In a completely random universe, if there are N choices, equally probable, the system tends to approach 1/N for each choice. We are seeing that in IT. Of the few most popular operating systems on client devices, web-stats indicate that that other OS with share much greater than 1/N is decreasing rapidly while the challengers generally have positive slope rapidly approaching 1/N. The only real question left is “What is N?”
Clearly, the big play is in smartphones. Android/Linux is king there with iOS pulling in behind. In tablets the competition between Android/Linux and iOS is more even. The desktop shipments have declined for 8 consecutive quarters so that other OS is dropping but GNU/Linux hangs in as does MacOS. That suggests that N will be 5, that other OS or what remains of it, Android/Linux, iOS, MacOS and GNU/Linux. Considering that other OS is declining despite no lack of salesmen and GNU/Linux persists despite a definite lack of salesmen, it’s reasonable to assume salesmen are not definitive. Eventually, OEMs and retailers will awaken to see that other OS is not sacred on the desktop. For the moment, a lowering of margins is all that affects M$’s bottom line, but all along the food chain, folks are realizing that offering more variety is a good thing. This will result eventually in more retail shelf-space and salesmen for GNU/Linux. No one seems to think that Android/Linux makes a great desktop OS. It’s just too limited a concept for large screens and powerful CPUs with huge RAM. ChromeOS is competitive but it’s just another distro of GNU/Linux, IMHO.
Thus, I predict that within a few years, GNU/Linux will have ~20% share of client OS. There has been great progress in government/education to provide diversity but tablets have taken a good share of educational clients and governments are severely locked in to that other OS. It’s the consumer space where GNU/Linux will make huge progress and the key will not be salesmen or retail shelf-space but price. M$ has already recognized that by cutting its price to $15 on the low-end devices which are good enough for most consumers. Given an equal price on low-end devices, will consumers choose an OS with salesmen and ads or will they choose something from a more diverse ecosystem and with less malware, re-re-rebooting and slowing down? I think a lot of consumers will choose GNU/Linux on merits. Consumers have seen Linux run their smartphones. They will be interested to see what it can do on desktops/notebooks/tablets. They will be pleasantly surprised what native code can do. Conversely, all consumers have seen that other OS deteriorate to the point where they wanted to buy a new PC just to replace it.