I would appreciate any help in trying to accomplish the following workflow:
Upon finding an article online that I would like to read, I would like to be able to identify it as something that I will read at some future point. This should save the article for viewing later.
Ideally, this should do so in a way that is independent of the original article. In other words, it should make a copy of the article rather than just link to it.
It should provide a mechanism of marking the article that works on both a traditional (desktop/laptop) computer as well as a mobile device.
This could easily take the form of a browser extension on a traditional computer and the use of the standard share functionality on mobile, but other implementations that are similar are acceptable.
In the saved form of the article, it should be possible to highlight text from the article, and annotate those highlighted sections, or otherwise make notes related to them.
Again, this should be possible on both traditional and mobile computers.
The highlighted text and the associated notes should be able to be exported into a separate file for additional editing and/or publication.
Plain text is fine, with markdown/commonmark being more desirable.
Finally, all of this software should be open source and, where applicable, self-hosted.
I’ve been using Obsidian for awhile now for notes, but I haven’t tried any of the plugins to pull in web pages, so not sure which is best or do/don’t work smoothly on mobile.
I used Obsidian for a while based on a recommendation from one of the Youtube channels that I watch. I had previously used Evernote, OneNote, and Notion, but tired of the proprietary nature and gated access to my own data. Unfortunately, I incorrectly assumed that Obsidian was Open Source when it isn’t. I ended up switching to LogSeq. It has many similar features to Obsidian, but it’s not a direct replacement, so I can’t recommend it unreservedly.
Slurp and ReadItLater are specific to Obsidian, so those are unfortunately out. I’ve looked at Omnivore before and it looks promising, I just need to work through how to self-host the server before I can use it in earnest. I might try it out through their hosted offering to see if it solves the challenge of getting article content into a format that I can annotate and prepare for publishing.
That walkthrough doesn’t actually look that bad, compared to some of the things I’ve tried self-hosting. Though it seems a little odd to go through the trouble of self-hosting the app itself, only to rely on a remotely-hosted postgresql and elastic setup.
That, along with a number of features that are not available in this setup (web app, pdf search, email, TTS) make it a poor choice as compared to their hosted setup. I would suspect that they were trying to push people to use their commercial service, but it’s not actually possible to subscribe in a traditional way. The only means of giving them money is via OpenCollective, so it feels like either lack of interest or lack of experience. If I can find or work out a proper compose stack that would provide feature parity, I may give self-hosting a try.
Another area of concern is that some features are currently iOS or MacOS exclusive, and that their Android app is lacking many features. This suggests that they are understaffed and underfunded, which is not encouraging.
I think I’ll stick with their hosted offering for now. My initial experiments last night worked pretty well.
I’m guessing you mean this service: https://dotepub.com/ LogSeq actually has pretty good support for ePub files, which can be handy if you want to mark passages in a book and add notes to them.
I misremembered. LogSeq actually has excellent support for annotating PDFs, but doesn’t support ePub directly. It’s fairly straightforward to convert an ePub to PDF, but that’s less than ideal. In any case, here’s what PDF support looks like:
This doesn’t negate the other flags you mentioned, but it sounds like that particular example was written to exclude some of the other services for simplification, and then they never got around to writing up a more complete example. The github implies that you can start up the web app and other services, though it mentions a need for a google cloud storage bucket for PDF search, and it doesn’t go as in-depth into configuration.
