Implicating holding_cup → holding_object
Link to implication

Reason:

Do we really need this tag? It feels way too specific. I think an alias would be better. So if someone would want to find a character holding a cup, then he could do so by searching for holding_object cup

Delian said:
Do we really need this tag? It feels way too specific. I think an alias would be better. So if someone would want to find a character holding a cup, then he could do so by searching for holding_object cup

It may work sometimes, but it's not airtight and it wouldn't work for the other 19 tags that implicate holding_object. There could be a cup on a table while a character is holding a leash. It'd be reasonable to hold a condom and have a sex toy laying around in the image, but this would be reached by holding_object sex_toy, despite them not holding the sex toy.

Plus given that there's a lot of usage for holding*, people seem to like the tags.

There are quite a few missing implications though, like holding_cup -> cup. I'll compile a list.

Edit:

I'm unsure about holding_glass, since the wiki for it is about drinking glasses, but glass is only for the material. Does glass need a disambiguation? Are all glasses cups?

Delian said:
Do we really need this tag? It feels way too specific. I think an alias would be better. So if someone would want to find a character holding a cup, then he could do so by searching for holding_object cup

Like rezi said, we have a lot of holding_* tags, ranging from around hundred up to a couple of thousands of posts.

rezi said:
I'm unsure about holding_glass, since the wiki for it is about drinking glasses, but glass is only for the material.

The glass tag seems to mostly contain drinking glasses, but there is still a lot of posts with glass windows/boxes/walls. Perhaps it's a good idea to disambiguate glass like you said, even though it's a pretty big tag, and implicate holding_glass -> drinking_glass.

Could we just use holding_drink instead of holding_glass and holding_cup? It doesn't seem like a useful distinction.

Glasses can be empty, so I don't think that'd work too well.

As for implicating holding_glass to holding_cup? Don't cups usually have handles? At least, I've never thought of glasses as cups...

I'm not sure how the holding_cup tag came to be, that's the first time I've seen it. But it does seem surprisingly common, and therefore fits the guideline for creating new 'holding'-tags (...that is, only common objects deserve separate subtags, everything else should simply be tagged as holding_object).

Genjar said:
Glasses can be empty, so I don't think that'd work too well.

As for implicating holding_glass to holding_cup? Don't cups usually have handles? At least, I've never thought of glasses as cups...

I'm not sure how the holding_cup tag came to be, that's the first time I've seen it. But it does seem surprisingly common, and therefore fits the guideline for creating new 'holding'-tags (...that is, only common objects deserve separate subtags, everything else should simply be tagged as holding_object).

Cups don't have to have a handle and some glasses do have handles.

holding_cup is more common than several already approved holding tags. +1 for implication.

BlueDingo said:
Cups don't have to have a handle and some glasses do have handles.

holding_cup is more common than several already approved holding tags. +1 for implication.

Fair enough, but since those terms have different meanings in different countries (see this ), it definitely needs a detailed wiki description. The current one is a good start.

Specifically, I'm wondering if we count mugs and tankards as cups?

mug is aliased to cup. No implication for tankard yet. coffee_cup is in use but refers to two different cup types: ceramic (ie. mug) and styrofoam.

I recommend splitting coffee_cup into mug and styrofoam_cup, and implicate styrofoam_cup and tankard to cup.

Also needed: Implicate teacup -> cup and alias tea_cup -> teacup

Genjar said:
I'm not sure how the holding_cup tag came to be, that's the first time I've seen it. But it does seem surprisingly common, and therefore fits the guideline for creating new 'holding'-tags (...that is, only common objects deserve separate subtags, everything else should simply be tagged as holding_object).

Just out of curiosity, how common would a particular held object need to be to deserve its own holding_* tag?

BlueDingo said:
Just out of curiosity, how common would a particular held object need to be to deserve its own holding_* tag?

Back when that group was made, we tried to keep the categories broad enough to have minimum of fifty posts each.

...but that's, uh, got a bit out of hand since then: https://e621.net/tag?limit=50&name=holding_%2A&order=count&page=2&type=

Genjar said:
...but that's, uh, got a bit out of hand since then: https://e621.net/tag?limit=50&name=holding_%2A&order=count&page=2&type=[/quote]

Possible suggestion: holding_tableware for characters holding spoons, plates, etc. Not sure if the numbers will be high enough yet but it might be worth looking into.

I also suggest changing the mug -> cup alias to an implication because a mug is only one type of cup.

EDIT (12 hours later):
Just did a search for plate -holding_* and got 11 pages of results. I spotted 10 images of characters holding plates on the first page and estimate there are about 100 images of such, so holding_plate can make the threshold.

spoon -holding_* had only 8 pages but 15 hits on the first page (plus several images of characters spooning each other) so holding_spoon could make the threshold as well.

I recommend increasing the threshold from 50 to 100 or possibly higher.

EDIT (3 hours later):
What if the character is holding the cup via telekinesis?
post #183199

Bumping this thread because holding_cup is now tagged more times than holding_glass (934 vs. 918) and the suggestions within haven't been processed yet. Things that also need to be implicated:

• wine_glass
• shot_glass
• disposable_cup
• stein
• Other cups I've missed...

Will this be resolved anytime soon?

JAKXXX3 said:
Will this be resolved anytime soon?

Probably not, since nobody else weighted in about the naming. Can't build a consensus with just two users.

Well, cup vs glass is pretty regional and varies around the world.

That said. wikipedia defines a cup as

A cup is a small open container used for drinking and carrying drinks. It can be made of wood, plastic, glass, clay, metal, stone, china or other materials, and it might have a stem, handles or other adornments.

So, it sounds like all glasses are cups, but not all cups are glasses.

So, I would argue for holding_cup over holding_glass :)

In my experience, "cup" is a real-world umbrella word for the daily drivers of sippable drinkware. "Cup" is what we say when we don't care to be more specific, and it damn well includes (drinking) glasses lol. It's the word that gets the job done, even when it feels wrong to say. If you can't recall a better word, you might say, "Get me the- cup," and I would dare anyone suggest they didn't understand that "cup" really meant "glass" or whatever else.

Holding_glass still provides a useful distinction, but does anyone really want to see

holding_cup
holding_glass
holding_object

together in a tag list? Saying that a glass is also a cup seems too redundant. At least say what kind of glass it is, which usually is a wine_glass anyway. I think, specifically, the nondescript glasses (post #1348599) are my problem, and saying that one is being held doesn't provide enough useful information. Perhaps holding_wine_glass, holding_shot_glass, holding_martini_glass, and so on are unneeded when the obvious theme is holding_alcohol_glass? Who cares what package it comes in? Tag it separately.

Coffee is another beverage worth distinguishing with a holding_ tag if the holding_* route must be plotted. holding_cup coffee has 200 results, and coffee is generally iconic and even vital for some people. That's enough reason to justify a tag.

abadbird said:
Holding_glass still provides a useful distinction, but does anyone really want to see

holding_cup
holding_glass
holding_object

together in a tag list? Saying that a glass is also a cup seems too redundant.

Nah, no one does. I was just saying it from a general sense: Cup will do to cover the majority of beverage-holders. Glass wouldn't.

That said, some things are with distinguishing. wine glasses, tankards/steins, certainly. Though, holding_cup wine_glass would probably get very few false positives. bt that doesn't mean it's not worth having a tag of it's own.

SnowWolf said:
That said, some things are with distinguishing. wine glasses, tankards/steins, certainly. Though, holding_cup wine_glass would probably get very few false positives. bt that doesn't mean it's not worth having a tag of it's own.

The false hit rate for holding_cup wine_glass is pretty much zero right now and will probably stay that way. There's no need to make a tag for that specifically.

In fact, searching holding_cup with any cup type reliably gives you images of characters holding that particular cup. Searching [[holding_object|holding_object]] with a cup type will give you false hits.

BlueDingo said:
In fact, searching holding_cup with any cup type reliably gives you images of characters holding that particular cup.

Yep, that seems fine.
And it's why we tried to keep the categories broad when the holding_* tag group was originally created: there's no need to tag every specific item separately, as long as combo searches work.

There's probably a whole lot of other tags in holding_* that could be safely combined. (There's no need for holding_sword when holding_weapon + sword does the same job, for instance.)

Genjar said:
(There's no need for holding_sword when holding_weapon + sword does the same job, for instance.)

holding_weapon sword has a higher false hit rate (especially when projectile weapons are involved) but it's definitely not high enough to need a holding_sword tag. What would be a reasonable threshold for false hits?

BlueDingo said:
holding_weapon sword has a higher false hit rate (especially when projectile weapons are involved) but it's definitely not high enough to need a holding_sword tag.

True, there's some false hits. It doesn't seem too bad, but adding some kind of broad holding_ranged_weapon subtag could be worth thinking about.

What would be a reasonable threshold for false hits?

I'd say that it varies. For smaller tags, it's easy to browse the results even if there's some mismatches. And on the other hand, it's important to consider how much work it'd take to fix the problem. If it wold mean that we'd have to disambiguate a common tag (10000+ posts), it's usually better to think of other solutions.