| > Writing a program for all of them meant writing it over and over Not quite correct. Companies ported programs, even assembly-language programs, to different processors all the time. Visicalc ran on z-80s, 6502s (on Apples, Ataris and Commodores) and 8086s. the hard part is to write it the first time. Porting then is more or less straightforward. OSs were so minimalistic at that time they didn't create any significant difference. > Businesses were confounded because some of the software they wanted ran only on one stack, some ran on another Most business software for microcomputers in the high 70's and very low 80's ran on CP/M. Even the first DOS blockbusters were straight ports of CP/M titles. CP/M (I am talking about CP/M 80) machines were all based on 8080-like processors, but, besides the common processor, had very different hardware - disks from one computer often could not be read on another. Transferring through serial ports usually solved this. The market that was more fragmented was the home computer market. There Apple IIs, Ataris, TRS-80s and Commodores competed with one another. Choosing a home computer could be confusing at that time. The higher end of the spectrum had mainframes, just as incompatible with each other as they are now (I think Unisys still makes them and they are incompatible with IBMs), minis that ran proprietary OSs or UNIX-based OSs. There were a couple high-end multiuser "supermicros" that ran Unix-like OSs. I have used Cromix for some time. Most of the software was not easily portable at that time because it was written in hand-tuned assembly language. I remember how revolutionary Unix seemed for being written in C and, thus, being portable across different architectures. Faster computers and HLLs (can't believe I am calling C a HLL...) eased the pain of porting. My bet? Were it not for Gates, today we would have a diverse computing environment, with different OSs and processors, with cross-platform software being used in combination with open formats (like GIF was intended to be) to exchange data between non-portable (or non-ported) programs. There would be large software companies but hardware makers would compete with far more latitude than they can now and those software companies would do their best to use the tricks those hardware makers would bring. > You may not like it, but IP suits and lobbying are as much a part of big business as dribbling is basketball You're right. I don't like it. I think software companies should gain market through technical merits that benefit their users, not backstabbing each other. |
My bet for what would have happened: one of two things.
1. Either someone would have done much of the same things Microsoft did and you'd be complaining about them instead. Except maybe you wouldn't be complaining because nobody would have written this article because whoever benefited from it would have hoarded their money as most billionaires do rather than becoming the greatest philanthropist of all time.
2. Hacker News wouldn't exist for you to comment on, because computers would still be something you maybe used at work and that's about it.
I don't think any other result is possible (and I don't think the 2nd is practical). Users simply aren't savvy enough to deal with fragmentation. You wouldn't have the web we know today without cheap home PCs anymore than you'd have man without the apes.