steve@kontron.UUCP (Steve McIntosh) (05/30/85)
MY theory of software evolution: (and what it is, too!) Recently, there have been many people fretting out load about how the Atari ST will HAVE TO have loads and loads of software available when it hits the streets in order to be successful. Every time I heard this, it provoked an immediate gut level response of "WRONG!" and finally it struck me what felt wrong about it. I present the following statements for your amusement: 1) There has yet to be a successful personal computer come out with more than minimal software at the time it gets on the shelves. 2) Having a set of "canned" software available hurts sales and popularity more than it helps. 3) The REALLY successful personal computers have become so due to third party software and hardware add-on's. My reasoning is thus: Looking back at the REALLY successful PC's (Thats Personal Computer in the generic sense) we have in order of appearance - The Apple II, the TRS-80 the IBM-PC the Commodore 64 and (perhaps) the Apple Mackintosh. With the exception of the Mac, NONE of these machines had much more than a BASIC interpreter available when they first hit the shelves. In the case of the Mac, it's worse. You couldn't even program it, just draw pictures. (I was there when the local Computerland uncrated the first IBM-PC they got for sale, and was helping them line up third party software support, so I know quite well what was ACTUALLY available at the time.) The reason the machines appeared without much in the way of software is that software development costs much more, and takes longer than hardware development. The companies needed to sell some machines to make enough money to afford to pay for software development. When a company develops a computer, they really cannot afford to wait the extra time required to develop software. This is due to the highly competitive nature of the business. Sales windows slam shut. Four examples that come quickly to mind are the Japanese MSX machines, the Coleco ADAM, the Commodore +4 and the Sinclair QL. The Japanese have spent so much time developing software that they think they needed to get into the American market, the market has passed them by. Even a top of the line MSX machine hasn't got a chance of selling if the Atari ST hits even near the quoted prices. The Coleco machine had a full range of your "basic business needs" software including a Letter Quality printer for a reasonable price. Everybody liked the price and the printer, but couldn't stand the software. (At least those associates of mine who, after a few drinks, would confess to having actually purchased one.) ((The Adam bombed.)) The Commodore +4 was to sell to people who weren't interested in games. It also had all the "basic business" programs built in. Too bad you could buy a C64 and whatever programs suited you for less than the cost of the +4. (It has also bombed out) The Sinclair QL comes with a suite of "basic business" programs, and form personal experience, I can say that it is a great little machine. Too bad the sales window is slamming shut on it - the phrase that comes to mind is 'too little too late'. (It may not get a chance to bomb out.) The Buyers want CHOICE! The thing that makes or breaks a PC is choice. Choice comes from third party support. If a machine comes out with a great raft of "approved" or built in software, for better or worse this software becomes the "standard" for that machine. No matter how good or bad the software is. Third party vendors have to sandwich their products in around this software and often meet active hostility from the manufacturer who sees them as unfair competition. (silly but true) This leaves the consumer with little in the way of choice, and throttles third party vendors. A machine that does not have "pre-approved" software gives third party vendors an open field and a much larger market for their programs. After a while, there is a vast choice of programs for users. More and more users buy the machine, because they can get software that they like (instead of what the manufacturer likes) and so even more machines are sold and more vendors support it and so on. == Third Party Support == The ONE factor that makes or breaks a PC is the third party support. A manufacturer just simply does not have the resources to provide everything potential buyers want. The more software and hardware available for your machine, the more likely it will suit someones needs and the more likely a sale. The likelihood of a vendor supporting your machine can be guessed at by how many units they expect to see out in the hands of users. That number can be guessed at by considering the unit price of the machine and the price/performance ratio. (The CRAY has a very good price/performance ratio but there are not very many of them due to a high unit price.) Of course, there are exceptions - the IBM PC has a (fairly) high unit price and a (fairly) low price/performance ratio. It sold because of the three initials on the front. And even that wasn't a sure thing, the current IBM P/X/At line is a fluke - remember the IBM S-100 box? And you do know that the PC failed in its target market - the home - do they still have the cassette and joystick ports on the PC? The Commodore 64 became a big hit because of the usual factors. For one third the price of an Apple II you could get a system that except for the lack of card slots is superior to the Apple II. Even at the initial price of $400 the price/performance ratio was high. (At $200 even better) The ST, IF it comes out near the rumored price and IF it has the features promised and IF Atari can keep from hindering third party vendors will be the Next Big Gold Mine and probably the most popular PC yet to hit the market. The lack of initial software will keep the bean counters from buying up the initial supply, leaving them available for small software houses and kitchen table hackers to buy so they can write the hordes of software the public needs to satisfy its needs. (figure 6 months - or about the time production really ramps up.) Enough time on the soapbox - opposing arguments welcomed (I like mail too!) [The above is One Hackers Opinion and not (yet?) the opinion of anybody or organization I know of. Keep computin' - Steve McIntosh @ Kontron]