All this talk of GNU/Linux not making it on the desktop is hypothetical. Where GNU/Linux was tried it has done well. In Portugal, some locally-built PCs were produced in several models. One of them had GNU/Linux and because of that had a lower price for software and better hardware. The result? It earned a decent share of the market, 10%. So, the fools who proclaim GNU/Linux has only 1% share due to geeks miss the effect of barring GNU/Linux from retail shelves, something totally on the supply-side. Consumers will choose GNU/Linux if it is offered.
- bundling of software and hardware with sales heavily dominated by the hardware retail channel
- connection between suppliers and consumers through an oligopoly of retailers
We should add another fact: the world wide Linux desktop market share, which includes both desktops and laptops, is nearly 1.5%, even though it is very impractical, if not impossible, to buy it preinstalled owing to the aforementioned properties 1 and 2. It was, however, tested on the Portuguese market via e.iniciativas, a Government project for the distribution of laptops to students, teachers and other citizens under state sponsored training. The project delivered around 600000 computers from 5 manufacturers, via 4 network operators who agreed to handle the logistics while supplying 3g Internet connections. Only one laptop model, made available by a single operator, had Linux preinstalled. This somewhat low profile presence of Linux was an imposition from the Portuguese Government. Given the much lower price per unit of the software,the aforementioned model had better hardware than its competitors. Even though the computer was supplied by a local, relatively unknown assembler and the software was not, and still isn’t, mainstream, the final market share of the Linux-based solution on the project was 10%. It is reasonable to think that having a similar option on additional models might have resulted in a larger share.
This remarkable achievement provides strong empirical evidence of the relevance of Linux on laptop computers. We should also like to add the results of an economic impact study that objectively deems this product viable for the entire supply chain, up to the consumer. In fact, due to the difference in software costs, there may be as much as 70 EUR per unit for a mixture of customer savings, retail margin improvement and hardware upgrade. A healthier market would, therefore, have a positive impact on the local IT economy”
The ~1.5% stuff is bogus, too, coming from a different oligopily on statistics.. Without running a parallel universe for the test, that’s the best you can get. It’s what I have been saying for years. The world can and does make its own software. It works. It’s GNU/Linux.
see Retail Oligopoly.