Bookmate VS PublicData

Create a list of bookmarks linking to websites you find interesting, and publish it for the world to see...The model is as recursive as it is simple. There is nothing preventing a list of bookmarks from linking to another list of bookmarks...The creation of a bookmark list is a surprisingly fun project.

I agree. I've often thought of people publishing a list of bookmarks in a way that everyone can see. I even created DownloadNet originally based on this idea. I wanted a way to publish one of my bookmark folders as a server for people.

But then, as so often happens, the simple idea evolved, and I got carried away by who knows what (technical challenges? I don't know) and ended up creating a personal archive and search engine with only a scant integration with bookmarks.

This article is a good reminder of what originally seemed to me a good idea. Perhaps I should add it there. Also, perhaps p2p could be an easy way to federate these things? Not everyone can just create their own server, nor do they want to host it on big providers always.

I've been tossing around the idea of p2p as a way to "solve" this, but it's still rather formless: new and vague. Over the last 3 days I created a p2p blog (and again, got carried away -- perhaps with technical challenges -- and added p2p chat). But I think there's something there.

Perhaps I should listen to that idea that keeps recurring for me. To that first version of it anyway.

Something simple, that unifies, publishing a bookmark folder (I have some chrome bookmark reading code^0), over p2p (I have janus^1), and possibly uses either the popularity of DownloadNet, or even some of the search/archiving stuff -- without getting carried away -- to assist in delivery or marketing.

I don't know. A clear synthesis right now escapes me, but that's OK. I think there's something there: bookmarks (maybe a special bookmark folder, something referential, like "/var/www/html"), into which bookmarks go and then become public; a lightweight p2p server (that perhaps in some limit future could be federates effortlessly for p2p discovery, but who knows how?). Ugh...still too complex perhaps.

Bookmark folder + p2p + transitive (my bookmark folder includes a link to another person's bookmark folder ~~ somehow).

So it's like that article recently on the homepage "We need webrings" or sth. I didn't think that was particularly a good idea, but now I see at least a partial appeal.

The "link" to another person's p2p bookmark "folder" will instead be a normal www hyperlink that links to the "signalling access point" where you can do the ritual to make the connection.

People may think the weirdness, unavailability (you have to be running the little service in your terminal or as a daemon), and difficulty makes it a non-starter. But I think these "backward" elements, could be a paradoxical strength.

I don't know. I think there's something there. I definitely want to keep pushing in this direction, anyhow.

0: https://github.com/00000o1/Bookmate

Yeah, they're selected based on something like "vibe". I've tried to make it a bit chaotic with a few wildcards and unexpected results, without the pure pandemonium you'd get with no curation at all.

The data set is available here: https://github.com/MarginaliaSearch/PublicData/blob/master/s...

I haven't built a clean workflow for that yet, but in the interim, make a pull request here:

https://github.com/MarginaliaSearch/PublicData/blob/master/s...

and I'll poke it into the DB.

If you don't want to dirty your hands with github, you can send me an email at [email protected] :-)

Not parent, but Marginalia's list[0] of blogs is a great aggregation. I've also stumbled upon some webrings through some of the blogs (have them open somewhere among my 3000 open tabs haha).

[0]: https://github.com/MarginaliaSearch/PublicData/blob/master/s...