This is an advantage for Linux over Windows? Buying Windows and all the software needed and dealing with any technical issues will be far cheaper than developing the software again in every scenario. People use computers for work, not to use computers.
For most countries, that is plausible, but not for Cuba.
We're talking about Cuba here. Buying Microsoft software is simply not an option anytime in the next few years; it's illegal under US law for Microsoft to sell to them. Or are you talking about the $1 needed to burn a CD with an unlicensed copy of the Microsoft software on it?
In the case of proprietary software from other countries, from companies that are willing to risk US trade sanctions, Cuba still has to pay for the licenses from its very limited foreign exchange reserves, acquired from its very limited exports of about US$220 per capita per year.
Cuba has about 11 million people, 20% under 14, which means 160 000 of them graduate from high school every year. If its leadership thought software was a priority, they could train the top 3% of each graduating class in software development, roughly the same fraction as programmers in the US work force, about 5000 people. If they were at all successful in this, then starting in 2013 or so, Cuba would have at least a few hundred good hackers every year.
Let's say that one average software developer can replace US$150 000 per year of software consulting services or licensing costs from abroad. (Outside of Cuba, they're paid US$50 000 and billed out to clients at twice that, but aren't as efficient as somebody local who already has both the hacking skills and the domain knowledge.) Under those assumptions, each graduating class provides software that would have cost US$750 million per year to purchase from abroad, if purchasing it from abroad is even possible.
Cuba's total imports are about US$7 billion per year. So in 2022, the value produced by these programmers each year would exceed Cuba's entire current imports. And that's with only 50 000 programmers out of 11 million people. A more aggressive program would convert 3% of the entire population to software development, or 300 000 people.
So it's quite likely that, for Cuba, developing software internally would be far cheaper than trying to buy it abroad.
Of course, the value produced by programming is very hard to predict, as is the success of an educational initiative, as are international politics.
