Topic: Tag BURs: Unimply color_fur/scales/feathers -> color_body

Posted under Tag Alias and Implication Suggestions

As discussed on topic #33456, a summary will be posted below.

Rationale

Currently <color>_fur/scales/feathers/skin/exoskeleton all imply <color>_body.
These are implied as such because (a) they are mainly used to tag the textures of a character's body and (b) for convenience to automatically tag <color>_body.
E.g., A character covered in brown_fur naturally has a brown_body.

However, often times, sections of fur, scales, or feathers do not necessary end up being part of the overall texture of a character's body, existing as part of a minor or trivial feature.
E.g., A character having fur_tuft, facial_scales, or feathered_wings, etc.

Thus, if someone tags the colour of these minor features (e.g., black furred arm_tuft), it automatically implies the <color>_body tag (i.e., black_body), which may not be entirely accurate in certain cases.
Leading to examples such as:

What is being changed?

Suggest to unimply all <color>_fur/scales/feathers (incl. features such as glowing, striped, etc.) from <color>_body, with the specific exclusion of skin & exoskeleton for now.
All colour wikis will be edited to reflect these changes if the BURs get approved.

How does this affect me?

Start tagging <color>_body separate of <color>_fur/scales/feathers, and stop relying on the latter as the go-to tag for body colour tagging.

Updated

The bulk update request #2541 is pending approval.

remove implication monotone_scales (370) -> monotone_body (12834)
remove implication monotone_fur (8725) -> monotone_body (12834)
remove implication monotone_feathers (388) -> monotone_body (12834)
remove implication two_tone_scales (12403) -> two_tone_body (185944)
remove implication two_tone_fur (126671) -> two_tone_body (185944)
remove implication two_tone_feathers (2984) -> two_tone_body (185944)
remove implication rainbow_scales (48) -> rainbow_body (869)
remove implication rainbow_fur (563) -> rainbow_body (869)
remove implication rainbow_feathers (137) -> rainbow_body (869)
remove implication multicolored_scales (17437) -> multicolored_body (324272)
remove implication multicolored_fur (227404) -> multicolored_body (324272)
remove implication multicolored_feathers (8514) -> multicolored_body (324272)
remove implication dark_scales (289) -> dark_body (33578)
remove implication dark_fur (4877) -> dark_body (33578)
remove implication dark_feathers (188) -> dark_body (33578)
remove implication light_scales (212) -> light_body (30081)
remove implication light_fur (4310) -> light_body (30081)
remove implication light_feathers (139) -> light_body (30081)

Reason: Monotone / Two-Tone / Rainbow / Multicolored / Dark / Light Body BUR

As per topic #33456 & topic #33859.

The bulk update request #2542 is pending approval.

remove implication glistening_scales (919) -> glistening_body (43054)
remove implication glistening_fur (3918) -> glistening_body (43054)
remove implication glistening_feathers (105) -> glistening_body (43054)
remove implication glowing_scales (157) -> glowing_body (3238)
remove implication glowing_fur (764) -> glowing_body (3238)
remove implication glowing_feathers (73) -> glowing_body (3238)
remove implication mottled_scales (40) -> mottled_body (1508)
remove implication mottled_fur (786) -> mottled_body (1508)
remove implication mottled_feathers (9) -> mottled_body (1508)
remove implication piebald_scales (25) -> piebald_body (472)
remove implication piebald_fur (355) -> piebald_body (472)
remove implication piebald_feathers (0) -> piebald_body (472)
remove implication spotted_scales (1426) -> spotted_body (52373)
remove implication spotted_fur (41958) -> spotted_body (52373)
remove implication spotted_feathers (340) -> spotted_body (52373)
remove implication striped_scales (1664) -> striped_body (57711)
remove implication striped_fur (43634) -> striped_body (57711)
remove implication striped_feathers (492) -> striped_body (57711) # missing
remove implication translucent_scales (5) -> translucent_body (9169)
remove implication translucent_fur (19) -> translucent_body (9169)
remove implication translucent_feathers (8) -> translucent_body (9169)

Reason: Glistening / Glowing / Mottled / Piebald / Spotted / Striped / Translucent Body BUR

As per topic #33456 & topic #33859.

The bulk update request #2543 is pending approval.

remove implication black_scales (18351) -> black_body (347115)
remove implication black_fur (242060) -> black_body (347115)
remove implication black_feathers (13676) -> black_body (347115)
remove implication blue_scales (25257) -> blue_body (344793)
remove implication blue_fur (181286) -> blue_body (344793)
remove implication blue_feathers (36148) -> blue_body (344793)
remove implication brown_scales (7043) -> brown_body (404789)
remove implication brown_fur (327249) -> brown_body (404789)
remove implication brown_feathers (10346) -> brown_body (404789)
remove implication brown_scales (7043) -> brown_body (404789)
remove implication brown_fur (327249) -> brown_body (404789)
remove implication brown_feathers (10346) -> brown_body (404789)
remove implication green_scales (30066) -> green_body (153516)
remove implication green_fur (40623) -> green_body (153516)
remove implication green_feathers (5232) -> green_body (153516)
remove implication grey_scales (12032) -> grey_body (310418)
remove implication grey_fur (227096) -> grey_body (310418)
remove implication grey_feathers (9844) -> grey_body (310418)
remove implication orange_scales (9608) -> orange_body (197520)
remove implication orange_fur (148945) -> orange_body (197520)
remove implication orange_feathers (5028) -> orange_body (197520)
remove implication pink_scales (5936) -> pink_body (133058)
remove implication pink_fur (72613) -> pink_body (133058)
remove implication pink_feathers (4216) -> pink_body (133058)

Reason: Black / Blue / Brown / Green / Grey / Orange / Pink Body BUR

As per topic #33456 & topic #33859.

The bulk update request #2544 is pending approval.

remove implication purple_scales (13366) -> purple_body (155868)
remove implication purple_fur (79316) -> purple_body (155868)
remove implication purple_feathers (8729) -> purple_body (155868)
remove implication red_scales (19443) -> red_body (156635)
remove implication red_fur (74971) -> red_body (156635)
remove implication red_feathers (8045) -> red_body (156635)
remove implication tan_scales (8281) -> tan_body (192435)
remove implication tan_fur (136943) -> tan_body (192435)
remove implication tan_feathers (2480) -> tan_body (192435)
remove implication white_scales (17268) -> white_body (691855)
remove implication white_fur (528792) -> white_body (691855)
remove implication white_feathers (27552) -> white_body (691855)
remove implication yellow_scales (14574) -> yellow_body (227236)
remove implication yellow_fur (145688) -> yellow_body (227236)
remove implication yellow_feathers (16933) -> yellow_body (227236)

Reason: Purple / Red / Tan / White / Yellow Body BUR

As per topic #33456 & topic #33859.

thegreatwolfgang said:

How does this affect me?

Start tagging <color>_body separate of <color>_fur/scales/feathers, and stop relying on the latter as the go-to tag for body colour tagging.

I think this is a bit much to expect from most taggers. As much as there is a small issue with some non-body uses of <color>_fur/scales/feathers being tagged <color>_body by implication, not having the implication will leave <color>_body next to useless. I also don't think I fully agree with your assessment:

thegreatwolfgang said:
Thus, if someone tags the colour of these minor features (e.g., black furred arm_tuft), it automatically implies the <color>_body tag (i.e., black_body), which may not be entirely accurate in certain cases.
Leading to examples such as:

The tail and ears are part of the body, so part of their body is brown and the brown_body tag is valid there.

thegreatwolfgang said:

The wings are also part of the body, as much as an arm or leg or other limb is, so these rainbow_feathers on their wings does mean the character has a rainbow_body for part of their body.

watsit said:
I think this is a bit much to expect from most taggers. As much as there is a small issue with some non-body uses of <color>_fur/scales/feathers being tagged <color>_body by implication, not having the implication will leave <color>_body next to useless. I also don't think I fully agree with your assessment:

The tail and ears are part of the body, so part of their body is brown and the brown_body tag is valid there.

The wings are also part of the body, as much as an arm or leg or other limb is, so these rainbow_feathers on their wings does mean the character has a rainbow_body for part of their body.

Except, generally the term "body" is referring to the big central thing that isn't limbs.

nystemy said:
Except, generally the term "body" is referring to the big central thing that isn't limbs.

That would be the torso, the area between the neck and the waist excluding the limbs. When the torso has a distinct color, <color>_torso would work, just as <color>_tail does when the tail has a distinct color, or <color>_wings, etc. But just because a tail can be its own color doesn't make it separate from the body. And I don't think we need to neuter the <color>_body tags just because some uses occur in error.

watsit said:
I think this is a bit much to expect from most taggers. As much as there is a small issue with some non-body uses of <color>_fur/scales/feathers being tagged <color>_body by implication, not having the implication will leave <color>_body next to useless.

An issue would still be present and will always be until something is changed.
The question now is whether or not people are willing to make the effort in tagging <color>_body separately.

I also don't think I fully agree with your assessment:
The tail and ears are part of the body, so part of their body is brown and the brown_body tag is valid there.

While technically true, I do not see it as practical as any minutely coloured feature can be tagged as <color>_body.
If the white_fur colour for their inner_ear_fluff is tagged, it would imply white_body. I don't believe anybody would be searching white_body for this post.

The wings are also part of the body, as much as an arm or leg or other limb is, so these rainbow_feathers on their wings does mean the character has a rainbow_body for part of their body.

I don't see it that way. I see monotone_body + white_body as opposed to multicolored_body + rainbow_body.
And again, while technically true, it isn't practical to tag as such since nobody would realistically be searching rainbow_body for this post.

Even if they did not have rainbow_wings, a single feather that is multicolored would be enough to say that they have a multicolored_body, which is the main argument for this unimplication.

watsit said:
That would be the torso, the area between the neck and the waist excluding the limbs. When the torso has a distinct color, <color>_torso would work, just as <color>_tail does when the tail has a distinct color, or <color>_wings, etc. But just because a tail can be its own color doesn't make it separate from the body. And I don't think we need to neuter the <color>_body tags just because some uses occur in error.

In general it is likely better if no limb/part related color tag implies the same color on the body.

Just because a character has a distinct colored detail, be it a tail, feathers, wings, etc, doesn't mean that their body in general has that color.

color_body tags is better if kept for the overall body color and not auto included due to a limb/part having said color.

thegreatwolfgang said:
Even if they did not have rainbow_wings, a single feather that is multicolored would be enough to say that they have a multicolored_body, which is the main argument for this unimplication.

Having minor distinctly colored details auto imply the color for the whole body is indeed a bit stupid.

(Similar to the stupidity I am trying to fix with the micro_on_macro tag auto implying micro and macro even if the micro_on_macro tag don't require a micro or macro to be present in the image... https://e621.net/forum_topics/33834 But at least the colored_X implying colored_body is at least implying a color that is actually there, even if it might be a tiny detail.)

thegreatwolfgang said:
The question now is whether or not people are willing to make the effort in tagging <color>_body separately.

It would be nice if e621 had a recommended tags list, where it knows that red_scales imply a recommendation for "red_body" as well. The list can be similar to how e621 already remembers tags that one has recently added to prior posts one has edited. And then we just have to bind tags together in a similar fashion to implying, but for recommendations instead where it just shows up on the list. (and yes, this list of recommendations might get long on a post with many tags.)

Updated

No where is written:"The body color has to cover at least X% of a characters body parts to apply". Or am I wrong?

thegreatwolfgang said:
While technically true, I do not see it as practical as any minutely coloured feature can be tagged as <color>_body.

I wouldn't consider the tail and ears to be a "minutely coloured feature".

thegreatwolfgang said:
I don't see it that way. I see monotone_body + white_body as opposed to multicolored_body + rainbow_body.

I definitely would not want or expect to find images like post #63199 under monotone_body, just as I wouldn't want or expect to find post #2416576 or post #3025724 under monotone_body.

thegreatwolfgang said:
Even if they did not have rainbow_wings, a single feather that is multicolored would be enough to say that they have a multicolored_body, which is the main argument for this unimplication.

I don't think a single off-color feather in an otherwise monotone body is that often of an occurrence. And I don't think the frequency it does happen is worth losing the implication to <color>_body, because the latter is too useful but rarely explicitly tagged.

dubsthefox said:
No where is written:"The body color has to cover at least X% of a characters body parts to apply". Or am I wrong?

Is this dragon yellow?
post #1822199

If we thoroughly tag an image by every bit that has a colored tag, then we don't generally want all those colors to also end up as the body color. Since most of them aren't.

Yes, there isn't really a "minimum percentage" required for the color_body tags, but common sense does generally say that minute details do not state the overall color of an individual.

watsit said:
I wouldn't consider the tail and ears to be a "minutely coloured feature".

I do not want to specify what constitutes as "minute" since features can be very subjective depending on the character and/or species.
A character can have very large ears or wings while others can be tiny or barely visible.

What I would say is if a colour is not a major part of the character's colour palette, it should be exempted when tagging <color>_body.
To quote multicolored_body, it is used for "a character ... with two or more distinct body colors". I don't think minor colour features would be considered distinct.

I definitely would not want or expect to find images like post #63199 under monotone_body, just as I wouldn't want or expect to find post #2416576 or post #3025724 under monotone_body.

I don't think a single off-color feather in an otherwise monotone body is that often of an occurrence. And I don't think the frequency it does happen is worth losing the implication to <color>_body, because the latter is too useful but rarely explicitly tagged.

Likewise, I would not expect seeing them under multicolored_body. They would more accurately be found under multicolored_feathers.

Your reasoning behind not having this unimplication is due to the supposed loss of people tagging <color>_body, rather than due to the very real problem of people unintentionally tagging multicolored_body on posts that otherwise don't warrant it.

thegreatwolfgang said:
Leaving that out for the time being, we can discuss what constitutes as part of <color>_body later.

Though it does influence how severe you see the problem to be. I consider the tail to be part of the body, so brown_fur on a humanoid tail implicating brown_body is perfectly fine, no problem at all for images like that. Whereas someone else may consider the tail to not be part of the body, meaning an implication with those images is a problem that needs addressing.

thegreatwolfgang said:
The question now is whether or not we should include every fur/feathers/scales feature as being part of the body.

The question is whether removing the implications would do more harm than good, and considering I don't see the implications creating as much as a problem as you do, as we have different standards for when <color>_body applies, we come to different answers.

watsit said:
The question is whether removing the implications would do more harm than good, and considering I don't see the implications creating as much as a problem as you do, as we have different standards for when <color>_body applies, we come to different answers.

To be fair here, people are bad at tagging in general and the upload page should likely just add "[random color]_body" to the example tags on "characters" part of the form. Though the form currently does ask if there is one or more characters in the image, if they are feral or anthro, among other details, so adding body color to that isn't impossible. (could even make it required, unless one selects "Zero Pictured") This would more or less make all new pictures have a mostly correct tag for body color.

In short, I do not think we should accept implications that leads to incorrectly tagged images. Even if it is a minority.
Especially since the errors can't be fixed since as long as the implication exists, then tag X will include tag Y whenever one saves the edit to the post. (if it were 1 or 2 posts we can shrug our shoulders, but when it is a couple of % it is a lot of posts.)

Or e621 could have a recommended tags list, where it knows that red_scales imply a recommendation for "red_body" as well. The list can be similar to how e621 already remembers tags that one has recently added to prior posts one has edited. And then we just have to bind tags together in a similar fashion to implying, but for recommendations instead where it just shows up on the list. (and yes, this list of recommendations might get long on a post with many tags, that is however not a problem.)

thegreatwolfgang said:
E.g., post #2647658 = monotone_scales -> monotone_body???

Honestly, I've been having my own issues with the monotone_* tags in general, between having an excessive number of them (monotone_eyebrows, really? are there so many multicolored_eyebrows that a normal eyebrows search isn't good enough to find monotone eyebrows?), and being used very incorrectly (monotone_eyes being filled with standard white-sclera + colored-iris + black-pupil eyes, very much not monotone, which makes finding actual monotone eyes, like post #2579502 or post #3345287, impossible; if post #3352220 is considered monotone_eyes because it's just the pupils that count for monotone eyes, post #3358262 and post #3355001 should count as well, but they're obviously not).

In this case, removing the implications for monotone_fur/scales/feathers -> monotone_body would make sense to me, because there's enough characters that have a mix of fur, scales, and feathers, that one of them being monotone doesn't make the body as a whole monotone. A character having <color>_fur does 99.9% of the time mean <color>_body applies to at least part of their body, so those implications should stay given how necessary <color>_body is, but this reasoning doesn't carry to monotone_fur so well.

watsit said:
I wouldn't consider the tail and ears to be a "minutely coloured feature".

Forgot to ask this, but would you be okay with implying <color>_tail or <color>_ears or any other coloured body part to <color>_body?

thegreatwolfgang said:
Forgot to ask this, but would you be okay with implying <color>_tail or <color>_ears or any other coloured body part to <color>_body?

Yes, if those tags aren't significantly used for fake or detachable accessories (e.g. a tail fashion accessory that's clipped to a belt, fake rabbit ears, etc).

watsit said:
Yes, if those tags aren't significantly used for fake or detachable accessories (e.g. a tail fashion accessory that's clipped to a belt, fake rabbit ears, etc).

So for fun I searched: white_ears black_body solo
The majority of the images do not contain a character with a black body, but rather a white one with some black marking, sometimes very subtly. (this is just ~600 posts, that a lot of them are incorrectly tagged but can't be fixed due to tag implications.)

One can play around with other color combinations on the other body parts and get similar results. There is literally thousands of images that are incorrectly tagged as something they clearly aren't.

Is this really something you consider worth while keeping?

All for the sake of "if a lazy person just tags the color of a body part it will imply the same color for the whole body.", most people do not tag colors to start with and there is better options for getting people to start. One being to include body color as part of the upload form. (and yes, I know that would be a project in and off itself.)

thegreatwolfgang said:
Even genitals or markings (e.g., multicolored_markings -> multicolored_body)?

Markings probably, not sure about genitals. There are plenty of cases where the penis appears to come from inside the body, rather than contributing to the epidermal coloring. There are cases where I would consider it part of the body color, like post #3305088 (as it's no different from any other limb, aside from being shorter), but also plenty of cases where I wouldn't, like post #3139322 (as it's more of an internal organ being exposed, like a tongue), so I don't think a <color>_genitals implication would work.

watsit said:
Markings probably, not sure about genitals. There are plenty of cases where the penis appears to come from inside the body, rather than contributing to the epidermal coloring. There are cases where I would consider it part of the body color, like post #3305088 (as it's no different from any other limb, aside from being shorter), but also plenty of cases where I wouldn't, like post #3139322 (as it's more of an internal organ being exposed, like a tongue), so I don't think a <color>_genitals implication would work.

Sorry, I meant external genitals. In that case, you would consider tagging the colours of furry_balls/ball_tuft as being part of the body colours (e.g., red ball_tuft -> red_body) to be accurate then?
What about fur_tufts appearing on other body parts, such as whisker_tuft?

In regards to markings, would you consider a rainbow_markings enough for a character to be considered rainbow_body?

thegreatwolfgang said:
Sorry, I meant external genitals. In that case, you would consider tagging the colours of furry_balls/ball_tuft as being part of the body colours (e.g., red ball_tuft -> red_body) to be accurate then?

Furry balls yes, ball tuft not really (at least when the tuft is a separate color from the rest of the balls, which makes it look more like artificial highlighting to me, like the hair highlights on post #357956).

thegreatwolfgang said:
In regards to markings, would you consider a rainbow_markings enough for a character to be considered rainbow_body?

If it's a normal part of their body, yes. A quick glance seems to indicate the majority would apply, with post #2970660 being an outlier that I'm not sure about since it appears more like an artificial paintjob/makeup.

nystemy said:
So for fun I searched: white_ears black_body solo
The majority of the images do not contain a character with a black body, but rather a white one with some black marking, sometimes very subtly. (this is just ~600 posts, that a lot of them are incorrectly tagged but can't be fixed due to tag implications.)

I can't say I'm seeing many I wouldn't tag with black_body. There's some I'd say are more grey or more green or yellow/tan/orange rather than black and white, but I'm not otherwise seeing posts that are outright wrong with the tag.

nystemy said:
most people do not tag colors to start with and there is better options for getting people to start. One being to include body color as part of the upload form. (and yes, I know that would be a project in and off itself.)

Would "remember to tag white_fur, black_fur, and striped_fur" or "remember to tag white_fur, black_fur, striped_fur, white_body, black_body, and striped_body" work better to get people to start tagging body color and type properly? There's only so many extra buttons that can be added to the upload form before it becomes unwieldy and people start ignoring all of them. And they're only visible when "Enable Compact Uploader" is disabled in your settings, which is an option because some people were already complaining the uploader was getting too bloated with all the buttons and separate fields.

As it is, some people are only tagging <color>_body because they don't know or don't care if it's fur, scales, feathers, or whatever else, and some people are only tagging <color>_fur, <color>_scales, etc, because the implications do the rest of the work. Removing the implications will spread body colors out among different tags so people will have a hard time searching for characters with particular body colors, because they don't know whether the tagger used <color>_body or <color>_fur for the character's furry body, etc.

watsit said:
Furry balls yes, ball tuft not really (at least when the tuft is a separate color from the rest of the balls, which makes it look more like artificial highlighting to me, like the hair highlights on post #357956).

I guess the main question I was trying to ask is if you consider tuft colours to be part of the body in general.
People can easily tag <color>_fur for the colour of tuft or balls and it would automatically imply <color>_body.

E.g., post #2849430 (bottom crocodile = brown_body)

If it's a normal part of their body, yes. A quick glance seems to indicate the majority would apply, with post #2970660 being an outlier that I'm not sure about since it appears more like an artificial paintjob/makeup.

So, nothing artificial then? In that case, if someone has a rainbow-coloured cutie_mark/marking on any part of their body, then it would be enough for it to be called multicolored_body/rainbow_body then.

E.g., post #632806 (ignore the hair and tail for the sake of an example)

thegreatwolfgang said:
I guess the main question I was trying to ask is if you consider tuft colours to be part of the body in general.
People can easily tag <color>_fur for the colour of tuft or balls and it would automatically imply <color>_body.

They can, and I'm sure there are cases where they do. I'm not saying there won't ever be a problem (I've brought it up before with an example). But I think those problems are relatively minor, compared to the problems there would be without the implications.

thegreatwolfgang said:
So, nothing artificial then? In that case, if someone has a rainbow-coloured cutie_mark/marking on any part of their body, then it would be enough for it to be called multicolored_body/rainbow_body then.

E.g., post #632806 (ignore the hair and tail for the sake of an example)

Artificial looking at least. If it looks like makeup or a stick-on/washable tattoo, I wouldn't consider it <color>_body (or really <color>_fur/scales/etc either, for that matter). Cutie marks do have an artificial look to them for me, though I understand not everyone may think so and may consider it part of the body. So if they think it warrants <color>_fur, then having <color>_body with it makes just as much sense. post #3295654 has the yellow fur looking like a natural part of their body to me, so yellow_fur+yellow_body would be applicable IMO, even though it's only part of their body (same as post #3307175 isn't a white-only monotone_body; despite the torso and arms/logs being primarily white, the pink and blue markings/tail/ears look like a natural part of their body, so pink_body/blue_body/multicolored_body makes sense).

watsit said:
Would "remember to tag white_fur, black_fur, and striped_fur" or "remember to tag white_fur, black_fur, striped_fur, white_body, black_body, and striped_body" work better to get people to start tagging body color and type properly? There's only so many extra buttons that can be added to the upload form before it becomes unwieldy and people start ignoring all of them. And they're only visible when "Enable Compact Uploader" is disabled in your settings, which is an option because some people were already complaining the uploader was getting too bloated with all the buttons and separate fields.

I won't say you are wrong in that assessment to be fair.
People are generally easily overwhelmed, confused and lazy. Making tag recommendations is however one of the better tools in the box, especially if one does it dynamically as new tags are added. But this is obviously beyond the scope of if we should implicate these tags or not.

watsit said:
As it is, some people are only tagging <color>_body because they don't know or don't care if it's fur, scales, feathers, or whatever else, and some people are only tagging <color>_fur, <color>_scales, etc, because the implications do the rest of the work. Removing the implications will spread body colors out among different tags so people will have a hard time searching for characters with particular body colors, because they don't know whether the tagger used <color>_body or <color>_fur for the character's furry body, etc.

I can agree that "fur" colors can more or less implicate body color too, since fur by definition is something that covers the body. Else it is hair. (fluff is also a description one can use.)
Scales likewise often covers the body, or at least belly. Seldom do a character have scales elsewhere. (unless it is differently colored scales for detailing reasons, and this isn't rear. Same applies for colored fur patches above.)

But a lot of tags like ears, paws/feet/hands, to a degree tails, and so forth generally shouldn't imply a body color, since it isn't uncommon for an appendage to have a different color than the overall body.

The body color tags shouldn't include detail colors, else we can then just remove the whole "body" part of those tags and just state what colors the character is having anywhere, in any amount, and this would be a useful tag in its own regard. But generally if searching for the color of a body, then preferably that color should be the majority of the character, or at least close to it. Since this is more or less what people think when they search for a body color.

Also, body color is something ALL images should have, even if only 0.1% of posts ends up incorrectly tagged due to the implications then that is currently about 3000 posts that are wrong. But there is plenty of characters with detail coloring for ears, tails, paws/feet/hands, wings, markings, and so forth that would increase the amount of incorrectly tagged images into a good few percent, ie tens of thousands of images.

In my opinion, implicating these tags to body color is just causing too many false positives. It muddles the usefulness of the tags and that were the main thing I pointed at with the search of: white_ears black_body solo since a rather huge portion of the images aren't characters with a black body, ie the search is having a fairly low hit rate, despite the tags technically being very descriptive about what one tried to find.

Currently, we can either choose between two solutions without having to remove the implications:
1. having them incorrectly tagged as black_body.
2. remove the tags for the details currently colored black on those posts as to remove the incorrect black_body tag.

And non of these two options are a good one. Ie, the current implication forces us to choose between two dumb decisions.
If all the implications weren't there, then all the images of characters that don't have a black_body wouldn't be in the search while still allowing those posts to have their black body parts correctly tagged.

(and if you think that white_ears black_body solo is a niche search, then let me remind you that this is just one combination of body parts and colors, there is hundreds of other combinations (actually way more, but lets not dump maths in here), and then there is group images too. As stated, tens of thousands of incorrect posts, some of which currently don't have any color tags but they should in the future. And when they do get tagged, they should preferably be able to be correctly tagged.)

Tags implicating other tags is a useful tool at times, but it is easy to get lazy and frankly abuse it to the detriment of search quality. As shown above.

Likewise isn't technicalities beneficial to search quality either. Yes, a paw is part of the body, and yes it is therefore part of the body_color, but it is generally better to have tags be more specific. And as I have said earlier, having a tag for the existence of "a color" in an image is in itself actually a useful tag, but when one adds a descriptor, like "body". Then one generally want to take one of the less technicality inclusive definition of the term as to improve the quality of the search. (this is however a simplified explanation about how to tag content in a dataset to achieve good search quality with few tags per search, and that is a much larger topic to be fair.)

Updated

Three potential solutions that everyone will hate:

1.
Perhaps the word “body” is the issue here? What if we just use the long-since-invalidated color tags for when any part of any character is a color? Ie. red_scales implies red. “(color)_body,” to me, implies that a character’s body is a majority (or at least a significant amount) of that color, but that is obviously impossible to use with implications of any kind. That way, red_body can be manually tagged whenever the character’s body actually has a significant amount of that color, and there is still an umbrella tag for any color as we have now, which will still be able to catch cases where users forget to tag *_body, even though the results will be pretty diffuse (as they already are).

Since I already know everyone will hate that idea, perhaps we could rename the tags (color)_body_part instead? Then it’s clear that it refers to any part of the body. Still, I think the plain color tags (red, blue, green, etc.) are a more elegant solution. I can’t think of any substitute for “body” that isn’t either just as vague or just plain clunky. And it’s not like those tags are being used for anything else currently.

2.
Alternatively, we could initiate some tags like mostly_red_body for what I imagine red_body to actually mean - that a character’s body consists of a significant amount of a particular color. If we would have to manually tag body color after this BUR anyway, might as well make a new tag to do that and start populating it. I think the red_body and etc. tags are perhaps beyond recovery at this point anyway, as they’ve been implied by these tags for such a long time.

My main issue with the *_body tags is that they just aren’t that useful for anything. Limiting them to the majority of the body would make them very useful, but the current results are fairly diluted by these implications, leading to a lot of black_body that’s actually a white_body with a slight bit of black on it somewhere. Which leads me to my third solution - and, in my opinion, the best…

3.
We could stop tagging red_scales, red_feathers, red_fur on posts where a majority of that character’s scales/feathers/fur aren’t of that color. If there’s a small patch of red fur on an otherwise green character, you don’t tag red_fur.

This would certainly be the least invasive solution, but the issue would be getting users to follow that line of thinking. But perhaps with some clarification in the wikis and diligent users correcting this when it happens, it could be manageable. Personally, I already avoid tagging small parts of the body with these scales/feathers/fur colors simply because I know the *_body implications are in place and I don’t like tagging (color)_body where it isn’t a significant part of the body. There are already other tags for colored markings, hands, feet, ears, tufts, etc. which can remain separate.

scaliespe said:
Three potential solutions that everyone will hate:

1.
Perhaps the word “body” is the issue here? What if we just use the long-since-invalidated color tags for when any part of any character is a color? Ie. red_scales implies red.

People would end up using it for anything that's red. A red apple, red hair, red fingernails, etc. I've seen/had to fix cases like that in the past.

scaliespe said:
Since I already know everyone will hate that idea, perhaps we could rename the tags (color)_body_part instead?

We already have <color>_tail, <color>_ears, etc. Of course, there's also the issue of overtagging -- tagging so many individual parts of the character that the tag list is just a mess of red_fingers red_hand red_arm red_torso red_ears etc... monotone_fingers monotone_hand monotone_arm monotone_torso monotone_ears etc... that it becomes impossible to notice any mistags or missing tags -- which is why having general <color>_body tags that can refer to any part of the body is also helpful in cutting down the monotony. The <color>_body tag is a catch-all for any part of a character's body so you don't have to tag each minor thing the same way.

scaliespe said:
2.
Alternatively, we could initiate some tags like mostly_red_body for what I imagine red_body to actually mean - that a character’s body consists of a significant amount of a particular color.

Define "significant amount". Everyone will have a different idea of what constitutes being "mostly" a color. It would also make it less useful for people searching when they know some part of the character was some color, but don't remember what appendage or how much. And there's the issue with clothed or obstructed characters, where a significant portion of their body is covered so you can't see that they're mostly some color.

scaliespe said:
My main issue with the *_body tags is that they just aren’t that useful for anything.

On the contrary, I see them as quite useful. They help you find characters that have part of their body with a certain color. If it's restricted to certain parts of the body, or only "a significant amount" of the body, it will be less useful since a person searching can't know if the tagger otherwise though it's "enough" to tag, rather than simply tagging what you see.

scaliespe said:
3.
We could stop tagging red_scales, red_feathers, red_fur on posts where a majority of that character’s scales/feathers/fur aren’t of that color. If there’s a small patch of red fur on an otherwise green character, you don’t tag red_fur.

Same as above, different people will have different ideas for what the "majority of that character" means. And be less useful for searching when you know some part of the character is that color, which may or may not be a large enough portion.

scaliespe said:
My main issue with the *_body tags is that they just aren’t that useful for anything. Limiting them to the majority of the body would make them very useful, but the current results are fairly diluted by these implications, leading to a lot of black_body that’s actually a white_body with a slight bit of black on it somewhere.

And this is the main reason no color tag should implicate the color of the body.

watsit said:
Define "significant amount". Everyone will have a different idea of what constitutes being "mostly" a color.

Well, everyone has different lower ends to the threshold. But that doesn't mean we can't use that definition.

Just like the hyper tag is for unrealistically large body parts, also something with a diffuse lower limit to when the tag is applicable.
Same for macro and micro.
Or the lines between erect, half_erect and flaccid.
Thick_tail likewise don't have a strict lower bound.

And for all of these, things gets even more muddy since the species do effect people's opinion on where the line should be for a given character.

But despite that, people can largely agree enough for these tags to be useable as far as searching goes. (Ie, the majority of the results are in line with what one tried to fine.)

Asking people to tag based on what a characters overall color is shouldn't be impossible for people to mostly agree on. There will be some edge cases resulting in disagreements, but there is always edge cases.
Some characters will need 2-3 color tags for the body, like a Zebra is applicable to tag with white_body and black_body since it is more or less 50/50 black and white. (Meanwhile a normal skunk should often be tagged: black_body white_stripes)

Going beyond 3 colors will be rare, and those times the multicolored_body tag is applicable. (and perhaps "multicolored" shouldn't start at 2, since a lot of characters have 2 primary colors that often covers a ratio anywhere from 30/70 to 50/50 of the body. It is practical if the "multicolored" tag is actually distinct in what one finds, instead of mostly redundant.)

If a lack of a well defined and agreed on line in the sand is a problem. Then we have a problem with the majority of all tags on this site. We just have to trust that the majority of users will see the term "When X color covers a significant amount of a character." and use it in mostly the same fashion. (If the occasional image has "a tag too many" it won't effect search quality that much, at least compared to the current "The color is somewhere on the character." And if one really wants to find a character mostly consisting of X color then one will have to manually look through the images since no current tags will aid in that mission.)

As an experiment, what do you consider to be the lower bound for "when a character is mostly a given color."? (Like 50%?, 30%?, 10%?, 5%?, 0.01%? Not that we should stick to a strict percentage as a definition, I am just curious. I myself would say around 20-25%)

watsit said:
It would also make it less useful for people searching when they know some part of the character was some color, but don't remember what appendage or how much.

This is indeed an edge case of searching. Here having a tag for a color just existing on the character could be useful. Though, I know a lot of people how has described an image they can't find. And it turns out they mixed up the color of the character with the color of something completely different in the image. The human recollection of images is a puzzling thing.

watsit said:
And there's the issue with clothed or obstructed characters, where a significant portion of their body is covered so you can't see that they're mostly some color.

Tag what you see. The most important rule of tagging on e621. Character lore doesn't apply when tagging.
Ie, can't see what the overall color of an individual's body is, then tag according to the parts you can see. (just like we tag ambiguous_gender on most scalies on SFW pictures.)
And if one can't see any of the body's color, then don't tag any color.

nystemy said:
Well, everyone has different lower ends to the threshold. But that doesn't mean we can't use that definition.

Just like the hyper tag is for unrealistically large body parts, also something with a diffuse lower limit to when the tag is applicable.
Same for macro and micro.
Or the lines between erect, half_erect and flaccid.

At least in those cases, there are different stages to try. If you thought a character was hyper and you're not finding it, maybe they weren't as big as you thought so try searching huge instead. Macro/micro have a lot of overlap with hyper/huge/etc and size_difference, which you can fall back to. Same for erection/half_erect/flaccid. If <color>_body isn't working because it wasn't "enough of the body", however, what can you fall back to?

nystemy said:
Asking people to tag based on what a characters overall color is shouldn't be impossible for people to mostly agree on.

You'd be surprised. People were saying in the previous thread that they couldn't even be consistent by themselves with different images, so expecting people to be consistent between themselves is quite a high bar.

nystemy said:
Some characters will need 2-3 color tags for the body, like a Zebra is applicable to tag with white_body and black_body since it is more or less 50/50 black and white. (Meanwhile a normal skunk should often be tagged: black_body white_stripes)

As an experiment, what do you consider to be the lower bound for "when a character is mostly a given color."? (Like 50%?, 30%?, 10%?, 5%?, 0.01%? Not that we should stick to a strict percentage as a definition, I am just curious. I myself would say around 20-25%)

That's exactly it, I don't know. "I'll know it when I see it", i.e. I'll just guess based on my intuition on a case-by-case basis. Hardly a good tagging standard for something like this. I mean, I'd consider white_fur/white_body to be valid for a skunk because of their stripes. Zebras can vary between a small amount of relatively thin stripes to thicker more prominent stripes, but in both cases part of their body is colored black. Adding more variability for species that "should" have some body color but the artist decided to use thinner lines instead of thicker ones that time, would make it even harder to properly search with the tags if they depended on some arbitrary amount of their body being that color.

nystemy said:
This is indeed an edge case of searching. Here having a tag for a color just existing on the character could be useful.

That's exactly what the <color>_body tags are now. Changing the <color>_body tags to be only for "a significant portion" of the body, then acknowledging we could use tags for any portion of the body like the tags are now, makes me question why change them at all then.

nystemy said:
Tag what you see. The most important rule of tagging on e621. Character lore doesn't apply when tagging.
Ie, can't see what the overall color of an individual's body is, then tag according to the parts you can see. (just like we tag ambiguous_gender on most scalies on SFW pictures.)
And if one can't see any of the body's color, then don't tag any color.

How about images like
post #3361592
do we assume there's not much of the body offscreen that this constitutes a "significant portion", or do we assume much of the body is offscreen and this doesn't constitute a "significant portion"?

Posts like
post #3362592 should not have black_body because we can only see just the head and some of the neck, not a "significant portion".
post #3354487 should not have brown_body because we can only see the head, hands, and part of the tail, not a "significant portion".
post #3353036 should not have grey_body or white_body because we can only see parts of their fur, and what's visible is split between grey and white, no one color being a "significant portion" of their body as a whole.
post #3361046 should not have blue_body or tan_body for the dragon/tobi-kadachi because most of their body is offscreen, and what's visible is split between blue and tan, no one color being a "significant portion" of their body as a whole.

How is this making the tags' usefulness better?

the *_body tags are supposed to be catch-all terms for any type of skin/coat, it's not meant to only define if a character is mostly one color or another. it's also used when the skin type is hard to define under existing tags, i.e. characters who are animate_inanimate, flora_fauna, goo_creature, etc..

I feel that there is two schools of thought with how people are approaching with using this tag, one is the "absolute" and the other is the "practical".

For absolute tagging, it is all-or-none. We follow tagging strictly and include *_body for any (or if not all) external body parts that are coloured/textured differently, regardless of whether it is significant or not.
For practical tagging, we view there as being some unspoken exceptions. Tagging based on how we feel best represents the definition of the tag, avoiding or tagging it on how we interpret it as being used.

However, the current issue stems from (a) the vague nature of body and (b) somewhat inaccurate naming/use of fur, feathers, & scales.
Putting aside how we interpret body, using the *_fur/feathers/scales tags automatically implies the *_body tag, which in most cases can lead to the unwanted tagging of *_body.

For example, having translucent_feathers (i.e., transparent feathers) would imply that there is a translucent_body present (i.e., transparent body), which can lead to cases such as post #3272989.
This also happens in other tags such as glowing_body, striped_body, spotted_body, etc., where we cannot differentiate between it as being a partial or full-body feature without the use of *_body_part tags.

As another (rather extreme) example, a post consisting of solely a severed blue feathered_wing would imply that a blue-coloured body is also present.
One can argue that said body part can still be considered as being part of the body, but in that case, they shouldn't have any qualms when it comes to implying any *_body_part with *_body then.

Again, everything boils down to how we see body, either as being a body part or the whole body itself. The subsequent fur, feathers, & scales implication argument also derive from the same problem.

nystemy said:
(Meanwhile a normal skunk should often be tagged: black_body white_stripes)

ok. Where do you have this from? I have never seen it written anywhere, besides this and the previous thread, that stripes aren't considered a part of the body color. Only once, when I have attempted to imply inner_ear_colors to body colors. topic #29553

Nystemy said:
As an experiment, what do you consider to be the lower bound for "when a character is mostly a given color."? (Like 50%?, 30%?, 10%?, 5%?, 0.01%?...)

I'd go with 1% even if it's just one finger that's different. It's a part of the body, and therefor should be tagged as such, in my opinion.

watsit said:
nystemy said:
This is indeed an edge case of searching. Here having a tag for a color just existing on the character could be useful.

That's exactly what the <color>_body tags are now. Changing the <color>_body tags to be only for "a significant portion" of the body, then acknowledging we could use tags for any portion of the body like the tags are now, makes me question why change them at all then.

Exactly this.

Btw. I still don't understand exactly why we should exclude wings/tails/faces from the body tag. Is this some sort of translation problem I have? I have seen that "body" can be used as a synonym for "torso" (midriff and chest). Perhaps, could it be that this is the problem, we have here? Because "body" can also refer to the whole organism.

darryus said:
the *_body tags are supposed to be catch-all terms for any type of skin/coat, it's not meant to only define if a character is mostly one color or another. it's also used when the skin type is hard to define under existing tags, i.e. characters who are animate_inanimate, flora_fauna, goo_creature, etc..

The issue is that *_body no longer functions as a skin/coat tag anymore with its implication with *_fur/feathers/scales, as the latter trio of tags are being used for both body texture and body parts/features.

So, in a TL;DR sense, if you see post #2849430 as being brown_body, then everything is working as intended. If not, please read the thread further.

thegreatwolfgang said:
I feel that there is two schools of thought with how people are approaching with using this tag, one is the "absolute" and the other is the "practical".

it seems like one of the two schools of thought have a misunderstanding on what the tag's for. *_body is used to define the color of any texture on a character's body, regardless of prominence. here, look at this post
post #1173371
how should that be tagged? it's not fur, it's not skin, what is it? the colors are certainly part of the body, though. and we're not going to create new tags for every potential body texture, it'd be a mess, both for tagging and searching.

thegreatwolfgang said:
The issue is that *_body no longer functions as a skin/coat tag anymore with its implication with *_fur/feathers/scales, as the latter trio of tags are being used for both body texture and body parts/features.

???
body parts are by definition, part of a body, though.

thegreatwolfgang said:
So, in a TL;DR sense, if you see post #2849430 as being brown_body, then everything is working as intended. If not, please read the thread further.

there's nothing even tagged brown in this post, what are you talking about? hair? because stuff like hair, facial hair, and markings are generally tagged separately from body colors.

thegreatwolfgang said:
The issue is that *_body no longer functions as a skin/coat tag anymore with its implication with *_fur/feathers/scales, as the latter trio of tags are being used for both body texture and body parts/features.

That's overstating the issue. Yes, some uses of <color>_fur/feathers/scales probably shouldn't have the corresponding <color>_body tag, but the vast majority should. And removing the implications would end up leaving <color>_body undertagged, making the tags a lot less useful. The counter to this is that <color>_body should only be tagged when it's a significant portion of the body as a whole, that it's basically being overtagged now. To which our response is that still ends up making the tags a lot less useful. Tagging a bit more than what's proper is better than tagging a bit less than what's proper, when it comes to something as basic/important as what colors are on a character. A small handful of false positives is better than a bucket load of missing results.

darryus said:
it seems like one of the two schools of thought have a misunderstanding on what the tag's for. *_body is used to define the color of any texture on a character's body, regardless of prominence. here, look at this post
post #1173371
how should that be tagged? it's not fur, it's not skin, what is it? the colors are certainly part of the body, though. and we're not going to create new tags for every potential body texture, it'd be a mess, both for tagging and searching.

I'm not removing the function of using <color>_fur/feathers/scales for tagging body colours with these BURs.
If it's vague, then you can tag <color>_body without any issue.
If it's specific, then you can tag both <color>_body & <color>_fur/feathers/scales seperately.

???
body parts are by definition, part of a body, though.

there's nothing even tagged brown in this post, what are you talking about?

I'm just using it as an example, specifically looking at the brown ball_tuft of the bottom crocodile.

hair? because stuff like hair, facial hair, and markings are generally tagged separately from body colors.

We have tuft. As well as head_tuft, facial tuft, and fur_markings, which are all bypassing your exception to body colours.

Updated

thegreatwolfgang said:
...and fur_markings, which are all bypassing your exception to body colours.

Glove_(markings) are markings. Those are tagged as body_colors. I have seen a ticket, quoted freely:"can't remove body_color" the answer was:"has glove markings". But I can't find it anymore.
And can someone finally tell me where this is coming from, that markings are not tagged with body colors, please?

watsit said:
Yes, some uses of <color>_fur/feathers/scales probably shouldn't have the corresponding <color>_body tag...

Here's a list of some probable areas for "mistagging", unless of course if one considers all of these to be part of <color>_body.
For those who don't see one of these as being part of <color>_body, then it should be of concern.

Eligible for <color>_fur/feathers/scales

Updated

dubsthefox said:
Glove_(markings) are markings. Those are tagged as body_colors. I have seen a ticket, quoted freely:"can't remove body_color" the answer was:"has glove markings". But I can't find it anymore.
And can someone finally tell me where this is coming from, that markings are not tagged with body colors, please?

generally markings are tagged with <color>_marking and which doesn't get extended to <color>_body or <color>_fur or whatever. this might not apply to all markings, however, glove_markings might be a significant enough part of the character's color scheme to be tagged as a body color, and not just a marking color.

darryus said:
generally markings are tagged with <color>_marking and which doesn't get extended to <color>_body or <color>_fur or whatever. this might not apply to all markings, however, glove_markings might be a significant enough part of the character's color scheme to be tagged as a body color, and not just a marking color.

Still does not explain where the exclusion clause (i.e., markings =/= <color>_body) originates from.
And there is no mentions of them in fur_markings, feather_markings, & scale_markings, with the fur_markings & scale_markings wiki even suggesting you to check out their other fur & scales (colour) tags.

watsit said:
People would end up using it for anything that's red. A red apple, red hair, red fingernails, etc. I've seen/had to fix cases like that in the past.

Hair and fingernails are part of the body, so they should still be included under this definition if we’re going to bother tagging body colors for really insignificant features, like fur tufts. On the rare occasion that someone wants to tag a random inanimate object in the background of a post with a color, those can probably be fixed manually, but I don’t expect that to be a common occurrence.

We already have <color>_tail, <color>_ears, etc. Of course, there's also the issue of overtagging -- tagging so many individual parts of the character that the tag list is just a mess of red_fingers red_hand red_arm red_torso red_ears etc... monotone_fingers monotone_hand monotone_arm monotone_torso monotone_ears etc... that it becomes impossible to notice any mistags or missing tags -- which is why having general <color>_body tags that can refer to any part of the body is also helpful in cutting down the monotony. The <color>_body tag is a catch-all for any part of a character's body so you don't have to tag each minor thing the same way.

Hence my suggestion (color)_body_part to cover all of those, or even to stand in as a substitute for all of those. It clearly refers to the fact that the color in question may be any small part of the body, unless body alone whose meaning is more ambiguous.

Define "significant amount". Everyone will have a different idea of what constitutes being "mostly" a color.

Define big_breasts. When are they huge and when are they medium? Define shortstack. How short and wide must a character be to qualify? Where do we draw the line? The answer to this and to all similar tags seems to be that we can all easily recognize this as a “significantly red” body and this as NOT a “significantly red” body. There will be a point somewhere in the middle where there may be some disagreement about whether it qualifies or not - but this is true of most tags used for describing characters. The answer is that we tag according to a general guideline and make judgement calls on edge cases.

It would also make it less useful for people searching when they know some part of the character was some color, but don't remember what appendage or how much.

But it would make it much more useful for people who just want to find characters that are mostly of a particular color. That seems like the more common use case, I think.

And there's the issue with clothed or obstructed characters, where a significant portion of their body is covered so you can't see that they're mostly some color.

I’ll specify that I think it should apply to the visible character. Like the examples you gave of headshot portraits - I’d just tag their body color according to what color you can see of their head. It’s the same way we tag clothing. If a bust_portrait of a character has no shirt on, do we assume they’re wearing pants? Do we tag pants? Do we tag topless (which implies clothing)? No, we just tag them as nude. We can assume from what is visible in the image about the rest of it. So if all of a visible character is red, we tag them as red. If the character’s body is canonically green, but only the head is red, we still tag it as red_body because the green body isn’t visible in the image and is therefore lore information.

On the contrary, I see them as quite useful. They help you find characters that have part of their body with a certain color. If it's restricted to certain parts of the body, or only "a significant amount" of the body, it will be less useful since a person searching can't know if the tagger otherwise though it's "enough" to tag, rather than simply tagging what you see.

But again, it would be much more useful if you’re just looking for artwork of characters with particular colors rather than trying to find a specific character that you don’t know the name of.

And anyway, this is why I was proposing having two separate tags. We could have both.

darryus said:
it seems like one of the two schools of thought have a misunderstanding on what the tag's for. *_body is used to define the color of any texture on a character's body, regardless of prominence.

Uhh… what? You can’t just say “you don’t understand what the tag is for” when the entire point of this discussion is about determining what the tag should be used for.

here, look at this post
post #1173371
how should that be tagged? it's not fur, it's not skin, what is it? the colors are certainly part of the body, though. and we're not going to create new tags for every potential body texture, it'd be a mess, both for tagging and searching.

I’d just tag multicolored_body and leave it at that. There are so many different colors there, and none of which comprise a significant portion of the body, that I don’t think having five or six different (color)_body tags for that one character is going to be of much use for anybody. On the one hand, if you’re looking for green-bodied characters, this would likely be an undesired result. On the other hand, if you’re trying to find this artwork from memory, you likely wouldn’t be using green_body to find it - you’d just use multicolored_body.

I'm for removing the implications. (specifically because I searched 'brown_fur wings wolf solo' and over half of the results are irrelevant, especially post #3304337 as it's tagged for the artist's signature/stamp/logo up in the corner)
What I'd like to know is- Will this be resolved anytime soon?
And how would be best to resolve this, a site-wide vote or?

Updated

lurkinglupinoyd said:
I'm for removing the implications. (specifically because I searched 'brown_fur wings wolf solo' and over half of the results are irrelevant, especially post #3304337 as it's tagged for the artist's signature/stamp/logo up in the corner)
What I'd like to know is- Will this be resolved anytime soon?
And how would be best to resolve this, a site-wide vote or?

It would not be that easily resolved for contentious suggestions like these. Votes are split at the moment but they don't usually have the final say.
The admins will have to decide on whether or not they want to approve or decline the BURs, often weighing in on everybody's opinions and the benefits/consequences involved.

rakustrike said:
To borrow from scaliespe:
What if we were to create new tags that highlight both?
As in:
<color>_body_[fur|feather|scales|exoskeleton]
so characters with black_body & black_scales -> black_body_scales
And if you want you can then imply:
black_body_scales -> black_body & black_body_scales

This way we can both ensure that current characters that fit the current implication is covered and those that don't gets to have their own distinction.

I don't think that is a good idea. By default, <color>_<texture> already serves that function.
However, it can also refer to the colour of the texture on individual body parts (e.g., black_feathers + feathered_wings = black_wings).

The thing people disagree here is what constitutes as a <color>_body.
Some believe it should only consist of distinct body colours that stand out the most while others believe that it should include any colour that appears on the character regardless of distinctiveness.

thegreatwolfgang said:
I don't think that is a good idea. By default, <color>_<texture> already serves that function.
However, it can also refer to the colour of the texture on individual body parts (e.g., black_feathers + feathered_wings = black_wings).

I agree overall with this point, but I also see the reason why the implication was made hence why I suggested a compromise.
I'll give a more detailed explanation just in case (not assuming you don't understand my point I need to stress, just to highlight my line of thinking):
afaik understand the current implication of <color>_<texture> with <color>_bodyis that it makes the assumption of anthros (mostly) with clear indication of said <color> covering the entire body, hence:
<color>_<texture> -> <color>_body.
E.g. take kindred_(lol) if she's depicted in her "vanilla" form then it's fairly reasonable to make the assumption of what the current implication implies:
kindred_(lol) -> white_fur -> white_body

But removing said implication would make having to add white_body manually which I can understand people want to stay given how common it can be to have anthros with a color covering the entire body.

However, said implication does collides with multiple scenarios such as:

  • animalistic humanoids
  • clothing
  • anthros with certain body parts such as wings

E.g. take ahri_(lol) if she's depicted in her "vanilla" form then it gets complicated:
She has white_fur on her tails but brown_fur on her ears.
So with the current implication she gets:
ahri_(lol) -> white_fur white_body + brown_fur brown_body.

When it probably should be:
ahri_(lol) -> white_fur + brown_fur.

Hence why I suggested we instead use:
<color>_body_<texture> as this ensure we still can imply <color>_body if we want or just use the new tag if said implication is redundant.
E.g. with kindred_(lol) it becomes:
kindred_(lol) -> white_body_fur + (white_fur -> white_body) # optional implication if wanted.

thegreatwolfgang said:
The thing people disagree here is what constitutes as a <color>_body.
Some believe it should only consist of distinct body colours that stand out the most while others believe that it should include any colour that appears on the character regardless of distinctiveness.

Wouldn't this be up to each post instead of what we think here? I don't see the unimply BUR trying to dictate what constitute <color>_body,but maybe I'm interpreting it wrong?

Updated

rakustrike said:
I agree overall with this point, but I also see the reason why the implication was made hence why I suggested a compromise.
I'll give a more detailed explanation just in case (not assuming you don't understand my point I need to stress, just to highlight my line of thinking):
...
Hence why I suggested we instead use:
<color>_body_<texture> as this ensure we still can imply <color>_body if we want or just use the new tag if said implication is redundant.
E.g. with kindred_(lol) it becomes:
kindred_(lol) -> white_body_fur + (white_fur -> white_body) # optional implication if wanted.

I get where you are coming from. Yes, it can make for a good tag for implications and making sure *_<texture> and *_body are both tagged at once.
However, with how difficult it is to make people use a new tag and the recent push to get rid of unneeded/unnecessarily excessive tags (see topic #43680), I don't see this as a good compromise.

Wouldn't this be up to each post instead of what we think here? I don't see the unimply BUR trying to dictate what constitute <color>_body,but maybe I'm interpreting it wrong?

I did implicitly state what constitutes as <color>_body on my first comment (i.e., "Thus, if someone tags the colour of these minor features..."), though I did not intend to start a debate on it.
However, since it formed the basis of my entire unimply argument, it became the main contentious point in the discussion.

The reality is that no wiki explicitly mentions what constitutes as <color>_body, neither in the body wiki nor in tag group:colors.
The only thing that remotely mentions that is in the number of body colours wiki (i.e., monotone_body, two_tone_body, multicolored_body) which classifies colours as "distinct body colors."

thegreatwolfgang said:
I get where you are coming from. Yes, it can make for a good tag for implications and making sure *_<texture> and *_body are both tagged at once.
However, with how difficult it is to make people use a new tag and the recent push to get rid of unneeded/unnecessarily excessive tags (see topic #43680), I don't see this as a good compromise.

Fair enough about making user use the new tag, that would make it difficult since *_<texture> and *_body would still be valid and I doubt admins/mods/janitors are too keen on having to babysit the new tags.

thegreatwolfgang said:
I did implicitly state what constitutes as <color>_body on my first comment (i.e., "Thus, if someone tags the colour of these minor features..."), though I did not intend to start a debate on it.
However, since it formed the basis of my entire unimply argument, it became the main contentious point in the discussion.

Ah right, maybe we should shift away from generic *_body (i.e. make it invalid tag) and instead use*_body_part since it seems to be quite divisive as it stands then?

thegreatwolfgang said:
The reality is that no wiki explicitly mentions what constitutes as <color>_body, neither in the body wiki nor in tag group:colors.
The only thing that remotely mentions that is in the number of body colours wiki (i.e., monotone_body, two_tone_body, multicolored_body) which classifies colours as "distinct body colors."

I assume it's because wiki feels more like explaining more abstract tags (say sex positions or what specie something is) and or lore character instead of a specific guideline - and the question arises if the wiki should even work as such since it's going to be subjective after a certain point.

would just like to point out that we have *_markings in addition to the *_fur/feathers/scales/body tags. I was under the impression that *_fur/etc was for when the color is a significant part of the design, but *_markings when it's a more minor part of the design.

rakustrike said:
Ah right, maybe we should shift away from generic *_body (i.e. make it invalid tag) and instead use*_body_part since it seems to be quite divisive as it stands then?

I believe we're currently considering doing away with at least a few of the *_<body_part> tags, though.

rakustrike said:
I assume it's because wiki feels more like explaining more abstract tags (say sex positions or what specie something is) and or lore character instead of a specific guideline - and the question arises if the wiki should even work as such since it's going to be subjective after a certain point.

That's the thing though. If these wikis classify colours as distinct body colours when it comes to counting colours, wouldn't it make sense that it should apply to the other <color>_body tags as well?
Say a character is tagged with:

However, when taken overall, it could only be tagged as monotone_body because only one colour stands out as being "distinct".

wandering_spaniel said:
would just like to point out that we have *_markings in addition to the *_fur/feathers/scales/body tags. I was under the impression that *_fur/etc was for when the color is a significant part of the design, but *_markings when it's a more minor part of the design.

I do believe *_fur/feathers/scales/etc. are tagged on *_markings, especially if the markings take up a major portion of the character's design.

In addition, *_fur/etc. are still tagged on any minute detail on the body (e.g., white_inner_ear_fluff).
However, since it also implies *_body, the latter also gets tagged on any minute detail on the body.

Some people on this thread argue that it is accurate and working as intended, I argued that it is impractical when it comes to searching.

  • 1