I recently read the following comment in the Zotero Forum: "Because I use a very involved annotation system, anything important that I read has to be converted to PDF to support handwritten ink, multicolour highlighter hotkeys, heirarchical [sic] bookmarks, etc. in bluebeam. Currently this is a very involved process ... I would desperately like to simplify this as it acts as a real disincentive for reading!"
There is no need of going into the details of this "involved process." The simple advice is to give it up altogether. Annotating source is at best a supplementary or preliminary tool in note-taking. It marks the things you want to take note of, but it is not a substitute for excerpting and reformulating the material in your own words and organizing it in such a way that it will be useful even years later.
It represents a rather superficial "surf and collect" attitude that mistakes the accumulation of information for the creation of knowledge. I would not go so far as to call it "epistemologically irresponsible," but it is "epistemologically challenged" because it involves the implicit belief that having marked some things somewhere (in disparate places) amounts already to having made connections between them or that annotating materials is sufficient for thinking about them. Well, it is not.
Even such programs as such programs as "Surfulator" represent a better approach to note-taking. They are based on the claim that we need "better ways to collect and manage the worthwhile information we find." This means that we should "permanently save anything you find on the web." While simply "collecting," "managing" and "saving" what one has "found" is rather less than note-taking, it is more than mere annotation.
In my experience, students who highlight their textbooks with markers—often with many different colors—often do not as well in exams and papers as they expect to do. In fact, I believe that the money spent on highlighters is money "thrown out of the window."
1. See also Note-taking versus Information-gathering.