I FIGURED OUT THE ATOM FEED (sort of)

[EssExTee]

If you can navigate to a personal blog (____.kinja.com) you'll be greeted with this page. Copy and paste the link it gives you into your Atom/RSS archiver. So for The Railyard it was:

https://therailyard.kinja.com/rss/export?t=6c53e767-34f6-4ffb-89b6-39766c65eac8-0-8528bda7780f105fcdcc60773f38f7d6db6e40f8

Unfortunately there's some caveats. I can't get it to work with users (kinja.com/____), just personal blogs. And this may just be my RSS reader but it's not downloading images, and on some posts it seems to skip the text completely and just gives me a link to open the page in a browser.

This isn't really useful for archiving right now but it's a start and I'll keep experimenting.

[FTTOHG]

You're at about the same spot I got to. It seems pretty typical for an RSS feed that the longer stuff just becomes a link... but maybe there is something more in there. Hopefully one of our mods is still in touch with someone on the Kinja developer side who can get us some info later today.

[EssExTee]

@fttohg they said the nerds had it "working well" so hopefully they have a full solution that will be shared with us at some point

[Jb boin]

Good news, the API seems to still work, for example : https://kinja.com/ajax/comments/views/flatReplies/1845683946?startIndex=0&maxReturned=5000&approvedOnly=false&cache=true&sorting=oldest

[Stef Schrader]

That's good. I wonder if it's just a front-end shut-off somehow? The data's up, but it's hidden because Jimmy Bedwetter over there has no spine?

[FTTOHG]

@jb-boin cool! ...how do we use it to actually archive things??

[Jb boin]

@fttohg well, the whole post and it's comments are in that JSON, you need to do some "code" to make it look like a post...

And download the images listed on it that are on Kinja servers before they might get deleted.

[FTTOHG]

@jb-boin thanks. I’ll have to play with it some after work assuming the content is still actually on the server by then. The images I think can mostly go... the comments and vote counts are what I’m really after.

[Jb boin]

Just in case, i included this JSON on the articles i backuped (only the ones between a few days ago and march 2018 for Oppo unfortunately) : http://jbboin.phpnet.org/oppo/extractor/archive/

[Jb boin]

And peoples might have done backups using the tool from Nantenne (@Discord) that is saving the JSON and can fetch images from posts (which i did plan to do on my script only once after i would have fetched the posts themselve... in a few days...) : https://github.com/crose7/Undertaker

[FTTOHG]

@jb-boin OMG thank you looks like a decent amount of my stuff made it into there... any easy way to download my folder from there for my local drive???

[Jb boin]

Well, there are many different ways, what is your login?

[FTTOHG]

It was alphaass - basically I'd like to pull a local copy of everything in here: http://jbboin.phpnet.org/oppo/extractor/archive/oppositelock/alphaass/

Looks like it even has the comments in the JSON which is awesome... maybe someday I'll be able to parse them out into something readable.

Any chance I can use that Undertaker thing to fetch older posts? Looks like your archive grabbed a lot of mine but there are still some to go and if that JSON is still there then maybe I can grab them. Sorry for so many questions - I know a little about programming but most of my day job is MATLAB and C and Javascript is not what I'm good at.

[Jb boin]

I only took up until march 2018 on Oppo unfortunately.

[Jb boin]

@fttohg http://jbboin.phpnet.org/oppo/extractor/archive/oppositelock/alphaass/alphaass.tar.gz

[FTTOHG]

Awesome thank you so much. Anything you need from me? Would a tool to better reconstruct these into something that looks like the full page with comments be helpful? Also... any idea if we can use the API to search back further?

[Jb boin]

I haven't re-checked but from what i saw with the atom feed and that the API seemed to be still working, it's possible to fetch the JSONs and images but re-create the pages as they were like on my archive are not realistic (at least if i would be the one doing it).

Re-creating comments would be easier as there is less complexity on what was possible (and on what was posted) but it's still a lot of work.

[FTTOHG]

10-4 Gotcha. I may still make it a project to build a script that at least somewhat approximates a page with comments as it was on Kinja Oppo. Obviously the formatting won't be perfect but I have some pages saved with the Save Page WE extension that I can use as a guide for parsing the JSON. THANKS AGAIN SO MUCH FOR THE DOWNLOAD!! For me it only went back to June 2018 (+ 2 random posts from 2017) but that was enough to get to a couple of posts that I really wanted back including a rental review.

[doodon2whls]

@stef-schrader said

Jimmy Bedwetter

I LOL'd