Danbooru

Requiring copyrights for all "connection" tags, even with "memes"

Posted under General

Because of discussion in previous forum threads (see forum #80841 and the threads cited there, among others), some taggers go out of their way to remove copyright tags from images considered "memes" when it doesn't involve characters from the series parodied. I don't think images that establish connections between series should be subject to this kind of removal. For instance, when I search chuunibyou_demo_koi_ga_shitai! seiyuu_connection, I should be able to turn up post #1277698. I shouldn't have to search ~chuunibyou_demo_koi_ga_shitai! ~sparkling_daydream seiyuu_connection to get all the results I'm looking for. I brought up this argument in forum #68716, yet the image I made an example of there apparently still had my tag edit reversed later. So, I thought it would be worthwhile to bring it up here for further discussion and a final decision so we have something to point to and avoid tag edit wars.

Updated by jxh2154

Well then, I tried.
Let's face the facts - this problem with current tools we have (ie. current Danbooru) is not 100% objectively resolvable. We need something like secondary tags - for everything with weaker or not so obvious connection, like, for example meme origins, cameos and few other things I can't even imagine right now. And with appropriate search choices implemented (by all tags, or only by "primary" ones).

Coming down to earth, we have following choices now:
1. keep de facto anarchy
2. use 'meme is not directly connected with its origin' rule
3. use 'meme is always connected with its origin' rule
4. depending of meme, deciding for every case separatedly (on what basis?)

Solution nr 2 gives us cleaner database and additional search possibilities. Of course, it's not flawless because of potential search problems mentioned here by EB or jjj14.

Solution nr 3 removes these problems, but it also removes any possiblity of doing search of "hard" connection between meme and anime it's been spawned from (see mirai_nikki yandere_trance example now). It also can "pollute" anime copyrights queries. Also, note that in such case there will be no reason not to implicate every meme to its origin(s). Please check this list if you really think it'll be alright. There are also previous discussions at this forum concerning this matter I'd suggest to check (especially if someone is relatively new here).

Finally solution nr 4 - I think we've been there already, and we've failed. The reason is simple - it's hard to think about objective rule when to do an implicaton and when not to. Consequently - good luck to expect an average danbooru user to learn using them correctly. Not to mention about (undoubtely charming) prospect of all these discussions about every new meme when it only arises.

Personally, I'm for solution nr 2. But it'd good if any solution, other than current anarchy was finally accepted. BTW, lack of definite resolution in this case would only mean staying by nr 1.

Updated

I'd lean towards #3, as EB, Log, & jxh2154 seem to support. Tag pollution can be a tricky issue, but for most copyrights it won't be an issue at all, and for those where it is a simple exclusion will almost always fix it in this case.

I'm not going within a mile of the broader meme issue, I was just speaking of connection tags.

The broader issue will never be solved to everyone's content and I've got issues with any manner other than one that's so stupidly complex nobody will agree to it anyways.

richie said:
Solution nr 3 removes these problems, but it also removes any possiblity of doing search of "hard" connection between meme and anime it's been spawned from (see mirai_nikki yandere_trance example now). It also can "pollute" anime copyrights queries.

Making a new tag for parodies within the same copyright would be one solution to that problem (I brought it up in forum #50670). Then you could search mirai_nikki meta-parody and get the same results. Those concerned with parodies of that scene overwhelming their results could exclude them. Provided parody and meta-parody are completely separate tags, you can even keep the in-series parodies in the results when searching mirai_nikki -parody.

I will admit that being able to separate out parodies within the same copyright and those involving other copyrights is helpful with some tags. cinderella_girls_card_parody really needs it for instance IMO.

I think it's better to only require addition of copyrights with connection and parody tags, but not memes that have their own tag.
Memes, after they take off, are quite independent from the copyright that spawned them. Some are becoming even larger than the original copyright, e.g. yaranaika. IMO, it's really wrong to force people to negate them every time they are checking up a copyright.

It is basically this:
With copyright tag required: You need to constantly negate the meme when you are not looking for it;
With copyright tag not required: You need to additionally check the meme tag, or add it to the search, if you are looking for it.
And I think that the latter is much more convenient than the former.

Edit: After looking at the yaranaika tag I mentioned before, I noticed that the way it is currently used would be impossible if "meme -> original copyright" rule is established. While without it one still can search for ~yaranaika ~kuso_miso_technique.
In other words, such rule would also remove options that we currently have.

Updated

In the case of Kuso Miso Technique, I wouldn't say that the meme is larger than the original copyright. The original copyright pretty much is a meme. There's barely anything connected to it here that's not a parody of some sort.

I've grown somewhat less receptive to the "If we add 'x' tag, 'x' will overrun the original 'y' tag." school of thought. More and more I find my reaction to that is "That just reflects the output of the fanartist collective. It is what it is."

I'd rather have the discussion focus on whether having copyright tags alongside their meme tags is good or bad, conceptually, without considerations of tag counts.

1