RISC OS news and information from Drobe


Beta! \| About us \| Contact \| RSS	Webspace \| Tech docs \| Downloads \| BBC Micro \| Gallery \| Wallpaper

News in brief

By Chris Williams. Published: 11th Jun 2005, 18:56:17.

Select for iyonix saga rolls on, plus software news

And now a round up of our news backlog that's built up over the past week or so.

ROL moves Select Iyonix goalposts
It's reported that existing Iyonix owning Select subscribers who want to see Select on their XScale powered computers don't count in RISCOS Ltd.'s eyes. ROL's Paul Middleton apparently explained before the weekend that he is now seeking at least 100 Iyonix owners who are not subscribers and want to see Select32 for Castle's machine to sign up before his company will commit to developing Select for the Iyonix. Until then, 32bit RISC OS 4 is for the A9home only although bits and pieces have been allegedly shown to work on the Iyonix in public. ROL surveyed its existing subscriber base in November last year to gauge Select for Iyonix demand, but (as far as we're aware) neglected to approach those who haven't bought into the Select scheme, the very people it wants - so should it come as no surprise that rumours of fast falling subscription sales are perhaps true? ROL also called for Castle's cooperation in Select for Iyonix in early 2004. Did they eventually get it?

More RISC OS websites perform vanishing acts
If you're wondering where riscos.org had disappeared to, then worry no more: its owner, Paul Vigay, says it was down to a glitch in the DNS. He told us earlier today, "I changed the DNS yesterday - ready for the closure of Argonet's DNS servers at the end of the month - and it's not propagated to Orpheus yet." Vigay runs start up ISP Orpheus and is currently end user support guru for Argonet, which will be shortly closing down its ISP business. Also, OreganoUk.net has slipped off the radar, although it's now known as GeneSys following a name change in April this year.

RiscCAD beta testers sought
David Buck, developer of 2D design software RiscCAD, has called for beta testers to test drive his software and suggest new features and other additions. David explained, "Interested parties can look at the RiscCAD website which shows new features currently being implemented for the next release. Currently, additions include extra toolboxes, flood fill with automatic detection of fill boundary, and, at last, bezier curves and revision clouds. Items already suggested include text areas - or multi-line text to make placing notes easier - and area calculation in addition to perimeter."

In brief
Buy Desktop Repton Plus and get a free badge - an official Superior Software Repton badge at that. Apparently at the Wakefield 2005 show, Liquid Silicon told punters that 32bit MIDI modules were in development, although LS haven't confirmed the specifics of this with us yet. MW Software has released AWViewer 1.76, which now supports rainbow blends.

ELS screenshot

Dave Lambley's Internet Printing Protocol (IPP) driver, which allows users to print to network based printer devices, is now up to version 1.04 with new documentation and an installer thrown in. Long time no see for search engine utility RiscSearch, updated with a load of bugfixes and improved decoders. Peter Everett, of KinoAmp and Shine fame, has released "Electric Light Show", a sound-to-light visualisation plugin for DigitalCD (pictured left).

NFS client Sunfish now supports NFS v3 with TCP and UDP protocols - boosting performance over NFS v2 - and is up to version 1.06. Stephen Revill has released SRDPq, a module to aid other programmers who want to use Wimp message queues. Graph plotting software Tau has been tweaked and also moved to a new address. GNU diff front end SideDiff has been updated as has Variations - with substantial updates including an auto-redeye and dodge/burn tools, splitscreen viewer to compare two images side by side and image rotation support for cameras with orientation sensors.

Links
News? Comments? Going to the Big Ben Club show?

Discussion

Viewing threaded comments | View comments unthreaded, listed by date

It's nice to see Drobe directing cynicism somewhere it's well overdue.

As an Iyonix owner and a Select subscriber (two computers, three years), I'm not easily fooled by the movable goalposts currently advertised.

If ROL want my money, they already have plenty of it. I'd avoid offering my opinions about their latest set of proffered excuses.

Equally, I'd avoid spending *any* of my money with some of the people now apparently involved. I would encourage people to make up their own minds - personally, I wouldn't touch any product from these organisations with a bargepole.

dgs

dgs on 12/6/05 2:31AM [ Reply | Report ]

With a business model seemingly focussed on the past is it little wonder that buyers of current machines are reluctant to put their cash on the counter. Even those wanting to support the system and get nothing in return would eventually grow tired of the split and stop stumping up their cash. Until there are definite signs of the forked road drawing closer to meeting again in a single modern RISC OS then little wonder that users of current hardware are reticent to spend money on nothing.

There is a limit to most people's generosity and paying for development that only seems to benefit those moving across to the IBM compatible world of Micro$oft (and now Apple) would stop me contributing.

rmac on 12/6/05 3:23AM [ Reply | Report ]

On the other hand there is no RISC OS 5 for the Risc PC either, let alone for my old A400.

Jaco on 12/6/05 9:31AM [ Reply | Report ]

I would LOVE to buy Select for ROS 5. No one has contacted me, despite the fact I used to be a Select subscriber. Ahh! That old problem of Acorn's ... marketing. When will some company's ever learn - if you want to survive in a difficult market place you have to explore every avenue. Certainly haven't been down my street yet.

DaveW on 12/6/05 9:38AM [ Reply | Report ]

And that should be companies and not company's!!

DaveW on 12/6/05 9:39AM [ Reply | Report ]

ELS looks very nice, although it's a tad stuttery on my SARPC

andypoole on 12/6/05 12:01PM [ Reply | Report ]

Nor is there Windows XP for a 286 or a 386 processor. Suspect that you can't run Mac OS X on a ten year old machine either.

My point was that the RISC OS world needs to look forward not back. How long do people expect support and development for hardware that is 10 years old in an industry where some equipment is virtually obsolete before it reaches the market.

Slowly but surely the RiscPC's and Axxxx models will fall over. If there are not enough Iyonix/Omega/A9home machines around then RISC OS is gone on original hardware. It will only exist in emulation.

rmac on 12/6/05 12:19PM [ Reply | Report ]

I may be in a minority - but I think RO5 works well enough anyway. And my spending on Select would simply divert my limited resources from supporting companies/developers that *are* willing to support Iyonix (e.g., Icon Technology, MW, Castle et al).

Select may well be a "nice to have" on Iyonix but (guess what), the Iyonix does all I *need* it to do, if ROL wanted to get me to even consider spending my money on their wares they'd have to do their port to Iyonix *first*.

Additionally I must admit that I am somewhat more wary now than before, ROL tend to do things to exclude people (DVD-ROM that when they were launched were said not to play on Iyonix . . . .) and documentation that is only available to subscribers (contrast that with RO5 who details are open to all on www.iyonix.com, you don't even have to buy an Iyonix). ROL need to convince people .... and putting arbitrary limits and conditionality and fudge in just doesn't cut it - sorry !!!!

AMS on 12/6/05 3:24PM [ Reply | Report ]

"if ROL wanted to get me to even consider spending my money on their wares they'd have to do their port to Iyonix *first*"

Agreed. The subscription thing is bad enough, but to want money and/or commitment to buy *before* they even have a product to sell is insanity. You make your product, then you sell it.

arenaman on 13/6/05 12:59AM [ Reply | Report ]

I have no problems with the 'subscription' approach, but after 2 years of paying subs as an Iyonix user, I feel it ROLs turn to bring something to the table.

markee174 on 13/6/05 8:09AM [ Reply | Report ]

I have to agree it looks like madness. I seem to remember a long time ago RISC OS Ltd said that if people wanted their version of the OS on Iyonix then they should subscribe. Anyone who has done so could feel a little aggrieved now I suspect.

But like other people said previously, does Iyonix really need it? (I don't have an Iyonix, but most people seem to think it works OK.) If v4 and 5 can be reconciled in a new version in all new hardware, that is probably enough.

Loris on 13/6/05 9:49AM [ Reply | Report ]

I'm a subscriber of select (& Iyonix owner) and I have never been contacted by ROL about wanting select for RO5 so I have no confidence in them doing anything for RO5!

Bassy on 13/6/05 9:53AM [ Reply | Report ]

RO5 is good enough: Except that it possibly isn't. That does not mean that it doesn't work and doesn't work well; it's simply about numbers and potential. One way to improve the chances of RISC OS working into the future is to make things easy for software developers to do two things: produce software for users of all modern machines; and produce more advanced software taking advantage of new provisions in the OS. That's only possible with a single stream OS. I don't see that there can be any possibility of RO5 for all modern machines. RO4 & Select/Adjust is technically possible for A7000s, RiscStations, RiscPCs, Iyonixes, Omegas, A9homes, virtual RPC - and probably whatever the next new piece of hardware happens to be. None of that suggests that you must move to RO4 and RO5 may do all that you want; it does indicate a direction for the future of the platform. It shouldn't be beyond the developers to get together to produce an OS for everyone combining the very best of Select and RO5. What I would agree with is that we need new ideas on marketing! ;-(

jc on 13/6/05 10:24AM [ Reply | Report ]

jc: "it's simply about numbers and potential"

I don't see how. There are currently more RISC OS 5 users than there are Select subscribers (though some people are still both). So it's not crystal clear that a RISC OS 4 derived OS is the way forward - especially with ROL promoting it in the way that's been very badly received in most of the comments above.

"some of the people now apparently involved" (I wrote earlier)

It's been pointed out to me that I should make clear I wasn't referring to Advantage 6 with that comment.

dgs

dgs on 13/6/05 10:37AM [ Reply | Report ]

We wish this fork would just *end* rage

- it was looking possible after wakefield :-(

ROHC on 13/6/05 11:09AM [ Reply | Report ]

In reply to dgs: I feel I should point out that I'm not a Select subscriber, but I am using ROLs latest version, Adjust 4.39. There may also be a number of other people in the same situation.

I suspect that the number of Iyonix users who /would/ buy Select if it were available are quite few in number.

Where would they stand with regards to warrantees if they replace Castles Operating system with one that Castle have nothing to do with? Cheers!

fwibbler on 13/6/05 11:22AM [ Reply | Report ]

In reply to fwibbler:

I wouldn't imagine the warrentee you get with the Iyoinix would be effected. RO Select is softloaded over you existing OS.

sa110 on 13/6/05 12:29PM [ Reply | Report ]

To dgs: I can understand your confusion. Comparing RO5 users and Select subscribers may (or may not) show what you say. But that's not the appropriate equation. Compare how many could use Select/Adjust with how many could use RO5 and the balance dramatically changes. If you can show me that the RO5 numbers approach those likely to be using A7000s, RiscPCs, virtual RO, Omegas, RiscStations, A9homes (and Iyonixes!) over the next year then I'm wrong. If you know of a way that RO5 can run on all those machine then I'm wrong. If RISC OS can cope with two divergent OSs then I'm wrong. If I'm right we can all choose to move to Select/Adjust and software developers can spend time using the new options available with that OS knowing that their customers can make use of that potential. If I'm right the numbers will be sufficient to generate new customers for OS, software and hardware. I do wish dgs would take more care with his criticism though. Whilst it would be right to take ROL to task over their presentation he should appreciate the work they have done over the past year to complete work on 32-bit RO4; the way he has chosen to voice his criticism is damaging to the market as a whole.

To fwibbler: If Select/Adjust is seen as being successful and software is developed to take advantage of the potential in RO4 then I'd be surprised to see any Iyonix users not wanting to opt in to that development. The advantages of one OS are large in what can then be made available to users. Up to recently it has been an internal squabble - we can move that to co-operation to get our OS up to date.

jc on 13/6/05 12:33PM [ Reply | Report ]

jc: "I can understand your confusion"

I'm not sure I can understand yours.

"the way he has chosen to voice his criticism is damaging to the market as a whole"

I'm far from the only one who has voiced criticism. If RISCOS Ltd are to succeed in increasing the uptake of Select/Adjust to a wider variety of users, then they need to understand why their constant fudging of objectives and requirements is so badly received by potential customers who have supported them in the past.

dgs

dgs on 13/6/05 12:53PM [ Reply | Report ]

The problem is the entanglement of the technological benefits of Select / Adjust for all machines with the company who produces it. A merger of the branches, or even just the universal adoption of one and the dropping of the other, is hard to argue against. The worry (not entirely unjustified IMO) is for ROL to deliver, considering that they have appeared in the past to be so scared of trying to look to the future. Perhaps that is unjustified, and they simply don't have the resources to develop for anything other than the RiscPC (or move on from it) without cash up front, but if so then a bit more honesty would put them in a much better light.

SimonC on 13/6/05 1:22PM [ Reply | Report ]

I am an Iyonix user, and will ned to be given a good reason to upgrade. RO 5 is fine, it does what I want and has several features that Select/Adjust doesn't. Since I am happy with RO 5 I will need a good reason to replace it. If there is any possiblilty that any of the RO 5 functionality will get lost in the process, ROL would have to pay me to run Select/Adjust. ISTM that a simply putting the existing Select functionality on the Iyonix is not enough, in fact it is a backward step. What we need is a new OS version that combines the functions of both RO 5 and 4.xx in a backward compatible way. It might involve supporting two alterative APIs for some functions for example. I can't see that ever being achieved unless ROL and Castle are prepared to co-operate in its development.

mrtd on 13/6/05 2:23PM [ Reply | Report ]

Does ROL expect Iyonix owners to give them money when they've not even explained the situation?

How does Iyonix Select cope with the castle usb system, graphics, sound?

If Castle want there to be Select on Iyonix then it will happen and it will be a good thing, without their cooperation it'll just fragment the market even more

For an OS variant to be targetted by developers it's got to be standard on most of the machines sold.

mavhc on 13/6/05 4:15PM [ Reply | Report ]

dgs: I'm sorry if I didn't make it clear enough for you. I was talking about potential and you seemed to think that there were more RO5 users than Select subscribers. That's irrelevant; I was talking about potential. RO4 Select/Adjust can technically (and mostly in actuality) run on A7000s, RiscPCs, RiscStations, Omegas, Iyonixes, A9s and probably the next RISC OS computer (and virtual RO). By orders of magnitude that is greater than the numbers of machines that RO5 can run on AND it includes all the machines that can use RO5. As mrtd rightly says it would make sense for Select/Adjust to include RO5 functionality as far as possible but that will probably require Castle to co-operate with ROL. That Castle have passed the Merlin project on to ROL should be an indicator of possible future (current?) co-operation. If individuals are happy with RO5 then there should be no problem - but it makes sense for software developers to concentrate on adding new features to their software where that is best enabled by the OS and where most future sales are likely to be generated. If that happens to be with Select/Adjust then I hope mrtd won't insist on those developers limiting their new features to the lowest commonality of RO4/RO5; it is essential that RO moves forward.

to SimonC Last November IIRC ROL announced that they had been working to produce a 32-bit happy version of Select/Adjust and Advantage 6 revealed that they were working on a machine that would use this version of Select/Adjust. At Wakefield the A9home was demonstrated and software developers are now testing their work on A9s running 32-bit happy Select/Adjust that can be made available for other hardware developers using ARM processors. How does that square with your comments about ROL being "scared of trying to look to the future", don't "develop for anything other than the RiscPC", &c.? Directly and in co-operation with other developers ROL have enabled a version of RISC OS that can be used with a range of current and future ARM processors, a version that runs in emulation on Windows and Macs, and even a version that runs on the Iyonix. I don't know which ROL you're talking about - but it isn't the one that produces RISC OS! I realisethat you are talking about the past but the 32-bitting didn't come overnight - or even over the last 12 months. It's based on the work that ROL have been doing for 4+ years. Arguably it could have been done faster had a hardware developer put up the money 4 years ago but, now that ROL and A6 have found a way around that obstacle and have produced the goods for the whole market, is not the time for complaints!

jc on 13/6/05 4:24PM [ Reply | Report ]

jc: I was referring to the stance that I perceive ROL of having, of not being willing to develop anything for any new machine speculatively, without being given large wodges of cash to do it. Of course, I've no idea of the agreement between Advantage 6 and ROL, so I don't know anything about why RO Ltd decided to go ahead with that. What I saw in the past is a lack of willingness to move RISC OS beyond what is really legacy hardware, which they would need to do to have more than a short term future, whether or not someone else was going to pay them to do it. I've not picked up much to suggest that that attitude isn't still there. I hope I'm wrong, and considering most of my information comes from Drobe and Usenet no doubt there is an inherent cynical slant to it!

I don't think they had to do anything from a programming point of view to get RISC OS to run under emulation, so that one should be left out of your list.

SimonC on 13/6/05 4:51PM [ Reply | Report ]

jc: I would be very happy to see a combined RO 4/5 for the Iyonix, one that would hopefully be fully compatible with and provide the same features in its versions for other machines like the A9 and Omega at least. But that new version would have to be mainstream. So it should be included with all new machines as the default operating system, and not regarded as an extra cost upgrade on new machines. Otherwise we still have the fork. I have no objection to paying for the upgrade for existing/old machines, provided that the new combined RO becomes the standard version. And yes, it should endeavour to include all of the features of RO 4 and 5, plus maybe some new ones, so that it becomes the definitive and most advanced version of RISC OS available. Anything less than that will probably make things worse.

mrtd on 13/6/05 5:24PM [ Reply | Report ]

jc: You keep going on about developers being more likely to use the feature set of the OS version with the most potential. It is interesting to note, however, that that OS version isn't RISC OS 5, nor is it Select/Adjust. It's RISC OS 3.7. The reason for this is entirely practical - the API provided in 3.7 is present in all subsequent versions (and are available for older versions through freely available API compatibility modules). The fact is, that making use of the Select/Adjust API changes/enhancements is all well and good, but the developer still has to duplicate all that functionality for all other OS versions. Therefore, there's utterly no point in using the new APIs. The changes in RISC OS 5 are minimal and, in the main, deal with underlying hardware changes so, unless the software is needing to make direct hardware access, they don't affect most developers. Until the vast majority of RISC OS users are using Select/Adjust it's highly unlikely that any developer would use the new APIs. The potential market in future is irrelevant, as it doesn't pay the bills _now_ As for merging OS versions, forget it. The distinct lack of movement from either party is evidence enough that nothing's going to happen in that area (I'd love to be proved wrong, but reality suggests otherwise). The best that can be hoped for is API compatibility, but there's already divergence and it's only going to get worse. Next we move on to the issue of API documentation. Documentation for (many of, but not all) the API changes in RISC OS 5 is publically available to all. For Select/Adjust, however, you need to be a Select subscriber to have access to documentation. The position of Paul Middleton at last year's South East show was that the documentation was only available to Select subscribers as the API was a moving target (which may or may not be understandable). I don't, however, see Adjust moving anywhere. As for the issue at hand - that of ROL requiring 100 _new_ Select subscribers before any Iyonix version is produced - fortunately, this doesn't affect me, as I can't afford it, even if I wanted to. However, a pertinent question would be "is ROL willing to provide a money-back guarantee to those new subscribers if the prerequisite 100 is not reached?". If not, then what's the point of any Iyonix user subscribing?

jmb on 13/6/05 6:23PM [ Reply | Report ]

to jmb The current "common point" is RO 3.7 because that is where RO4 and RO5 diverge. many of those wanting and able to afford new hardware and software forced that fork by opting for the only new machine that happened to use RO5 whilst the majority used RO4 in the other fork. Had Castle funded ROL to produce the 32-bit friendly version of RO the minority of consumers pushing development and the majority would both be using the same RO version. We've had the fork but, thanks to ROL and A6, we have the chance to converge and all get onto the same side of the fork. In the meantime we need to push for proper documentation in preparation for that convergence rather than use it as a reason not to converge. And if Iyonix users don't want to risk their money they can simply write to ROL and say they intend to buy Select as soon as it's available for their hardware; that at least would ensure that the demand could not be ignored.

jc on 13/6/05 6:43PM [ Reply | Report ]

jc: Erm. Either I've read this wrong, or you're barking. How can you say with a straight face that the OS fork is the fault of Iyonix _owners_? Presumably, you'd have preferred it if everyone ignored the Iyonix and carried on using their decrepit Risc PCs or waiting for Omega?

As for 3.7 being the common point, it is true (well, not quite, as the divergence began at the 3.8 stage) that it's the last released OS version before the fork. However, I don't see that as being the reason why it's still the most commonly targeted OS version. Even before RO5 appeared, few (if any) developers were making a move towards using new APIs present in Select. The reason for this? The number of Select users was greatly outnumbered by users of other versions of the OS. Not much has changed in this respect, afaics (although the A9 may make some difference here)

The documentation issue is entirely orthogonal to convergence. It is, however, an extremely important issue, which far too many people ignore or dismiss as inconsequential.

Seeing as ROL have been very good at ignoring the demand for Select on Iyonix for the past 3 years (you'll note that the vast majority of features requested for Merlin were already present in Select), I don't really see what writing them a letter stating intent can hope to achieve (especially without this magical figure of 100 new subscribers, which is clearly arbitrary).

jmb on 13/6/05 7:18PM [ Reply | Report ]

to jmb: You're quite right - I did not mean what you read! ;-)

All you Iyonix owners could have refused to buy into RO5 but that probably wouldn't have been a good idea... ...of course it's certainly no fault of the owners but their (very reasonable) action did produce the fork. You mention 3.8 rather than 3.7 but the numbering gets very muddy after 3.7 and there are repeated sets of numbers for quite different versions on different forks. AFAIK no version of any 3.8 was ever officially released so it's safe to ignore it. 3.7 has remained as the common point for software as those pushing for new, faster, better are separated from the majority with 3.7 as the nearest ancestor. With the new, faster, better groups and the majority all able to use the latest version of Select there will be greater pressure for software writers to release new versions using that latest version. I can see "PowerData v2.8 : needs RO4 (full features only available using RO Select/Adjust 4.8) - RO3 users should purchase PowerData v2.2"

Documentation? Essential.

As for writing to ROL - do you really want them to say there is no demand from Iyonix users - or are you going to make sure they have no excuse?

jc on 13/6/05 9:44PM [ Reply | Report ]

jc: RISC OS users can only purchase products that are put in front of them. If you have an issue with the actions of Castle (or anyone else, for that matter) take it up with them. Blaming the users only antagonises people.

Beta versions of 3.8 were made available to registered developers. In the general case, it is safe to ignore it, providing, of course it is acknowledged that the current OS trees split at the 3.8 stage, rather than immediately after 3.7 was released. Not doing so just adds to the confusion

As I said originally, "until the vast majority of RISC OS users are using Select/Adjust it's highly unlikely that any developer would use the new APIs". This is simple economics. Multiple versions of software may occur in the long-term, but in the short term, do you really believe there are enough developers to sustain development of multiple, parallel versions?

As for ROL, I'm not entirely sure they have an excuse. They've moved the goalposts so many times wrt Select on Iyonix that until they have something to offer, I'm not interested in what they have to say.

jmb on 13/6/05 10:39PM [ Reply | Report ]

Please login before posting a comment. Use the form on the right to do so or create a free account.

Create a new account
Forgot your password?

Search this website