Here is the washington post doing this with TikTok users to reverse engineer the algorithm!! https://thewashingtonpost.formstack.com/forms/help_investiga...
they've got data from 800 users so far, with watch data on 55 million videos
Twitter shut off API access; users volunteering their own data for an open API
91–100 of 106 posts
Re: Twitter shut off API access; users volunteering their own data for an open API
#92Earlier quoted context omitted.
Not parent, but I did use Mastodon for about a year. I ended up moving back to Twitter because: 1. I just couldn't vibe with the culture there. From my POV, Mastodon is made out of pearl-clutchers and politics. 2. So much drama. The FediSearch drama. The Raspberry Pi incident. It's just so tiring and you feel like you need to walk on eggshells all the time. 3. So much drama. You would just pray that your admin didn't…
>4. So much drama. Pray that your server doesn't shut down, because you can't import your posts elsewhere. I would have thought a well-designed decentralized system would allow you to, for instance, download/export all your posts, and then import them all to a new server, in case you have to move servers for some reason.
The server-oriented aspect of Mastodon was the second-worst decision ever (the first one was the name). Mainstream social media doesn't ask you to pick a server (nor deal with the consequences of it going down), its replacement shouldn't either.
Re: Twitter shut off API access; users volunteering their own data for an open API
#93Earlier quoted context omitted.
>4. So much drama. Pray that your server doesn't shut down, because you can't import your posts elsewhere. I would have thought a well-designed decentralized system would allow you to, for instance, download/export all your posts, and then import them all to a new server, in case you have to move servers for some reason.
The server-oriented aspect of Mastodon was the second-worst decision ever (the first one was the name). Mainstream social media doesn't ask you to pick a server (nor deal with the consequences of it going down), its replacement shouldn't either.
I agree about the name: they shouldn't have stolen the name of a great metal band.
Re: Twitter shut off API access; users volunteering their own data for an open API
#94Here's a thought: someone "trustworthy" should maintain a Chrome extension or Tapermonkey script that automatically scrapes data from various social media sites in a fully anonymized fashion. As people browse Twitter, Reddit, or XYZ, the posts/comments are sent to some aggregation system. It might be against TOS, but certainly far less than scraping, and you couldn't tell, as it's the user driving what gets scraped.…
Reminds me a little of RECAP ( https://free.law/recap ), an automated scraper/saver/sharer for PACER (the US court electronic records system). Obviously the content is very different, but the technology is basically doing what you’re talking about, minus anonymizing the data.
I use RECAP constantly and try to always upload everything I grab from PACER.
I think this administration finally passed a law to make PACER free, which is good.
Re: Twitter shut off API access; users volunteering their own data for an open API
#95I hate to be the one that says this, but what's to stop someone poisoning this and uploading a file of someone else's fake tweets?
Re: Twitter shut off API access; users volunteering their own data for an open API
#96Earlier quoted context omitted.
Reminds me a little of RECAP ( https://free.law/recap ), an automated scraper/saver/sharer for PACER (the US court electronic records system). Obviously the content is very different, but the technology is basically doing what you’re talking about, minus anonymizing the data.
I use RECAP constantly and try to always upload everything I grab from PACER. I think this administration finally passed a law to make PACER free, which is good.
Re: Twitter shut off API access; users volunteering their own data for an open API
#97Earlier quoted context omitted.
Yeah, something generic to work for any use case would be nice, but privacy becomes more difficult as you need to tailor the situation to each site to maintain privacy (i.e. only pull the information on the apartment/flat listing, or public tweet or reddit comment, etc...)
I wonder if those websites as a response might start adding fingerprint code in the source to indicate whose user session it was.
The city of Berlin wouldn't care. The housing website would probably start with lawyers.
Re: Twitter shut off API access; users volunteering their own data for an open API
#98Earlier quoted context omitted.
>4. So much drama. Pray that your server doesn't shut down, because you can't import your posts elsewhere. I would have thought a well-designed decentralized system would allow you to, for instance, download/export all your posts, and then import them all to a new server, in case you have to move servers for some reason.
The server-oriented aspect of Mastodon was the second-worst decision ever (the first one was the name). Mainstream social media doesn't ask you to pick a server (nor deal with the consequences of it going down), its replacement shouldn't either.
As for the server-oriented aspect, what's the alternative? There's only two ways I see to run a social media service like this: centralized and decentralized.
The mainstream ones are centralized, of course: FB, Xitter, etc. You just connect to their web server and do everything there. Any centralized alternative would have be a well-financed corporation able to set up a huge IT infrastructure to accomplish the same task. But the whole idea of Mastodon is to not be a big, for-profit corporation, and to be in the hands of the users instead, which I don't think is possible with a centralized service.
So the only alternative is a decentralized service. But this means you have to have multiple servers, leading to the whole "federated" approach, so users can run their own servers and control things themselves, or use accounts on servers of their choice. This of course leads to all the consequences you named, and more: servers going down, servers defederating other servers, etc.
But again, what's the alternative? If the requirement is "not run by a single big for-profit company", I don't see one. If you think that having a decentralized service actually be competitive with a centralized one like Xitter simply isn't feasible, that's fair.
My prior complaint still stands though: if you want users to have freedom (of which server to put their account on, etc.), then why would you not have a way of migrating their account/data to another server, and force them to lose everything if their server disappears?
Re: Twitter shut off API access; users volunteering their own data for an open API
#99Re: Twitter shut off API access; users volunteering their own data for an open API
#100I hate to be the one that says this, but what's to stop someone poisoning this and uploading a file of someone else's fake tweets?
The system should require multiple identical copies uploaded independently.
The "independently" part is hard, could require some kind of web of trust.