I was browsing through Hardiman's work and came across the following image:

post #39497

In my opinion the tagging for this image has multiple problems.

First of all "assault" is invalid, here, because there is no actual assault happening.

Secondly, I'm not sure how "hypocrisy" and "irony" got in here, unless it was one of Hardiman's detractors hinting at the rumors that he's a pedophile.

Thirdly, there's no molestation going on in this image, although the image implies that in a (non-existant) previous image this did happen.

Fourthly, there's no rape going on here.

Fifthly, that's not a "rape face" on the bear. She's about to carve him up into little tiny pieces.

Finally, how did "toothy" become a character label?

I'll wrap this up by noting that the "penance" tag is not present, even though this is clearly Hardiman's character, Penance.

DIBS

Ok, I think I fix'd the tags. Tag what you see dude.

Yes, just hit the edit button and it'll bring up the tag edit box for you to change the tags

Well, I didn't feel like I was quite ready, yet, to jump into the deep end of the pool, although I'd really like to help out with some of the tagging problems, here.

Tell you what ... After browsing over the entire Hardiman collection (he's probably my favorite furry artist, so go figure), I've concluded that it's in bad need of a serious and intensive tagging overhaul. Especially the earlier posted works are almost devoid of tags for his most well known characters, Lori, Onyx, and Natasha, and they've been on the site for three years.

I'll see what I can do to fix some things, but some of you might want to keep an eye on me. If I screw up, just give me sharp kick in the pants and I'll mend my ways. Seriously, guys. I'm just starting to do this kind of stuff so I'll probably make mistakes, but I'm a learner.

However, I've been up for going on 27 hours and I need a couple hours of sleep before I can even think of looking at the computer screen without it going on blurry on me, so I'll put off starting this until tomorrow -- unless someone objects to me doing this project.

You go and do that, as soon as you think you got everything you bump this thread and we take a look, no one will jump on your back if you state that you are new at this and warn everyone beforehand.

You know, I was pretty certain I'd run into some quandaries when I started this project. I need some advice on the "tag what you see" rule as tempered with a little common sense. The case in point is the pool "Lori".

This is a fourteen page "story" that relates a particular dating experience in the life of our favorite Hardiman slut, Lori. So ... it's not just drawings. There's text in some of them. Now, I "see" that text, so can I use it to tag the images? Let me explain.

Hardiman introduces three new characters with this story -- Todd (a fox), and two donkeys by the names of Brian and Keith. Now, it's very clear from the text who is who in some of the images because Lori, herself, is explaining who is doing what. It seems to me that I can legitimately tag those images with the character's names.

However, what about other images in the series where, if you're reading the story panel by panel, it's trivially-obvious who is who, but you can't prove it from a given image taken all by itself?

This is why I said "tag what you see" but "tempered with a little common sense". Consider. How can anyone tag any image with a character's name unless the image, itself, explicitly identifies that character? If you take the "tag what you see" rule literally, then you can't, and that invalidates most of the character taggings which have been done on this site. I've already seen comments where people have been confused over the identity of look-alike and near-look-alike characters, so how can you be certain that the mink who looks like Minerva Mink is, in fact, Minerva Mink? Again, you can't, unless the artist, himself, has put her name into the image.

It seems to me that if you recognize a character from distinct features then you tag that character, whether or not the character is explicitly identified in the image. That's the way you guys have been doing it for years on this site. However, in the case of clearly-related series of images, posted together and grouped into pools, I think that context-related information should be used in tagging, as well.

For the case in point, I can identify who is Keith and who is Brian in every single image in the series, except for the second one. Todd, being the only fox in the series, is easily identifiable, of course, and someone has already tagged him as such in two of the images, even though he's not explicitly identified by Lori in one of them. Should I, then, go ahead and use what I think is already a widely-used "common sense corollary" to the "tag what you see rule". If so, I can properly identify the characters in this pool based upon other images in the pool. If not, then this brings into question the entire process of character tagging on this site.

Again, I know. Someone is going to say "tag what you see and not what you know", but almost every single case of character tagging on e621 is based upon what people know and not what they see. I "know" a skunk is Lori when she was drawn by Hardiman, has a white face, green eyes, and is sitting on a large insertion device. If Hardiman doesn't explicitly identify the character as "Lori" in the image, though, I'm simply assuming that what I know is actually true for the image. This is an invalid assumption if you take "tag what you see" literally. I didn't actually see it, unless I saw the name "Lori" in the image.

Yes, this opens up a can of worms, because I'm asking for official permission to be a little lax on "tag what you see", although not to any greater extent than has already been done extensively on this site, already.

Eh, tag what you see is the rule. However, like all good rules, comes with a bunch of exceptions.

The most important exception is probably the one containing additional and implied knowledge about a picture, better yet, the content in the picture.

What does this mean?
For characters in a picture it simply means that if looks like Lori, barks like Lori and behaves like Lori it gets tagged as Lori.
To identify them it is best to use the most commonly known things about a character like furpattern, clothing style, typical accessories and the likes.
This also applies to relations between certain characters, for example the incest tag is based nearly solely on outside informations.

However, this does not work with genders. These are solely tagged on the what you see rule.
The main reason why the gendertags for artica_sparkle are all over the place.

The bottom line is, if it is something tagworthy that you simply can't draw you use outside knowledge aka common sense, everything that is or could have been drawn follows the "tag what you see and not what you know" rule.

dude

you can write a three page essay for your posts

but you can't tag an image?

I have been tagging images. If you read what I wrote you'd understand why I wrote it. I encountered an issue in which the tagging rules conflicted with common sense and I reported that issue and asked for administrator permission to go lax with the "tag what you see" rule, rather than just follow the general rule with casual taggers on this site and throw the rules out the window.

There ... I summed it up for you, so you don't have to bother reading my "essay". We should all be happy, now. Have a sweetroll. It's the one somebody stole, but I found it and stole it back.

Anyway, thanks, NotMeNotYou for actually reading, understanding, and responding to my post. I'll still wait for permission from an admin to actually tag that particular pool. For now, I think I'll finish reading the forum posts and then get some shuteye, seeing as how my Internet has been flaky all day and has been down more than it's been up.

RedRaven said:
I'll still wait for permission from an admin to actually tag that particular pool.

That is a good idea.

However, sometimes our admins are a bit lazy in just jumping in. and saying "uh-huh, what he says." or "you crazy, stop spreading lies D:<".
Well, they do the later but sometimes you may need to poke one with a stick and/or hit one with an mail.
I'd recommend Riversyde , Ippiki ,Toaster or our evil Overlord .
They are all active in posting, tagging and shit.

With that said, this Wiki-Page confirms what I'm saying and uses less text for it.

Riversyde is actually Ippiki's profile.

Edit: Riversyde's profile

http://e621.net/user/show/12705

furballs_dc said:
Riversyde is actually Ippiki's profile.

Edit: Riversyde's profile

http://e621.net/user/show/12705

Damnit. :X

Thanks for catching that.

Hey I do things too, and I read your whole post, give me a bit to look up some sources to write up a proper reply, or just pm me

OK, Rainbow. I've got no life, so take your time. Uh, wait. Did I actually say that?

Anyway, standing by ...

Okay I'm home now, anyway

This wiki for tag what you see is a litle bare but still useful

Point is, yes we sometimes use outside information for a few things that need to be tagged but otherwise can't be drawn into an image. These exceptions include the date a post was made (often only found by looking at when it was originally submitted to it's first source), the artist (obviously this is really important), if characters are related to eachother or incest (often times this is identified by looking at similarities in appearance but can sometimes use outside information within reason, example if a furry and a scalie are engaged in intercourse and the artist says they are related, this is too far a stretch for our tagging purposes unfortunately) and any distinct characteristics that can can be identified as belonging to a certain character

Obviously this last one can be a little ambiguous at times, especially regarding comics/pools. Just try your best at only tagging what you think would be very relevant to the image at hand when it comes to character tagging. We just try to tag it if it appears as though it has enough characteristics of a certain character to be associated and easily and consistently identified with said character.

As it comes to pools, well think of it this way; often times when people search for images, they are not concerned about story or context, so they usually would disregard if it's in a pool or not. That said, I see how that would be handy to have characters mentioned in context to be tagged, but as it stands I honestly see it doing more harm than good in terms of regular searches, of which is the main target of our tagging system

I do appreciate you bringing an in-depth discussion to the table, and I apologize if they weren't very well received before, but...
[EDIT]
...I asure you I do read all your posts and reply as best I can

Rainbow, everything after "but" got cut off. I feel as if the Sword of Damocles is hanging over my head, now, and I don't know why.

Anyway, back to seriousness. I'll quit being so hypothetical and get directly to the point at hand. I know what follows will be regarded by many Internet denizens as a "wall of text" (don't people actually read any more?), but bear with me. I think my reasoning, here, is important in understanding why I think it's appropriate to tag each image in the pool with the proper character names.

The fox character, Todd, is mentioned by Lori in the second panel. She doesn't actually say that Todd is a fox, and part of the head and face of an equine also appears in the image. She does mention that Todd invited two of his friends along, and he mentions that they're "jackasses". In Lori's words, "I didn't realize he meant 'jackass' as in 'mule.'" This, and the fact that "todd" is a British synonym for "fox", pretty much nails the fox as "Todd", yes? This is what I meant by using text, and not just art content in an image as tagging references.

In the third panel she menions giving Brian (newly-introduced) "a little mouth work". We see her perfoming fellatio on a rather large penis (which she also commented on in relation to Todd's friends), in an inset to the image, so that's probably not Todd's penis. In the image, itself, Todd is shown spreading her vulva and peering inside, but she doesn't actually mention Todd. In addition, an equine is shown extending his tongue toward her anus.

Following my logic of contexting tags within pools, I can legitimately tag Todd and Brian. We know that Brian is at her head, so that must be "the other guy" behind her, but we don't know his name, yet.

In the fourth panel we see "the other guy" and Todd clearly portrayed, as well as Brian's penis (which is all we've seen of him thus far). This is our visual confirmation that the inset in the third panel is of Brian and not the "other guy".

In the fifth panel we see "the other guy" performing analingus on Lori, and a black hand/forearm in the process of fisting her vaginally. Since Todd was already fingering her and "the other guy" doesn't have any black fur, we can safely assume that's Todd, even though we don't get confirmation of that until the sixth panel.

It isn't until the eleventh panel that Lori mentions the name of "the other guy" (Keith). At this point, since there is very good continuity in the events depicted in all the panels when they are viewed in proper order, we have all the information to deduce the names of every character (or character part) in every image, except the second one, although I'd bet that was Brian, since he's shown in front of Lori, and Keith is likely out-of-frame to the right.

Anyway, that was my reasoning, and in pools like this one which are obviously telling a story because they are associated with text relating Lori's own words about the events, it seems reasonable to consider the pool as a whole for tagging purposes, rather than just each individual image. I seriously doubt that many people chancing upon one of the panels with text wouldn't follow up by looking at the entire pool.

Furthermore, proper tagging would require the creation of the following tags:

todd_(jmh)
keith_(jmh)
brian_(jmh)

This is just to avoid confusion and is in keeping with the way we have explicit jmh tags for some of Hardiman's other characters.

See the thing is, a lot of people actually do just search for individual images, otherwise I can see using a small amount of outside information to tag characters mentioned, but only if they are present

So, based upon my logic in acquiring that "outside information" directly from the horse's (er, skunk's) mouth, and relating it to other images in the story, then I can tag Todd, Brian, and Keith in all the images?

That's what all this boils down to, and may be relevant to the other two portfolios of the skunkettes (Natasha and Onyx pools), since these are written in a similar manner. If the answer is "yes", then I need those tags created.

If "no", then I'll go with that, even though it'll mean removing one of the existing "todd" tags. That tag may be currently doing multiple duty, but seems to be used mostly for a character from Daniel P. Mannix's The Fox and the Hound and Disney's spin-off of that novel. This is why I'd like a separate tag for JMH's "Todd" character.

hmm, here link it so I can see exactly what you mean. I don't like making a call unless I am absolutely sure

pool #956

As a related aside, I have a black and white version of this in which the cover panel (which I refer to as "panel 1") is missing and the second panel takes its place as the cover, but modified with the title "Three's Company". This was on a computer I inherited from a friend who recently passed away and she wanted me to have her furry art collection, so I got the entire computer, minus some of her more private stuff. I also have alternate versions of several other JMH series and a bunch of miscellaneous images he did that don't seem to be on here.

Anyway ... I'm not quite sure what to make of that. I assume the stuff that's already on this site is OK to be here or Jim would have protested about it, seeing as he's a member. I'd upload this other stuff, but I really don't think that would be ethical, since I don't have his explicit permission to do so.

The issue is that I don't know which came first -- the Lori portfolio or the "Three's Company" story. Were we to have both one would have to implicate, or perhaps even be an alias of, the other, I'd think, perhaps with images of both series merged into a single pool.