and Psion. Although your answer amused me, Psion has pretty much divested itself of its OS business and is no longer making consumer PDAs - in some ways resembling Acorn's licensing of the OS to ROL and getting out of the desktop market. Any consumer models available from either are now rather long in the tooth and no longer manufactured.

It was also to illustrate the point that, once upon a time, I was rather keen to "buy British". Neither company could now provide me with a new model. I feel somewhat cautious about buying British again lest the same fate befall the new one.

jms on 23/05/04 1:24PM
[ Reply | Permalink | Report ]

I can assure you that Psion is still making consumer PDAs. Psion hasn't divested itself of its "OS business", merely of EPOC. Unfortunately, I can't say a great deal more about that at the moment, apart from its public statements about Linux.

mrchocky on 23/05/04 2:29PM
[ Reply | Permalink | Report ]

bundled software with the computer package. A standard set of software (bundled) with level of computer and purpose use (homeRiscPC or businessRiscPC). If practical to do so, maybe offer packages like: HomeRiscPC 1, 2, 3, 4, 5 - that has several choice options of preferred software. The same idea for the business package? The choice of software options keeps the competitiveness and development market open. I don't feel that the Microsoft no software choice bundled packages are the way to go. Overall, the whole idea I'm getting at (with other people's opinion and refinement added??), is that I remember when I bought my RiscPC600 in 1995, it came with no word processor or other important basic software like you get now with Windows XP Professional (of course many other computers didn't include basic software then or in the earlier years). For those who still prefer the unbundled software packaged computers, then maybe HomeRiscPC 1, could be basic with virtually no DTP or optional software in the deal. Something basically raw to start with in the level 1? Level 2 could be the existing Iyonixs package (EasiWriter, Fireworks etc.) Level 3 could have Level 2 software plus OHP, CinoDVD (when available?), a Mediaplayer, a WebPage maker. Level 4 could include OvationPro or ImpressionX, Resultz, (and other alternatives etc). Then Level 5 be the same as Level 4 but include OHP, CinoDVD etc.... I understand many RISC OS users may prefer an unbundled choice and building their own software choices, but I am merely concluding that to attract some of the non RISC OS user market, this may be a way to help encourage more sales. In my experiences of many schools, a majority of common MS PC users are far from being skilled (when compared to the average RISC OS user) in using their computers. I think that offering some packages from a basic to a higher level (more software that is) may help this end of the "Dummies" market. Like myself when I started in RISC OS, I did not have a clue what optional software was available and neither did the PC company that knew very little about Acorns who sold me the RiscPC600 in Australia. I did ask them what software can you get for this machine, when my computer arrived from Melbourne to my home country town's computer shop (3 hours north of Melbourne), they thought that the !Edit was the word processor, so did I. It was 6 months later when I came to Auckland and got on to a real Acorn Agent and I quickly learnt what I needed. Cheers for now, STEVE. :-)

Sawadee on 29/05/04 08:05AM
[ 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

The new apple of my eye
Would you swap your dusty Acorn for a polished Apple computer? Martin Hansen has been checking out the world of Steve Jobs and his range of shiny kit.

15 comments, latest by adh1003 on 6/1/09 1:06PM. Published: 17 Nov 2008

Random article

News in brief
RSS client, programming tools and more

11 comments, latest by mavhc on 19/5/05 9:43AM. Published: 16 May 2005

Useful links

News and media:
Iconbar • MyRISCOS • ArcSite • RISCOScode • ANS • C.S.A.Announce • Archive • Qercus • RiscWorld • Drag'n'Drop • GAG-News

Top developers:
RISCOS Ltd • RISC OS Open • MW Software • R-Comp • Advantage Six • VirtualAcorn

Dealers:
CJE Micros • APDL • Castle • a4 • X-Ample • Liquid Silicon • Webmonster

Usergroups:
WROCC • RONE • NKACC • IRUG • SASAUG • ROUGOL • RONWUG • MUG • WAUG • GAG • RISCOS.be

Useful:
RISCOS.org.uk • RISCOS.org • RISCOS.info • Filebase • Chris Why's Acorn/RISC OS collection • NetSurf

Non-RISC OS:
The Register • The Inquirer • Apple Insider • BBC News • Sky News • Google News • xkcd • diodesign

Â© 1999-2009 The Drobe Team. Some rights reserved, click here for more information
Powered by MiniDrobeCMS, based on J4U | Statistics
"This poll is not indicative, and doesn't reflect public opinion"
Page generated in 0.2759 seconds.