Do we really want mob_x_character

BUR #47433 has been rejected.

nuke mob_x_character

Previously: topic #30098

Now, I actually do get the argument for this tag, I really do. But at the same time, it seems like its a tag that's just made to be a >1M posts tag.

It also covers a wide range of concepts: mobs with barely any presence in the image (post #9946200), mobs with lots of presence in the image (post #9896311).

In addition to this, it also looks like it's applied to ambiguous cases like post #9260808, where the identiy is obscured so we *technically* can't tag the character, despite the artist having an obvious preferred pairing (siy_(shengxiu760722) kawaragi_momoka yuri). Personally, finding a solution to tagging in these cases seems more valuable, but my earlier concerns regarding this being an "everything" tag still stands.

my opinion hasn't really changed since the original thread. as much as populating this tag would be a project of colossal proportions, it is mildly annoying how searching for pairings of two canon characters is anywhere from a little confusing, to borderline impossible, depending on how you look at it.

something like chartags:>1 hetero brings up an infinite amount of cases of "character with implicated costume, therefore chartags>1, with a mob character", and that's not to mention untagged original characters.

i've filled in the wiki that was agreed upon from the previous thread just in case.

If this tag doesn't get approved within 45 days or it gets rejected manually, I assume the tag would stay. It's up to NNT or Evazion. I'm totally indifferent to either choice.

I do think @wingdings concerns are valid too. If this tag is nuked, I hope that we at least find some alternate solutions.

I tried my damn best to pass a BUR that would make this into a workable concept, and alas, nothing worked. Seeing this tag crop up with no input feels...Odd. On the one hand, I still stand by my thoughts in the prior hand and previous efforts. But on the other hand...This sort of thing always felt massive enough to require discussion. Especially due to the name.

I wish someone didn't populate it manually, as I knew if attempted we'd end up right here.

We may as well ask the tagger, @danbooru02, why they decided to make the tag after the BURs in the original topic got auto-rejected.

BUR #47459 has been approved by @evazion.

deprecate mob_x_character

If we think it's bad intrinsically, rather than just being misused at this time, we should make it so it can't be built back up again with a deprecation.

CrossbowArcanePlus said:

BUR #47459 has been approved by @evazion.

deprecate mob_x_character

If we think it's bad intrinsically

i'm not sure anyone here is arguing that.

wingdings said:

i'm not sure anyone here is arguing that.

I can see at least a few points in this thread about the tag itself being overly broad and so applicable to things outside of what a user actually wants when searching it, or not being able to reach a workable definition for the tag, which both concern the tag rather than its current usage.
I do sympathise with what you want from the tag, and I personally have no strong opinion on it, but there's the option to vote for deprecation now if people do feel that way.

i do believe the broadness is a concern that would have been solved had the tag been given some more discussion.

if workload is a deal-breaker (which it shouldn't be, workload didn't stop anyone from, say, deprecating tags like looking away and eyebrows, both of which are still in the tens of thousands range years post-deprecation), then creating a tag for pairings of two canon characters sounds like it may be a more tangible task.

i don't have any numbers on me right now, so that's just a gut feeling, but it's strange to me that we're discarding a potential solution to a real tagging problem ahead of time.

wingdings said:

i do believe the broadness is a concern that would have been solved had the tag been given some more discussion.

if workload is a deal-breaker (which it shouldn't be, workload didn't stop anyone from, say, deprecating tags like looking away and eyebrows, both of which are still in the tens of thousands range years post-deprecation), then creating a tag for pairings of two canon characters sounds like it may be a more tangible task.

i don't have any numbers on me right now, so that's just a gut feeling, but it's strange to me that we're discarding a potential solution to a real tagging problem ahead of time.

Though the thing is, we're still beholden to (an arbitrarily-decided) oldbooru rule that pairings shouldn't get gentags (that was how we also nuked idol and vtuber duo tags in the past, which was a sore point of contention since I was in favor of keeping these tags for record-keeping). If you want to argue in favor of pair gentags, we'd have to do a lot of convincing for us to throw away this old rule, especially since there's actually sizable opposition to such.

Wrt to the current "mob x character" thing, it really does feel like a rather botched alternative to duo gentags.

wingdings said:

if workload is a deal-breaker (which it shouldn't be, workload didn't stop anyone from, say, deprecating tags like looking away and eyebrows, both of which are still in the tens of thousands range years post-deprecation), then creating a tag for pairings of two canon characters sounds like it may be a more tangible task.

By workable I just meant that it can be consistently applied without having to have too much guesswork. I'm talking about Knowledge Seeker's quote and I assume that's what they meant too, though I may have misinterpreted.

wingdings said:

it's strange to me that we're discarding a potential solution to a real tagging problem ahead of time.

There's a month and half on the clock, and only a nuke is looking likely right now anyway. That'll just reset the tagging done after the discussion seemed to suggest it wasn't where the forum wanted to go rather than discard the solution entirely. Even the deprecation would be reversible if a good finished solution is agreed upon.

CrossbowArcanePlus said:

By workable I just meant that it can be consistently applied without having to have too much guesswork.

with "workload" i was referring to the actual amount of work needed to be done to populate this tag into a usable state.

anyway, coming back to this thread with a refreshed mind, i agree that being disappointed in the outcome in advance is not productive. i'll wait and see where this goes for now.

Damian0358 said:

We may as well ask the tagger, @danbooru02, why they decided to make the tag after the BURs in the original topic got auto-rejected.

There's no harm in making a temporary tag. We can nuke or deprecate it whenever.

You should be using a favgroup if you want to populate something for discussion to create a tag based on such a widely-reaching concept. You shouldn't be creating tags if a BUR gets rejected, especially if it's explicitly rejected moreso than autorejected, because then that's defying admin (in)decision about a tag being on the site.

Personally: I like the tag in concept but I just don't think it's... you know, useful when it applies to most posts.

danbooru02 said:

There's no harm in making a temporary tag. We can nuke or deprecate it whenever.

But when you do it like this, you effectively create a crap ton more work, not to mention it opens any thread with people poised against the tag's existence (which is exactly what happened here). I'm saying this moreso because this approach was absolutely not the approach I think should've been taken.

What I think would've been better and would've created a better first impression would've been to use the favgroup approach (like I did initially) or at least make an attempt at reviving the old thread. I still want to exclude mob characters in my searches, yes, but not if the tag is going to be at risk of becoming nigh-unusable due to too wide a range (as pointed out in the opening nuke BUR).

ArcieA said:
Though the thing is, we're still beholden to (an arbitrarily-decided) oldbooru rule that pairings shouldn't get gentags (that was how we also nuked idol and vtuber duo tags in the past, which was a sore point of contention since I was in favor of keeping these tags for record-keeping). If you want to argue in favor of pair gentags, we'd have to do a lot of convincing for us to throw away this old rule, especially since there's actually sizable opposition to such.

Wrt to the current "mob x character" thing, it really does feel like a rather botched alternative to duo gentags.

Not the same thing. The difference here is that those duo tags are easily replaced with a [character] + [character]" search. As it stands, removing mob characters from searches has zero good alternative searches, even with unlimited search privileges.

Removing mob characters from searches is still very much something desirable. I just don't think it was done right. And until we find a way to create a way for me to search for characters having sex with each other and not get slammed with faceless self-inserts, we're just going to keep going circles like this.

Even if the tag size is massive, I think this tag would be an understandable exception a la solo, because I have no clue what a good proposal for the exception would look like. We do need to put better parameters on it though (as I don't think cases like post #9260808, especially with artists fond of certain pairings, really belong in that tag).

it's been a month now, and with this bur expiring soon, it's just as impossible to search for or exclude this stuff as it was before. my initial proposal for fixing this tag is to restrict its scope to strictly NSFW (meaning rating:q or rating:e) situations. yes, it makes the tag slightly less useful, but if it helps make the concept more palatable, then i'm all for it.

i'm also going to totally go out on a limb here and remove post #9260808 from the tag. everyone in this thread including myself agrees that it's not what the tag is for and that it's ambiguous and that it being in the tag is pedantry, so let me force everyone's hand a little.

wingdings said:

it's been a month now, and with this bur expiring soon, it's just as impossible to search for or exclude this stuff as it was before. my initial proposal for fixing this tag is to restrict its scope to strictly NSFW (meaning rating:q or rating:e) situations. yes, it makes the tag slightly less useful, but if it helps make the concept more palatable, then i'm all for it.

As someone who just wants to see two characters I like fucking without getting loads of generic male self-inserts, I still would really like to have a way to filter them out without sacrificing anything I want to see.

I do think restricting it to NSFW posts will help cut down on nonsense like post #9260808 ending up in that tag (especially when the artist is a well-established shipper where you are meant to assume said other character), as people really want to use it to filter out NSFW posts and no one wants it in the SFW posts (outside of "lead-up" scenarios that just end in porn I assume).

i'm also going to totally go out on a limb here and remove post #9260808 from the tag. everyone in this thread including myself agrees that it's not what the tag is for and that it's ambiguous and that it being in the tag is pedantry, so let me force everyone's hand a little.

Please do. No one wants that post in there, regardless of what happens to the tag.

wingdings said:

i'm also going to totally go out on a limb here and remove post #9260808 from the tag. everyone in this thread including myself agrees that it's not what the tag is for and that it's ambiguous and that it being in the tag is pedantry, so let me force everyone's hand a little.

One of the main reasons I don't like the tag was because I had this mass-added to a bunch of POV and solo focus posts that are almost exactly like this one and it felt like it completely reduced the value of the tag.

I'm in tune with the intention and desire for having this tag. I am not with the way it was created nor how it was assigned which were all at the whims of a single person who came up with their own definition and pedantic qualifications.

WRS said:

One of the main reasons I don't like the tag was because I had this mass-added to a bunch of POV and solo focus posts that are almost exactly like this one and it felt like it completely reduced the value of the tag.

i'm curious to know where exactly people draw the line as to where the tag is useful and where it becomes an overreach so we can hopefully come to some kind of consensus about the guidelines.

should povs not count? well, i can see how it may be a different category, but it still operates on the same principle.

how about a fully in-frame, but faceless, clearly intended to be a self-insert character like in post #9306230? and if we take that approach - how in frame can a mob be before it's no longer a mob? i think this is something the present guidelines handle perfectly well, pedantic as they are. if we open the door to questions like that, i think everyone will come up with a slightly different answer and we'll get nothing done.