[PeterWhittaker]

Well, it may be a case of what Jobs referred to in his quote about being able to connect the dots only when looking backwards, that in the present moment, one has to hope that the dots will connect.

What he refers to as IBM's RedHat strategy may have been their motivation, but it's hard to discount the value of a proven, growing revenue stream; they may simply have thought "hey, revenue, plus some possibilities!" Likewise commoditizing the PC peripheral market: I could buy it, if they had made big money in that market. The fact they went on to spend gazilladollars on OS/2 (which had its moments) and got M$ to write it (and gave Gates an opportunity to have his engineers learn the 386 in detail while not providing any transferable knowledge to IBM, since OS/2 was written in assembler...) makes me think Joel was cherry picking.

That said, it is an interesting argument: If you do a thing that makes something else desirable AND you can make decent margins on the other things, especially if they are consumables or have a repeating revenue stream, it may be worth quite a lot to do the thing.

I'm not wholly convinced, but I am intrigued.