This is the third of three articles published in the Wakefield RISC OS Computing Club newsletter last year. The articles focussed on three key years for the Acorn Archimedes and its successors - their original launch (with the Arthur OS) in 1987, the introduction of the A3000 (and RISC OS) in 1989, and finally, the belated attempt at the wider UK home computer market with the A3010 in 1992.
But what if...? (cue wibbly wobbly screen effect...)
1992
Despite the growing popularity of games consoles such as the Sega Mega Drive and Nintendo SNES, Commodore are going just great, at least volume wise, with the UK arm expecting total Amiga sales of 300,000 machines that year. A600 sales are a bit meh, and certainly not at the numbers of its predecessor. The A500 averaged 25k units per month, the new kid on the block only 16k per month between May and August of '92 (5). With the announcement of the A1200 in the UK on October 21st, the Amiga has a much needed tech upgrade, especially pertinent since the core A600 spec is dated and not far removed from 1987's A500. That model takes a price cut from £399 to £299, and legacy high end models see prices slashed: the base A3000 can now be had from just £1,299 instead to its previous price of £2,999! (6) It all looks great, except for one thing...
The A1200 is delayed, with only limited shipments reaching the UK by December. A combination of quality control issues and a failure to contract manufacturing in Scotland due to a liquidity squeeze within the main company means that instead of the expected 70k units, Commodore UK has just 500 to sell before Christmas. The presence of the new model, alongside some remaining stock of the much more expandable A500 Plus (a model whose production ended months prior), means that the A600 withers, with fewer than 50,000 machines making it off the shelves in the three months running up to the festive period. The A1200 situation improves slightly in the new year, but missing the all important Christmas period is a huge blow not only to Commodore UK, but also to the parent company.
Atari are unable to capitalise of their traditional foe's woes. Their next generation machine, the Falcon, also misses the Christmas window. That it also happens to be nearly twice the price of the A1200 matters not. For that amount of money, you are also looking at a reasonable 386 DOS/Windows PC. Not even a pre-emptive price cut of the steadily aging STE to £199 can save them. For less than that, you can buy a SNES or Mega Drive if gaming is your thing. There are even plans to re-introduce the practically archaic STFM into production for an RRP of £159, with wildly ambitious/what the hell were they drinking sales figures of 150,000 mentioned. Never mind that they'd stopped making the STFM back in early '92 (7).
Acorn, however, do step in to the breech. The A3010 is a 1992 Christmas success. A solid effort by Dixons and other retailers desperate for some sort of cheap-ish home computer gives the machine a high street push, whilst advertising in ST Format (8), CU Amiga (9) and Computer and Video Games (10) targets those looking to either buy their first home computer (but not wanting re-heated mid-80's spec kit) or those wanting to upgrade said kit but can't buy it for neither love nor money. It's not huge numbers compared to the competition's usual figures, mind you, some 120,000 units by the time the dust settles(11), but it's enough to dent even Commodore's seeming omnipotence. Even the negativity of having an order backlog in the New Year doesn't hurt, with some mildly successful spinning of the shortages adding to the computer's allure.
1993
The fallout from Christmas 1992 is apparent almost immediately. There is a rush of ports from traditional Amiga and ST developers to RISC OS as another source of games buyers arrives on the market. At a time when selling 10,000 copies of a game meant it was a big hit (and even 5,000 was acceptable) (12), getting even the latter figure on top of existing sales (and on a format that is relatively starved of games) is worth chasing. It's also a less capital intensive publishing environment than the console market. Following on from that, publishers begin to include RISC OS versions of new titles, some even taking advantage of the platform's capabilities, especially 256-colour graphics, aiding many a PC conversion. Much is also made of the fact that the Acorn's pack a 1.6Mb high density floppy which certainly helps with multi-disk releases.
Despite a rush of new sales and the aforementioned £159 STFM, Atari's bullish talk is beginning to sound like marketing bollocks. 100,000 Falcons (13) in 1993 is as believable as the 150k new build STFM's, and as the summer draws to a close, quality control and supply issues mean that only 7,500 Falcons are in the hands of users (14). The financials for Q2 fail to mask just how far the company has fallen: net sales of $5.7m leading to a net loss of $6.6m compared to the same quarter in 1992 (15). Q2 was always expected to be a quiet business period, but even so, the silver lining of $35m in cash reserves doesn't hide the fact that it won't take long to burn through that. The company focus shifts to the Falcon and their new console, the Jaguar. Even that doesn't go well, with only around 1,000 consoles hitting the UK in December (16). The STFM revival is never mentioned again.
Commodore, well, it's just a shit show and a half. The results for 1992 are dire. Revenue drops from that billion dollar mountain to $725m, and the corresponding operating loss of $86m stuns pretty much everyone outside of the C-Suite. The decision is made (once again) to double down on the Amiga, with every effort going towards the A1200/A4000 duo, as well as the CD32 console. That doesn't go well either. Having sat for too long on the A500/B2000 platform, the company are way behind the technology curve and having to push to make up for it. Already in a rushed development cycle, and with ongoing arguments about whether to hold off the launch until 1994 as per the planned roadmap (17) in order to have a decent software line up or scramble for whatever revenue they can get, the console is announced in May 1993, with a trickle of production units reaching stores that September.
It's a disaster. What units that make it to stores do sell, but there is a dearth of CD32 specific games. A1200 ports make up most of the announced titles, yet there are precious few of those as the A1200, already struggling, now faces competition from its console brother as well as the A3010 and the PC. Not that this matters, as by November 1993, Commodore's problems finally catch up with it. The parent company is declared bankrupt after finally running out of cash and the disposal sale is completed early the following year.
Acorn, however, are holding their own. A revitalised software stream and steady sales means that their financials announced mid-93 show a profit of £12.6m from a turnover of £72.1m, and a growing realisation that the platform isn't just for games. Windows PC's are a threat, but even now, a decent spec'd 386 is still the best part of a grand. The decision to standardise on a 2MB A3010 model in mid-93 pays off for Acorn, as does a price cut to £349. Margins are tighter, both for dealers and Acorn itself, but in the UK at least, sales of the 32-bit machine rockets passed the 350,000 mark by the end of the year.
1994
As the new year hangovers clear in 1994, there is much for fans of RISC OS to rejoice, some good news for Amiga users, and little comfort for ST owners.
Atari are desperate and the fire-sale of STFM's and STE's adds pittance to the coffers. The Falcon hits £495 but there is little software available and the company's eyes its dwindling resources as it tries to make a success of the Jaguar. Despite lip service to the contrary, Atari as a computer manufacturer is all but done. The computer range is sold off and, in an effort to define the term "Forlorn Hope", the last roll of the dice is made for the Jaguar.
There's better news for Commodore fans as the UK management team buyout succeeds in gaining control of what assets are left. A crash program is put in place, doubling down on the CD32 and the A1200, whose base price for Commodore manufactured stock hits £249, whilst the last of the A600's go for £149. The A300 is announced, coming in at £249 for an AGA-equipped CD-capable (an add-on drive appears from third party suppliers for CD32 compatibility). The CD32 continues to sell for £249, and a new licensing program is launched to bolster revenue. At the higher-end, the Amiga 4000 is retained. The future looks promising, as long as they can sell hardware, and there is bullish talk of a next-gen chipset. Paging Mr Osborne...
Acorn begin the year caught somewhere between shock and swaggering relief. The A3010 is selling as fast as they can make them, forcing some cannibalisation of the A3020 supply. That doesn't affect school sales too much, with the (Acorn) A4000 picking up the slack despite initially being advertised purely for secondary schools. A brief diversion in to making a home console is considered before sense returns to the room and a proposal is made to support the 3DO. This isn't as daft as it sounds, as the launch of the Risc PC in April means that there is now an expandable, CD-equipped tower running an ARM 610, for £1,500. Ouch for the pocket, but a hell of an upgrade for early A3010 adopters, aided by a trade-in scheme to push existing owners towards the higher spec. The presence of a guest CPU slot (and some fancy engineering) makes the 3DO card possible, and in November 1994, RISC OS users have the option of playing the (limited but slowly growing) range of 32-bit console titles.
However, the company isn't resting on its laurels, as it is apparent that the A3010 will struggle in 1995, not least because it will fall foul of new electronics legislation. They are also aware that the traditional market for home computers is slowly disappearing. The rise of the Windows PC, initially 386 and now 486-powered, has altered the dynamic of buying a computer. Budgets now stretch into the low four figures for a capable gaming PC, and the market is awash with consoles. There is also an acceptance that your bog-standard television is of no use as a serious computer display and that, just like the PC, future machines will require a dedicated monitor. As 1994 concludes, the consequences of this realisation reshape the home computer arena once more.
Acorn strike with two ARM 7500-equipped desktops. One offers a desktop case with space for a CD-ROM, but includes a hard drive and monitor as standard. The other matches the styling of the A3010, but with included hard drive and TV/VGA connectivity. An external CD-ROM available for £150 proves a popular add-on. Existing owners have a clear upgrade path, whilst developers have hardware that can sustain ports from the PC world as well as console conversions. Indeed, throughout 1994 and 1995, RISC OS and the 3DO see a cross-pollination of titles. Rather than just Star Fighter 3000, there is enough software support to benefit both platforms. The lack of a dedicated 3DO card for the A7000 series does hurt the concept, but remains a handy upselling tactic for the Risc PC. Not enough, of course, to save 3DO from the Sega Saturn or Sony's PlayStation, but every little helps.
 |
| What could have been... |
To boost sales, there is a trade in scheme that takes £150 off a new A7000-series machine, with the older computers handed in re-cycled for emerging markets. Acorn now have a trio of machines aimed at the low, mid and high level user, and news of a laptop based on an existing Olivetti design (18) demonstrates that the once majority shareholder still supports it's British off-shoot, and profits of £15.5m from a turnover of £85m are ample evidence of the company's success.
1995
After another brisk Christmas period where Acorn announce their 1 millionth RISC OS machine sold, it becomes apparent that the pace of sales is beginning to slow. The breakneck speed of the PC's advance has meant that by the summer, even the billy basic 486 can be had for the same as the low-end A7000 CD-equipped model, and as Windows 95 arrives, it becomes clear that Acorn no longer have the upper hand when it comes to a price/power/capability comparison. There is little despondency though as the platform appears to have more than enough momentum at present.
The Risc PC700 arrives to pep up the hardware offerings, and discussions begin about the future of the platform. The new laptop proves a hit, and whilst there are signs that the company is beginning to lose the education market to the Wintel combination, the relatively buoyant home market in the UK brings in the money. It is acknowledged though that this is a shrinking percentage of an ever growing pie of general users. Even so, profits of £13.2m off £76m revenue are a reason to celebrate.
Less good news for Amiga fans as "Neo-Commodore" closes its doors in September. Despite valiant efforts with the CD32, it becomes clear that it isn't a competitor in the wider console wars, selling less than half the numbers of even Atari's ill-fated Jaguar. Against the new offerings from Sony and Sega, it isn't even a joke. There are rumours of a successor which would beat the newcomers, but time (and finance) isn't on the their side. The A300 similarly failed to sell against the vibrant Acorn machines, and despite a licensing scheme that saw the Amiga brand on everything from mugs to towels (the PJ set was the nadir), the company's biggest problem was that it just could sell enough computers. It didn't help that the AGA tech underpinning the range was only just competitive in 1992 and looking increasingly antiquated in 1995. Talk of more advanced machines dampened enthusiasm for the current range, and squeezed by more successful consoles at the low end, Acorn in the middle, and by PC's at the top, it was doubtful even with that £50m investment fund they could have succeeded.
1996
It has been a good run, but the seasonal sales figures for 1995 come as a bit of a disappointment to the Acorn board. The company has still sold some 70,000 machines as a good proportion of adopters from 1992/3 decide to upgrade, but there were some heady sales projections being bandied about in the autumn prior. Acorn's share price takes a hit, and a company-wide re-organisation is put in place. The primary consequence is the complete decoupling of chip designer ARM. What's left of Acorn is then organised into two distinct divisions: Workstations, which continues the hardware business, and Open Acorn, aiming to further develop and expand the reach of RISC OS. The latter hints at early talks of the "Network Computer" leading to rumours of an Acorn designed box with a custom RISC OS ROM, whilst the NewsPAD, leading a European initiative for a portable multi-media viewer, becomes something of a minor success.
Development of the successor to the Risc PC is begun, and there is an urgency about the project as the venerable slice-based desktop began life with a 33MHz bus and the latest StrongARM processors run at 233MHz, hampering its performance. Project Phoebe promises a revised internal architecture and more developed StrongARM processors, as well as utilising current PC standards. The four PCI slots will come in useful...
With the internal reorganisation and increased research and development, although revenue dips to £65m, profits are down to £1.2m, albeit with much ink spilled proclaiming it's all about investing in the future.
1997
There is an air of optimism abound in the world of RISC OS as the launch date for Phoebe is set as September, just in time for the European Computer Trade Show at Olympia. And there is even better news. After a chance encounter at the show (19), it is soon announced that 3dfx will support RISC OS and, indeed, by the Christmas of 1997, the first Voodoo graphics card arrives. In the short term, this places the Acorn range on a par (in capability if not quite in price) with regards gaming compared to the Windows PC, the consequence being that there is a minor surge in PC ports to the system.
 |
| Ironically from the issue of Acorn User that said it was cancelled. |
It's not all good news though as the successor to the A7000 series (imaginatively called the A8000) are announced, again with two models, and it's almost instantly apparent that the base machine is being ignored. There isn't much of a requirement for a hobbled £500 general computer that really shouldn't be plugged into a TV, and the better spec pushes the price close to £900 that also includes a monitor. There is a general shedding of users as they jump to Windows and, later, the revitalised Mac, but enough remain on RISC OS to make Phoebe, at least, a relative sales success.
1998
Financial results for 1997 reveal some good and bad news. The A8000 is a flop, whereas Phoebe pushes past 50,000 units. Revenue is down to £51m although profits are a relatively healthier £4.1m, reflective of the high end desktop sale and a steady push of RISC OS into embedded markets. The last of the tail end early 90's adopters upgrades to the latest hardware and it's becoming clear that the heady days of six figure sales numbers per annum are long gone. To those in the know, there is a general feeling of managed decline creeping in, and before the end of the year, the A8000 is cut to a single spec only, just to retain a sub-£1,000 machine.
1999 to 2002
And managed decline it is, as although there are processor upgrades for Phoebe, there is no replacement desktop planned. The failure of the Network Computer initiative isn't a stunning blow to Acorn, and the decision is made to sell off the set-top box unit to Pace Plc, also giving them a licence for the OS. It becomes apparent that the future of the company is more assured via software rather than hardware as the company posts its first annual loss in nearly a decade in 2000. In early 2001, what's left of the hardware team is disbanded, and the user base declines further into enthusiast niche territory. New machines do arrive from small scale manufacturers who can cater for sales in the hundreds, and it's only as the third year of the new millennium arrives that the last of the large games publishers abandons the platform, leaving the leisure market to solo and indie teams.
As for the OS, in early 2002, the decision is made to open source RISC OS. There isn't the revenue to support even the cut down Acorn that remains, and it at least assures the future of the operating system as there is renewed interest in a free, low footprint OS.
(wibbly wobbly screen effect)
Plausible scenario or complete bull hooks? It was reported at the time of Acorn's sudden hardware volte face in 1998 that the core business had essentially lost money since 1993, so any success with the A3010 would have had an impact. I don't believe it would have been enough to see off the competition coming from PC or Apple (post 1998), but it might have changed the circumstances of the company's withdrawal from desktop hardware. It would also have affected moves like the co-founding of Xemplar and the cosying up to Oracle. Maybe...
When it comes to counterfactuals, your original decision point, the locus of change from actual history to the alternate version, has to be believable. So it is that I have chosen Christmas 1992 as that point. Without the success of 92's festive season, Commodore would have collapsed even sooner, and given that almost every computer manufacturer has suffered from quality control issues and shipping delays then, well, why not? The company was fleet of foot when it came to financial efforts, sometimes desperately so, but it's not too implausible that they might have struggled to pay for components and set up a manufacturing contract in mid-92. It truly was a make or break event. The success of the A1200 briefly rejuvenated the Amiga range and bolstered the coffers of the financially mis-managed company, giving it breathing room to continue the CD32's development and sustain the Amiga brand. And yet it mattered not a jot past 1993.
According to David Pleasance (20), the decision to launch the CD32 in that year was made purely for revenue reasons and it hampered sales of the A1200. The issue here that even with the CD32, Commodore was failing, the console merely extending its death scene. That and without the 1993 release, there might not have been a Commodore left in April 1994 to launch anything. Nor would the CD32 have been the silver bullet even if it had made it to the US. The 1994 console market was rammed with new hardware: the 3DO and Jaguar were already out, Sega's Saturn and Sony's PlayStation were incoming, and I can't help but think that the CD32 would have looked extremely tired in comparison by the end of that year.
With regards to the era after the collapse of Commodore, I'd like to think the UK management buyout could have worked, albeit for a limited period. The CD32 would have needed a replacement within a year or two, and the A1200 wasn't exactly cutting edge in 1992. It most definitely wasn't when Amiga Technologies were trying to flog them. Would the new UK/Germany-centric company have had the resources for the full-scale development of both a new console and an affordable home computer? It's not beyond the bounds of reason, and maybe David Pleasance will tell us in his latest book, but having experienced the "joys" of the Kickstarter journey for From Vultures to Vampires" (the tale of the publication of those three volumes is almost as convoluted as the story they tell!), I will not be finding out that way, at least not yet. And if you have seen the updates on Kickstarter, you'll know about the one dated 6th September, he's already running behind and "we might be late reaching the printing stage." Cynical me now awaits the announcement that due to the staggering amount of material, a second volume priced at £35 plus postage will be coming...
For Acorn, it's 1990's behaviour smacked of that must abused term (guilty, your honour) "managed decline", which is why it comes into use as the decade closes in the counterfactual. The ubiquity of x86 hardware and Microsoft's business practices meant that the DOS/Windows compatible was the general computing platform of the decade. Add in SVGA, soundcards, CD-ROMs etc, although you'd spend more money, you also gained greater capability. Even Acorn noticed this with the introduction of the A7000 range - these were priced similarly to low and mid-range PC's, following the pattern that the entry price point for a home computer had moved from a few hundred to nearly a grand, and you had to include a monitor as resolutions had advanced way beyond what your standard TV could handle. You could spend less, but that could often be false economy. The RISC OS high end was catered for by the Risc PC which for a short period could go toe to toe with the PC. However, the rapid development in the PC world meant that window of opportunity was exceedingly small. It would also have taken a degree of prescience few people have for their management team to make the "right" decisions every time.
As the decade closed, only Apple remained as an alternative to the Windows PC, rejuvenated as it was by the iMac (as well as a change of leadership. $150m from Microsoft and the timely sale of some ARM shares), demonstrating what a combination of in-house design and careful tech choices could achieve (excluding that pucking mouse - not a typo!). Acorn had proven in could do the same (my affection for the Risc PC design is well known, at least by regular readers), but Acorn was too small, and even the once hugely successful Apple flirted outrageously with financial destruction as the decade progressed.
Anyway, none of this really matters now except as the occasional day dream and pondering. I have more important things to do like maybe fire up my RISCOSbits PiRO Noir or Pinebook Pro and enjoy the best of what RISC OS still has to offer. Might I humbly suggest that you give the OS a try too...
1) ST Format Issue 35 - June 1992
2) www.dfarq.homeip.net/commodore-financial-history-1978-1994/
3) Acorn User Issue 119 - June 1992
4) Acorn User Issue 150 - Christmas 1994
5) Commodore: The Final Years
6) Amiga Format Issue 38 - September 1992
7) ST Format Issue 46 - May 1993 - in this counterfactual, Atari pretty much continue as they did.
8) ST Format Issue 40 - November 1992
9) CU Amiga Issue 32 - October 1992
10) Computer and Video Games Issue 132 - November 1992
12) Amiga Format Issue 53 - December 1993
13) ST Format Issue 46 - May 1993
14) ST Format Issue 50 - September 1993
15) ST Format Issue 52 - November 1993
16) ST Format Issue 58 - May 1994
17) Amiga Addict Issue 24
18) This would be similar to the Stork design from 1996 but two years earlier.
19) Interview with Andrew Rawnsley of R-Comp Interactive 04/02/2023
20) Amiga Addict Issue 24
