egloos has been shut down

BUR #22564 has been rejected.

mass update source:*egloos* -> bad_egloos_id
create implication bad_egloos_id -> bad_id

shut down june 2023, completely purged december 2023

If every post from a certain source is tagged bad_id, then this tag serves no purpose for this source.

hdk5 said:

If every post from a certain source is tagged bad_id, then this tag serves no purpose for this source.

Bad BCY ID (closed)
Bad CGHub ID (closed)
Bad Drawcrowd ID (closed)
Bad drawr ID (closed)
Bad DrawTwit ID (closed)
Bad Google+ ID (closed)
Bad Miiverse ID (closed)
Bad Picasa ID (closed)

thelieutenant said:

Bad BCY ID (closed)
Bad CGHub ID (closed)
Bad Drawcrowd ID (closed)
Bad drawr ID (closed)
Bad DrawTwit ID (closed)
Bad Google+ ID (closed)
Bad Miiverse ID (closed)
Bad Picasa ID (closed)

My point still stands. I see no point in all these tags that are replicated by source: search.

hdk5 said:

My point still stands. I see no point in all these tags that are replicated by source: search.

The reason we have individual bad ID tags is that often the link is not uniform. Consider twitter, where posts are hosted on twitter.com and images are hosted on twimg.com, or pixiv, where images are hosted on pximg.net while posts are on pixiv.net. A lot of these tags can't be replaced with a single source search.

nonamethanks said:

The reason we have individual bad ID tags is that often the link is not uniform. Consider twitter, where posts are hosted on twitter.com and images are hosted on twimg.com, or pixiv, where images are hosted on pximg.net while posts are on pixiv.net. A lot of these tags can't be replaced with a single source search.

That gives me even more reason to disagree.
First - like you said, searching just for egloos substring in the source field is not a proper way to check if the image was sourced from a certain site.
Second - why only allow searching for manually tagged bad ids? Since danbooru already can recognize a lot of site URLs, why not just make it possible to search for site names it specifically coded to know (1, 2, 3), or by automatically chosen site names for other cases? Wouldn't it even allow to abolish all the bad_*_id and *_sample in favor of site:Pixiv bad_id kinda search?

hdk5 said:

Second - why only allow searching for manually tagged bad ids? Since danbooru already can recognize a lot of site URLs, why not just make it possible to search for site names it specifically coded to know (1, 2, 3), or by automatically chosen site names for other cases? Wouldn't it even allow to abolish all the bad_*_id and *_sample in favor of site:Pixiv bad_id kinda search?

That's the idea, but someone has to implement it first. Ideally such a system could even replace pixiv:<id> with custom site IDs. Even more ideally, we'd just store all source-related data in a source table and search source:pixiv source:deleted source:id:1. But for now, we only have tags.

BUR #22895 has been rejected.

mass update source:*cgsociety* -> bad_cgsociety_id

died jan 8th

thelieutenant said:
The BUR #22895 is pending approval.
mass update source:*cgsociety* -> bad_cgsociety_id

Seems that you missed the bad_id.

I'd change both BURs to an alias

BUR #23019 has been approved by @nonamethanks.

mass update source:*egloos.* or source:*cgsociety.* -> bad_id

I'm gonna add bad id to these instead of custom tags. It's true that if a site is completely dead and there's so few results, and they're all under a single source search, then there's no utility in the specific tag. These tags are more useful for large sites and sites with multiple CDNs, and for automated updating of live sites.

The bulk update request #23019 (forum #267564) has been approved by @nonamethanks.

The bulk update request #22564 (forum #265712) has been rejected by @nonamethanks.

The bulk update request #22895 (forum #267138) has been rejected by @nonamethanks.