A friend wrote: “I’m unable to even start comprehending in what kind of world this VisualMeta could make sense. Please apologize if that reads harsh or negative or something, I instead would like to learn and understand it, but from my technical perspective (which you’re free and invited to dismiss/ignore of course), I’m just so puzzled to find a spot to start. It’s not because it’s complex or something, but more of the nature of showing a sailor a solid cube as being the “ship”, and sure, it floats somehow by mere mass displacement, but from just looking at it, it’s not clear what the engineering and operating principles are according to which that thing can navigate the seas.
Again, I really want and have to stress that I’m entirely prepared and available to explore my difficulties, so it might constructively help my understanding. It’s probably not that this is an amazingly new innovative paradigm change, and more of the nature that for many, many years I work on and with that kind of stuff (and aren’t that much of a hard-core expert even) including inventing, implementing, parsing, augmenting some of these myself, so I mostly wonder about the particular decisions and how you would expect it to work. Of course I generally agree with what you’re trying with it and what it is for, sure, there are a few questions around that as well, but even if I wanted to make use of VisualMeta, let’s say for my hypertext book list or the Peeragogy Wraps I’m increasingly producing, I would have a really, really hard time to figure out how to potentially relate these, while there are a bunch of other options which don’t come with that kind of trouble.
There’s so much, to constructively provide one and also important question: the “Visual-Meta puts this metadata into an Appendix” is most likely not JSON at all, it’s so not JSON if I didn’t miss a major, world changing update which for some reason went unnoticed to me. So OK, I could imagine this is what’s merely printed on a paper page, but then, if that’s the case, and there’s actual JSON in the data file for the machine to read and another different variant for the human reader, which is not machine-readable, then why have any kind of strange symbols in the version for the human at all instead of just English? These are things that are very strange to look at, I think I never have seen anybody doing something like that, and that’s not because of innovation I fear, but it appears as if it’s designed to not work (I can be entirely wrong of course, and then there’s probably a good reason or explaination). Another constructive text for this question: have you or anybody else tried to deserialize that VisualMeta example? In what kind of computing environment are these things expected to be read successful? In only your own?
I now feel terribly sorry, because I end up in this kind of dilemma, of either staying away form discussing VisualMeta, or on the other hand are forced to describe some questions that might be read dismissive, while to the contrary it is again and again an attempt to learn and understand how things are intended to work, also because I genuinely want all these hypertext things to work, but there always needs to be some explaination, how. Apologies for asking, I’m just really, really dumb in that regard :-(”
And this is my response:
Why ‘Meta’ would be a surprising question. I think you know why meta is useful, considering your background and I also think you are familiar with the potential for richly interactive text and hypertext in specific.
The ‘Visual’ part then is quite easy to explain: Meta in the document at a data level, and not on the same level as the content, can get lost over time, particularly as the document format gets revised or lost in time. Visual-Meta solves this by keeping all the metadata and special formatting instructions on the same level as the text itself so that if the PDF document survives in a basic form or is even converted to plain text or printed and later scanned, there will be no loss of metadata. That is the essential point. Look at the graveyard of old document formats and look at the huge expense many have invested in new document formats. Visual-Meta takes an established format (PDF) and crudely (to be robust), bolts on new functionality.
As for JOSN or BibTeX or anything else, I really don’t have a preference. I chose BibTeX since it’s a well known standard in academia and we use JSON for headings as it seems a good fit. It truly doesn’t matter–as long as a person (or soon, AI) can read the Visual-Meta and understand it, they can employ it/deserialize it in code for the users benefit. I already do this with the Future of Text book, which is created in InDesign and has hand-coded Visual-Meta which works in Reader.
Does this address your concern for whether it’s useful?