In the past few years I noticed a significant speedup on the iteration of tech stack. Meanwhile, I observed that many legacy tech stack (already deprecated by their original authors) are still widely used, but few fresh blood have related skills/willing to learn them.<p>The similar situation of supporting short-lived software for a longer time has been proved to be profitable by Red Hat's business model. I was wondering if there could be similar business for things beyond system libs like old PHP frameworks, perl libs, Ruby on Rails (sorry to mention it here, but it's impacted a lot by Node.js) and even deep learning frameworks in older versions.