Standardization is a special ingredient that makes human intelligence possible. Standards are the structure onto which consensus clings. They are the scaffolding that hold language as meaningful, turn the wildest mathematics into domestic shorthand, and transmute humble science into monolithic Theory. Computer science is no exception: it is more standards than substance.
Standardization understands that there is diversity in the universe. However, to communicate information, we need to build tunnels of consensus between unique nodes. Standard and consensus are sympoietic: they make-with each other. Standards are born from consensus, but they also shape present and future consensuses.
Consensus is a byproduct of shared exposure to the same universe. It is enabled by a common ground beneath our feet, mediated by sense organs, and shared into reinforcement through communication.
Standardization is a byproduct of society and civilization. It stands as an object of organization created by a group of people. It implies not only community, but compact and enforcement. Those who exist outside of a set of standards risk exile from the institution that created them.
The difference between standards and consensus is that standards are policed. Some group needs to make decisions about how consensus shapes a standard. More importantly, a group needs to see that standard is enforced. Policing can be organic: there are standards in our daily life that are unwritten, we all co-police these standards, built on millennia of bio-social consensus around what makes group-life possible. But we mostly care about the standards that are policed through bureaucracy.
There is a special type of revolutionary dissonance when the standards and the consensus deviate. The policing party has no ethical, moral, or epistemic gravity to legitimize enforcement. It can lead to the adoption of a new standard.
Because standards and consensus are sympoietic, a schism in standardization can forgive schisms in consensus. This is something to avoid at all cost: a schism in consensus is epistemic dissonance, and looks like confusion over truth. This post-truth world that we live in is a symptom of divided consensuses and of multiple mutually exclusive standards.
Remembrance is a tool to shape standard adoption in a way that does not shatter working consensus. By building remembrance into a new standard, an adoption is possible that does not alienate those on the periphery of consensus. By keeping a body-of-consensus as large as possible around a new set of standards, the designers can maximize the quality of that standard as a communication pipeline, the policers have widespread democratic backing, and the core epistemic structures supporting the standards are truer and stronger.
Failure to remember results in standardization that is increasingly narrow. A narrowly adopted standard is self-defeating and doesn't do the important job of rallying diversity around agreement to get work done.
Apple's premature abandonment of USB-A for "Thunderbolt 3" is a wholesale assault on remembrance, even if it is the true future standard. The result is a schism of consensus that disrupts daily life. Seen in isolation, it is an act of minor inconvenience, and a transgression remedied by consumer choice in a free market. Seen in context, it is yet another assault on group consensus. A jettisoning of rememberance to make way for a factional sub-standard that excludes more than it includes. We see here the material embodiment of the Post-Truth Era, and all of it's anti-democratic consequences.