What about creating a more "pleasant" export format?

[chmac]

Firstly, thanks for this, it's frickin AWESOME!

Seoncd, I'd be happy to put some effort into this idea.

I want to turn the JSON that's mostly useless into something a bit more useful. Maybe in markdown. Potentially with each item being a file in some kind of sensible directory layout. I like CouchSpinner, but I'd rather be able to export and have something in plain text files which I can store and reuse without relying on a specific tool.

If it's something you're interested in supporting, happy to chat and see how we might collaborate on it... :-)

[simison]

@chmac Some kind of "local export" would be awesome! 👍

Any ideas welcome! Feel free to just jump in and implement something. :-) Happy to just give you repo access so that you can push code here.

Personally I'm not super convinced that markdown is best there, because it's still rather technical format and not accessible in readable format to most.

How about HTML? Since the page already is in HTML, it would be more straightforward to output what's already in the browser. Either as working React app, or output of ReactDOMServer.renderToString().

Markdown works too, of course. :-)

[simison]

Another idea could be letting folks download the website app and run it locally. The zip they would download could come with their profile already automatically loaded in.

[chmac]

We could render something into HTML statically and then have folks download that. But doing all of that in the browser seems ambitious. First it loads a zip, then unpacks it, then does all kinds of stuff, then renders react to string in the browser, then creates files, and finally zips up the result. Phew, maybe browsers are better with memory and so on these days than I'm used to, my assumption would be that this is just too much work for a browser, unless we use a webworker or something equally advanced (and beyond my knowledge!).

What I like about markdown is that you don't need anything to read and understand it. I can open the files on my computer in anything that reads text, and see what's happening. I could browse my messages. I could put it into a git repo and view it on my phone. I could copy it to nextcloud and see it from any of my devices. That's a lot harder with HTML. I need to actually run the HTML somehow. I can't do it in a console. I can do it on localhost on a computer, but not really on a phone.

What I had in mind was an output that created some folder hierarchy with plain text files that I could easily view. Maybe if they have images, then looking at it with some kind of markdown viewer would make it cooler. But either way, I could still see it all. Maybe something like friends/simison/index.md and friends/simison/profile.jpg. Or all in one folder, so I can just browse it on my filesystem and see what's going on.

But from my side, not sure how much work I'll put into this. With the hacky scripts I have now captured the data, so my data hoarder has been satisfied. At this point, that might be as much as I end up doing...

[alooshu]

Hi, awesome. and super glad chmac is here. I'm coming from a 'political' rather than a technical perspective, I would like to see this come together, maybe at some point, the data itself, hosted and correlated again, which is how it is most useful. But I'm also looking for bringing the debate to one place, and I'm hoping for a good linkeable summary that could serve for the orientation of users as well as campaigning/trying to get media (incl. social media) attention. I'd love to see opencouchsurfing.org unfrozen, as one option, and in any case, it's old partner irc channel, #opencouchsurfing on freenode, is still existing, currently owned by the freenode-placeholder-account, but useable and potentially, recoverable for active moderation.

[chmac]

@alooshu I think the opencs project is pretty dead by this point. Great that you have motivation and are impassioned about the topic, my suggestion would be to get involved in TrustRoots as its somehow the closest "spiritual successor" to CouchSurfing that I'm aware of...

[alooshu]

there are good reasons why campaigning for not only the (ex-) CS user's right to their data, but to the context they have created, and that this data is useful in, in a platfotm independent fashion, running couchspinner on its own domain, was the right decision, IMO. TrustRoots, at this point, doesn't have a group/forum structure I'd be aware of, BW has, but I still couldn't publicly link the info surfacing in the groups.

opencouchsurfing.org is different in that regard, it has already often helped me to raise attention even in its archived state, and in its day, it was one of the better quality places wrt discussion and info surfacing - there was tim loal's, far too 'mastermind-y', person-centric, small, there was BSR, susceptible to moderation and diluted by what today would be called 'SJWs'.

The opencouchsurfing.org domain, does already exist, even using a different subdomain, a wealth of relevant info would be nearby. The goal, though, would need careful reformulation, you're right in that this wouldn't have much to do with the 'opencs project' in the form in that it existed, and whose goals probably got obsoleted with the CS sale.