Posted under General
BrokenEagle98 said:
Just throwing that idea out there to see what other thought about it...
Okay. I'm against splitting up md5 mismatch, at least.
If someone decides to move all 340k+ bad id to something like bad pixiv id, I suppose now would be a good time, since I don't plan on running any of my automated routines while I'm on vacation anyway. Just PM me a reminder, kthxbye.
RaisingK said:
Okay. I'm against splitting up md5 mismatch, at least.
I updated the wiki then to account for more than just Pixiv.
RaisingK said:
If someone decides to move all 340k+ bad id to something like bad pixiv id, I suppose now would be a good time, since I don't plan on running any of my automated routines while I'm on vacation anyway. Just PM me a reminder, kthxbye.
I plan on doing a scan through the posts later to verify that all of the bad id tagged posts do indeed have only Pixiv sources. Unfortunately, I haven't been able to find a "-source:" metatag that works properly for this, so I'll be scanning them with a script.
Once that's done, I plan on submitting a BUR to change bad_id -> bad_pixiv_id, and then implicate both bad_twitter_id and bad_pixiv_id to bad_id. Because of the implication, the BUR should be able to be done like:
update bad_id -> bad_id bad_pixiv_id imply bad_pixiv_id -> bad_id imply bad_twitter_id -> bad_id
Although I believe that the update and implications will have to be done in separate steps, since the order of the operations isn't necessarily done in the order that they are written...?
Please chime in if there's anything wrong with the above...
Just a quick status update...
For those last two, I've added an additional tag category, bad source, for posts that did not have a valid Twitter post address.
Example:
An invalid equivalent for Pixiv would be...
Someone's gonna go on a flagging spree… :)
If it happens frequently enough though, do you think we should have a temporary thread for requesting entry flagging for these kinds of things (so that those with approval privs can take care of)? Like in the case that builders start to sweep through everything in twitter sourced md5 mismatches to upload the originals since it's relatively easy to do (copypaste tags, upload, set mismatch to parent, flag mismatch).
There are frivolous flags being thrown around all the time, but in cases like these it would actually be a very legitimate reason.
I'm trying to think of the best way to do that without it being exploitable...
I initially thought about using the search twitter_sample parent:any, however that could be exploited by a malicious user, or it could also happen by accident by a well-intentioned user.
Would starting up a "Twitter post delete" thread be the best method...?
At least with that, if a user that an approver trusts posts in that thread with the post # of a Twitter sample that has been properly parented with the higher resolution/greater filesize, then it would be safe to flag and/or delete.
If a new user posted that an approver was unsure of, I would recommend a manual verification, at least of the first couple of posts until trust is built up...
Thoughts?
Actually, RaisingK flags with his own account since the RazingK account is member-level only and so is limited to 1 flag per day. Even so, his normal account is limited to 10 flags a day.
I'm guessing that he does 10 uploads with RazingK and then flags 10 times with RaisingK a day, and slowly catches up over time...?
Hmm... I just had a thought, and I wonder how it would sell, but what about a new Special Permission for Unlimited Flagging...?
This would be for those users that don't need approval privileges to do the kind of automated post gardening as discussed in this thread, however are limited by the 10 count flag limitation...
Updated
Hmm some positive developments I see.
Anyway I've always wondered why we can't just merge the antecedent post with the better one? Like move the votes and favs to it, and make it "aliased" to the better one. Duplicate posts seem a bit excessive since we have to host those files, the tagging usually slumps on them compared to their parent, and there's no rational reason for anyone to prefer them.
Also, I guess just tag post #2489489 as bad tumblr id? I uploaded this after I moved from Gelbooru and at that time the artist's tumblr went inactive however I was the original uploader at Gelbooru.
BrokenEagle98 said:
Hmm... I just had a thought, and I wonder how it would sell, but what about a new Special Permission for Unlimited Flagging...?
This would be for those users that don't need approval privileges to do the kind of automated post gardening as discussed in this thread, however are limited by the 10 count flag limitation...
Unquestioningly endorse this. It's obvious to everyone the positive work he is doing, I see no likelihood of abuse from such users as long as they are as reputable as RaisingK.
sweetpeɐ said:
Also, I guess just tag post #2489489 as bad tumblr id? I uploaded this after I moved from Gelbooru and at that time the artist's tumblr went inactive however I was the original uploader at Gelbooru.
Yeah, I plan on doing Tumblr, Nicoseiga, and then Nijie eventually, so might as well get started with what you can find.
BrokenEagle98 said:
Actually, RaisingK flags with his own account since the RazingK account is member-level only and so is limited to 1 flag per day. Even so, his normal account is limited to 10 flags a day.
I'm guessing that he does 10 uploads with RazingK and then flags 10 times with RaisingK a day, and slowly catches up over time...?
Indeed. No tricks, just patience.
Status update...
I'll start working on Tumblr next...
BrokenEagle98 said:
Status update...
- source:http://pbs.twimg.com/* (Finished)
- source:https://pbs.twimg.com/* (Finished)
- source:http://twitter.com/* (Finished)
- source:https://twitter.com/* (Finished)
I'll start working on Tumblr next...
You're doing great work :) Enjoy your Christmas
Just want to mention the issue also exists on
As mentioned on its thread.
