No.44749
(By AI translate)
I have a solution, and I will now demonstrate the new approach.
I have subscribed to a certain creator on Patreon, and I can use a browser plugin/software provided by Kemono (hereafter referred to as Kemono Downloader (KD)) for batch downloading updates from the creator. At this point, KD will choose whether to upload the resources and update the page data based on my privacy settings.
Implementation Idea (using Patreon as an example below):
Stateless Data:
The server will create a simplified post model (service, post_id, title, published) for different providers.
Whenever the KD sends an upload request, the server will visit the specified URL to gather this information and compare it with the database for data initialization/updates.
Stateful Data:
The stateful data will be collected by the KD (the uploader), and the scope of data collected in one go is determined by the uploader.
Stateful data for a single post includes: image resources, attachments, comments, etc.
There are further details about the implementation, issues, and points to note, which I won’t go into here.
If the admin sees this and has any ideas, feel free to DM me at pikzyb@hotmail.com. If I don’t respond, please check your spam folder.
Just my personal suggestion
No.44751
Additional information:
The following is the excerpt from the beginning
I believe everyone has been troubled by the inability to use the Fanbox and Patreon importers. I have a temporary solution, or rather, a brand-new permanent solution.
Is Kemono fundamentally different from other resource sites? Yes, Kemono offers complete, highly credible, and traceable content (resources) from OnlyFans.
How do traditional websites obtain resources? That's right, through the selfless uploads of numerous users.
If Kemono also accepts user-uploaded resources, wouldn't that solve the issue the importers cannot handle? However, this would make it difficult for Kemono to assess the reliability of the resources, and the cost of maintaining these resources would be very high. Plus, Kemono doesn't just provide resources; it also offers comments, DMs, etc.
No.45400
>>44751This is a nice idea, but again it would rely on a "trust system"; which i'm pretty sure we are all in agreement.
No.45426
That's a bit too convoluted. Can I just send my payment info and my ID card to you instead?
No.46461
This would create a possibility for interference on the side of the client. The plugin/application solution is the way to go, yes, but it should rather move the burden of trust to the user rather than the service. That is, have the plugin provide credentials to Kemono and act as a proxy so that the Kemono server can connect to the service using the client's IP and scrape everything as if it were the client doing it from their machine. Since client MITM-ing the Kemono scraping process would result in a TLS handshake error rather than corrupted data having the potential to be submitted to Kemono, this is more optimal.
No.46462
>>46461Additionally, this plugin/application could just be open sourced and built from source so the users aren't concerned about Kemono admin stealing their shit. It wouldn't be that complicated to make. Just have a Wireguard (or otherwise) tunnel to Kemono main server and on the side of the backend just add logic for proxified scraping.
No.46463
So you don't know english but plan to implement a totally brand new idea (manual uploads)?
Also this stateless/stateful distinction is pretty retarded. How are posts, literal entities in any database of worth btw (aka dynamic stateful values), "stateless" but comments are not?
No.46610
>>46463The fundamental difference between stateful and stateless is whether user data is needed.The kemono server can also add credibility to evaluate a post/resource uploaded by a user. If the user thinks there is an error in the resource, they can submit it, and the server will evaluate it and start a low-speed check thread. The process of users submitting to the server can be similar to version tools such as git.
No.46631
>>46610it's better to go with a scraper if you have to validate the posts with a subscription when it' being uploaded by users. it's just redundant.
the only positive is an on demand update(IF someone does import it on your request that too with questionable integrity) with innumerable problems you'll have to address to facilitate it.