This part of the annotation should be put in the metadata for future versions of the exporter

This could be both good and bad.

potential downside: If people only fix things locally, then they may be less inclined/likely to actually/also submit a merge request, and therefore it may be less likely that this actually (ever) gets fixed upstream. Which is kind of ironic, considering the stated goal "No more waiting around for pull requests to be merged and published." But if this obviates the need to create a pull request (does it), then this could backfire / work against that goal.

Requiring someone to fork a repo and push up a fix commit -- although a little extra work compared to just fixing locally -- is actually a good thing overall, for the community/ecosystem.

Ah, good, I see they touched on some of these points in the sections:

• Benefits of patching over forking
• When to fork instead

While I'm reading this, I can't help but wishing that Hypothes.is would add a redaction functionality to their product. They could potentially effect it by using the highlighter functionality, but changing the CSS to have the color shown be the same as that of the (body) text instead of being yellow.

Cool idea how they show which ideas they are considering, planning, and have already launched.

Nice to see hypothes.is off the ground. One thing that genius has that would be welcome is the ability to up/downvote comments. A casual glance at the annotations on this page and you'll see a lot of cruft that should be hidden by default, perhaps the way reddit hides posts below a certain threshold.

The question "So what's the relationship to h?" is a good one here; in particular, where does h end and other services/apps begin? I have quite a few thoughts in this area, including possible h spin-off companies, but my first interest in thinking about integrating it with other services is more from a strategic engineering perspective: what are the best places to focus h development so that it fits that composable unix-y philosophy of "do one thing well"; and I translate that thinking from tool to person... how can h help me do one thing well? As an end-user, even though I am admittedly a power-user with a lot of tools, I actually want to use as few tools as possible. The browser-extension part of h is the single most important part of the project from my end-user perspective-- the back-end infrastructure is there to support the browser-extension doing one thing well.

The one thing I want h to do for me that I can't do with any other tool that I know of is to allow me to rapidly track my reading and thinking and note-taking habits together. I want to be able to quickly select multiple portions of text and apply commentary and tags to the text within particular activity-based or goal-based contexts. The last part of that thought is the essential element I need that is missing. Speeding up the text selection would be very helpful in making it a tool I want to use on a daily basis for everything I do, but the contexts feature is what will make h a killer app for me.

I need a way to add tags to a document that will apply to all annotations in a particular document (except where explicitly canceled).

The problem is that I often want to query all annotations related to a specific document, collection of documents, or type of activity.

Type of activity requires further explanation: Given a document or collection of documents I may annotate the document for different reasons at different times.

For example, while annotating the reading materials, video transcripts, and related documents for the CopyrightX course there are certain types of annotations that may be "bundled together" so that when I search for those things later I can easily narrow my searches to just that subset of annotations; but at the same time I need a way to globally group things together.

While reading judicial opinions the first activity/mode of interaction with a particular document may be to identify the structure of the judicial opinion (the document attached to this annotation describes the parts of the judicial opinion I might want to identify: *caption, case citation, author, facts of the case, law of the case, disposition, concurring and/or dissenting opinions, etc).

The above-described mode I may use for multiple documents in one session related to the course syllabus for the week.

To connect each of these documents together I might add the tags: copyx (my shorthand for the name of the course, CopyrightX), week 1 (how far into the course syllabus), foundations (the subject matter in the syllabus which may span week 1, week 2, etc), judicial opinions (the specific topic I am focused on learning at the moment (may or may not be related to the syllabus).

Later on another day I might update my existing annotations or add new ones when I am preparing to study for an exam. I might add tags like to study, on midterm, on final to mark areas I need to review.

After the exam I might add more tags based on my test score, especially focusing on areas that received a poor score so I can study that section more or, if I missed some sections so didn't study and it resulted in a poor score in that area, add tags to study for later if necessary.

I have many more examples and modes of interaction in mind that I can explain more later, but it all hinges on a rich and flexible tagging system that:

• allows tagging a document once in a way that applies to all annotations in a document
• allows tagging a session once in a way that applies to all annotations in all documents connected to a particular session
• allows tagging a session and/or a document that bundles together new tags added to an annotation (e.g. tags for grammar/spelling, tags for rhetological fallacy classification, etc)
• fast keyboard-based selection of content
• batch selection of annotation areas with incremental filling-- I may want to simply select all the parts of a document to annotate first and then increment through each of those placeholders to fill in tags and commentary
• Mark multiple sections of the document at once to combine into a single annotation
• Excerpting only parts of a text selection, but still carry the surrounding textual context with the excerpt to easily expose the surrounding context when necessary
• A summary view of a document that is the result of remixing parts of the original document with both clarifications or self-containing summary re-writes and/or commentary from the reader
• structural tagging vs content tagging