Topic: [Request] “Anti-implication” for tags

I believe there’s an automatic way to remove conflicting tags (i.e duo with solo) without having to do it manually

If a user accidentally add a tag that’s conflicting to the original/first tag, it wouldn’t have add it unless they specifically remove it first then replace it with a new one

Another thing to add is a new “request anti-implication” or something so we could make a list of what should/shouldn’t go with

Also to make sure the user know whether their tags was added or not, a pop up will notify them (like how removing implied tags was notified)

If this request sounds too ambitious/difficult, then maybe limit this feature for top level users

There are always situations where mutually exclusive tags can be used, see multiple images and multiple scenes

There are so few genuinely exclusive tags (human only vs. human on anthro) that I don't see much dev time getting put into something like this.

donovan_dmc said:
There are always situations where mutually exclusive tags can be used, see multiple images and multiple scenes

lafcadio said:
There are so few genuinely exclusive tags (human only vs. human on anthro) that I don't see much dev time getting put into something like this.

Seems like we need some consensus to agree on. Here’s a (non-exhaustive and short) list that I think should go first (for beta testing) since it’s worth a try

• solo x duo, trio, group, or zero pictured
• human only x human on anthro, anthro, human on feral, feral, or their respective implication (like bestiality)
• thumbnail x low res, hi res, absurd res, or superabsurd res
• fully clothed (if solo) x nude or it’s respective implication (like nude anthro)
• long playtime x short playtime
• headshot portrait x full-length portrait, three-quarter portrait
• black and white x colored or any colored theme (like green theme)
• 1:1 x any other aspect ratio (like 16:9)
• year A x any other year
• ancient art x anything made in or after 1980

That’s all I could think of right now, any more suggestions would be great

whatismyname1234 said:

• solo x duo, trio, group, or zero pictured

multiple images & multiple scenes

whatismyname1234 said:

• human only x human on anthro, anthro, human on feral, feral, or their respective implication (like bestiality)

These are extremely rare and definitely not worth it, I'd also argue that multiple images and/or multiple scenes could apply here despite the wiki definition not mentioning them

whatismyname1234 said:

• thumbnail x low res, hi res, absurd res, or superabsurd res

Already handled by the site

whatismyname1234 said:

• fully clothed (if solo) x nude or it’s respective implication (like nude anthro)

multiple images & multiple scenes

whatismyname1234 said:

• long playtime x short playtime

Already handled by the site

whatismyname1234 said:

• headshot portrait x full-length portrait, three-quarter portrait

multiple images

whatismyname1234 said:

• black and white x colored or any colored theme (like green theme)

Not worth it

whatismyname1234 said:

• 1:1 x any other aspect ratio (like 16:9)

already handled by bots

whatismyname1234 said:

• year A x any other year
• ancient art x anything made in or after 1980

Not worth it

The majority of metatags are not mistagged often enough or mistagging doesn't cause enough detriment to warrant a whole system to prevent this
This isn't just flipping a switch, this would need a whole new fleshed out system which would touch every corner of the site
We get what you're suggesting but it's really not worth anyone's time, developer or otherwise

whatismyname1234 said:
Seems like we need some consensus to agree on. Here’s a (non-exhaustive and short) list that I think should go first (for beta testing) since it’s worth a try

• solo x duo, trio, group, or zero pictured
• human only x human on anthro, anthro, human on feral, feral, or their respective implication (like bestiality)
• thumbnail x low res, hi res, absurd res, or superabsurd res
• fully clothed (if solo) x nude or it’s respective implication (like nude anthro)
• long playtime x short playtime
• headshot portrait x full-length portrait, three-quarter portrait
• black and white x colored or any colored theme (like green theme)
• 1:1 x any other aspect ratio (like 16:9)
• year A x any other year
• ancient art x anything made in or after 1980

That’s all I could think of right now, any more suggestions would be great

I think of these the only ones that are actually a problem right now are human_only ~anthro ~feral ~semi-anthro ~humanoid ~taur and black_and_white colored. Everything else either isn't really mutually exclusive (e.g. fully_clothed and nude) or returned no results/so few I was able to fix them while writing this post (e.g. long_playtime short_playtime).

FWIW a few users are running bots that automatically edit tags on posts based on things that can be analysed by a computer script, like image dimensions or colour. There are also tagging projects that can be arranged to clean up certain tags. I think it would be a good idea to invest in these methods of cleaning up tags rather than putting already sparse development resources into a new feature.

The difference between the resolution and playtime examples, and the other examples you're providing, is that the server itself can verify the correct value. There is no real way for the server to tell whether a post should be human or humanoid, and the thresholds for what constitutes a specific size or duration are not likely to change later, so there doesn't need to be any system for configuring these rules, deleting them, or adding new ones. Editing the source code is sufficient.

black_and_white colored examples could be spot color instead, or black and white and red, etc.; it is much more realistic to inform the problem user instead of messing with their tag input, and as-is people already get confused by what should be obvious implications like sheath → animal genitalia.

Implement this yourself and make a merge request if you want, but otherwise this isn't something I expect anybody to spend serious time on.

Welp at least I know this suggestion isn’t likely to be useful in the long run

eightoflakes said:
There are also tagging projects that can be arranged to clean up certain tags. I think it would be a good idea to invest in these methods of cleaning up tags rather than putting already sparse development resources into a new feature.

I think I’m going to join that site, thanks for recommending this

I've been mulling over a system that validates these kinds of cases & pops up a message w/ suggestions or something if the tags don't make sense for months; it's just a matter of getting the time to do it.