* Ihor Radchenko <yanta...@gmail.com> [2020-12-14 15:55]: > Jean Louis <bugs@gnu.support> writes: > > > * Ihor Radchenko <yanta...@gmail.com> [2020-12-13 03:39]: > >> Jean Louis <bugs@gnu.support> writes: > >> I have hypothes.is installed inside docker container locally. No serious > >> protection is required in such case (at least, no more than one would > >> use to protect private files from dangerous software like browsers). > > > > I can install it on VPS which is definitely in plan. Locally I do not > > think so, as locally I have dynamic knowledge repository that may > > export to Org if necessary or accessed by collaborative group of > > people. > > I am actually just trying hyposes.is now (after you reminded me about > it). For me, the main advantage is not for pdfs, but rather the ability > to have pdf-like annotations in web-pages: highlights, comments, etc. > Combined with local ArchiveBox [1] storage, I can get annotations for my > local web archive. > > [1] https://github.com/ArchiveBox/ArchiveBox
I have seen it, good tool and it makes sense to have one's own archive as web pages really disappear. You reminded me of so many references that it helped me streamline my workflows for soon future and new projects. > >> I am not sure how it is different from using hypothes.is for the same > >> purpose. Note that hypothes.is uses pdf fingerprinting, so you don't > >> even need to store pdf on server side. If user can open the pdf > >> (obtained from you directly, for example), hypothes.is will > >> automatically show the up-to-date annotations shared via public > >> hypothes.is instance for that particular user. > > > > The difference is that annotation is separate from file, and there is > > no need for Javascript. Hyperdocument may contain the PDF file and the > > annotation together, dispatched to somebody, or referenced from WWW > > page. It is lightweight. HTML file can be very small and speedy > > loaded. > > Hypothes.is does not store the file - just file fingerprint and > information required to identify and annotation positions within the > file. OK and not that I meant it stores files. I was rather referring to collaborative work within a room or distant servers over VPN where people collaboratively open references to PDF files. Such PDF files can be stored on a local computer, could be fetched from server, but not from public server. This is more privacy issue. Hypothes.is as public online server must have access to files to show the annotation as that implies that for example those 1300 files here would need to be placed online where they by their nature do not belong. They could be placed on a computer within a course room where each student may access them. Hypothes.is as online instance is then useful for those online files, and WWW pages, but the approach of having private archive and then annotating such is even better. Still the hypothes.is is separate dynamic knowledge repository for annotations. Different database, different set of rules but same open hyperdocument project set of principles. So I better stick to one database, not to two. And I just guess that hypothes.is could be invoked from hyperlinks to show annotations even if not stored yet. That would be great feature, to just provide section of text with few hyperlinks where user may start to read the annotation and then open the PDF file to see the context around the annotation. Jean
