pdf.js+hypothes.is is not architected to be stand alone. Rather, it's basically an example repo to show how pdfs can be used with hypothesis on top of them. Hypothesis is loaded in using embed.js, a small file that basically loads in an app hosted on hypothes.is in a sidebar. That app uses all of the usual hypothes.is code, and isn't easily extendable in this instance (it's all minified and just brought in via script tags).
Instead of integrating with pdf.js+hypothes.is, it would be smarter to see whether we can separate parts of hypothes.is into modules - reusing as much code as possible - and put a different backend on, possibly a couchdb app, so that we aren't forced to go through Hypothes.is's user authentication and through their silo. So, really, this is a non-issue. @jbenet let me know if this makes sense. Spent a couple of hours browserifying pdf.js+hypothes.is yesterday, before I ran into this larger issue.
pdf.js+hypothes.is is not architected to be stand alone. Rather, it's basically an example repo to show how pdfs can be used with hypothesis on top of them. Hypothesis is loaded in using
embed.js, a small file that basically loads in an app hosted on hypothes.is in a sidebar. That app uses all of the usual hypothes.is code, and isn't easily extendable in this instance (it's all minified and just brought in via script tags).Instead of integrating with pdf.js+hypothes.is, it would be smarter to see whether we can separate parts of hypothes.is into modules - reusing as much code as possible - and put a different backend on, possibly a couchdb app, so that we aren't forced to go through Hypothes.is's user authentication and through their silo. So, really, this is a non-issue. @jbenet let me know if this makes sense. Spent a couple of hours browserifying pdf.js+hypothes.is yesterday, before I ran into this larger issue.