Eikonos/Ideas: Difference between revisions
Antonizoon (talk | contribs) No edit summary |
m (Reverted edits by 149.88.104.21 (talk) to last revision by Antonizoon) Tag: Rollback |
| (One intermediate revision by one other user not shown) | |
(No difference)
| |
Latest revision as of 23:22, 4 September 2024
Auto-tagging[edit]
https://github.com/rezoo/illustration2vec
(neural training for tagging,has some basic premade neural maps for animu/hentai,maybe we could train our own for chan stuff)
Monetization[edit]
well i'd suggest strongly to have 3add types - popup add, banner add and banner add which is shown to users who have adblock enabled
Referral link: https://www.popads.net/users/register/577070
Dagobah SWFs[edit]
We need to have a reliable way to generate SWF thumbnails though.
4ca.org[edit]
This contains a list of all directories (find their page numbers here too):
Increment page numbers:
http://4ca.org/show.php?dir=epic&page=0'
Then just extract all the full image urls.
Macrochan[edit]
The Macrochan Image Database, based on SQLite, is a longtime archive of 4chan images from 2006-2012. Since it's tag system is actually a good basis to work off of, it would do us good to import it.
Using Macrochan[edit]
There are already existing tag databases that we can build off of, since it's easier to start from something than it is from nothing.
Recently, we've scraped the entire Macrochan database of classic 4chan images, 45,575 of them.
It used a hierarchical system of tag inheritance, but that kind of relational system can get too rigid and complex, because a tree system necessitates that each object has no relation to another. That is rarely the case.
Thus, we will need to convert that hierarchy into a different, tag cloud methodology.
Macrochan Tag Tree[edit]
I've already implemented tag inheritance on the tree.
http://macrochan.org/tagTree.php
PyBooru[edit]
The API Wrapper you use to send data is PyBooru.
https://pypi.python.org/pypi/Pybooru/
Source Data and Image Packs[edit]
Source Data and Image Packs to upload as seed data. Someone has to tag all of these though.
"checkItems": [
{
"state": "incomplete",
"idChecklist": "565cc9bcc6f6a6cd233d2e2f",
"id": "565cc9c8da0808edf4c46824",
"name": "http://www.threadbombing.com - An ol' meme database.",
"nameData": null,
"pos": 17211
},
{
"state": "incomplete",
"idChecklist": "565cc9bcc6f6a6cd233d2e2f",
"id": "565cc9cfa5231749a69a6871",
"name": "knowyourmeme.com - actually is a good source of tans,memes,threads and other *chan media",
"nameData": null,
"pos": 34269
},
{
"state": "incomplete",
"idChecklist": "565cc9bcc6f6a6cd233d2e2f",
"id": "565cc9f1a05b292bada97935",
"name": "https://archive.org/search.php?query=4chan - has many uploaded databases and 4chan folders thanks to gaz and other anons",
"nameData": null,
"pos": 51085
},
{
"state": "incomplete",
"idChecklist": "565cc9bcc6f6a6cd233d2e2f",
"id": "565ccc565209d1127c4bc7e0",
"name": "imgur.com - there are countless 1000's of images we should import,like the /r/4chan, /t/4chan and the such",
"nameData": null,
"pos": 68334
},
{
"state": "incomplete",
"idChecklist": "565cc9bcc6f6a6cd233d2e2f",
"id": "565ccc94fcf952fa5af05c8c",
"name": "Our Trello 4chan Historical Data board - Time to sort through the donated 4chan folders.",
"nameData": null,
"pos": 85394
},
{
"state": "incomplete",
"idChecklist": "565cc9bcc6f6a6cd233d2e2f",
"id": "565e04c9fc823a9c1bea21af",
"name": "other booru site - we may be the first to have a *chan specific content booru,but we are not the first one with *chan content,might want to look for tags related to chan culture",
"nameData": null,
"pos": 102341
},
{
"state": "incomplete",
"idChecklist": "565cc9bcc6f6a6cd233d2e2f",
"id": "5670685337e05b451fe0508c",
"name": "http://x.oranse.net/b/",
"nameData": null,
"pos": 118725
}
]
"checkItems": [
{
"state": "incomplete",
"idChecklist": "560d73251d908636717e0611",
"id": "560d733203b16e2d4d98ba72",
"name": "[2006 4chan Image Pack](http://shevchenko.duckdns.org:8080/webdav/back%20up%20of%20old%20pics.rar)",
"nameData": {
"emoji": { }
},
"pos": 16647
},
{
"state": "incomplete",
"idChecklist": "560d73251d908636717e0611",
"id": "560d7344301009befaec12d2",
"name": "ThrustVect's Image Pack",
"nameData": null,
"pos": 33263
},
{
"state": "incomplete",
"idChecklist": "560d73251d908636717e0611",
"id": "560d7350213c7bc5b27afbd6",
"name": "Yotsuba Society Image Collection",
"nameData": null,
"pos": 49763
},
{
"state": "incomplete",
"idChecklist": "560d73251d908636717e0611",
"id": "560d735999a857f0370ddc2c",
"name": "The Information Library 2.0+2.1",
"nameData": null,
"pos": 66437
},
{
"state": "incomplete",
"idChecklist": "560d73251d908636717e0611",
"id": "560d73648f1ac07bfa64a475",
"name": "2011+2014 GET Pack",
"nameData": null,
"pos": 83059
},
{
"state": "incomplete",
"idChecklist": "560d73251d908636717e0611",
"id": "560d736d7e518b27aba8dd23",
"name": "Pokeball99's Image Pack",
"nameData": null,
"pos": 100448
},
{
"state": "incomplete",
"idChecklist": "560d73251d908636717e0611",
"id": "560d7374f97b36caf65c1cf3",
"name": "Antonizoon's Image Pack",
"nameData": null,
"pos": 117850
},
{
"state": "incomplete",
"idChecklist": "560d73251d908636717e0611",
"id": "560d737c0cd01a7796c95b62",
"name": "Hagamablabla's Inconspicuous Folder",
"nameData": null,
"pos": 134912
},
{
"state": "incomplete",
"idChecklist": "560d73251d908636717e0611",
"id": "5668b9df150d8d67025d0999",
"name": "https://mega.nz/#F!Y0pTnZiY!g0ALqjpneHo-Lr2cJJaBuA",
"nameData": null,
"pos": 151322
}
]
}
Everyboty.net[edit]
Really shitty site design.
But there seems to be a way around it:
What to scrape with Grab-site[edit]
The image gallery can only be viewed upon scroll down, but grab-site can be used on some areas, such as the forums.
Site Structure[edit]
The image id is thankfully, sequential.
- Permalink/Details -
http://everyboty.net/?perm=49166How do i parse this javascript bullshit - Image URL -
http://trolledbot.net/shared/post_media/images/full_sized/30928.jpgIncrements up. Simple.
Image Count[edit]
IDs go from 1 to 100042. Simple. It also implies that there are 100,000 images on the site…
trolledbot.net/?perm=100042
Everyboty[edit]
It seems like everyboty is an aggregate of all of them?
Shutdown of the Bots[edit]
After talking it over, it seems that the best thing about Everyboty is the community is the community itself. This makes some sense, our community has been declining both in size and quality for some time now. Less people are uploading new content, less people are voting on new content, and there are more outsiders here than there used to be (or maybe they just seem more numerous because the tide of normal users has receded so much). The discord can be seen in the constant comment battles.
Let’s talk a little about the big players: reddit, imgur, and tumblr.
They have a much larger communites. “But bigger isn’t better,” I have often heard (and said). Here are the important counter points: 1) We’re not “cozily small” anymore. We’re a ghost town. 2) Big communities have smaller, cozy communities hidden inside them. You just have to look for them.
There are some things we do (or used to do) better than the other sites, but by and large they have us beat on feature implementation. Sure, some of their interfaces might feel awkward or even annoying, but these issues fade out of perception faster than you might think, especially when you have so many positive aspects to distract you. “why not just fix the sites so that we’re more competitive again?” I don’t have that kind of time or desire; I can’t spend all my time on a failing hobby.
Options going forward:
I’m seriously considering shutting down the bots, but I’ll keep the lines of communication open for a while and keep thinking things over. If I did shut down the bots, I would probably leave them online in read-only mode or something. I’m also thinking about starting a sub-community on one or more of the big sites. I’ll open the floor now.
