I recently noticed @aaronfranke added a bunch of wiki edits which felt a bit excessive. These edits mainly surround subtags for the teacher_and_student relationship, which they have also started populating by themselves.

• Form: anthro_teacher/anthro_student, humanoid_teacher/humanoid_student, human_teacher/human_student, feral_teacher/feral_student, taur_teacher/taur_student
• Gender: male_teacher/male_student, female_teacher/female_student, intersex_teacher/intersex_student, herm_teacher/term_student, gynomorph_teacher/gynomorph_student
• Size: larger_teacher/larger_student, smaller_teacher/smaller_student

I know there is precedence for *_<form> (e.g., male_anthro, larger_feral), *_<gender> (e.g., larger_male, nude_female), and in a limited sense for *_<age> (e.g., larger_young, muscular_young).
There is even precedence for specific interests such as predator_and_prey, with basic descriptive tags (e.g., larger_prey, male_prey), unique subcategories (e.g., dominant_prey, willing_prey, struggling_prey), and even a convergence with other interests (e.g., farting_on_prey).

Now the question that I want to ask is: Should we allow for more *_<role> tags? If so, how far should we allow it?

Teacher/student does seem a bit excessive to me. What about rider/mount? Driver/passenger? How far would it go?

As it is, it seems teacher/student/teacher_and_student are being used very loosely:
post #4710384 post #4674607 post #4597557
there's no indication these are teachers and students aside from the age_difference.

Sometimes without even being able to tell who's the teacher or student:
post #4633393 post #4591907

In many cases, it's only based on dialog (which isn't otherwise supposed to be considered for tagging context):
post #4708299 post #4642902
You can't tag a character as male when the dialog says they're male, so you shouldn't tag them teacher or student based on their dialog saying or implying they are.

Most of the time it just seems to be synonymous with a school or classroom setting, though not always. When it's not for a school setting (which we already have tags for), student, teacher, and related tags seem at best lore info.

I started this tagging project so that I could tag and later search for feral teachers and smaller teachers, because I wanted to see more of those. Plus the male/female ones seem generally useful. The rest are mostly just there for completion's sake, since this way I can do a search for teacher size_difference -larger_teacher -smaller_teacher and get posts that should be categorized (either adding larger teacher, adding smaller teacher, removing size difference, or removing teacher), same for teacher student feral -feral_teacher -feral_student.

As for whether these should be lore tags, I don't know what's best. In many cases it is evident from the image (classroom setting, appropriate attire, age difference combined with other context, etc), but in other cases it cannot be determined by the visuals of the image (and relies on either the dialogue or lore). If we made these lore tags, would we want to move all uses to it or have 2 sets of tags?

Note that for the tagging I've done myself, for the most part, I have been adding *_teacher and *_student tags to posts that already have teacher and student, so I am building upon the precedent for what someone else has already tagged as teacher or student.

watsit said:
In many cases, it's only based on dialog (which isn't otherwise supposed to be considered for tagging context):
post #4708299 post #4642902
You can't tag a character as male when the dialog says they're male, so you shouldn't tag them teacher or student based on their dialog saying or implying they are.

when visuals contradict the text like for gender tags you're totally right, but I feel like it's not helpful to ignore text in situations like this. People looking for student/teacher images would want to see those, excluding them only makes searching worse.

wandering_spaniel said:
when visuals contradict the text like for gender tags you're totally right, but I feel like it's not helpful to ignore text in situations like this. People looking for student/teacher images would want to see those, excluding them only makes searching worse.

Problem is, text on an image isn't the word of god (in literary terms). Text can say incorrect things, characters can lie; a character creeping on another can claim they're a teacher to get into a student's pants, but the context for that will be missing in most pages. If we took text into consideration, such cases would be tagged teacher and student because text says it and it's not contradicted in the image, resulting in a tag that's not visually apparent and is wrong. If it's not a visible trait, it would need to be the author's stated intent that characters are a particular thing, which is the role of lore tags.

Being completely honest, I don't see much use for teacher and student tags given posts being properly tagged with a school or classroom setting. But if there should be such tags, I think only a single teacher_and_student_(lore) tag would be necessary. You wouldn't tag a character as a teacher if the student isn't also present, and vice-versa. Making all the <form>_teacher_(lore), <gender>_teacher_(lore), <form>_student_(lore), and <gender>_student_(lore) would be overkill. Lore tags aren't meant as a dumping ground to put tags that aren't TWYS, they're supposed to be a carefully curated set of necessary tags that can't work in TWYS, and having gender and form variants would not be a small list.

aaronfranke said:
Note that for the tagging I've done myself, for the most part, I have been adding *_teacher and *_student tags to posts that already have teacher and student, so I am building upon the precedent for what someone else has already tagged as teacher or student.

I understand that teacher_and_student is valid tag and interest. However, I personally feel that having additional subtags for them seem a little excessive and might set a precedent for other interests.
On one hand, predator_and_prey is given special status with their subtags since it is a rather complex interest.
On the other hand, I don't think I would want to see the same treatment being given to other interests such as mount/rider or brother_and_sister.

I can already see things getting out of hand for the clothed/nude interest, with nonsensical tags such as clothed_anthro_nude_male or clothed_male_nude_anthro being created.

Imagine if someone were to create an entire list of subtags like you but for their own interests (e.g., larger_brother or anthro_mount), would you support or oppose having more of these tags?
I also kinda hope that I would hear opinions from more people in regards to this question, because it is slowly developing into a problem with users following by example.

thegreatwolfgang said:
nonsensical tags such as clothed_anthro_nude_male or clothed_male_nude_anthro being created.

Yeah, I think those tags are bogus. Anthro and male are completely different and tangential concepts. I could maybe see clothed_feral_nude_anthro or similar, because those are the same concept, but really I do not have strong opinions for these "double-combo" tags. However in general I support "single-combo" tags like nude_anthro and clothed_male.

thegreatwolfgang said:
Imagine if someone were to create an entire list of subtags like you but for their own interests (e.g., larger_brother or anthro_mount), would you support or oppose having more of these tags?

I can see larger_brother being a good tag, I would support it. I'm not sure about anthro_mount but if someone had a justification then I would lean towards supporting that too.

aaronfranke said:
I could maybe see clothed_feral_nude_anthro or similar, because those are the same concept

Even those I find quite excessive. They won't be terribly much difference from clothed_feral+nude_anthro most of the time. We're trying to avoid more x_verbing_y tags because of the explosive tag growth it causes, and tags like that are worse since the x and y can have different adjectives associated with them (x_clothed_y_clothed, x_clothed_y_nude, x_nude_y_clothed, x_nude_y_nude, for all combinations of x and y). In addition to the confusion regarding that ferals shouldn't be tagged nude, causing some of those tags to stand out or be oddly missing.

Tag groups exist as a guideline on to suggest how to pair adjectives/roles with *_gender and *_form.
tag_group:index

Here are some strange things to note:
tag_group:anatomy is generally not used with *_gender and *_form tags. I think it's obvious why.

tag_group:body_types are generally used as the form for *_form, but according to topic #29463 , while human is a body type, it's also a species. There are also uncommon body groups that are not used this way.

tag_group:figures are often tagged this way (e.g. muscular_female, athletic_anthro), but some are not and don't lend themselves to the *_form or *_gender format (e.g. musclegut, pear-shaped_figure).

tag_group:genders has a list of the main genders, as well as what I think are poorly generalizable precedents like clothed_ambiguous_nude_andromorph and the predator/prey tags mentioned above. I think this is a mess and needs more open discussion. I think clothed_ambiguous is fine to tag, but people are not tagging things like skimpy_ambiguous or topless_ambiguous.

tag_group:character_relations is another grey area because it's even less TWYS.

mature_female gives an "age tag group" which seems pretty well developed, other than the fact that disambiguating "mature" from "older" is going to be a mess. Older is meant to apply when the mature female is with a "younger character," which raises the same issue I have with tags like "larger_female" (larger than what?)

I think these wiki pages need a lot of work, since they can lend legitimacy to what I think are bad tagging practices. Here's a list of attributes I think should be followed for *_gender and *_form tags:
The tag * should be:
Clearly TWYS: Body shape is an example. Character relations are not an example. This is just in the interest of taggability, since I think there's a lot of demand for these kinds of tags but not a lot of inertia behind tagging them, and being able to do it quickly is important to set a precedent.
In a small-ish enumerable category: I think basically all characters fall into fully_clothed/partially_clothed/skimpy/nude, so clothed_male or skimpy_anthro make sense. This is not the case for something like fur color, where characters are often different combinations of colors. This may require some significant tag reorganization, since many tags are (imo) ridiculously specific.
An independent attribute of the character: It's clear what a muscular_anthro should look like, whether or not the piece includes multiple characters. It's less clear what a larger_female should look like in a trio of characters.

tags are generally a mess. They're manageable for duo, but as the number of entities in a piece increases, the number of "verbing" tags grows combinatorially. The e621 tagging system is not designed to handle complex relationships between entities in a piece--it's barely equipped to handle describing those entities. I don't think there's an easy solution here.

xerxes_i said:
...

Yeah, I do agree that some form of consensus or guideline should be established before things gets out of hand.

I was never convinced in the need for clothed/nude's long/combined tag combinations (e.g., clothed_male_nude_female) when short/individual tags already accomplish the same thing (e.g., clothed_male, nude_female).
This becomes especially messy when you consider people aren't just tagging clothed and nude, but also the various stages of nudity (e.g., partially_clothed, mostly_nude) and creating absolute tag abominations such as partially_clothed_male_nude_male.
And like you said, they are just about manageable for duo scenarios but a world of mess when there are more characters with various genders and states of undress.

On the other hand, one could argue that this is the same for *_verbing_* tags (e.g., anthro_penetrating_feral is just anthro_penetrating & feral_penetrated) or any other interests that involves 2 parties (e.g., predator/prey).
Why should this tag group get preferential/exclusive treatment while mine doesn't?
As a result, people start to implement what is already established for other tags and include them into their own tag groups.

That is what I mean about precedents setting other precedents. For example:

• We have female_on_feral, and now we have female_on_human.
• We have clothed_male_nude_female, and now we have partially_clothed_male_nude_male.
• We have male_penetrating_female and anthro_penetrating_feral, and now we have male_rimming_female, female_facesitting_male, prey_penetrating_predator, etc.
• We have *_pred & *_prey, and now we have *_teacher & *_student.

Should we try to do something to stop this cycle, or should we turn a blind eye to it since it does not directly affect oneself?
Most of these "new" tags are generally created and maintained by those who have an interest in them.

thegreatwolfgang said:

• We have *_pred & *_prey, and now we have *_teacher & *_student.

*_pred and *_prey seem like way more "core" concepts. those are to vore what *_penetrating and *_penetrated are to "normal" sex or like dominant_* and submissive_* are to bdsm and other dom/sub situations.

dba_afish said:
*_pred and *_prey seem like way more "core" concepts. those are to vore what *_penetrating and *_penetrated are to "normal" sex or like dominant_* and submissive_* are to bdsm and other dom/sub situations.

So, no to *_teacher & *_student? What makes a certain interest a "core" concept?
I have to say that teacher_and_student is one of the most popular sexual roleplay scenarios out there, similarly to rape_play.

Don't get me wrong, I do feel that the above is a bit excessive, but I am also conflicted by how other more established interests are allowed to keep their subtags.

I think decomposing x_verbing_y into x_verbing and y_verbed is the most scalable way to handle these interactions.

I'm not sure about role tags. Roles are not clearly enumerable and lend themselves to TWYK, but at the same time I think tags like female_cop or anthro_teacher are/would be used a lot. However, I don't think tags like farting_on_prey (great example) are worthwhile to have. *_prey farting_on_* is probably sufficient for this specific request.

Obviously that query may not return the exact results people want, but that's because I don't think there's a perfect solution with the danbooru tagging system. There's a tradeoff between ease of tagging, manageability of tags, and specificity of tags. I think specificity is less important than making searching easy for the average user and tagging easy for the average uploader. Making taggers' lives easier makes searchers' lives easier. Encouraging taggers/searchers to use extremely specific tags through the wiki may be nice for a few dedicated fetish taggers, but makes it harder for everyone else to use the site.

^^^ I agree with Xerxes I's above post all the way. And this example...

thegreatwolfgang said:
On the other hand, one could argue that this is the same for *_verbing_* tags (e.g., anthro_penetrating_feral is just anthro_penetrating & feral_penetrated) or any other interests that involves 2 parties (e.g., predator/prey).
Why should this tag group get preferential/exclusive treatment while mine doesn't?
As a result, people start to implement what is already established for other tags and include them into their own tag groups.

...feels incredibly pertinent, especially for recognizing this as a general pattern with how new tags come about, and these awkwardly specific (albeit definitely interesting!) conversations have to happen. The problem isn't that a specialized conversation is unproductive, it's that similar conversations have to happen a lot, and what doesn't get addressed... just doesn't get addressed. It's really worth considering the root of how this comes about, IMO.

Tag decomposition for the benefit of easier tagging and necessarily more powerful searching is probably always the best way forward, IMO. That goes for the existing tags, too. anthro_penetrating_feral should be turned into anthro_penetrating and feral_penetrated—no special casing just because the tag's been around for a long time, when the existence of exceptions (and very commonly used ones at that!) is exactly what causes new tags to be made in bad example.

dba_afish said:
*_pred and *_prey seem like way more "core" concepts. those are to vore what *_penetrating and *_penetrated are to "normal" sex or like dominant_* and submissive_* are to bdsm and other dom/sub situations.

Not the way they're currently used. "male_pred -vore" alone currently has 5 pages of results (noting that tags like male_pred may be underutilized while vore is overutilized), with plenty of images taking "pred" to mean a social role of the character, ala a teacher, not the particular activity the character is engaging in, ala penetration. The predator/prey wiki even says:

Images or animations that depict two or more different species who, in real-life, would be the predator or prey of the other(s). Note that this label by itself doesn't imply anything about the characters' interactions.

These are about the characteristics the species may have in real life, and nothing about what the character is doing in the image. It's like teacher_and_student in that regard, except worse (teacher_and_student at least depends on the specific characters being teacher and student in-universe for the image, whereas predator and prey depends on the species' real-life counterparts completely divorced from what the image depicts).

watsit said:
Not the way they're currently used. "male_pred -vore" alone currently has 5 pages of results (noting that tags like male_pred may be underutilized while vore is overutilized), with plenty of images taking "pred" to mean a social role of the character, ala a teacher, not the particular activity the character is engaging in, ala penetration. The predator/prey wiki even says:
These are about the characteristics the species may have in real life, and nothing about what the character is doing in the image. It's like teacher_and_student in that regard, except worse (teacher_and_student at least depends on the specific characters being teacher and student in-universe for the image, whereas predator and prey depends on the species' real-life counterparts completely divorced from what the image depicts).

That is... strange.
I was under the impression that predator/prey was the only tag to describe interactions between "typical" pred and prey species, while all the other *pred *prey tags were for actual vore content.

xerxes_i said:
I think decomposing x_verbing_y into x_verbing and y_verbed is the most scalable way to handle these interactions.

everybodyknowsuradog said:
Tag decomposition for the benefit of easier tagging and necessarily more powerful searching is probably always the best way forward, IMO.

While a few of us do think that tag decomposition is the way to go (especially for removing any unnecessarily long combo tags), it does come with its own problems and should be put up to debate on their own dedicated thread when requesting for their invalidation.
Say we invalidate tags like male_rimming_female, people will still continue to tag male_rimming and female_rimmed because they follow the precedent of *_penetrating & *_penetrated.
IMO, *_rimming and *_rimmed are excessive and should not be used, but now they are "needed" in order to fill in the gap left behind by the invalidation of *_rimming_*.

What is more important though is that we address the crux of the issue here, i.e., which *_role/role_* tags should we allow and which should we remove for being excessive?

There are a few points in this thread that are solved in different ways. Let me know if these are all of them:

BUR/Tagging Projects:

• Decompose relationship tags like <form>_verbing_<form> (there are many of these that have inconsistent implications. I think this would be a pretty significant project.)
• Invalidate <form>_verbing_<gender> or <gender>_verbing_<form>

Wiki:

• Formalize which tags do and do not work for *_form and *_gender tags. Should *_age be included? I disagree as age is more ambiguous.
• Formalize the point of lore tags, which are entirely used for incest and gender at the moment.

I can move the first two points into a separate thread (topic #43775).

watsit said:
Not the way they're currently used. "male_pred -vore" alone currently has 5 pages of results (noting that tags like male_pred may be underutilized while vore is overutilized), with plenty of images taking "pred" to mean a social role of the character, ala a teacher, not the particular activity the character is engaging in, ala penetration. The predator/prey wiki even says:
These are about the characteristics the species may have in real life, and nothing about what the character is doing in the image.

predator/prey is a character dynamic/pairing tag, it's supposed to be for romantic or sexual interaction between characters whose species would be predator and prey in real life. it's an entirely separate concept from *_pred, *_prey, and the other vore-related tags.

if predator/prey is being used on vore posts or *_pred and *_prey are being used on predator/prey posts, the posts are mistagged

dba_afish said:
if predator/prey is being used on vore posts or *_pred and *_prey are being used on predator/prey posts, the posts are mistagged

IMO, they should be given different terms, then. Predator/pred and prey are common terms and shouldn't have different semantics for different tags (having one meaning for predator/prey, i.e. referring to the species' real-life status as natural hunter/prey regardless of in-image activity, and a different meaning for *_pred and *_prey, i.e. vore activities in the image regardless of the real-life status for the species). That's like saying male/female means something different for "male" and "female" compared to male_anthro and female_human. It would be very easy for people to think these tags should follow the same groundwork by using the same terms.

I guess I should respond to the thread topic before going off. Invalidate the entire group. Easy. Some of the potential combinations are amusing, but the "problem" they address is so niche as to matter only to the tags' creator. teacher_and_student is fine as a theme tag. I don't think this teacher/student thing has anything to discuss. This is public shaming. I wish users wouldn't go off creating 20 tags or whatever for their personal projects.

thegreatwolfgang said:
Should we try to do something to stop this cycle, or should we turn a blind eye to it since it does not directly affect oneself?
Most of these "new" tags are generally created and maintained by those who have an interest in them.

I would cull a lot of combo tags that have popped up, and the users making them should get neutral warnings.

everybodyknowsuradog said:
Tag decomposition for the benefit of easier tagging and necessarily more powerful searching is probably always the best way forward, IMO.

Decomposition can only produce less powerful searching, assuming that the proper implications are in place. Exact result = powerful search. Mix-and-match tags = larger pile of less relevant results. But wait, a third benefit to *_*ing_* tags is an easy default way of classifying every situation that guarantees the correct implications get tagged, because relying on a pair of "decomposed" tags to actually get tagged often results in one side getting missed. Things that should have a tag pair very often miss one half of the pair (cannot understate), so diverting the taggers' thought process to a single *_*ing_* tag ensures both sides of the tag pair actually get tagged.

I don't want these tags either, but yall are saying things that are very untrue. And now my post is too long. I've spent an outsize amount of time on this thread compared to what I think this will accomplish, and I do not wish to return.

everybodyknowsuradog said:

Tag decomposition for the benefit of easier tagging and necessarily more powerful searching is probably always the best way forward, IMO. That goes for the existing tags, too. anthro_penetrating_feral should be turned into anthro_penetrating and feral_penetrated—no special casing just because the tag's been around for a long time, when the existence of exceptions (and very commonly used ones at that!) is exactly what causes new tags to be made in bad example.

To use your example, anthro_penetrating feral_penetrated will mostly find anthro_penetrating_feral but will also find things like post #3363210. How is that more powerful searching?

abadbird said:
...

I found this post needlessly snarky, but it raised points that are worth responding to.

xerxes_i said:
I'm not sure about role tags. Roles are not clearly enumerable and lend themselves to TWYK, but at the same time I think tags like female_cop or anthro_teacher are/would be used a lot.

Honestly, I agree. I personally want these tags, and I expect they will be useful, but it is hard for me to justify them since they do lend themselves to TWYK.

abadbird said:
You realize that's the opposite of "scaling?" More complex tags scale with more complex situations that the simpler tags cannot hope to organize. That's their main reason for existing.

By "scalable", they mean system growth vs manageability. Given x_verbing and y_verbed tags, adding a single new gender, for example, only creates two additional tags: <new>_verbing and <new>_verbed. This is fairly slow growth as new things are added while staying manageable, it scales pretty well. However with x_verbing_y, adding a single new gender creates substantially more tags; x_verbing_<new> and <new>_verbing_y for all defined genders in x and y, in addition to <new>_verbing and <new>_verbed. The latter growth is greater and less manageable, each new thing creating more tags with associated aliases/implications than the thing before it, it scales worse.

abadbird said:
The huge point in favor of keeping *_penetrating_* is that posts often do feature multiple penetrations, upwards of 5 or even 10.

I'd say it's more about them being ingrained in most taggers' tagging behavior. We're more accustomed to just tagging male_penetrating_female (with simple shorthand aliases like m/p/fm) to automatically add all the other related tags (male_penetrating, female_penetrated, penetration, male. female). If male_penetrating_female were aliased to male_penetrating or male/female, a number of those automatically added tags would end up missing as we slowly changed our behavior to tag male_penetrating and female_penetrated more consistently instead. In contrast, tags like x_rimming_y or x_fingering_y are much less ingrained, significantly fewer people use them let alone rely on them, making them better targets to alias away without harming quality tagging as much.

watsit said:
By "scalable", they mean system growth vs manageability.

I don't think there's a real chance of adding more "genders" or body types. However, there are 8 commonly used "genders" (male, female, intersex, gynomorph, andromorph, ambiguous, herm, maleherm) leading to 64 pairwise combinations. There are 4-5 body types (anthro, feral, semi-anthro(?), humanoid, taur) leading to 16-25 pairwise combinations. If we want to go the route of implementing these tag pairs for every action (which I agree is a bad idea), that leads to 80-90 new tags per verb. There are even more if gender_verbing_form or form_verbing_gender are allowed. This would lead to a very powerful search, but is the tradeoff worth it compared to having just 24-26 tags that "generally" capture most duo interactions in a way that's stronger than what the site has now?

I created https://e621.net/forum_topics/43781 to discuss getting rid of double-double tags.

watsit said:
I'd say it's more about them being ingrained in most taggers' tagging behavior. We're more accustomed to just tagging male_penetrating_female (with simple shorthand aliases like m/p/fm) to automatically add all the other related tags (male_penetrating, female_penetrated, penetration, male. female). If male_penetrating_female were aliased to male_penetrating or male/female, a number of those automatically added tags would end up missing as we slowly changed our behavior to tag male_penetrating and female_penetrated more consistently instead. In contrast, tags like x_rimming_y or x_fingering_y are much less ingrained, significantly fewer people use them let alone rely on them, making them better targets to alias away without harming quality tagging as much.

it's also worth noting that penetrating is a way broader concept than a lot of the other verbs knocking around. it covers nearly every case of any body part or any object being inserted into any orifice, while stuff like rimming is significantly more specific of an action, fingering a bit less so, but still.

EDIT:

xerxes_i said:
I don't think there's a real chance of adding more "genders" or body types.

you say that, but honestly, the five existing form tags have kinda needed to be added to for ages. there are mountains of characters who sit in no-mans land, stuff like dire_machine, amorphous, and non-humanoid elemental_creature, and more.

On the topic of whether or not the *_role tags are appropriate or not, I agree that it is a bit questionable because it is not always easy to tell from the image, and I know e621's policy is that the dialogue should not count. My own take is that as long as the visuals do not contradict the dialogue I think it should count. But, we can continue the discussion if people think *_(lore) is more appropriate.

On the topic of whether or not the *_role tags are bloat, I really think they are not. They are only combining together 2 things, which is a lot smaller of a combinatorial explosion than other things like the double-double tags. Aside from combinatorics, I would like to point out the usefulness of them. We have so many useless tags on this site. I recently came across a post that had monotone_anklewear monotone_arms monotone_back monotone_balls monotone_beard monotone_belly monotone_body monotone_bottomwear monotone_butt monotone_cape monotone_chest monotone_cloak monotone_clothing monotone_face monotone_facewear monotone_feet monotone_footwear monotone_fur monotone_genitals monotone_glans monotone_hair monotone_hands monotone_head monotone_horn monotone_inner_ear monotone_legs monotone_mane monotone_mask monotone_mouth monotone_mustache monotone_neck monotone_neckwear monotone_pants monotone_pseudo_hair monotone_tail monotone_text monotone_wristwear as just a small percentage of its massive amount of tags. I ask you: Who is searching for monotone_inner_ear? Probably not many. However, there are actually people interested in seeing specifically a feral_teacher or a male_teacher or a female_teacher, and these tags are genuinely useful for this. I think it is absolutely NOT bloat to have several dozen genuinely useful teacher/student role tags, which will only have a few of them per post (like 2-8), when we have so many other useless tags that are bloating the post already.

The post I mentioned actually has nearly 300 color tags: https://e621.net/forum_topics/26928?page=1#forum_post_400909

Why are people fine with many thousands of <color>_<thing> tags, but when I propose a few dozen role tags it is considered bloat?

aaronfranke said:
On the topic of whether or not the *_role tags are appropriate or not, I agree that it is a bit questionable because it is not always easy to tell from the image, and I know e621's policy is that the dialogue should not count. My own take is that as long as the visuals do not contradict the dialogue I think it should count. But, we can continue the discussion if people think *_(lore) is more appropriate.

On the topic of whether or not the *_role tags are bloat, I really think they are not. They are only combining together 2 things, which is a lot smaller of a combinatorial explosion than other things like the double-double tags. Aside from combinatorics, I would like to point out the usefulness of them. We have so many useless tags on this site. I recently came across a post that had monotone_anklewear monotone_arms monotone_back monotone_balls monotone_beard monotone_belly monotone_body monotone_bottomwear monotone_butt monotone_cape monotone_chest monotone_cloak monotone_clothing monotone_face monotone_facewear monotone_feet monotone_footwear monotone_fur monotone_genitals monotone_glans monotone_hair monotone_hands monotone_head monotone_horn monotone_inner_ear monotone_legs monotone_mane monotone_mask monotone_mouth monotone_mustache monotone_neck monotone_neckwear monotone_pants monotone_pseudo_hair monotone_tail monotone_text monotone_wristwear as just a small percentage of its massive amount of tags. I ask you: Who is searching for monotone_inner_ear? Probably not many. However, there are actually people interested in seeing specifically a feral_teacher or a male_teacher or a female_teacher, and these tags are genuinely useful for this. I think it is absolutely NOT bloat to have several dozen genuinely useful teacher/student role tags, which will only have a few of them per post (like 2-8), when we have so many other useless tags that are bloating the post already.

I don't think that pointing out that worse tags exist is suitable justification for keeping unrelated tags.
There has been some effort to eliminating monotone tags, with the recent influx of 12500 posts into ~ears ~nose ~eyes because of the aliasing away of their monotone counterparts.
The monotone tags you listed are from a (now banned) enthusiastic overtagger, I'llKogYourMaw (I didn't even have to look at the post to know who was behind it)

Why are people fine with many thousands of <color>_<thing> tags, but when I propose a few dozen role tags it is considered bloat?

We are not, and that's why it has been pretty discouraged (see: Support for invalidating tag panic, most tagged; banning of I'llKogYourMaw; and the banning of Fluffball from tagging(? I'm not sure what went down there))

aaronfranke said:
On the topic of whether or not the *_role tags are appropriate or not, I agree that it is a bit questionable because it is not always easy to tell from the image, and I know e621's policy is that the dialogue should not count. My own take is that as long as the visuals do not contradict the dialogue I think it should count. But, we can continue the discussion if people think *_(lore) is more appropriate.

On the topic of whether or not the *_role tags are bloat, I really think they are not. They are only combining together 2 things, which is a lot smaller of a combinatorial explosion than other things like the double-double tags. Aside from combinatorics, I would like to point out the usefulness of them. We have so many useless tags on this site. I recently came across a post that had monotone_anklewear monotone_arms monotone_back monotone_balls monotone_beard monotone_belly monotone_body monotone_bottomwear monotone_butt monotone_cape monotone_chest monotone_cloak monotone_clothing monotone_face monotone_facewear monotone_feet monotone_footwear monotone_fur monotone_genitals monotone_glans monotone_hair monotone_hands monotone_head monotone_horn monotone_inner_ear monotone_legs monotone_mane monotone_mask monotone_mouth monotone_mustache monotone_neck monotone_neckwear monotone_pants monotone_pseudo_hair monotone_tail monotone_text monotone_wristwear as just a small percentage of its massive amount of tags. I ask you: Who is searching for monotone_inner_ear? Probably not many. However, there are actually people interested in seeing specifically a feral_teacher or a male_teacher or a female_teacher, and these tags are genuinely useful for this. I think it is absolutely NOT bloat to have several dozen genuinely useful teacher/student role tags, which will only have a few of them per post (like 2-8), when we have so many other useless tags that are bloating the post already.

The post I mentioned actually has nearly 300 color tags: https://e621.net/forum_topics/26928?page=1#forum_post_400909

Why are people fine with many thousands of <color>_<thing> tags, but when I propose a few dozen role tags it is considered bloat?

Monotone tags must burn I swear to god

nimphia said:
Monotone tags must burn I swear to god

yeah.
also, a bunch of the other <color>_<thing> tags are borderline pointless. like, I'm not even sure if <color>_<body_part> do anything other than fill up the tag list.

dba_afish said:
yeah.
also, a bunch of the other <color>_<thing> tags are borderline pointless. like, I'm not even sure if <color>_<body_part> do anything other than fill up the tag list.

I think it only makes sense if the body part is a different color than the rest of the body, but obviously people won't use it that way. I like color tags for a lot of things, like clothing and the like (I often search by pink_clothing, striped_socks, blue_dress, etc - I'm really into fashion and character design stuff) but then I see stuff like this and I'm like... Yeah, okay, we don't need to tag every color of every part of this character's body. They're all brown.

Sure, some body part color tags can be useful... But if I search by a body part color (usually for identifying characters), it's almost always going to be eyes/hair/nose/pawpads/claws/tongue, which are usually stand-out elements of a character's design.

dba_afish said:
yeah.
also, a bunch of the other <color>_<thing> tags are borderline pointless. like, I'm not even sure if <color>_<body_part> do anything other than fill up the tag list.

hate the monotone tags so much and I think there need to be better rules about color tags, but I actually do use some in searches mostly [color]_blush, [color]_pawpads, [color]_nose, and [color]_tongue. I like when characters have unusual colored parts like those. Just thought I'd chime in as a user of them

nimphia said:
(I often search by pink_clothing, striped_socks, blue_dress, etc - I'm really into fashion and character design stuff) but then I see stuff like this and I'm like... Yeah, okay, we don't need to tag every color of every part of this character's body. They're all brown.

Hey me too :)

wandering_spaniel said:
hate the monotone tags so much and I think there need to be better rules about color tags, but I actually do use some in searches mostly [color]_blush, [color]_pawpads, [color]_nose, and [color]_tongue. I like when characters have unusual colored parts like those. Just thought I'd chime in as a user of them

Basically this ^

I honestly feel like blush, pawpads, nose, tongue and claws work because they're closer to accent colors usually and they typically don't match the rest of the character's body. That, to me, should be why a color is tagged, rather than just that the body part is there and is a color - if it's the same color as everything else it's not notable!

nimphia said:
I think it only makes sense if the body part is a different color than the rest of the body, but obviously people won't use it that way.

I feel like it'd be better to just have a tag to define when a character has a limb/tail/ear/head/whatever that is a different color than the character's main color. then just tag both the main color and any secondary color with fur/scales/feathers/skin/body color tags.

we already kind of tag markings like this, one tag for marking location, maybe a tag for marking type/shape, one tag for marking color, done. _{although we do have <color>_tail_tip for some reason...}

I think it's fine to keep the more specific color tags for stuff that's a different color "by default" like hair, eyes, lips, mouth, pawpads, etc. but what we've got is just too much and, at least from my perspective, it actually makes me want to tag character colors less in general because there's just overwhelming.