In terms of defining the “open” in open web annotation, I tend to take a standards approach: the Hypothes.is tool is built upon, and our organization advocates for, open standards in web annotation.
Comment by chrisaldrich: This explanation also highlights an additional idea of open itself. I have heard many in the W3C space criticize the open standard of web annotation arrived at because of the ultimate monoculture of the space. Most of the participants of the process were all related to Hypothes.is in some way and the result was a single product that implemented the standard. To my knowledge no other companies, groups, or individual programmers have separately implemented the standard.
In this sense, while the "standard" is openly defined, it isn't as open as other standards which were mote slowly evolved and implemented gradually and more broadly by various programming languages and disparate groups.