The RISC OS Time Machine - the rescue that never was
We re-visit the Peter Bondar rescue plan for Acorn Computers...
In the aftermath of the collapse of Acorn Computers in September 1998 there were a number of proposed rescue plans being considered. One of these was being fronted by Peter Bondar. Richard Hallas managed to get hold of Peter on his mobile phone on the 6th of October to ask some questions. Later that evening Richard spoke to Peter again, whilst Peter was reading an e-mail from Acorn confirming some of the details of the proposed rescue bid.
This interview makes fascinating reading and gives you some idea of what was going on behind the scenes in the closing months of 1998 in the months leading up to the formation of RISCOS Ltd. What is also interesting is the number of accurate predictions made by Peter in his responses to Richard. The meeting referred to in the second answer took place in Cambridge and was organised very quickly. There were around 100 Acorn dealers and developers present, including Aaron and Dave Holden. At this meeting Peter Bondar presented the basics of his plan, which involved the community paying him £10,000 to talk to Acorn. Luckily a number of dealers and developers decided to by-pass Peter and to approach Acorn themselves, the eventual result was RISCOS Ltd...
Peter Bondar - the Interview
How did you first hear of Acorn's decision to close its Workstations division, and how did you get involved in the plan to rescue it?
I decided to meet one of my other long-gone ex-colleagues to have a quiet discussion about what we should do next, because in fact he decided he was going to finish the work he was doing with another small company, and the two of us got together to decide what we should try and do, if anything, together, and we decided to meet on a Thursday lunchtime in a quiet pub on the outskirts of Cambridge, where we wouldn't be pestered. This guy, by the way, is Chris Stott, a software engineering manager for Acorn about three or four years ago, who went to work for Tau, the operating systems people. Anyway, we decided to meet, and when we got to this pub, we found it full of Acorn people, about fifty or sixty of them, and one of them said to me as I walked through the door, "Give us a job, then!" and I laughed because I thought he'd seen the two of us and put two and two together or something. And then one of the other guys came up to me and said, "Have you heard?" And I said, "Heard what?" Then he said, "We all got fired this morning," and showed me the letter; and that's the story.
A meeting was arranged between Acorn developers in the week following Acorn's decision. Were you involved with that from the start?
No, all the credit for that goes to Mr Todd. From my own viewpoint, the chain of events was that, having heard about this, I thought I'd go and subscribe to the newsgroups again. Having never expected to be involved in Acorn again, I'd de-subscribed from them, so I watched the furore break over the newsgroups. And then someone was saying about Peter Bondar in the middle of this newsgroup. So I thought, well, there's no point in getting involved in this if I'm seen as part of the reason why it all collapsed, or if I'm seen as part of the enemy, so I gently posted something that said, well, I am here, but would it matter? Would you care? Or something. And then there was this slightly unrealistic/sycophantic outburst, but I thought, well, at least my name isn't mud with the majority. Then I observed the various discussions that people were having about the way forward, and there was a naivety caused by people who (a) don't understand business processes, and (b) don't understand the politics taking place within Acorn. So, I thought, if I can help people salvage something from this wreckage, I'd be happy to give some words of advice, so I contacted Bob Todd, and he said, well, why don't I come to the meeting and explain my views on things? I was invited to the meeting, but it was very much Bob Todd's idea, and it was in that meeting that I presented my analysis of the situation, and that's when we got into discussions about me being involved on a full-time basis.
So, what is your analysis of the situation?
My analysis was that, for political reasons, the product business had for a long time not fitted into the strategic direction as perceived by the major shareholders and non-executive directors, and under David Lee's and my tenancy, we had allowed it a high degree of room to manoeuvre as long as we could show it to be profitable. Under Stan Boland's regime it was no longer to be tolerated because it didn't fit the plan, and Stan was focused on a very directed and specific target, which was part of the argument between the directors as to what the strategy should be. David Lee and I always argued for a portfolio basket' of activities, with different levels of risk and rewards. Stan decided to bank everything on the future of this Digital TV revolution and opportunity, so in that sense I wasn't surprised at what he'd done. The surprise was that he decided, at all times, then. I'd expected him to have done it within days of David Lee and myself leaving; or alternatively, at the very least, to get Poebe out and get some sales back to recover some of the research and development that had been put into it. But I knew the writing was on the wall when I saw, at the half-year, that he had written off the R&D on Poebe, as part of the write-down at the half-year. That's the classical financial indicator that says, this investment is a liability.'
What plan did you put forward at the meeting?
Well, the first step is that the Acorn market is not large enough to stand several fragmented groups operating in it (how true - ED); any plan would have to be unified and coordinated, so my suggestion to everybody was that for negotiating reasons with Acorn, we had to be unified and coordinated, otherwise Acorn might either (a) do nothing, or (b) simply do a fire-sale, which resulted in the assets being disposed of in a haphazard fashion, and the whole thing disappearing into the ground like water into the sand. (At this stage Acorn had already sold some of the rights to various parties - ED) So that was part one. Part two was to suggest the creation of a new start-up company that would take the assets, get Phoebe out and basically build a business, based around products, that didn't have the political overhang caused by a group of investors and non-executive directors who didn't want to participate in a product-based business.
How have the discussions with Acorn gone, and what is the position now?
Well, we met them on Friday and presented them with a 26-point Memorandum of Understanding (MoU). They agreed 24 of them; two of them they said they wanted to come back and make a counter-proposal on, and in a few minutes' time, when I get home, hopefully I will have that counter-proposal waiting for me. It'll be one of three things: either broadly they'll fine-tune what we've done, in which case we can quickly find an MoU, hopefully tomorrow, which says that the details are published within 48 hours. Or they could come back with such financially stupid numbers that it basically kills the deal. Or they could come back and be slightly greedy but be subject to some negotiation. In fact, I'll be home before we finish the interview, so I'll fire my computer up and we'll see what the email says!
Where is the financial backing coming from?
There are a number of different options, but the one that I recommended to the steering group (the team that was organising the meetings, at this point it consisted of a number of Acorn dealers including Paul Middleton and Chris Evans - ED) at this stage is to focus on a specific one, and what we have is a series of Dutch venture capitalists who are interested in coming in. By introducing a Dutch product-based manufacturer as a majority shareholder, that's where the initial financial impetus would come in, subject to our ability to then raise the rest of the finances which would come from 'the great unwashed'. So the major chunk of short-term finances would come from this Dutch venture capitalist acting on behalf of a product-based manufacturer (Tulip Computers - ED). So the plan is that, assuming the MoU does get signed, then we would bring the Dutch VC and the Dutch product manufacturer in to see Acorn early next week, and basically finalise the deal. And then the idea would be to offer shares in the company to existing Acorn people and to other people who want to come in on a start-up type basis.
Is the Phoebe technologically ready for release, or does it require any further development effort?
It requires on a relative scale a relatively minor amount of development effort (actually it needed a lot more work than Acorn had been prepared to let on, there were significant flaws in the design - ED), so it's not ready for release yet, but the effort measured in percentage terms is probably only another two or three percent to go. I've done an unofficial walk-through with the engineers; the belief is that, obviously provided we have the necessary code and everything else, that we could finish that off and release it to manufacturing within another four weeks from now (six months to a year would have been more realistic - ED).
Are you effectively going to buy RISC OS from Acorn, or will it involve some sort of partnership?
Well, the original idea was that we were going to buy the code from Acorn on what's technically described as a non-exclusive basis, but on the understanding that they wouldn't be using it in any product-related activities apart from digital television. And the deal, broadly speaking, is still the same, but the twist in the tail from the Friday meeting was that Acorn suggested that they actually take an equity position in this new company. That was an amusing thing, I thought! So we will have a non-exclusive worldwide agreement, so we can basically do whatever we like with it, and Acorn can do whatever they like with their version, and if we want to work together we can, and if we want to go our separate ways then so be it. My belief is that Acorn's strategy is not to consider RISC OS in anything but the short term anyway.
Is it true that there are problems with the IOMD chip?
Yes, in the sense that every chip has problems; but the problems are truly of a minor nature in most areas, and could typically be worked around with software or relatively minor extra hardware (the design of the chip had inherent faults, a re-design wasn't considered cost effective, so what was being considered was a bodge - ED). That's the view that everyone involved in the project, before it stopped, had of the situation. One of the ironies was that they only got the chip back two days before the project was cancelled, so given that it takes a while to get a chip debugged or worked around, for two days' effort, it's in, from what I think most anoraks' perspective would be, astonishingly good shape.
Assuming all goes to plan, when would you expect Phoebe to be launched, and how much will it cost?
Well, I think we're still trying to get the first ones into the shops before Christmas. I think our provisional plan is to have an Acorn show in early December somewhere, (the show went ahead in Birmingham and was organised by The Arm Club -ED) eventually launch the product at that show, and either be in a position to sell some, obviously a limited quantity, there, or if not, at the very least dispatch the first ones just before Christmas. We believe that's realistic because of the fact that there are sufficiently large numbers of long lead-time parts already in existence, so it's not as if we've got to buy a whole pile of bits and plug it together. Pricing is looking like being around the £1500 mark; effectively the same as it would have been previously. Packaging and pricing are not yet decided. One of my own ideas is that, if we can, we should actually ship every product with two CPUs, to start off with anyway, which will get the market going straight away for developing applications that can use that second CPU (something that's now common but would have been ground breaking at the time -ED)
Is that something you definitely expect to do?
It's my thought, and if I'm the boss, I'll do it! Technically there's no reason why it can't be done; the reality is that it'll be of little value without any additional software, but my belief is that if we can get a significant number of Phoebes out there with two processors, then people will work out what to do with it if they've got it.
Will you be using the 233MHz part, or will you have access to a faster processor?
There may be faster parts. It's just one of those things: whatever's going fastest at the time. At the moment it's 233MHz, but because of the way it's taken out of Acorn's hands and put in a smaller company, we may well take the risk of running them at 266MHz anyway, or maybe a bit faster, even if it's a standard part. And, in addition, there are faster StrongARMs coming out anyway.
There has been talk of the StrongARM 2 and 3, but they are incompatible with RISC OS. Is that correct?
Yes, they're still just dreams on bits of paper, but in the way they're envisaged, certainly relative to the current operating system stack they're incompatible (because they lacked a 26bit mode - ED), but what has to be done to make them compatible is a job for the new engineering team. Again, we rewind history back, and the viewpoint when StrongARM 1 came out was that because of its Harvard cache architecture and the relative speeds of the buses, there were dire predictions about performance and compatibility, but it didn't work out like that in the end.
What are the plans for the new company once Phoebe has been launched? Assuming you own RISC OS, do you plan to develop it further, and if so in what ways?
What we'd like to do is basically develop or refine and evolve some new low-end machines for the education market, where I believe either a better A7000 or a more in-tune network computer could be well received (which is what RISCStation decided to do - ED). So that's plan one; plan two is to license RISC OS relatively freely to other developers who want to develop additional RISC OS-based hardware. There are a couple of suggestions around from other people. Then there are people like Forbidden Technologies' Mr Streater discussing the road map forward for RISC OS and its successors. So we'll be a product-oriented company rather than a licensing-oriented company. The idea is to make RISC OS work on all the hardware that we envisage developing, which would involve StrongARM 2-based systems and ARM7100, etc., and to add development tools around it, so it's a case of basically working with the existing Acorn community to take it forward.
There has been talk of a RISC OS PCI card to put in PCs and Macs. Is that a good idea, or do you think it could damage sales of Phoebe machines?
No, I don't see it as threatening. People buy the products for a variety of reasons, and I think that most Phoebes would go to a group of people who want a machine; not simply to run RISC OS per se. For those people, a PCI card wouldn't be of interest. For the ones who want to run RISC OS, but aren't bothered about the machine, then a PCI card might have some use. If a third party wants to make a card, and can find justification for it, then so be it; certainly, if the new company was in a position to be able to grant licenses and rights to allow it to happen, then it would do so willingly and freely. (the RISC OS on a PCI card proposal was the basis for VirtualAcorn, I thought it was a great idea, but the cost of the hardware was prohibitive, hence the idea of doing the entire job in software - ED)
Another idea is to develop a RISC OS-style front-end to Linux. What do you make of that idea?
My view is that Linux represents a much larger opportunity outside of what you might call the RISC OS anorak community, and that's one reason why I'd like to work with Stephen Streater (EIDOS - ED) to see if we can build what might be described as attractive geek hardware that happens to have a RISC OS API sitting on top of a Linux multi-tasking, multi-processor kernel. That may well be an interesting option, and certainly one I'd like to explore. Again, that's something that under the old regime couldn't even be really entertained.
If you are producing Phoebe in the next few months, who do you expect to sell it to in the immediate future, other than the core enthusiasts?
Well, basically there are 45,000 Risc PC users (more like 30,000 RiscPC and A7000's - ED), and another say 50,000 users of A5000s and their predecessors (a very optimistic figure - ED), who represent a legitimate target for such a product, and the plan would be to market to that group of people, who are related to the Clan anyway. The other thing is obviously to persuade the vertical market people who are currently using Risc PCs that they should take these things; Eidos, AVS Graphics, etc. etc.
What will the company be called?
It's our desire and intention to call it Applied Risc Technologies, and the version of the MoU that we have already signed with Acorn allows us to use that name, subject to the final contract.
Who will the company initially employ? Will most of the former Workstations people be involved?
It's going to have a mixture of permanent and contract staff initially. It's likely to involve some people who are still within Acorn at the moment, with Acorn's permission, and it's likely to involve some ex-Acorn people as well.
Is the new machine still going to be called Phoebe?
My view is that we should paint it black, with an orange badge, and call it Risc PC II.
What happens to the Clan?
We hope to take the Clan over and continue with it, as it will be a key part of our communications with the market.
Do you intend to pursue new markets?
Yes; as I say, the discussion about going back into things like the geek market on a broader basis is one of the options open, and that's part of the discussions with people like Stephen Streater and others; ideas how to do that. I think trying to get RISC OS per se into a wider geek market could prove diffcult, but a RISC OS API on top of what's seen as something that's more politically acceptable, such as Linux, is an option.
How successful do you predict Acorn itself will be, now that it has shed its Workstations division?
Well, my personal view is that they've embarked on a strategy of extremely high risk, where the rewards may be significant but the risks are also extremely large. (as we saw, the only company that made significant profit from Acorn's demise was Broadcom - ED)
Finally, what about the news on the Memorandum of Understanding that you mentioned earlier?
Well, I'm reading the agreement while talking to you, now that I've got back in, and it doesn't appear that stupid, so I think we could be on with a deal here!