Tag implications: panties

@Shinjidude: Hmm, please don't take this as a personal...

2012-02-01T07:24:54-05:00

Hmm, please don't take this as a personal attack, because it's not intended as one, but a six screen comment with a preamble seems excessive, especially when you are more concerned with the way you were responded to rather than the matter at hand.

It would be more constructive to explain why you think forced recursive tagging would be useful.

As for "A depiction of something is different than an instance of that thing.", I don't really see how this was that unclear or ambiguous. You probably understand what I meant by now, but to be completely transparent:

In a given reality a depiction is a visual reference to something (an object that can be named), wheras an instance of something (an object that can be named) is the actual object itself. Those pretty obviously are not the same thing (you can't eat a depiction of a strawberry). Perhaps, you didn't realize I was talking about the reality within the context of the image (which is pretty much the only thing we tag). Everything on Danbooru is a depiction within our own realities, but that's sort of meaningless as far as tagging goes.

As for being "wrong", I think that's also tough to defend. I said such implications would cause tags to be used in ways they were not intended to be used, and that it would by necessity dilute parent tags (for most users as far as I can tell, strawberry_print panties in a query for food would simply be noise and would dilute the actual instances of food that tag was meant to refer to). You could argue that my opinion of how tags are meant to be used is different than yours, or different than that of Danbooru users as a whole (though I don't think it is). It's really hard to defend this sort of statement as being *wrong* though. Especially to the point of insisting that fact after it has been explained.

Anyway I'm not sure I really see the point to continuing this. We are discussing semantics and rhetoric (which really isn't useful to anyone), rather than matter related to the site. If you notice people abandon a thread after a long response such as the one I am responding to, that is probably the reason. I probably shouldn't have responded myself.

@Danielx21: So the implication strawberry_print ->...

2012-02-01T04:33:39-05:00

So the implication strawberry_print -> strawberry won't exist.

I give in either to the democratic consensus of everybody else here, or to the dictatorial decision of jhx. (Sometimes, I'm not sure which.) Both seem reasonable to some extent, especially since even Alignn — the first to defend the implication in the thread — eventually changed his/her mind.

Now I'm going to reply to NWF, since he/she did three things in this thread that caught my attention: 1) honestly, raised some interesting and reasonable points; 2) other times, I'm still not sure what's the point, so I'll ask; 3) really annoyed me with their tone and accusations.

I do defend that implication, this time not hoping that it'd be done, but to continue a good discussion to be archived; and to clarify my thoughts and even question the current outcome.

I don't always have time to dedicate to conversations on Danbooru, but when I do I'm prepared for potentially very long discussions. Sometimes, people share that determination and together make threads with multiple pages; most often, they abandon threads with none or few replies, even when the subject is controversial or when I raise a point. (I chose the first-person pronoun in "when I raise a point" because I'm focusing on my own experiences, but maybe "when points are raised by any person" would be equally informative.) Let's see what happens this time.

Now to my replies.

Reply I

NWF_Renim said:
For example if you attempted something like strawberry_print implies strawberry, then somewhere along the line you're going to need to break an already existing implication, such as breaking the implication strawberry -> fruit or fruit -> food.

Both the original projects are simple: either we do implicate "strawberry_print -> strawberry" or we don't. What you said is so complex it's horrible. It's apparently based on the opinion that prints and "real" things should be kept as SEPARATE tags, while they are forcibly MERGED nonetheless. Having no "strawberry_print -> strawberry" implication at all is really much better than cultivating that mess. On the other hand, if the implication had been done, then I'd expect the other implications to be kept: post #974144 would be tagged strawberry panties, strawberry print, strawberry, fruit and food. (And a search for food would bring that post up, among many others.)

Reply II

NWF_Renim said:
At some point we need to draw the line of what is "real" in the image and what isn't

Which of these is the apple of post #735008?

Why someone would conceivably want to avoid seeing post #743257 when searching for cat? Or post #697883 when searching for strawberry? Sometimes, the print is more-or-less realistic or is the focus of the image. Would these be "grey areas"? If not, are there any grey areas for cats/cat_prints and strawberries/strawberry_prints?

I suppose you could always narrow down your search to find what you want. (and maybe you should sometimes, since some of these tags are flooded with pages and pages of posts) Do you know of any trouble with print/nonprint happening to someone before?

Reply III

Shinjidude said:
A depiction of something is different than an instance of that thing.

Danielx21 said:
That wording is somewhat confusing, but I disagree on principle. In drawings, all instances are depictions.

NWF_Renim said:
You choose to word this so it does not match up with what Shinjidude said.

No. I quoted exactly what Shinjidude said and said exactly what I thought. I still see that the association of words "depiction = print; instance = nonprint; something = either a print or a nonprint" is weird (thus a confusing wording).

Being "confusing" is not a crime; confusing words can simply be clarified. But, anyway, I understand the discussion well enough, so I don't really require an explanation.

Reply IV

NWF_Renim said:
Yes, all instances are depictions, but not all depictions are instances.

Yes, that's why I proposed an IMPLICATION "strawberry_print -> strawberry"

It'd naturally work this way:

strawberry_print -> prints/depictions
strawberry -> nonprints/instances & prints/depictions

Now, here's a crazy logical reasoning, that probably has no practical use but serves right as part of my line of thought:

"If all depictions were instances and all instances were depictions, then depiction would be a synonym of instance and strawberry_print would be an ALIAS to strawberry."

Reply V

Danielx21 said:

I understand you have great knowledge of Danbooru's tags and practices, but your statement "it would screw up the way the tags work and unnecessarily dilute one to include things that don't quite apply to that tags' intended usage" is incorrect, because it implies we have a standard, popular, usage criterion and perhaps a huge number of separated "print" and "nonprint" images to be diluted somewhere. We don't.

NWF_Renim said:

What a nice way to sugarcoat what is pretty much you saying "for all the experience you have, you're wrong, and I'm right."

That's not sugarcoating, just good manners. And applicable context. Shinjidude has the merits I mentioned. You, too, NWF. IIRC, I've been in good terms with you both, and talked nicely with you, until you (NWF) recently started to see malice in my words and attack them. Please stop, by giving me the benefit of doubt when we have a disagreement.

As to Shinjidude's statement that I said is incorrect, it really is. Or, rather, it was incorrect when he/she said it. Please don't take it personally.

1) That practice has never been never discussed on the forum. (but now it is, on this thread)
2) The tags in question (strawberry_print, etc.) have few posts, which suggests that the tag is relatively recent and/or unused. (well, I know that "few" is subjective; more on that later)

Reply VI

Danielx21 said:

Almost all instances of types of "print" ("strawberry print", etc.) have 152 posts or less, which is a small number. Most of these tags have 35 or less posts.

NWF_Renim said:

Whoever said 152 posts is a small number? I don't consider it a small number.

It seems obvious to me that 152 is a small number. But, here on Danbooru, I already had hostile disagreements about the meanings of subjective quantitative words such as "many". Taking back that word is impossible because the whole context is subjective as well, so I'll just elaborate.

The (approximately) 5,200 most populated tags simply have more than 150 posts.
Strawberry_print has 150 posts, or 8 pages of posts. Seeing them all at once can be so easily and quickly done, because "strawberry_print" has few posts.
Strawberry has 2433 posts. Strawberry_print has 6% as many posts as strawberry, which I believe is a low number in this context. The practical result is:

If strawberry contains both PRINTS ("depictions") and NONPRINTS ("instances") together and someone, for whatever reason, wants to look for NONPRINTS only, he can just ignore the few PRINTS that would appear. They don't make enough noise to warrant a full separation between these two tags. And you wouldn't even need to consider that print/nonprint distinction.
With that in mind, don't forget that I said Most of these tags have 35 or less posts, so these arguments would apply to them as well, in larger scale.

Reply VII

NWF_Renim said:

It's also interesting that you're bringing up post numbers here, when in another thread you bragged on about how tags were underpopulated until you started going through and populating them. If it's true that tags are underpopulated, then you using the current number values now is rather pointless as because that just means that there are plenty of images that aren't tagged. So which is it? Do we have an issue of underpopulated tags or not?

1) Tags that have been created relatively recently or are relatively unused (like strawberry_print and leaf_print) often have a small number of posts, which is a key fact of my conclusion that the "print" tags had not received enough attention of the community. It does not matter, for this line of thought, that they may be better populated in the future. That said...
2) True enough. If there are drawings of strawberry prints NOT tagged strawberry_print, then they may be populated. I'd like to see that happen.

Reply VIII

Danielx21 said:

Point C:

The exceptions to that "152 posts or less" rule are:

paw print, floral print, cow print and tiger print

These exceptions have 640 or more posts. They do not suggest that we should have a distinction between "depictions" and "instances". (or "print" and "nonprint") They have entirely different meanings.

paw print does not mean "a drawing of a drawing of a paw"; it means "the mark a paw makes when moist with ink and pressed on the ground"

floral print does not mean "a drawing of a drawing of a "flower"; it means "symbolic bidimensional patterns based on real-life flowers"

cow print does not mean "a drawing of a drawing of a cow"; it means "the black/white pattern of spots found in cow fur"

tiger print does not mean "a drawing of a drawing of a tiger"; it means "the black/orange pattern of stripes found in tiger fur"

The meanings were typed by me, for this thread, based on the contents of each tag.

NWF_Renim said:

We've known for awhile that not all prints depict the actual item. Trying to compare things that have a similar name but a whole different patterning type is silly. Don't try and use that as an argument against what Shinjidude said, because you're trying to use something that is completely different. Stick to the pattern types that are actually relevant.

1) You seem to have read something completely different from what I wrote there. I'm curious: how did I tried and failed to use paw_print, floral_print, cow_print and tiger_print (let's call them PFCT) against Shinjidude?
2) You are completely right that PFCT are different from cat_print, strawberry_print, leaf_print and the like (let's call them PP = print patterns). That was EXACTLY my point: these are different groups of tags, so don't confuse them with each other. I had to say that because:

The PFCT have much more posts (640 or more), in contrast with my conclusion that the PP have few posts.
In fact, the PFCT are older than ALL the PP, indicating that having only the PFCT was the standard practice before the PP were introduced.
Now that I mention it, I could even request moving cow_print to cow_pattern, to look for some consistency. This idea is very on-topic, but it's off-topic enough to be better discussed elsewhere. Although, after this discussion that was initially about panties, I'm not caring much about choosing the right thread with surgical accuracy, at the moment.

@jxh2154: Prints of concrete physical entities like bears...

2012-01-28T14:08:34-05:00

Prints of concrete physical entities like bears or strawberries will not get implications to the objects themselves. We might only do that for more abstract/geometrical concepts like "stripes" or "polka dots" or "stars" (sort of its own special case) whatever.

Anyway, did the list in the first post.

@Alignn: > NWF_Renim said: > [...] all instances are...

2012-01-26T12:30:34-05:00

NWF_Renim said:
[...] all instances are depictions, but not all depictions are instances. [...]

Hm, that all makes a lot of sense. I'll side with that approach.

@Shinjidude: > NWF_Renim said: > For example if you...

2012-01-26T10:57:51-05:00

NWF_Renim said:
For example if you attempted something like strawberry_print implies strawberry, then somewhere along the line you're going to need to break an already existing implication, such as breaking the implication strawberry -> fruit or fruit -> food

This is one major issue. If I am querying food I certainly am not interested in, nor expect to see strawberry_print panties.

NWF_Renim said:
Yes, all instances are depictions, but not all depictions are instances. If you only want to tag instances that the item exists in an image, that is completely different from tagging all depictions of the item that exists in an image. If the bear tag is defined as all instances of a bear existing in the picture, than depictions of a bear (photo with in the image) would not be tagged, because they aren't an entity within the image. They're a depiction of the item, but they aren't an instance of the item.

This is the crux of my objection.

BCI_Temp said:
If I'm searching for an animal, a fruit, or a flower, I don't want images that don't actually contain any of the above. That's it, pure and simple.

This is my opinion as well, and is in line with Danbooru's tag use thus far.

We certainly *could* decide to tag depictions within depictions recursively, but I would argue that doing so would not be useful, and would simply add noise to the majority of queries.

I think it is more useful to say that we chiefly tag only instances within the outermost depiction (the image drawn and posted on the site). Things that actually exist in the reality depicted in the image.

There will always be grey areas and exceptions. I'd probably tag a character depicted within a depiction (say someone on a TV screen). However, by and large I'd suggest we tag only things in that outermost reality, and certainly suggest we not make implications that out of necessity make that jump down to depictions within depictions.

@BCI_Temp: > Danielx21 said: > If you see fit introducing...

2012-01-26T08:29:31-05:00

Danielx21 said:
If you see fit introducing the new rule of absolutely distinguishing between bear print and bear, I'd like to know why, please. Especially, why anyone searching for the latter would not want to see the (currently few) instances of the former.

If I'm searching for an animal, a fruit, or a flower, I don't want images that don't actually contain any of the above. That's it, pure and simple.

The print tags are for representing patterns on clothing. Some of these patterns represent an actual image of a full item, some do not. That is not the point. The point is that they are clothing patterns and not objects physically present in the scene. A picture with bear print panties doesn't necessarily have a bear in it.

While I personally can't see *myself* actually searching for any of the above, I'd rather have the descriptive distinction present in case others have a need for it. Sometimes they're useful for finding a character whose name you don't know but whose appearance you do.

@NWF_Renim: To be honest, such implications in the end...

2012-01-26T05:36:39-05:00

To be honest, such implications in the end result in a necessity to break or prevent higher level implications to be able to enable proper sorting of images. For example if you attempted something like strawberry_print implies strawberry, then somewhere along the line you're going to need to break an already existing implication, such as breaking the implication strawberry -> fruit or fruit -> food. At some point we need to draw the line of what is "real" in the image and what isn't. In the case of the bear tag, having something like bear_print imply bear means that we need to make sure then that we do not create an implication like bear -> animal. If such an implication is made, then the more general tag gets flooded with non-"real" animals and devalues the tag. It also forces unnecessary propagation of tags, because then someone will go "well we need a tag for real animals" which could have been accomplished simply by restricting the animal tag to real animals to begin with.

Danielx21 said:
That wording is somewhat confusing, but I disagree on principle. In drawings, all instances are depictions.

You choose to word this so it does not match up with what Shinjidude said. Yes, all instances are depictions, but not all depictions are instances. If you only want to tag instances that the item exists in an image, that is completely different from tagging all depictions of the item that exists in an image. If the bear tag is defined as all instances of a bear existing in the picture, than depictions of a bear (photo with in the image) would not be tagged, because they aren't an entity within the image. They're a depiction of the item, but they aren't an instance of the item.

Danielx21 said:
I understand you have great knowledge of Danbooru's tags and practices, but your statement "it would screw up the way the tags work and unnecessarily dilute one to include things that don't quite apply to that tags' intended usage" is incorrect, because it implies we have a standard, popular, usage criterion and perhaps a huge number of separated "print" and "nonprint" images to be diluted somewhere. We don't.

What a nice way to sugarcoat what is pretty much you saying "for all the experience you have, you're wrong, and I'm right."

Danielx21 said:
Almost all instances of types of "print" ("strawberry print", etc.) have 152 posts or less, which is a small number. Most of these tags have 35 or less posts.

Whoever said 152 posts is a small number? I don't consider it a small number. It's also interesting that you're bringing up post numbers here, when in another thread you bragged on about how tags were underpopulated until you started going through and populating them. If it's true that tags are underpopulated, then you using the current number values now is rather pointless as because that just means that there are plenty of images that aren't tagged. So which is it? Do we have an issue of underpopulated tags or not?

As for your point C, I'm not seeing the relevance. We've known for awhile that not all prints depict the actual item. Trying to compare things that have a similar name but a whole different patterning type is silly. Don't try and use that as an argument against what Shinjidude said, because you're trying to use something that is completely different. Stick to the pattern types that are actually relevant.

@Danielx21: Point A:> Shinjidude said: > A depiction of...

2012-01-26T04:26:05-05:00

Point A:

Shinjidude said:
A depiction of something is different than an instance of that thing.

That wording is somewhat confusing, but I disagree on principle. In drawings, all instances are depictions.

Both these hypothetical images would be drawings of bears:

1. A drawing of a bear chasing a girl.
2. A drawing of a girl wearing panties with a drawn bear.

Relatedly, Alignn said above when talking about strawberries: "The image [post #974144 or post #957006] contains a depiction of food (for danbooru's definition of food, which includes fruit and strawberries), so unless we specifically don't tag the contents of pictures in pictures, it's accurate."

If bear print specifically means "a bear drawn on some drawn object (e.g. a bear drawn on panties, on shirts, etc.)" then I'd expect all instances of it to hav the tag bear, too.

Point B:

I understand you have great knowledge of Danbooru's tags and practices, but your statement "it would screw up the way the tags work and unnecessarily dilute one to include things that don't quite apply to that tags' intended usage" is incorrect, because it implies we have a standard, popular, usage criterion and perhaps a huge number of separated "print" and "nonprint" images to be diluted somewhere. We don't.

Almost all instances of types of "print" ("strawberry print", etc.) have 152 posts or less, which is a small number. Most of these tags have 35 or less posts.

If you see fit introducing the new rule of absolutely distinguishing between bear print and bear, I'd like to know why, please. Especially, why anyone searching for the latter would not want to see the (currently few) instances of the former.

Point C:

The exceptions to that "152 posts or less" rule are:

paw print, floral print, cow print and tiger print

These exceptions have 640 or more posts. They do not suggest that we should have a distinction between "depictions" and "instances". (or "print" and "nonprint") They have entirely different meanings.

paw print does not mean "a drawing of a drawing of a paw"; it means "the mark a paw makes when moist with ink and pressed on the ground"
floral print does not mean "a drawing of a drawing of a "flower"; it means "symbolic bidimensional patterns based on real-life flowers"
cow print does not mean "a drawing of a drawing of a cow"; it means "the black/white pattern of spots found in cow fur"
tiger print does not mean "a drawing of a drawing of a tiger"; it means "the black/orange pattern of stripes found in tiger fur"

The meanings were typed by me, for this thread, based on the contents of each tag.

@Shinjidude: I think a bear print → bear implication would...

2012-01-25T07:44:41-05:00

I think a bear print → bear implication would be a bad idea. A depiction of something is different than an instance of that thing. It's similar but not quite the same as manga vs manga_(object) or marker_(medium) vs marker.

If we had an implication either way for either of those, it would screw up the way the tags work and unnecessarily dilute one to include things that don't quite apply to that tags' intended usage.

@Danielx21: > BCI_Temp said: > That would then defeat the...

2012-01-25T07:08:31-05:00

BCI_Temp said:
That would then defeat the entire point of the *_print tags which was to avoid implicating everything the tags for the objects printed on an item implicates.

e.g. strawberry print does not implicate fruit and food. This is by design.

I know. But I'm not interested in contesting that practice now. "Most certainly I'll even ask for implications of bear print -> bear, etc. in the future." means that most certainly I'll contest it in the future.

@Shinjidude: > BCI_Temp said: > Out of curiosity, how do you...

2012-01-24T09:08:34-05:00

BCI_Temp said:
Out of curiosity, how do you tell the age and use of a tag anyway?

Experience in using the site over the years.

Specifically in this case c-string (and strapless bottom) were brought up in the forum at least a year ago in forum #36187.

Maebari was established much earlier (before I joined, I believe, which is before this forum was created). It is first referenced in the current forum in forum #882 over four years ago.

@BCI_Temp: > Danielx21 said: > True. Most certainly I'll...

2012-01-24T07:45:50-05:00

Danielx21 said:
True. Most certainly I'll even ask for implications of bear print -> bear, etc. in the future. Meanwhile...

That would then defeat the entire point of the *_print tags which was to avoid implicating everything the tags for the objects printed on an item implicates.

e.g. strawberry print does not implicate fruit and food. This is by design.

@Danielx21: > EB said: > If these are implicated to...

2012-01-24T07:14:35-05:00

EB said:
If these are implicated to anything, it should be *_print tags instead (for instance, bear_print and strawberry_print which already exist).

True. Most certainly I'll even ask for implications of bear print -> bear, etc. in the future. Meanwhile...

Edited the intro to redo the system of "print" panties, this way:

The old replaced system was:

And I added:

@BCI_Temp: Out of curiosity, how do you tell the age and...

2012-01-24T06:38:33-05:00

Out of curiosity, how do you tell the age and use of a tag anyway?

I've done some reading up and tried to update the Wiki for it to help better tell the two apart, but I ran into strapless bottom, which muddies the waters further, and it looks like images that belong better with that tag are going into c-string as well.

As for its relevance to panties, I would say it should not implicate them since it seems to be classified more as a thong bikini than as panties. You wear it to avoid tanlines. I doubt it's comfortable enough to use as underwear.

@Shinjidude: Yes, they have a metal frame, so provided the...

2012-01-23T20:52:29-05:00

Yes, they have a metal frame, so provided the artist knows how they work, they should be mutually exclusive.

In general if it has a rounded tapered shape or flutes outward, it is not maebari which are usually rectangular

As for the tags being created, they have both been in use for a long time already.

@BCI_Temp: So, are they mutually exclusive with a...

2012-01-23T11:53:33-05:00

So, are they mutually exclusive with a cameltoe? And is there a good rule of thumb on the size in an image that doesn't contain both to tell them apart?

Personally, I think the two tags are just going to create a mess of confusion.

@Shinjidude: > BCI_Temp said: > I saw that tag yesterday and...

2012-01-23T10:48:09-05:00

BCI_Temp said:
I saw that tag yesterday and was wondering just how the hell you tell it apart from a maebari anyway.

They are usually slightly larger, and have more shape to them. If they are not being worn (which they always are), they are rigid rather than being adhesive strips.

@BCI_Temp: > kittey said: > The c-string disagrees with...

2012-01-23T04:45:46-05:00

kittey said:
The c-string disagrees with you and might need an implication to panties as well. Or at least to underwear if it doesn’t count.

I saw that tag yesterday and was wondering just how the hell you tell it apart from a maebari anyway.

EB said:
If these are implicated to anything, it should be *_print tags instead (for instance, bear_print and strawberry_print which already exist).

Ditto.

@EB: > Danielx21 said: > > * bear panties -> bear >...

2012-01-22T11:37:59-05:00

Danielx21 said:

bear panties -> bear

bunny panties -> bunny

strawberry panties -> strawberry

If these are implicated to anything, it should be *_print tags instead (for instance, bear_print and strawberry_print which already exist).

@kittey: > Log said: > Strapless panties doesn't even...

2012-01-22T09:52:42-05:00

Log said:
Strapless panties doesn't even exist.

The c-string disagrees with you and might need an implication to panties as well. Or at least to underwear if it doesn’t count.