Upload sniping / Uploading with few to no tags

IMHO this can only be properly addressed by moderating. An automated solution might guarantee things like a minimum number of tags on a post, but someone still has to judge whether those tags were appropriate and/or adequate, and if that is consistently not the case for some users, then act upon it.

...To think though, that nowadays, there's no better part of my Danbooru browsing routine -apart from looking at the actual images- than the feeling of relief I get when iqdb tells me that someone else already uploaded what I was going to upload and saved me all the time and trouble, haha.

Hmmm... what about some kind of reward/penalty system for tagging of uploads. Have a standard cooldown period between uploads for all users (e.g. 3 minutes), and the amount of tags that get added will affect the cooldown period.

http://isshiki.donmai.us/user-reports/taggers/%21CURRENT.html

Seen in the above, a tag count of ~30 could be decided to be the standard.

Scenario #1.

15 tags get added to an image = 30 /15 = 2 x multiplier
Cooldown period = 3 minutes * 2 = 6 minutes

Scenario #2.

45 tags get added to an image = 30 / 45 = 2/3 x multiplier
Cooldown period = 3 minutex * 2/3 = 2 minutes

Note: The tag count of 30 and cooldown period of 3 minutes in the above example were just for illustration purposes.

Ehh, not a fan of that... I burst upload a lot (finish tagging everything then uploading with very short delays), and I know a number of other users do it too so to have that hamper me would be pretty annoying.

A cooldown before is fine, but after is... sketchy, to say the least. Annoying if I'm being honest. I still prefer kittey/evazion's suggestions.

I agree with Mikaeri.
I guess that will have the opposite effect: Screwing away people.

This is not a solution of any kind, but rather something that might be helpful.

What about not merging/combining the very first revision in the tag history? So that it would be very easy to see what tags were added on upload.

I know that it is already possible to find this info, but it's not readily available, at least for older posts. This should help to notice users that consistently undertag their initial uploads. Investigating older posts by a particular user to check his tagging habits will also be easier.
On a related note, this particular issue was previously brought up in topic #9127/p38, albeit for a different reason.

Didn't @BrokenEagle98 want to install a report for this kind of thing?

MyrMindservant said:

This is not a solution of any kind, but rather something that might be helpful.

What about not merging/combining the very first revision in the tag history? So that it would be very easy to see what tags were added on upload.

I know that it is already possible to find this info, but it's not readily available, at least for older posts. This should help to notice users that consistently undertag their initial uploads. Investigating older posts by a particular user to check his tagging habits will also be easier.
On a related note, this particular issue was previously brought up in topic #9127/p38, albeit for a different reason.

I know that, when the API was acting up, people were tagging more before uploading because their 1-2 tag uploads would be exposed for a longer period of time, so apparently they do have some shame about min-tagging.

Provence. said:

Didn't @BrokenEagle98 want to install a report for this kind of thing?

Ah, you mean forum #119287...? I got considerable pushback when I first introduced that concept, so I stopped all work. Also, it would be a bit unlike all other reports, since the availability of the data needed is short-lived as upload records get deleted after 36 hours. Therefore, it would be impossible to prove that the report was in fact genuine. Though I could give it a go again, if people are interested.

Edit:

Just ran the report again for those that want to see a more recent copy.

As long as uploaders tag their posts, does it really matter if they do it before or after clicking "submit" on the upload page? What's the damage?

Yes, it's frustrating when you type a bunch of tags, only to find that it's already been uploaded by someone who typed less tags then you. It feels like lots of effort gone to waste...but that could be fixed by adding the tag merging function back in.

I think that instead of asking how to slow down quick uploaders, it's more important to ensure that uploaders who want to tag after uploading actually follow through with the tagging.

(This paragraph is going to stray off-topic, so if anyone wants to argue about it we should probably do a new thread, but I think it's important to point out: I'm also skeptical of the premise that "wanting to be first" leads to poor tagging. Almost all of the poor tagging I've seen is either from new users, or contributors who prefer "sufficient" tagging to the the meticulous & detailed tagging that's become popular in the forum. Keep in mind how few gentags howto:tag checklist calls "an acceptable level of searchability." Does that need to change? Does every user need to be a supertagger?)

Casual user here.

One of the best things about danbooru (besides being the OGbooru) is the exceptional tagging efforts that enrich the images.

Therefore I agree that proper appreciation of the tagging efforts is in order.

Right now the emphasis on upload stats overshadows other stats such as tagging, so I understand the annoyance of having tag work absorbed under the 'uploaded by' credential. People want to be recognized for their contributions after all.

Same goes for translators and wiki/tag editing efforts.

I have no solutions to offer. Just adding my 2 cents.

If glory is what's important, maybe we should add a "Tagger(s):" list next to the "Uploader:" name, sorted by most tags changed

This is the dumbest thread I have seen in a long time.

Join the Discord chat for better community responses than this, then.

Randeel said:

This is the dumbest thread I have seen in a long time.

Guess that what happens when there is an upcoming witch hunt against certain users. It seems to concern so much that it is starting to hail bad talk, sadly. And before this goes on and on, a discussion is still the best option.

Anyways, I think that the problem here is not really the sniping of posts. It is what you are doing with those posts. And one has to tag the images appropriately. Yes, the threshold of minimal tagging is according to the guidelines really low, but when looking at the reports, the Contributors seem to hit the 20 tag mark constantly (which is a whole lot more than what is "required").

Well, I am really in favor of testing something what @Type-kun has proposed: If you have a lot tags and you got "sniped", then it might be a good option to just merge your tags with the upload. @CodeKyuubi seems to do this manually for example when he got sniped.
And since lost work is actually more frustrating than not getting an image, I think it is worth to give this a shot. Because right now, when you are sniped and don't transfer your work manually, the whole effort you put into tagging the image is lost. And this looks way more frustrating than just getting sniped when you also have only added 5 tags.
Like I said: Everyone can upload, but the tagging work behind something is probably way more personal than just having used the bookmarklet on an image from an artist you follow.
So I guess that this is probably what @fossilnix had in mind. I'm sorry if you didn't, but that's how I understood your post and that is something way more logical to get frustrated about than just not having the image; we are an archive board, so one should be happy when an image gets uploaded as fast as possible and with good tagging (which seems to be the case, see the reports) so you can search a post.

Randeel said:

This is the dumbest thread I have seen in a long time.

And this is one of the few things I refuse to back down on -- Just because a problem isn't readily apparent doesn't mean it won't become one in the future. There are some solutions that I think won't work at all (as I've echoed above), but for some I think they will.

The reservation proposal makes sense to me, because it essentially says "alright, I'm uploading this, it's already on the server, and if I don't finish upping it with proper tags in 3-5 minutes then you are free to nab it from me", and if someone else manages to upload a file with the same exact md5/source link then their name goes on it instead, with very effort wasted on the site's end. At least that way we have a proper establishment of dedication to uploading a post well-tagged from the get-go.

That also means that if you're busy tagging a post, you finish tagging that post before you attend to any other uploads -- you don't let go of your opt-in until you're finished uploading what you chose to upload. Think of it as a kind of lock. Some uploaders have literally no shame when it comes 00:00 JST and will just 1-3tag snipe everything just to update it with marginal tags 10-20 minutes later, this will prevent them from being an upload whore.

BrokenEagle98 said:

Ah, you mean forum #119287...? I got considerable pushback when I first introduced that concept, so I stopped all work. Also, it would be a bit unlike all other reports, since the availability of the data needed is short-lived as upload records get deleted after 36 hours. Therefore, it would be impossible to prove that the report was in fact genuine. Though I could give it a go again, if people are interested.

I'm putting in a word for this -- this will be pretty useful for determining who is tagging what with due diligence, and to prevent poorly considered promotions for unlimited upload privs.

Anyways, I'm going to echo again that the problem with this perception is that upload sniping doesn't become a problem until it does. DakuTree was pretty dumb years ago to attempt taking matters into his own hands without proper input from a higher-staffed member, but he brought an extremely important concern to the table that has gone overlooked for years, and that is this:

• If it doesn't matter who uploads what, then why does it matter when I automate the process to level the field for everyone?

And besides, at least DT had the audacity to publicly document that his uploads were coming from a bot. If a user is uncanny enough, then as I said before -- they could just code a bot/script to monitor trusted artists' Twitter/pixiv accounts and upload posts literally the very same minute they do, and padding the tags only after such and such human user gets back and tags it properly. And they wouldn't mention this to anyone, simply claiming that they saw a good upload and upped it by hand to claim they'll 'tag it later'.

Not to mention the fact that it'll look like they are amazing taggers too. If such a user was smart enough, heck, they could even monitor the upload listing just to snipe tags from other users that uploaded duplicates of the post they sniped.

And then you wonder, are these the kinds of users that really deserve the unlimited uploads permission? Do we want these kinds of users to be representative of the tagging and uploading behavior that goes on the site? What's it say about the way we consider potential users to be promoted for "high quality uploads"?

These propositions are things that will help prevent a problems like these from happening in the first place. I don't see it as dumb at all, and just waving it off and saying "it's not really a problem" is plain ignorant IMO. If it's not really a problem then why did so many people start giving a shit when a highly trusted user started doing it, when absolutely nowhere did it say he couldn't? A Moderator-level user had to dictate that matter to him through a negative feedback, but the possibility is there that it may happen again whether anyone likes it or not.

Fred1515 said:

IMHO this can only be properly addressed by moderating. An automated solution might guarantee things like a minimum number of tags on a post, but someone still has to judge whether those tags were appropriate and/or adequate, and if that is consistently not the case for some users, then act upon it.

...To think though, that nowadays, there's no better part of my Danbooru browsing routine -apart from looking at the actual images- than the feeling of relief I get when iqdb tells me that someone else already uploaded what I was going to upload and saved me all the time and trouble, haha.

I understand your relief, and I feel the same way (I will sometimes post stuff to the danbooru discord for other users to upload when I'm lazy), but this isn't exactly about that. If you're an altruistic person about your uploads, excellent -- you have nothing to worry about on this front. But some users, if not many, get too selfish to consider doing what is good otherwise, and simply just bend the rules to their liking.

If such a solution can properly be addressed through moderation, then someone has to make noise about it -- and it'll usually only come from a few users, never enough to be seriously considered until it actually gets on the nerves on a great number of regular power uploaders. That, and it could easily be misconstrued as an abuse of power by a Moderator+ if such and such user looked like he was unfairly demoted because of poor/unfair uploading habits. I tend to err on being more strict, but even I know that as much as I would like to see some users demoted because of it, it is not within my responsibility to do things on my own without a proper discussion about it first. Perhaps a private forum for Moderator+ to discuss user demotions (with the relevant snipe reports that BrokenEagle98 has provided) could help towards that effort.

I'm also in favor for tag merging when upload a pic, but you got sniped for being "slow".
I'm also in favor what @fossilnix proposed about able to see who tagged it without the need to go through the tag history

Question about the reservation: When are you "finished" tagging a post? With 8 tags? With 20?

Ideally, you're finished when you tag as much as you can. Other users just put like 10 tags and that's it. I try to tag about the image as much as possible before I upload.

It's more of a qualitative instead of a quantitative thing. If the tags fully and accurately describe the post, then it's fine, regardless of how many there are. There really isn't a set number threshold on how many tags it is acceptable for a post to have.

E.g. there are posts under everyone that have close to a hundred tags, but might still have important tags missing, and there are posts under minimalism where 4 or 5 tags is the absolute maximum.

This is the problem with implementing a minimum required number of tags per upload, or a cooldown time based on the number of tags added.