jrista said:
At least, they didn't figure it out until much too late in the game. A significant part of that problem was developing the AmigaOS...there were difficulties in developing that for the RISC platform, which lead to very long development cycles, ultimately resulting in AmigaOS falling far behind Windows on the PC. The other problem with Amiga was the simple fact that it WAS built on RISC processors...Motorola RISC processors specifically. When Motorola left that market, Amiga was left high and dry. The only other real option a the time was PA-RISC, but given the difficulties in developing AmigaOS in the first place, a move to PA-RISC ultimately never occurred. Amiga management missed their window of opportunity, their product was selling extremely well in Europe until the bottom simply fell out, and they never really got a solid foothold in the US. Amiga management did not take the PC competition seriously until it was too late, then they were too inflexible, because of processor architecture, poor product design, etc. to be able to compete with the lightning pace at which the PC evolved from the late 80's/early 90's through the early 2000's.
Well I think that is a bit distorted. The initial most serious problem was way back in the Atari days when Atari had the basic tech for the Amiga. The Atari engineers wanted to move forward with the advanced tech they had under the wings, but Atari management said there was no need. We just wanna make more money and keep milking what we have. The Atari 8bits are selling well, we don't need to spend any money, why bother moving things forward?
But imagine if they had listened to the engineers begging them to move forward ASAP. Suddenly this crazy powerful computer drops on the market and actually has a shot despite the dirty tricks of Apple/MS/IBM, it probably would've simply too much for those companies to obfuscate the truth over. Instead by the time Atari lost the tech and CBM managed to snare it and get it finished and out, the MAC had already landed. Now sure the Amiga was better six ways to Sunday, but with the story that Jobs and Gates had woven to the public about how 'toy' companies like Atari and CBM are nothing to be bothered about and how one should pay more money (for less) to get a 'real' computer if one wanted to be 'real' and 'serious' simply having a machine out there with a fancy GUI interface to an OS and a fast 68000 chip inside and the Apple name on front (and Gates name and a new semi-decently pseudo, a little bit fast Intel chip on the other side) they had enough to play the game and, aided by just generally poor CBM management, keep the Amiga (and later the inferior but still better than the other stuff, Atari ST) hidden under the covers as it were. BYTE magazine made a huge deal about the Amiga in a 1985 issue and that was about the only splash the machine was ever allowed to get in the U.S. The power players squashed it and quieted down the press and the sheep made up a large proportion of the computer store salesforce and it it never caught on to the degree it needed to in the U.S. (although it managed to do fairly well in Europe, at least solely as a home computer, in time).
I'm not sure if you want to really call the 68000 a RISC CPU, although it was a bit more like one than say a modern Intel/AMD found in most machines today.
But then even still the engineers wanted more down with the machines and eventually wanted to move to a new updated custom chipset sooner, but dither, dither, dither and eventually they just got a half-baked intermediate chipset out rather late. Still more impressive than the competition, but the war had so been lost by them, they needed the earth-shattering design and needed it sooner.
By the time they eventually ported AmigaOS to the PowerPC chips after the MC680x0 line ran it's course that was so, so late in the game, most of the battle had long been shot. It is a bit unfortunate that the Intel chipsets and the interrupt methods and etc. etc. made porting of certain types of advanced OS over to Intel architecture trickier, it could be done, but would take some time and money and they didn't make a start and didn't spend. Although with fantastic management it's not impossible it may have been able to rise. People did get sick of Windows and Apple and Linux rose up to an extent and the AmigaOS had within itself a lot more promise than Linux (so did BeOS, have a lot more promise than Linux as an alternative). But again, so often in tech, the best doesn't become mainstream or make it.
The PC clones and Apple actually evolved at a snails pace. Just remember Amiga already had a full GUI OS with a power shell interface to an advanced pre-emptive multi-tasking OS already back in 1985. It tooks years upon years for Apple and Microsoft to finally manage to put decent multi-tasking into their OSes. And it took years for the clone hardware makers to finally push past the primitive Apple II-like basic hardware conceptions and move to custom graphics bus and autoconfig hardware systems and advanced DMA controllers and get the mish-mash of sound and graphics third party hardware organized in a way that could be reliably controlled almost as if all machines used the same custom hardware chipsets and it took a lot of power for the non to the metal programming through graphics libraries to overcome the huge speed penalties by not writing straight to the hardware (OTOH the freedom of the non to the metal let a few third party graphics guys then explode forth and have the cash and sales to then drive graphics hardware forward at a terrific rate and at some point that proved to be a bit better than the main maker using sole proprietary hardware system to drive things forward and eventually, now we have a bit of the best of both worlds, lots of the fancy system architecture originally imagined by the original atari/amiga kinda guys mixed with advanced custom chipsets but not proprietary and locked into a single set or two (although today we are basically down to Nvidia or AMD so it's almost proprietary in a sense, but they can drive many different levels of chipsets and old and new all at once through the libraries in uniform fashion which is different).
And it also took crazy fast CPUs to overcome the hideous bloated programming used to produce Windows OS. at one point in time it was said that I think just to do a single task switch Apple OS and Windows OS had to run through 4x and 16x times the code just to do the same thing as in AmigaOS, I forget at his point whether it was Apple or Windows that was the 4x vs the 16x). Heck a 16Mhz 68020 based AmigaOS machine had much of the OS feeling at least as snappy as a 130Mhz Intel Windows box circa 1999 (although obviously stuff like decoding a jpg would be much fast on the 130Mhz machine). And we are stuck with the nasty registry system on Windows, the source of much of the horror and nightmare of a Windows box. And the archaic dynamic library system and some other core components of Windows, especially, but even with Apple/Linux.
However, failing to be competitive because you built a rigid system architecture and did not really recognize your most significant threat until it was too late, is different than purposely gimping your products to "bring your customers back for more in the 'next release' of Product X".
They didn't fail to realize the threat. APple and IBM clones and such were selling better from day 1 and they were a threat from before it ever got released. And they did gimp things. Atari put them off and put them and off and wanted to milk the 8bits and they dribbled out cheap to produce little 8bit updates while sitting on advanced stuff. CBM kept putting off updated the chipsets and carrying out other things and sitting on stuff that should've been done too. And they didn't make a good go at swapping certain architecture elements at the end either.
The former is just bad management...and that does happen. The latter is just plain idiotic and terribly bad economic and business practice, and is GUARANTEED to ruin your company. It would take the most incompetent of management staff to come up with an idea like that, to purposely withhold features in a COMPETITIVE MARKETPLACE with the unproven hope that you'll somehow keep your current customers and bring them, as well as new customers, back for more with the next round of releases. In reality, the exact opposite is going to happen...a competitor is going to leverage your idiocy for their own benefit, and steal all your customers away.
And yet it happens. Atari did that to a huge degree. And they did more or less go out of business. CBM did it to some extent. And they are out of business. Nokia sat on tons patents and ideas. Kodak sat on a lot of stuff.
etc. etc.
Canon is sitting on lots of patents (granted some might take a good deal of money to be able to implement even more than some of the stuff mentioned above would have).
And they certainly play all sorts of silly little games with things like AutoISO and MFA and basic video usability features and it is certainly curious why magic lantern can get radically more detailed video out of the 5D3 than the 5D3 is able to produce with the firmware as shipped (although maybe that is simply down to DIGIC being utterly abysmal at processing images to high quality, it's hard to say).
To be strait up, I DO NOT believe that Canon management is incompetent on that level.
Maybe not to that level and it may take them more expense and risk to implement more stuff than it would have some of the other companies.
But they have even made statements, on video tape and shown on youtube and such, where they have been caught saying stuff like why in the world do we need to bother putting out a high performance FF body, sure we can we are kings, but we are kings so we have no need, Nikon doesn't even have a FF so why do we need to bother, we will sit, we have no need. I mean they obviously could've charged forward back some years ago and just made Nikon look beyond silly, but they played conservative instead (maybe it's just as well though as Nikon might have been barely around by now and maybe with little pressure it would be ages for Canon to ever think about improving DR and such).
Canon isn't a petty corporation. They are not a corporation utterly driven by the short term (if they were, they wouldn't be one of THE MOST innovative companies in the world.)
the little MFA and AutoISO games and such demonstrate a bit otherwise
I don't think Canon is another Amiga...Amiga really had terrible management.
Yeah I'm not saying that and the camera business is a bit more locked in and safe so even acting like that it's harder to get hit like that, plus it's tricky for other companies to buy out press and pull snow jobs over the public when it comes to camera performance the way Apple and IBM and Microfsoft were able to do, so it's certainly quite a different scenario. But there are light hints of it.
However Canon may be too comfortable, they may just be riding the wave of past success...and that could be a problem. (However, that STILL doesn't mean Canon's management is sitting in their corner offices plotting ways they can keep their customers coming back for more by withholding features...that would be SUICIDE for them in the current market environment!)
you can see little hints of the latter though in the dribblings out of certain minor features and talk of well if we do the sensor this time we can hold back on the body performance and if we do the body performance this time we can get away with not spending to go to new fabs yet, etc. and yeah the $$ calculations are different than in the examples above though
and they certainly could've charged the video farther, faster, they really caught the film (movie) guys by storm and they begged for them to charge forward and dominate, but they played the regular game and
anyway I'm certainly not saying that Canon will be a mere shell in a couple years as happened with Atari
i am saying that I don't think it was the engineers who wanted to put in silly little limits on min auto iso shutter speeds and so on and so forth, it's the other guys who tend to order that kinda stuff to be done and in some cases the other also do tell them to sit on stuff so they can milk the current stuff more (it's all a balance, don't sit long enough leave a bit of money on the table, sit way too long and totally blow it long or even semi-short term)
Management people at competitive companies always have endless lists of things they want their engineers to do. It's a never ending parade of "Why isn't this done yet? We demanded that feature six months ago?! And where is the one from six months before that?!?!?" All the while, most of these corporations, all but those with the most ingenious and clever people in management who know how to actually manage and get things done, the management at these companies fails to realize that the continual overload of demands is what ultimately mucks up the works. One week Item A is the top priority, it absolutely without question MUST get done! Of course, the next week, Item B is the top priority, completely trumping Item A, and of course, it absolutely without question MUST get done! And so on, and so forth, week after week, month after month, year after year, companies that can't quite manage their goals, their demands, their production flow...never actually get anything done. Hmm...Atari? Amiga? Kodak? Nokia?
