Question: Uploading with minimal tags.

Shinjidude said:
By that I mean it sets up it's own "mod queue" that you would have to go in and vet everything that passes through it.

This was actually my initial idea. I can essentially set up the script to poke me if any non-whitelisted artist pops up, but that can be a bit problematic in some cases.

If you really cared to, you could even open it up to other members to approve and add tags to automatically generated posts. In either of these cases you could then allow your script to process artists that aren't virtually automatic Danbooru quality, and allow the script (with multiple users) to have a much greater throughput.

Although I think this could be useful, it might just be better to look for images.
I do have something similar that could possibly be used instead, but it doesn't help much for actually getting users to look for content.

Perhaps it would be better to simply watch uploads and tally users who leave their uploads undertagged beyond some reasonable threshold (say a half hour or so), and then periodically publish a list of the worst offenders so that the Mod staff can see and make judgements to determine if their behavior is detrimental enough for a warning or a negative report.

I've been very slowly trying to automate this entire process for quite a while, as going through everything manually would honestly take to long. Script is here if anyone is curious (It's a tad buggy though).

Wypatroszony said:
I absolutely despise of downvoting good images, because someone else got it, but I can't say that who actually did get the upload doesn't have an impact on that... I think.

I think everyone does, but it doesn't change the fact that I can name "at least" 2-3 users that I know are doing this because they didn't get first. There is more, but being able to tell who is doing it can be a tad difficult.

The problem would persist anyway, people may or may not jump on blogs/twitters/nicoseiga of famous artists and stalk there, still uploading with little tags. Which would deem your method just as ineffective.

Not really. It isn't exactly very difficult to move this script to other sites.

On the other hand, DakuTree, you've proven to be quite a skilled coder. Would you give it a shot to actually code something like you suggested earlier and what Schrobby did bring back a little while ago?

Less skilled, more "I have too much free time". Ruby isn't exactly my forte either. But I've already suggested this and it was a no.

Ars said:
The fact that you've been doing this for months without telling anyone, not even the admins, just makes me question your motivations.
...
This is something that you're supposed to discuss with the community and admins openly before actually doing it.
...
Hell, I'm going to go as far and say that you probably don't actually deserve that contributor/janitor position, since it's likely you got it through your script and not actual hard work.
...
It's a fucking insult to everyone else that actually put in more time and effort to get to where they are now and didn't rely on some bot with a database.

I haven't been doing this for months. Every single thing I tag is manual. I did expand the danbooruup script to make things a bit easier, but I actually posted that on the forums for other people to use.
I've honestly held back so much from actually posting this thing since I know it will piss off uploaders.
I could have very well just run this thing and it wouldn't have been a problem since "this is a non-problem".

And no offense, but you really have no right to actually say this. You are one of the worst users I've seen for this problem, albeit not as bad as a few certain people, saying that it's an insult while you continue to waste peoples effort by minimally tagging for "first" is just a tad stupid.

I'm not buying for second your whole "oh I know it's a dick move but it had to be done" act. I can't see this as anything other than a thinly veiled attempt to justify screwing over other users in order to make yourself look better.

How on earth does this make me look better? If anything it makes me look worse. I know for a fact there is uploaders that don't like me, and honestly I don't blame them. But if I started caring about that, this thing would never be fixed, nor even attempted to.

You've been here barely a year and have been a Janitor for even less and you think you have every right to try and do whatever you want and feel is best without any consent

Because thinking of getting more useful content for the site is a bad thing? We have too many users that stalk pixiv. People obviously have enough time to stalk pixiv. Yet people cannot seem to spare a bit of the time stalking pixiv to maybe look through a few blogs, hell even search for a few things on pixiv. Yes, this may not be the best way about doing it, but as long as "easy" content exists, people will continue to stalk that.

glasnost said:
Breaking news: humans are not mathematical models, and human effort is not freely redistributable. If you remove the 'simple' task of uploading from Pixiv, people will not move on to more complex tasks, they'll spend their time elsewhere.

True, but at least it's an attempt at fixing the issue. We could very well just continue to bury our heads in the sand about the entire thing, or we could at least put some effort into trying to fix it.
This also being said that this is uploaders we are talking about.

EB said:
I just don't get the fury over being "beaten to the punch" on an image.

The problem is less that, and more "wasted effort". Personally if I see someone out-uploading me who actually tags well on upload, this is a nice thing to see. Uploaders are either improving, or thinking outside the box. Both are good things.
On the other hand, when you notice things like this where the uploader didn't even put the effort into tagging afterwards (which kind of shows it was blatent "first" uploading), it is extremely demotivating. Eventually after this happens enough, people will eventually give up and leave.

Anyways since people seem to be complaining about it enough, I guess I shall put it on hold on running the script at least until we can have more a discussion on this.

Okay, after rereading your initial post, it seems I misread it. Here I thought you had been using this bot of yours for months when it seems you're only just started/considered starting using it today. Or maybe not. I've no way to verify whether or not you've been using it before now, but whatever.

I flew off the handle like an idiot and for that I apologize.

I grow slightly tired of this.
You are not going to FIX anything with this. DESTROY is more likely term to refer to your idea. Most likely output will be: Good taggers shoo'd, bad taggers remained. And who needs taggers anyway, since everything would lead to full automatization, given your attitude.

I think everyone does, but it doesn't change the fact that I can name "at least" 2-3 users that I know are doing this because they didn't get first.

And you say you don't care about numbers.

Not really. It isn't exactly very difficult to move this script to other sites.

Which you WOULD do, no matter what would've happened.

Every single thing I tag is manual.

Somehow, I got unable to believe you on this.

On the other hand, when you notice things like this where the uploader didn't even put the effort into tagging afterwards

Why tag something that got tagged the same minute with what [uploader] would've tagged it? I too, would drop on tagging, when a considerable amount of tags appeared from other uploader. Beside the fact that my uploads have hardly ever less than 10.

Eventually after this happens enough, people will eventually give up and leave.

(Rude! May result in me feeling bad after a longer while.) Pity it did not happen to you, would save a lot of people nerves and time discussing some absurd initiative. Only because you stalk pixiv. Seriously, stop that yourself and the problem won't affect you. Especially since "first upload!!11" "hurts" for atmost 20 minutes afterwards and that's at most. I didn't do any research, but I am fairly certain that no one aside you, carries an eternal grudge against users who dare tagging after they upload. And as I think I said earlier, you are a motor to shitty tagging on uploads right now (used to be between each other).

The only good thing coming from all this, is Shinji's tally suggestion.

DakuTree said:
True, but at least it's an attempt at fixing the issue. We could very well just continue to bury our heads in the sand about the entire thing, or we could at least put some effort into trying to fix it.

Give me a break. You're a programmer; you know as well as anyone that a solution that accomplishes nothing is not to be praised just because "it's trying".

Besides, it's still not entirely clear to me what the nature of the problem you're trying to fix is.

Is it 'inefficient use of uploader time'?

DakuTree said:
We have too many users that stalk pixiv. People obviously have enough time to stalk pixiv. Yet people cannot seem to spare a bit of the time stalking pixiv to maybe look through a few blogs, hell even search for a few things on pixiv. Yes, this may not be the best way about doing it, but as long as "easy" content exists, people will continue to stalk that.

Besides the fact that this is not your problem to fix, and that it is questionable that this is indeed a problem that requires fixing, your solution will not actually fix that problem, as Log, ShadowbladeEdge, and I have all pointed out.

Is it poor tagging?

DakuTree said:
On the other hand, when you notice things like this where the uploader didn't even put the effort into tagging afterwards (which kind of shows it was blatent "first" uploading), it is extremely demotivating. Eventually after this happens enough, people will eventually give up and leave.

People leaving posts untagged is a separate problem that your script neither solves nor attempts to solve. We already have the record system in place to deal with people who do this.

Is it community unrest?

DakuTree said:
This an uploader problem. This is caused by uploaders, and it effects uploaders. It causes good uploaders to leave, regardless if the whole reason of "first" is stupid (well less "first", and more "wasted effort"), having good uploaders leave because of bad ones it not something that should be ignored.

I would be fascinated to talk to any 'good' uploaders that decided to leave because of this 'bad' behavior, given that under your definition of 'good', a 'good' uploader would be too busy uploading underappreciated stuff from blogs to care about what Pixiv stalkers do. If I could, I would probably ask why they chose to address the problem by leaving instead of leaving records for poor taggers, bringing the problem up in the forum, PMing an admin, or any of the many and myriad other measures available to them.

DakuTree said:
And no offense, but you really have no right to actually say this. You are one of the worst users I've seen for this problem, albeit not as bad as a few certain people, saying that it's an insult while you continue to waste peoples effort by minimally tagging for "first" is just a tad stupid.

Well, whatever the 'problem' actually is, apparently it is a name-and-shame offense, punishable by judge and jury DakuTree. Could your horse possibly get any higher?

In terms of your reasoning, you argue that this will encourage uploaders to look elsewhere for uploads? Most of the art from twitter, nijie, blogs and nicovideo are reposts from pixiv and those that aren't are still being poached seconds after being posted.
If I don't find revisions or sketches worthwhile what's left? If I wanted to upload scans I would do it on a more specialized imageboard.

Western art is generally fairly well-represented by users like Kikimaru and Animeboy12, and there's little inclination for me to join them as it's generally not very popular or appreciated by the userbase.
When I have found new sources they've generally been swiftly taken over by other users and if it's you I have no chance because I don't use a notifier.

Uploading's just been incredibly unsatisfying and overly competitive the last few months and I think completely removing any chance to upload from pixiv would probably be the last straw for me.
Do I count as a good uploader? I'm curious.

DakuTree, if I may emit an opinion, I think that you jumped a little too hastly to conclusions. I don't think the routine of most uploaders who tend to camp art sites can be broken by simply making it impossible to upload the popular art.

Maybe the opposite approach would be more effective? Instead of punishing 'campers', why not reward more frequently the people who take the effort to thoroughly search for overlooked art? Why not encourage the people who look into blogs and relatively unknown sites by giving them positive records more frequently than they would get if they simply camped Pixiv?

I think that the 'bot' approach is simply too extreme, and would harm the community more than benefiting it.

Wypatroszony said:
And who needs taggers anyway, since everything would lead to full automatization, given your attitude.

Nothing can be fully automated. Tagging can't be fully automated, upload to a certain extent can.

And you say you don't care about numbers.

I'm honestly confused here. All I did was point out how many people I see.

Which you WOULD do, no matter what would've happened.

Actually no, because more often than not there is no need to. If it ever got the case of where it was actually needed, yes I would.

Somehow, I got unable to believe you on this.

Right, because a bot obviously can tag nearly everything on an image on upload. If such a thing were possible I would have ran this months ago. Auto-tagging anything that pops up on pixiv is completely limited to the info it provides. Artist tag, pixiv tags, tool tags (watercolor, tradmedia etc.) & related tags. Even then that requires the actually translation for the tags and related tag info.

Why tag something that got tagged the same minute with what [uploader] would've tagged it? I too, would drop on tagging, when a considerable amount of tags appeared from other uploader. Beside the fact that my uploads have hardly ever less than 10.

The point here is the uploader shouldn't see this in any case. When you upload something, you tag it. There is no reason to upload then refresh the page, which would show any tags added by anyone else. The simple fact that there is no additional tag edits in the history shows this is exactly what happened.

And as I think I said earlier, you are a motor to shitty tagging on uploads right now (used to be between each other).

Right, because this didn't happen at all before I started uploading? And it was bad after Mr_GT eventually got banned. At least from what I've seen from looking at older posts, it was much worse than it is now. This post in particular is a brilliant example of what I mean. Notice how we have 5 uploaders who didn't even tag worth a damn, not even tagging characters?
Even if I stopped uploading, it would continue, and it would continue to get ignored.

glasnost said:
Besides, it's still not entirely clear to me what the nature of the problem you're trying to fix is.

Two issues. Issue with users minimally tagging for "first", which gradually causes good uploaders to leave.
Issue with people somehow having enough time to stalk pixiv yet never look for "useful" content.

Is it 'inefficient use of uploader time'?
...
Besides the fact that this is not your problem to fix, and that it is questionable that this is indeed a problem that requires fixing, your solution will not actually fix that problem, as Log, ShadowbladeEdge, and I have all pointed out.

True, it isn't my problem, it's an uploader problem. I brought up the issue simply because it is related with what it can fix.
The whole idea behind running the bot fixing things is simple, we have two types of uploaders. We have those who don't give a damn about first, then we have the ones who do.
The latter are what are causing the problem. You make impossible for this to happen by essentially moving the entire issue to a bot then does one of two things. It causes the "first" uploader to stop uploading, or causes them to start looking for images.
Neither is a loss. People who don't give a damn about first can still look for content if they actually want to upload, not exactly hard.
I agree it's nowhere near a perfect solution, but any other ways to fixing this issue have already been completely ignored earlier in this thread.

Is it community unrest?
...
If I could, I would probably ask why they chose to address the problem by leaving instead of leaving records for poor taggers, bringing the problem up in the forum, PMing an admin, or any of the many and myriad other measures available to them.

People do. Half the threads about this issue, including this one, are often killed with "this is a non-problem". I've already given a neutral record to someone I consistantly saw doing Mr_GT esque uploading a few months back, did it fix it? Nope.

Well, whatever the 'problem' actually is, apparently it is a name-and-shame offense, punishable by judge and jury DakuTree. Could your horse possibly get any higher?

I did kind of go a bit over the top there, sorry about that Ars.
This problem really just makes me a tad annoyed..

There is no reason to upload then refresh the page,

Wrong. I did that all the time when typing in my tags when uploading first with "some" tags. And even if there appeared some tags, I usually submitted them, even for posterity. Again, you are making a problem out of something that is not necessarily one.

Right, because this didn't happen at all before I started uploading?

I know that reading through tons of letters in short periods of time dampens concentration, but you even quoted an answer for this.

Even if I stopped uploading, it would continue, and it would continue to get ignored.

Then stop, because your uploading changes nothing and you'll make a handful of users happier. :)

Issue with people somehow having enough time to stalk pixiv yet never look for "useful" content.

Kill and remake them as sentient machines. That's the only remedy to your issue.

I've already given a neutral record to someone I consistantly saw doing Mr_GT esque uploading a few months back, did it fix it?

Would it work on you if you got one? I don't think so. Negatives are where it's at.

This problem really just makes me a tad annoyed..

Averaging at 150 posts a day and being "tad annoyed" over not getting 5 is immature.

If I can say something on this topic.

To my mind the thing which tour bot does it actually the opposite : uploading every bit of worth uploading piece of art. From my perspective it's frickin annoying to see every time when you try to upload something new that lovely yellow bar informing you that such pic already exists in DB. What's more check out recent list of popular images, DakuTree's posts totally dominated that place, so if we have such awesome bot uploading pictures and tagging them nicely, why we should give a damn and try to upload something, knowing that the bot will do it for us faster/better/more accurate.

And yes, it's not so hard to notice that you used some tool to speed up your pace of uploading.

Ars said:
I can't agree to automating uploads using a script at all. Danbooru has been functioning just fine for years without it and I don't see any reason to change that now.

and

Ars said :
Telling people how and/or where they should find their uploads and actively trying to discourage others from camping pixiv in an attempt to push this ideal is just sickening.

+1

Wypatroszony said :
And you say you don't care about numbers.

It reminded me about Mr_GT...

Now to be more constructive in critic and make some use of your code, it might be a good idea to put let's say 24 h time-window on your script after which it would check for overlooked art works.

This thread has become very alarming, largely because of the potential reaction of the community to this bot. Right now, danbooru offers 5 services:

• A hand-picked high quality art source (Japanese-oriented)
• A detailed and extensive user-tagged search function
• English translation/localization of uploaded content (primarily from Japanese)
• Image search of it's database (when you include iqdb)
• A great deal of fun and satisfaction in contributing to those services

and if a large chuck of contributors stop uploading/approving because they feel this bot makes them redundant, then this bot threatens to change the nature of the first and last services drastically.

I might be being a bit meta here, but communities have absolutely no inertia by themselves; if people stop seeing a reason to participate, or think what they saw as the community is gone or changed beyond recognition, then the community often dries up overnight or at least takes a large hit to it's growth. So I think it's important that this is handled carefully so that things don't change too much to put the community at risk.

Anyway, it seems to me that this sort of bot was perhaps inevitable; sources that always produce images that pass approval arguably don't need to waste human red tape on getting them onto the site and it only takes one user to implement such a bot.

Whether we should ban, allow, or modify this bot because of this isn't clear at first.

• Adapting the site and this bot to work, like some sort of "who goes on the upload-bot's list?" thread, would be a lot of effort and might not work as well as the current system in the short or long run.
• Also, if most of the uploading community dies, who's going update the list or improve the bot?
• This bot does potentially remove some red-tape for images that would've been uploaded anyway.
• Then again, there's lots of other red tape that could be more helpful to remove from the site.
• Also the fun in uploading these popular images is what pushes some users to help out with the other parts of the site (like editing a wiki page when the find a poorly written one whilst tagging an upload)
• After reading Log's post, I realize that the act of manually uploading art does tend to help one in finding more good art and in that way, this bot would be tremendously destructive in limiting the chance of uploaders to stumble on good art whilst visiting pixiv.

Edit: Actually what the fuck? Putting it that way, an auto-upload-popular-images bot will try to do a job that humans do better in a way that also threatens parts of the community. My vote is to turn this off and look at fixing the grievance that motivated the creator to make this bot.

Edit: And as for checking for "missed" popular images:

• If it's implemented at all then it removes an easy way that new users can get into contributing.
• If it uploads them anyway, then it becomes impossible for the community to avoid uploading one-off bad images.
• If not, then it's going to need supervision, in which case you may as well assign yourself a list of sources to check, and dispense with the bot.

I don't know if I've misunderstood something, but you're concerned about "good" uploaders leaving because of "bad" pixiv-stalking uploaders, and your solution to this is to become the ultimate "bad" pixiv-stalking uploader? It doesn't seem to me like much of a problem in the first place, and this certainly doesn't seem like much of a solution.

If there's one thing that's clear to me from all of this discussion, is that this bot will cause a whole lot of drama and butthurt while not definitively fixing the so-called "problems" it was supposedly created for, as glasnost and others have already pointed out, and you have admitted yourself.

Wouldn't it be better to just stick to actual moderating (PMs, records, promotions, etc.) instead of using a bot and adopting yourself the very same behaviour you wish to discourage in others?

Most of these issues are a bit over my head, but for what it's worth, these are my two cents.

It seems like a lot of people feel that their personal satisfaction and increasing a number next to their name to boost their ego trumps having good art on the board, which seems counter to our purpose.

Like I mentioned earlier, the only reasons I would have a problem with DakuTree's bot would be: first, if it uploaded mediocre or poor art (like prevous mass uploaders did); but by whitelisting only 99% likely to be approved artists, that simply is not happening. Second, it would be a problem if the script mistagged, or undertagged its uploads. Dakutree is stating he personally tags everything himself, which I find hard to believe if his bot is automated, but in any case, a spot check shows about 10 tags on the initial upload, which is worlds better than most people that camp for "first" provide. Though I don't have a problem with that behavior either so long as the tags get flushed out in a reasonably short period of time after upload. By all counts, DakuTree seems to be helping Danbooru's collection, and not hindering it. If either of those points were or became a problem, I'd say we ought to shut down or tweak the bot.

Right now though, the only thing shutting the bot off would seem to do (to the site's collection, that is) would be to potentially miss completely approvable posts that manual posters might not find or otherwise desire to post. For that reason, I'd argue against turning it off completely.

While I'm not sure I buy the "I must have my name next to posts" mentality as a valid argument, I can think of two ways to fix this. The first way would to use the suggestion I mentioned earlier and have the script queue everything up with appropriate tags, and then let various users vet and approve the uploads as they came in as themselves. This potentially would have the benefit of allowing users to turn down any bad post that might somehow would get through the white-list, and provide tags the script/bot would have no way of knowing. It would also spread the credit around to assuage people that are unhappy about that factor. It wouldn't solve the core problem of lazy tagging, and would just move people from camping Pixiv to camping your bot, but by providing a list of suggested tags initially it would solve the problem of under-tagged posts.

The other option, and one I might suggest we run in any case, would be to set a time delay to your script (as
HNTI suggests above), so that anything that gets missed after say 12 or 24 hours of availability on Pixiv gets uploaded anyway. That would give Pixiv campers first bite at the posts to get their ego boost and name recognition, and still make sure everything worthwhile that your script screens gets posted. It could also clean up after lazy taggers by supplementing existing tags with the suggested tags from the script.

It seems the "first" crowd would be happier to be named and shamed by listing poor posters as I suggested previously. If I get some time, I may try coding this myself as I think it would be a good way of keeping these people in check. Perhaps we are all wrong, and it's not as big a problem as it seems, in which case such a list would exonerate people by remaining blank.

These are all just suggestions. Like I said before, I personally don't have any problems with this current situation. If you wanted to be conciliatory though, DakuTree, and keep people from growing more upset with you, it may be good idea to heed one or more of them.

Skingraft said:
In terms of your reasoning, you argue that this will encourage uploaders to look elsewhere for uploads? Most of the art from twitter, nijie, blogs and nicovideo are reposts from pixiv and those that aren't are still being poached seconds after being posted.

There is over 75K artists on the database with links to unchecked blogs, twitter, etc etc. There is also places that have been checked that have had things overlooked. I know for a fact that all of this hasn't been checked yet. I know both seiga and nijie both have a large amount of stuff that is duplicates, that isn't being said there isn't artists who only upload on this sites.

Western art is generally fairly well-represented by users like Kikimaru and Animeboy12, and there's little inclination for me to join them as it's generally not very popular or appreciated by the userbase.

Western sources like deviantart & hentai-foundry is always a good place to look though. There is a huge amount of content there that simply just gets missed. I'm actually amazed I haven't seen anyone watching the "popular" feeds for the anime/manga category on DA, or actually looking for content there. (That being said, I'm sure the site has over 300M images in total..)

Do I count as a good uploader? I'm curious.

I could honestly ramble on about "what is a good uploader", but I'm not one to judge.

Deelles said:
DakuTree, if I may emit an opinion, I think that you jumped a little too hastly to conclusions.

I have been holding off posting this thing for quite a long time now, really don't think it's to hastly.

Maybe the opposite approach would be more effective? Instead of punishing 'campers', why not reward more frequently the people who take the effort to thoroughly search for overlooked art?

Honestly I did think of taking this approach, and probably still will at some point. Problem here is "what is a good uploader". Then there is the case of, if you reward them would they stop being a "good" uploader? It's unlikely they would be rewarded again, so why continue?

Records could help to a certain extent, but you'd have to get people looking first.

I think that the 'bot' approach is simply too extreme, and would harm the community more than benefiting it.

It is extreme, but I really don't see anyone else suggesting anything else to solve the problem. There is other ways that have suggested earlier in this thread, most of which were simply just ignored.

Wypatroszony said:
Averaging at 150 posts a day and being "tad annoyed" over not getting 5 is immature.

I thought numbers weren't an issue? Also since when did it matter who uploaded what?
I'm not complaining about "first", I'm complaining about wasted effort.

Fred1515 said:
Wouldn't it be better to just stick to actual moderating (PMs, records, promotions, etc.) instead of using a bot and adopting yourself the very same behaviour you wish to discourage in others?

I have already said that doesn't do a damn thing. Even then, this is a a "non-problem", so doing anything about the user side of the problem can't exactly be done directly.

Serlo said:
Edit: Actually what the fuck? Putting it that way, an auto-upload-popular-images bot will try to do a job that humans do better in a way that also threatens parts of the community.

And the whole problem about minimal tagging doesn't? It affects uploaders, which last time I checked are part of the community. I complained about it months ago, and the issue essentially got ignored "since it doesn't affect majority of the site". Hell, jxh said himself that it causes good uploaders to leave, yet it continues to be ignored. I could keep complaining about this issue, but nothing will ever be done about it.

Shinjidude said:
Second, it would be a problem if the script mistagged, or undertagged its uploads.

The bot uses any available info on the pixiv post to tag, which in most cases is more than the users I actually see causing this issue. Albeit the idea was that the post would be fully shortly afterwards.

Dakutree is stating he personally tags everything himself, which I find hard to believe if his bot is automated, but in any case, a spot check shows about 10 tags on the initial upload, which is worlds better than most people that camp for "first" provide.

I was meaning before, there would be no need to use a bot if I was manually tagging everything.

Right now though, the only thing shutting the bot off would seem to do (to the site's collection, that is) would be to potentially miss completely approvable posts that manual posters might not find or otherwise desire to post. For that reason, I'd argue against turning it off completely.

At the moment it is actually turned off completely, that being said though, I might have something else that could be run..

The first way would to use the suggestion I mentioned earlier and have the script queue everything up with appropriate tags, and then let various users vet and approve the uploads as they came in as themselves.
...
It wouldn't solve the core problem of lazy tagging, and would just move people from camping Pixiv to camping your bot, but by providing a list of suggested tags initially it would solve the problem of under-tagged posts.

This exact problem is why I'm rather hesitant to actually do something like this. It helps, but does nothing in actually getting users to look for images, instead increasing their options of "easy" content.

The other option, and one I might suggest we run in any case, would be to set a time delay to your script (as
HNTI suggests above), so that anything that gets missed after say 12 or 24 hours of availability on Pixiv gets uploaded anyway.

In which all of the content it actually sees is already uploaded anyway. I do have a script which can essentially do this live to a certain extent. Shall see if I can maybe expand that a bit..

If you wanted to be conciliatory though, DakuTree, and keep people from growing more upset with you, it may be good idea to heed one or more of them.

I've already put the bot on hold, at least until we can have more actual discussion on the matter.

Since people seem to have a massive problem with the bot (which does something which has been happening manually for years may I add), how else can we fix these issues?

DakuTree said:
The really bad part about this is that the worst people doing this are janitors. I mean really.

You hit the nail right on the head. There's a certain janitor shoving his epeen in everyone's face because he can't stand other people flaunting their epeens.

I do agree with DakuTree's goals, but not necessarily with his methods. Personally, I couldn't care less who uploaded what, as long as it is uploaded and properly tagged.
It is rather disheartening to see how many people care about that stupid "competition". But auto-uploading doesn't really solve the problem of poor tagging, it would be much better to somehow kill the "competition" itself.

DakuTree said:
... or we could you know, just remove the uploader status entirely (or make it invisible to everyone below mod, including the uploader) so there isn't an issue, and people can still be promoted if they actually do a good job uploading.

IMO, this is a good idea. Maybe except "including the uploader" part, it's rather convenient to be able to see what you personally have uploaded before.
At least removing "by user xxxxxxxx" and "user:xxxxxx" things from the interface, for everyone below Mods perhaps, would already help somewhat.

Alternatively, we could make the uploader status dynamic, so that if someone has uploaded already present image with twice as many general tags it had before, he would be credited instead of the first user. Or even just adding more new gen-tags than the image already had could do the same thing.
However such functionality has a large drawback, some people may use nonexistent or even gibberish dummy tags just to steal the uploader status. If this problem can be solved somehow, then such functionality would be a good incentive to tag posts properly.

DakuTree said :
In which all of the content it actually sees is already uploaded anyway.

(facepalm)
Sorry, it just made me laugh a bit. It sounds like you feel sorry for your bot, because it won't get a post or two -_-.

Today, quite wacky idea came to my mind, what if somebody else try to compete with your bot and make its own bot and Danbooru will be just a playground for them, huh ? Programming isn't a rare skill anymore and it's just a matter of frustration cumulated over time...

MyrMindservant said:
...

The problem is that removing the uploader's credit does the same thing that the auto-upload script does in taking away the competition, the prestige, and the ego boost that comes from being most successful at uploading images. It would make the same people just as angry, and by what they are saying, drive a large number of otherwise active members away from the site.

Dynamic user credit wouldn't work either as the rules for setting it would be mostly arbitrary, and easily exploited as you note. It's also a waste of Albert's time to implement, and I couldn't imagine him choosing to do so over Danbooru 2 development, or a huge number of more useful feature requests in the queue.

I'm not sure of a great way to eliminate the "problem" of people fighting over Pixiv posts that will inevitably be uploaded anyway. Even the list I suggested wouldn't keep people from doing it, rather it would just call attention to people who do it badly (by leaving posts undertagged). The same as I don't mind DakuTree's auto-uploader, I also couldn't care less if people choose to spend their time doing what could be done automatically, and claim posts for themselves. Again with the provision that quality and tagging is done satisfactorily (which it sometimes isn't by these users).

If there is a problem, and we did want to change it without driving users off, we'd have to somehow make sure people who violate policy get punished and people who go out of their way to find good content from less easily parsed and discovered sources get rewarded. Both are difficult to do automatically, and automatic is the only way that would really work with a userbase as large as ours. Moreover, any reward would have to be greater than the perceived reward of being first, which we don't have many means to provide. An account upgrade can only be done with very good posters, and is a one-time thing. A positive record is more or less a pat on the head, and also can't be given out continuously.

MyrMindservant said :
Alternatively, we could make the uploader status dynamic, so that if someone has uploaded already present image with twice as many general tags it had before, he would be credited instead of the first user.

I'm afraid to think about amount of overhead put on server to deal with it. In some time window it might work, but not for the whole database.

MyrMindservant said :
IMO, this is a good idea. Maybe except "including the uploader" part, it's rather convenient to be able to see what you personally have uploaded before.

And how do you imagine user interface, without its history of uploads, to get a promotion ?