DakuTree said: Although, considering this could happen in the future with any other pool, wouldn't it be better to see about getting it fixed?
Well obviously, but that would require work from albert. For now, either it gets abandoned for not working or a workaround is needed to keep it working.
As well, removing ordering from pools would completely defeat the original purpose of pools, which was to provide a way to easily browse comics and such, but then the current usage already ignores that original purpose, so maybe this entire paragraph is invalid. (Though I still stand with any suggestion for a more dedicated solution to subjectivity and tagging.)
Serlo said: Well obviously, but that would require work from albert.
Danbooru is on github, which means albert doesn't exactly need to add it, someone else could do it and all he needs to pull the request.
As well, removing ordering from pools would completely defeat the original purpose of pools.
For comic pools yes, but there isn't any need for pools like Disgustingly Adorable, Pixiv 1000, Hot Legs etc. to be ordered. The idea was just a simple option to turn ordering on or off, meaning that pools that don't need to be ordered won't cause extra strain on the DB ordering them.
The Pixiv 1000 pool page has links to other pools for groups of images that qualify. If the Pixiv 1000 pool became a tag, someone would just need to make sure that those links got moved to the wiki page for the Pixiv 1000 tag.
john1980 said: The Pixiv 1000 pool page has links to other pools for groups of images that qualify. If the Pixiv 1000 pool became a tag, someone would just need to make sure that those links got moved to the wiki page for the Pixiv 1000 tag.
While the links are neat and all it is debatable if they are on topic to the concept and it is kinda pointless to copy them over if pixiv1000 becomes a tag.
Being a pool it does not matter but as a tag linking to images sets where only one images has 1000 bookmarks is clutter at best.
Pyrolight said: While the links are neat and all it is debatable if they are on topic to the concept and it is kinda pointless to copy them over if pixiv1000 becomes a tag.
Being a pool it does not matter but as a tag linking to images sets where only one images has 1000 bookmarks is clutter at best.
I'm not sure how it would be pointless to copy them to the wiki page.
I'm not sure what you're saying is clutter. Having those links in the pixiv1000 tag's wiki page? Or are you remarking on changing pixiv1000 into a tag in general?
A significant problem with the definition of pixiv1000 that was brought up before still remains: it's based entirely on a favorites count. This means the following will occur:
Extremely popular artists will be biased towards having nearly every single post of theirs favorited.
Posts that are funny will get a large number of favorites because of the content rather than art quality.
Even if pixiv favorites follow a relatively stable pattern for a minimum threshold of quality before it gets 1000 favorites (which doesn't happen), the growing number of users on pixiv means that 1000 eventually becomes a non-notable number. It would be similar to someone compiling a "Youtube 1000" list when the site first launched and trying to maintain it for perpetuity.
Switching to 5000/10000 won't solve anything in the long run, and switching the pool to a tag does not address the aforementioned fundamental problems either. Not only it is a ton of work for the taggers/pool gardeners to review every post, but eventually we'll run into the same problem with Pixiv 5000, Pixiv 10000, and even Pixiv 1000000.
What's more, strictly adhering to a pixiv1000 definition makes the pool useless. The pool was filled with good art before, but now that people are adding every picture with 1000 favorites into the pool the art quality has degraded significantly.
I'm not sure how it would be pointless to copy them to the wiki page.
Nothing useful is actually gained by having links to image sets unless the whole set since you if it is a tag the whole set is already marked.
I'm not sure what you're saying is clutter. Having those links in the pixiv1000 tag's wiki page? Or are you remarking on changing pixiv1000 into a tag in general?
Tag wikis should be simple and to the point with a few examples.
You might have some related links but pages and pages of links to image sets (that will all be tagged already) is to me clutter.
Aristocrat said: What the hell? It's basically useless now. I might as well search pixiv instead of navigating the pool one post at a time.
+1. Seriously, I'm astonished this matter has to be so thoroughly disscused. It's just another hopeless "vip quality" pool. The only difference is instead of some danbooru anons, now it's bunch of pixiv anons who decides which post deserve to be "vip quality" and which not. And I wouldn't be surprised if the responsible ones for most of these pixiv points(favs) are literally few zealous otas who have mastered good old socket puppetery skill and got too much free time on their hands. Don't pollute tags by adding this one. And if keeping this pool produces any extra strain for danbooru servers just nuke it. To be honest, I can't see why we should wait with nuke any longer.
Danbooru isn't a democracy. We need to reach a general consensus instead of carrying out something because it got 51% of the votes.
That aside, allowing Pixiv 1000 to stay as a tag sets a dangerous precedent for "DeviantArt 1000" or something even worse. The pool was great when it first got started; but the current state of that pool makes it essentially worthless, the reasons for which I have highlighted in my prior post.
Either clean up the pool and remove the 99% of posts in there that are not up-to-par with the art quality of the initial posts added, or nuke it altogether. Switching to a tag addresses only the technical issue and not the fundamental worthlessness of the current incarnation of Pixiv 1000.
Actually it is very much a democracy. If it was not there would be no discussion at all. Jhx or Albert would decide everything.
We need to reach a general consensus instead of carrying out something because it got 51% of the votes.
Depending on the definition of consensus, it can mean the whole or majority.
I assume you mean the former.
It won't happen for this topic.
If we try, either it will just die with nothing being done (happens a lot) or the few that disagree will be ignored (also happens a lot) and the action will proceed.
Pyrolight said: Actually it is very much a democracy. If it was not there would be no discussion at all. Jhx or Albert would decide everything.
They do have the final word on everything though, so in a sense, they indeed decide everything. They simply do so after taking into consideration the arguments put forth about a potential issue and the proposed solutions.
And in this case, it still seems pretty clear to me that killing it is the best solution. Nothing of value will be lost.
Fred1515 said: They do have the final word on everything though, so in a sense, they indeed decide everything. They simply do so after taking into consideration the arguments put forth about a potential issue and the proposed solutions.
I would say that in general we come to a general agreement on something and Jhx decides whether to enact it or not.
In a sense like the electoral college or even the President aka veto power.
It's the fact we have a say does seem democratic in essence.
The other situation would be use simply bringing the issue to light and any sort of agreement we came to would be totally irrelevant.
Personally, I don't think it's needed at all. We aren't Pixiv, so why are we compiling their data? The only real point is to say, "here's a good image by a good artist," which one should be able to tell from Danbooru's stats even if their eyes can't manage it. If anything, it really ought to be a tag.
Although it's still a somewhat decent filter for good art for the time being, the fact that the art is becoming more and more popularity/humour based rather than quality based means that it'll eventually just die out and turn into nothing.
Might as well kill it now before having to deal with it in the future.
I might add that deleting no longer destroys the information, so I think I'm in favour of marking it deleted so the quality can't drop any further, and those that still like to browse the earlier posts in it still can.
That's going to confuse the heck out of people without the greasemonkey script (lol).
Serlo said: I might add that deleting no longer destroys the information, so I think I'm in favour of marking it deleted so the quality can't drop any further, and those that still like to browse the earlier posts in it still can.
Even for pools? That's good; there's no compelling reason to not delete it then.
