[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Settings Mobile Home
/mlp/ - Pony

Thread archived.
You cannot reply anymore.

[Advertise on 4chan]

File: op.png (3.79 MB, 1903x2959)
3.79 MB
3.79 MB PNG
This is the general thread for the alternative fan sites of the My Little Pony fandom. Discussion, development suggestions, and criticisms are all welcome here.

>Why alternate boorus?
Recent changes to Derpibooru have made the site a less than ideal location for pony fan art. So, several new boorus have arisen to provide alternatives. Many of the altbooru owners post in thread and listen to questions and comments from anons. Note that this is NOT the Derpibooru drama general. Please keep posts about Derpibooru in the thread on mlpg.co.

>Why alternate text sites?
Pastebin.com has deleted, and continues to delete, greentexts on the site. FimFiction does not allow stories in the greentext format, so a new home is needed.

Overview: altboorus.org
>General purpose boorus
twibooru.org by Twifag [!twifag.HOY] and Not Twifag/Floorb [!!H3PJaMJmY66]
ponybooru.org by ZizzyDizzyMC [!DizzyMC4pc]
ponerpics.org by Lotus [!!WYvH32WjaoR]
rainbooru.org by Raindev [!RaInbOWink] (dev MIA)

>Specialized boorus:
booru.foalcon.com by Pathos [!!arjIjTdCTdq] - Underage
lyrabooru.org - Hot glue

>Greentext sites
ponepaste.org managed by Not Twifag/Floorb
IAS (derpy.me/1yd6A) (Text search)

>Video sites

>File sharing

>Latest Developments:
-Anon releases searchable DNP list (derpy.me/mxM0C) (derpy.me/LSj5y)
-An anon is scraping pastebin links (derpy.me/zFWPk) (download green.zip)
-[Dead] Mediafire rip now available on IPFS (>>36440908)
-IAS Anon needs help seeding IPFS files (>>36700329)
-Reminder to backup Fimfiction (derpy.me/fPTg1)
-yuki.la likely dead, some other /mlp/ archives (>>36764936) (>>36764947) (and of course Desuarchive)
-yuki.la backup (>>>/t/1033260)
-Ponemusic archive is dead, latest torrents (>>37005785)
-Poneb.in dump (>>37346011)
-Userscript to load GDrive resource keys (>>37536855)
-FimFic story filter (>>37512031)
-WiP userscript for offline "accounts" on Philomena (>>37705442)
-Twibooru still looking for help(?) (>>38113999)
-New Ponepaste dump >>38375740
-Old ponibooru.org database found (>>38882814) + beta v. of Ponibooru.org Restored (>>38872905)
-Search millions of images by MD5 or SHA1 hashes, as of 2022-08-07 (>>38927537 >>38947506)
-pony.tube (running PeerTube) updated to a newer software version (>>39172527)
-Someone is working on a new booru website, which should be cool (>>39189860)
-Missing pbooru.com images and maybe also web pages were found (>>39190286). Some users were going to share some files, but didn't yet
-Archiving stuff like old pastes
-List of alt boards https://ponepaste.org/8236
-Some scripts for scraping youtube comments
-smutty.horse went down due to content reports on the amazon servers, is up now >>39579066
-smutty scraping

Last thread: >>39579729
Well, have a bump.
>New thread
I guess now is a good a time as any for a progress report on Sunbooru. I often find these type of things interesting, maybe you do too.

NEW FEATURES (or at least, features that are now working)
>Added tag info boxes to the tag view.
Still somewhat incomplete, but now lists any tag aliases and implications as well as the usual watch/hide interface. Still need to do tag descriptions and example images, but I need to make the tools to create those first.

>Add tag categories and prefixes
Tag categories are as you'd expect. In Sunbooru they are not hardcoded (don't know if that's the case in other boorus). You can add as many as you want and in any order, though I will say the css to stylize the different categories only exist for about a dozen or so. Prefixes are how I'm handling things like artist: and comic: tags, any tags that start with the set prefix will automatically have the category applied.

>Add ip logging
I am storing IPs in a separate table from the user posts and instead referencing them with an id. The idea here is to be able to track ips separate from the ips themselves and possibly in the future be able to prune ips after a while. We'll have to see how that goes, if that would create other problems or what.

>Add alias and implication functionality
Both features work as you'd expect. Might be some minor stuff left to do but they are basically done.

>Tag edit history
Tags edits are now recorded and can be viewed.

>Source edit history
Same deal as above but for source history.

>Add quick link to moderation tools for staff

>Markdown/up now added to image description
The usual text styling. Linking support has not been added yet.

>Add permission check to description editing

>Add favoritor listing to image pages

>Add (some) dropdowns to the nav bar at the top

>Add image reporting functionality
Currently mod tools can only dismiss these reports. Can't actually apply them until I get further on the ban tools.

>Fixed no featured image breaking home

>Fixed missing permissions for mod tools

>Fix videos not showing on image page

>Fix unhide list attempting to show deleted images

>The full image page now loads while image is processing

>Video and gif thumbs now generate more reliably

>Record and use mime types to more accurately determine media type

>Fix deleting current filter

>Fix tag counts being inaccurate (probably)

>Move media processing to it's own queue/thread
Database won't be locked while waiting for media to process.

>Major restricting of of data.py
Split into many separate files. The original idea was to contain all the sql commands in one place, but one file was clearly not enough.

>Split modForms into separate files
Same deal, should be more sensible to work with now. I'll probably end up splitting views and forms in the future as well.
Will link one more time. >>39644104 all youtube comments on sawtooth waves channel just before it was cleared out
And another thing happened >> 39652881 anything with this musician in it is probably going away
Nothing booru related, but reminding for anyone interested
I've suggested various levels of this several times at Twibooru, from the idea of re-importing artist tag aliases and implications, to the idea of re-importing tags on images from Derpibooru as well.

It's always met with resistance, generally with a justification like "Derpibooru can use it to intentionally break our site." I don't agree with that justification, given it would be intentionally breaking their own site as well and I know that precisely nobody at Derpibooru other than me cares anything about Twibooru... But I just haven't had the energy to argue it.
>argue it.
Derpibooru is already breaking the site by giving it incomplete tags. Maybe not intentionally, but the result is the same. Tags have to be imported at some time, might as well wait until they are correct.
Although I agree tags and general metadata (derpi votes and favs) of images should be updated, how long is "some time?" I don't browse derpi so I am lacking in the 'tags are finished' department.
Just handle a few tags specially. Like the artist tags. Whenever someone wants to nuke themselves the artist changes to "php:*" which is fucking annoying.
I don't know. A few hours should be enough. Maybe a day.
Just make it so the new tag scrape can't override the first one. New tags will be added, but existing ones won't be removed.
The issue I'm most frequently encountering is not wrong tags, but missing ones. I use the FiM species only filter and still see lots of humans and anthros.
In this case, if artist:somefaggot became artist:php123, you would end up wit both tags. This would also be a way to link the two identities together, in case you were wondering who some php artist used to be, and can't recognize them by the art style.
File: 1657628977448.png (431 KB, 838x884)
431 KB
431 KB PNG
This will never not be unf.
wget: "2023-02-23 15:59:03 (2.81 MB/s) - ‘heinessen-mlp-images.tar’ saved [238256291840/238256291840]"
-> (Downstream) file https://archive.org/download/heinessen-mlp-images/heinessen-mlp-images.tar matches 100% to this torrent:

Status of older 4chan /mlp/ images and newer ones:
* From its start to about 2014-12: see the above magnet link
* From 2014-12 to about 2019: some images might have been lost due to desuarchive.org deleting images that have not been posted for one or more years!?
* From roughly 2019 to about now 2022: that Synthbot guy from those PPP threads put the images in IPFS; hopefully someone else also pins that set.
* From about 2022 to current day (2023-02-24 UTC): images are live in Desuarchive, maybe Synthbot is still downloading newer image posts, or he has "enough" for the AI model or whatever.

About 12 to 16 hours ago I restarted Brave Browser and it fixed the "banned from archive.today" problem: might have been caused by a bug in their report system.

>And another thing happened >> 39652881 anything with this musician in it is probably going away
That post is:
>Sawto0th Wave5 was last week. This week it's Agat4n. P0nies at Dawn has removed anything with him in it from youtube and bandcamp. I'd expect other things elsewhere to be removed too.
>His youtube account if the videos haven't been saved yet https://www.youtube.com/channel/UCkAqNterdS3UOETZxewaLvg
>Does he have a bandcamp/soundcloud?
Correction: "From roughly 2019 to about 2022"

>Synthbot downloaded /mlp/ images
In https://ponepaste.org/7515 :
>As of Jun 25, 2022, the following data is available
>/mlp/ images - 930 GB
>derpibooru images - 3,520 GB
>animation render traces - 141GB
>From 2014-12 to about 2019: some images might have been lost due to desuarchive.org deleting images that have not been posted for one or more years
Unfortunately true:

Next snapshot = alive (was posted in previous thread in this general). Unix Epoch 1426675273 = Wed Mar 18 10:41:13 2015 UTC.

Therefore, /mlp/ images between ~2014-12 and ~2015-03 were never in Desuarchive or were in that website but were deleted off of it because the image was not reposted later enough. In the case of the latter, as of 2022-12, all 2015 "non-reposted" /mlp/ full images were mass deleted.
This is a 404 page. The next snapshot is 2023 and contains the full image (not 404).

I wonder if Desuarchive mass deleted /mlp/ images which were first/last posted in 2016.
>delay importing of images further
Not really any problems with a delay of 30 minutes or less for each new image, a longer delay might be bad. What could be done: make user d_tag_import_bot add tags from Derpibooru about 30 hours after each new image post.
>Page 10
Tag monkeys on Derpi tend to either do their job immediately or not at all. Also, automatic tag fixer is warranted, e.g. stripping "foal" because derpi has it implied by "filly" and "colt" (i.e. "girl" and "boy") rather than keeping it exclusively to pony-toddlers like the Cake twins.
Derpi does not have to break anything on itself for it to cause issues on importing boorus. Stupid aliases like the "foal" one are just one example.
>This would also be a way to link the two identities together
Twibooru already aliases artist:php* back to the real artist tag.
Updated https://github.com/ProximaNova/ia-to-ipfs
MediaFire being gated+shitty:

>-List of alt boards https://ponepaste.org/8236
You could add 7chan /co/ to that, which had/has some MLP-related threads:

7chan started in 2008 or 2009, I think because 4chan /b/ became lame back then too. I am pretty sure 7chan never had dedicated archives/mirrors to any boards. It used to have /be/ - Bestiality which had this one FIM-related image. It is kinda crazy to think, but I first posted on 7chan almost ten years ago now. (Or maybe it is 10 years ago now.) 7chan is lame in a way now since catbox.moe links cannot be posted there and result in a ban of one or more hours if you try to post such a URL.
I just checked the archive and found out that Agatan's songs were deleted. I PAID FOR THOSE SONGS and didn't save them, so Ponies At Dawn can go fuck themselves. Stupid bandcamp and its music licensing. I can't download them anymore.

If anybody gets it and shares it here, I'll upload them to the bandcamp. Please, the highest quality possible would be much appreciated.
The songs from Jyc Row's bandcamp would be nice too.
/be/ closed forever(?) is like 2016.

>I just checked the archive and found out that Agatan's songs were deleted. I PAID FOR THOSE SONGS and didn't save them, so Ponies At Dawn can go fuck themselves. Stupid bandcamp and its music licensing. I can't download them anymore.
Guess you didn't read >>39628932
>>It sucks that Bandcamp deletes the albums for the people who bought the songs too.
>Didn't know that. So they didn't own them, and were merely licensed to have access to that digital content. (Didn't buy the songs, but bought a lame license to them)

Which "archive" did you check and see that the songs were deleted?
>MediaFire being gated+shitty
Strange. I rarely have that problem with other mediafire links. Maybe it's just that link.
>/be/ - Bestiality which had this one FIM-related image
Found it in the 4chan /mlp/ mirror/archive, pic related, don't know which /be/ post it came from.

No, I got the same results with 2 or 3 other links, these ones:
Then I don't know. Mine are okay. Example: https://web.archive.org/web/20220512064800/http://www.mediafire.com/file/ngt13jx0yyaye8w/Tavi_and_Scratch_acapella.zip/file
May have began in 2007, picrel

That is an older link, I was posting about recent additions (today/yesterday). Update: looks like web.archive.org does not get Cuckflared if one changes http to https in the Mediafire link before saving. Links tested:
Didn't see " http://archive.heinessen.com/ ", "/mlp/ - Clopper's Closet", or "Heinessen" here (in "⁄mlp⁄"):

Or the link which is linked here:

About that >200GB file:
SHA1 of "heinessen-mlp-images.tar" per "heinessen-mlp-images_files.xml": 4b03a47fa51c53ee17ee6f96caf7cb9191a9dc4d
MD5 of that tarball per heinessen-mlp-images_files.xml: 93e3b666c44dbee648a489ec9ec6d6c3
CRC32 of it via that XML file: dfb5bd01
Filesize of heinessen-mlp-images.tar (from "heinessen-mlp-images.torrent"): 238,256,291,840 bytes
mtime>1422325746 , filecount>635770

Info on Heinessen 4chan archiver: >>39645699

I guess I will upload files from archive.org/download/heinessen-mlp-images to IWIFTP.
Someone showed up with 100% of torrent "My Little Pony 1986-1987 Disc 1 ISO" (previously only 73.1% available, now 83.1%), but he abandoned us :(
A Russian peer showed up with 100% of "My Little Pony 1986-1987 Disc 1 ISO", spotted 2023-02-25 05:58:41 UTC; this is the first time in years that that torrent has been 100% available/fully seeded!

ISO files of G1 MLP discs 1 through 4 - all 100% available torrents for now:
D1: see above
D2: https://ponepaste.org/8450#21
D3: https://ponepaste.org/8450#23
D4: https://ponepaste.org/8450#25
File: 1402192651593.png (649 KB, 1600x1300)
649 KB
649 KB PNG
TL;DR: recovering from a crash which happened when tar was still extracting files.

I was extracting files via "tar -xf heinessen-mlp-images.tar"; while the command was running (wasn't finished) my computer soft crashed. I assume and hope the answer to each of the following questions is yes. When the tar program is extracting files, it extracts them one at a time (a concurrency of one)? When tar is extracting files it extracts them in the same order as they are listed via "tar -tvf $file" (so working from top to bottom in regards to the files listed to stdout via that tar -tvf $file command)? I have some proof to support those two answers: next sentences. "tar -tvf heinessen-mlp-images.tar" -> working from top to bottom, check files to see if they were extracted or not (exist outside the TAR file or not), narrow down to where they stop existing. At that point a part of the line-delimited file list looks like this: [10 files which do exist, important file "image/1402/19/1402194305613.jpg", 7 files which don't exist - "ERR_FILE_NOT_FOUND"]. The important file "1402194305613.jpg" only about halfway loads (it should fully load if it had the full data), so I think that is where tar was at when the computer crashed; c.f.
https://archive.org/download/heinessen-mlp-images/heinessen-mlp-images.tar/image/1402/19/1402194305613.jpg - takes a long time to load

With tar, this "will not overwrite existing files": "--keep-old-files". The other consideration is to delete 1402194305613.jpg (where tar left off) so tar can fully extract it next time.
>Page 10
See "overwrite control" at
I'm downloading/seeding those because why not?

I've just expanded my local storage capacity by a lot (think ~100TB) on a NEET budget (despite no longer being a NEET) so I've gotta fill it with something.

Also, this effects like 3 people and I think most of them already know, but I fixed custom badges (eg: for winning a Pastejam) not showing up on PonePaste. I'm putting more and more work into the new version of the site and trying to fix the remaining bugs. It will be the same in visual appearance and functionality, but I've removed a lot of the dependence on external JS, and also made the site work a lot better without JS. And yes, for the 4 people who care, I have fixed deletion of pastes.

BTW, I am aware of the issue with the xdelta patches not being properly updated.
File: 1402199939386.png (100 KB, 800x600)
100 KB
100 KB PNG
Speaking of 7chan, I was listening to these audio files: https://7chan.org/ch7/res/471.html#1030 . They are OK to pretty good. This one has MLP-related music at 1:22:50 I think (lyrics "I don't mind cause this is something that I have to find after all this time I'll be OK ... sky turns black ... floor starts to crack"):

Extracted this image which is the attached image, it is from Unix Epoch 1411650017 / Thu Sep 25 13:00:17 2014 UTC and is Anonfilly or proto-Anonfilly.

>looking to fill unused HDD space
Could you please pin/seed this: >>39659900 . I am interested in the 930 GB of /mlp/ images part especially.
- description: ponepaste.org/7515
- download link: ipfs://QmdMjH7EsHdd4gGgCUnssDWndf54rVXQANvaSZFnhp5Tnw

It might be offline now, again, might have to ask Synthbot in the PPP threads.
That song is "All This Time" by Fox Stevenson. Nothing to do with MLP I guess. Might have gotten it confused with similar-sounding fan-made music from the 2010s.
>It might be offline now, again
Online today, the "images" folder is taking a while to load.
>Index of /ipfs/bafybeig7eygqnpyle2sspx44xut3xmbupwv4e6gukhtsvncq35l6lwvb4a
>4.6 TB
>animation QmXr…ZzXb 142 GB
>images QmU9…2etR 4.5 TB
Hope to download a folder in "images" or all of directory "images". Current situation described in PPP thread: >>39666621. (Also, post >>39663567 in other thread links to >>39663553)
I have the heinessen file already, it is not on there since I ran out of space again. It'll be back when I have more money to buy some SSDs (or you can donate some ¯\_(ツ)_/¯) and stop spending it all on new MacBooks and fox tails.
Is the copy I have on my TOR site incomplete?
>I have the heinessen file already
OK. Would the SHA1 hash of heinessen-mlp-images.tar show up in
I might download that .sqlite file, it is like 70 GB.

>Is the copy I have on my TOR site incomplete?
Yes, some of the ISOs are (check against the torrents). Only like 2 or 3 out of 4 G1 MLP ISO files were 100% complete in that onionsite. More info in the previous threads in this general which I might post again here in a condensed form or something (about other incomplete files).
File: 1648665020036.png (755 KB, 854x673)
755 KB
755 KB PNG
Here's the MD5 since that's all Google Drive supports:
93e3b666c44dbee648a489ec9ec6d6c3 heinessen-mlp-images.tar
I'll replace them if they're incomplete.
Or I was thinking of this PMV which I probably first saw months ago - "All This Time - PMV Collab":

Only has one comment:

OK. (That hash matches with >>39663208)
*/be/ closed forever(?) in like 2016.

>Which "archive" did you check and see that the songs were deleted?
What was that one website? project-vinyl.com or something? I looked up project vinyl mlp in DDG and saw this link on the first page - did not see the website:
Suspected it was .net - CWC has more SEO than it:

No results here:
I think agatan is a bit too new for whatever was dumped on project vinyl
Oh great the spam is back.
More smutty.horse links: >>39667188

Good to know, so it is similar to TPA in that it mainly has pre-2022/2023 stuff.

Have you applied to become some sort of Internet janitor? I am sure you would fit right in!
>Good to know, so it is similar to TPA in that it mainly has pre-2022/2023 stuff.
Even earlier than that. Reminder that it isn't really supposed to be an archive, rather that someone decided to dump thousands of videos there several years ago.
File: 1674696810917.png (195 KB, 1000x1000)
195 KB
195 KB PNG
latest list of every known smutty.horse URL, deduped, removed some stray characters, and changed to the y (Storj) subdomain:

(fixed two file extensions and removed 404 link https://y.smutty.horse/smutty_scrape1.ahk)
Yes please.
Shit, 2016 images mass deleted off of Desuarchive /mlp/ also:

Image from Unix Epoch 1464435048
2016-05-28 11:30:48 UTC
= dead as of
2022-06-17 22:27:09 UTC

> -> now alive via repost at https://web.archive.org/web/20230227072808/https://desu-usergeneratedcontent.xyz/mlp/image/1464/43/1464435048856.gif
I might have made a post like that, but you already did so thanks. (I will look into that text file maybe.)

Were 2017 "nonreposted" images also deleted off of that /mlp/ mirror?
"Email me a WACZ file with the results"
They also comply with DMCA takedown requests:
>404 Not Found
>This domain contains user submitted media for desuarchive.org
>Send takedown requests to >admin@desuarchive.org
Bronibooru was a Danbooru-based MLP booru. Bronibooru was an imageboard that died. The website at http://bronibooru.com/ now has the text:
"UPDATE (March 17, 2021): There is now a torrent for the Bronibooru image archive. This contains the entire 100+GB image archive as a tarball, as well as a sanitized SQL dump of the Bronibooru database. Just make sure you have sufficient room to extract the image tarball!"

That text links to:

Links to that torrent (with the same info hash):
- magnet:?xt=urn:btih:16bab76f8fde6732aeeca6132d7a4770f4b80189&tr=udp%3a%2f%2ftracker.opentrackr.org%3a1337%2fannounce&tr=udp%3a%2f%2ftracker.coppersurfer.tk%3a6969%2fannounce&tr=udp%3a%2f%2fopen.stealth.si%3a80%2fannounce
- NOT in itorrents.org at https://itorrents.org/torrent/16bab76f8fde6732aeeca6132d7a4770f4b80189.torrent

Since it is not in iTorrents.org, that seems to mean it was not a very popular or seeded torrent. I have 100% of that 101.07 GiB torrent, and qBittorrent says "Last Seen Complete: Never".
Dream I had which might have included Twifag: >>39672562

/fl/ - Flash:
>Furry flash is still b& from /fl/.
>/fur/ accepts .swf uploads.
This could include pony SWF files. Some MLP flash animations might have been grandfathered in. Also, whoa:
>3054 posts and 3047 images omitted. Click Reply to view.
>kinda crazy to think, but I first posted on 7chan almost ten years ago now
Thread I created 9 years ago -- https://7chan.org/fl/res/15490.html -- when I saw it today it was the last thread on page 12; page twelve is the last page, so it could have been bumped off.

MLP-related /fl/ thread:
>Brother, I saw this pony show for the first time the other day.
>It's badly animated and everything seems to be on fucking MDMA, in a bad way. Granted it's one of the better budget animations but it really doesn't deserve the praise these manchildren shower on it.
Why prune IPs?
Was just a thought I had about somewhat limiting data collection. We'll see how it goes.

More updates:
Works as you'd expect, shares markdown with description. Posters and staff can edit or delete comments. Comments can also be reported with the same report system, but the state of the report system is still the same. The comment overview on the home page now shows actual comments on the site.

>Add ability to set minimum tag category requirements
Mainly for enforcing a minimum number of rating tags, but requirements can be added for any created tag category.

>Add ability to set minimum number of total tags

>Add expand video button to images page for videos

>Add featured image link
Basically a link that will always redirect to the current featured image.

>Implement spoilers in search
They work as you'd expect, though there is no spoiler images yet, just text. With this filters are basically fully functional, though the sharing functionality of them still needs some work.

>Add ability to set a custom favicon

>Quick navigation dropdown for mod links

>Fix tag order in search
This should mean that all tags are now displayed in the proper order. First sorted by tier and then alphabetical order.
While working on the 4chan /gif/ 2023-02 full images torrent >>>/t/1210103 I have some time (waiting for computations and stuff) and am copying these archive URLs into archive.today: https://files.catbox.moe/932sw1.json
In one of the 4chan /gif/ torrents I saw a video kinda which made me want to throw up. It is not an illegal video, but it made me sort sweaty and black out a tiny bit (still conscious, just felt bad). People can be mean.
File: twibooru-186301.jpg (363 KB, 1247x1272)
363 KB
363 KB JPG
I am basically finished adding sawto0t hwaves webpages to archive ph since they are all pretty much there now


Shouldn't pic related from
be tagged as "meme"?
typos show I wasn't in a great state of mind
*video which kinda made me want to throw up
*made me sorta sweaty
File: mpv-shot0001.jpg (35 KB, 640x360)
35 KB
Scrape poneb.in outlinks if I or someone else wants to - lot of different links, including about 1,600 i.imgur.com URLs:
>links only: https://u.smutty.horse/mlbpztkvgnq.txt
>full text: https://u.smutty.horse/mlbpzgpmudt.txt
>in dweb link: ipfs://QmP4htCjj2dpFuA7zH1uA21rKavFaXSaQNB96sUkSha73r

Not in TPA, incl. low-quality VR Trixie Lulamoon as seen in attached image - "Death Grips - Hacker (Music Video)":

Neat video, didn't have me blackin' out like >>39676410, but still cool.

Yes, Vomit-chan is probably a meme
It was similar to the lost safeponi.com booru, but Bronibooru actually cared about their website. Hence, why there is an archive of bronibooru.com and not SafePoni. Lame rules. One of the rules banned furry, so a plus for some:
>You have 56 download credits remaining. >We'll bump that back up to 200 on 3/15/23.
Bandcamp only allows 200 free downloads/month (in total, not per account), so the number will be back up at 200 in 2 weeks.
Forgot the "maybe". Sorry
200 free downloads per month per IP address?
Or maybe like soundcloud which turns off the download button after like 100 downloads (not streaming).
Neat video and music; it's 480p.
File: bronibooru-4.png (531 KB, 1200x800)
531 KB
531 KB PNG
In previous and related threads there was posts about uploading stuff to archive.org (IA). Below is part of log https://catalogd.archive.org/log/3606422814 which has UTC timestamps:
>2023-03-01 19:52:36,443 INFO Maximum item size is 1099511627776 bytes.
>2023-03-01 19:52:36,443 INFO Total current item size is 8306034 bytes.
>2023-03-01 19:52:36,443 INFO Space left in item is 1099503321742 bytes.
>2023-03-01 19:52:36,443 INFO Loading torrent or magnet.
>2023-03-01 19:52:36,453 INFO filename: bronibooru.torrent
>2023-03-01 19:52:36,453 INFO type: torrent

I am pretty sure I have seen uploads to IA from one or more years ago where the file(s) uploaded to one item were significantly more than one terabyte. So about "Maximum item size is 1099511627776 bytes" above: somewhat recent change I guess. 1,099,511,627,776 bytes is like 1 terabyte; it is 1100 gigabytes. The torrent for "WARC: pbooru.com" had a total payload size of 903 GB = 90,300,000,000 bytes. WARC: pbooru.com was uploaded to IA via HTTP, not BitTorrent. Trying via an uploaded .torrent file failed twice, and IA would stop downloading it at about 400 GB both times. Unless the max item size was smaller back then or there is some other torrent-related limitation, I guess said upload initially failed because the "pbooru.com" torrent contains zero-byte files. Zero-byte files can't be uploaded to IA via "Internet Archive HTML5 Uploader [version # to #]", I guess that caused the failure. (Empty files can be uploaded to IA via the python script uploader thing -- would result in "Scanner" not containing "Internet Archive HTML5 Uploader" -- and/or TubeUp.)

BTW, "Daring Do and the Ruby Eye" by /mlp/ - offline for now:
>In the wake of the utter failure that was “Daring Do and the Endless Wastes”, a sequel was
>promptly commissioned by 4chan’s /mlp/ board. We had nowhere else to go but up.
"Chapter VI: Volleyball Gets All The Bitches" was kinda funny.

Internet Archive has a broken new search thing - also mentioned in the previous thread:
>new (not fully working):
> https://archive.org/search?query=%22Daring%20Do%20and%20the%20Ruby%20Eye%22
>> The search engine encountered an error, which might be related to your search query. Tips for constructing search queries.
>> Error details: The search backend returned an invalid reply (secondary request is missing collection title data)
> https://archive.org/search?query=%22Daring+Do%22
>old search thing (working):
> https://archive.org/details/@oddgrenadier?query=%22Daring+Do+and+the+Ruby+Eye%22
File: ponies_in_bat_country.png (214 KB, 500x500)
214 KB
214 KB PNG
*by /mlp/ - online for now

This image is from http://www.bronibooru.com/posts/5 and it is curiously still online at
Captcha: DSHGM8
Is Derpiboorus background pony comments truly anonymous? I always wanted to know just in case I felt like commenting.
>to random people
Yeah probably.

>to staff
Someone is currently live on https://www.pon3.stream/ !
Video not loading in Safari, iPhone
Shouldn’t have anything to do with IP. This limit should be tied to the bandcamp account.
So it’s IP based like 4chan? I thought so. I know the same background pony post number keeps in the thread your posting in. It’s probably that.
Thanks for checking on these. I suspected that some web.archive.org captures had files with HTTP 206 Partial Content. I was wrong. All 26 URLs returned HTTP 200 OK. There is an older set of 13 links and a newer set of 13 links; they map one-to-one. When comparing the two sets all of their filesizes match (units: bytes). However, their hashes do not match when comparing the older set and newer set. 2 sets + 3rd set, which is the live version at y.smutty.horse. Newer set compared to the 3rd set: hashes match. In what way or where do the bytes differ when comparing the older set and the newer one? Dunno, but it might be interesting to find out; maybe there is a bunch of binary zeros where they shouldn't be in the incomplete older set. All 26 files were larger than 30 megabytes.

I learned that incomplete snapshots in the Wayback Machine do not always show up with an obvious "206 Partial Content". Or they are actually HTTP 206 but were incorrectly recorded or playback as HTTP 200. Good thing I downloaded the live version (not the archive links/files) of everything in the web.archive.org index of smutty.horse; further reason to download it and store it locally being that web.archive.org does not always capture everything correctly.

Proof, hashes - in w3s:

Other text in post >>39605256:
>.json-s, .js-s, .svg-s and .html-s are modified to include web.archive.org in urls
This probably refers to files where links were modified by web.archive.org.

>https://files.catbox.moe/5ddgau.gif is empty even though 2023.02.02 at 05:56:01 it was not
>This one does not play in chrome but does play in ff.

>Theese do not play all the way through. Had to re-archive them.
>>This one does not play in chrome but does play in ff.
Does not play in Brave Web Browser, which is Chrome/Chromium. It does play in Safari Browser in iPhone 6, but it shows up as "Live Broadcast" instead of a normal MP3 file. I think "Live Broadcast" shows up for live-streamed audio and video files.
File: 1651541030867.png (1.39 MB, 5409x7500)
1.39 MB
1.39 MB PNG
Still wonder how pbooru.com (Gelbooru) created certain hexadecimal numbers >>39683734

This is another reason why closed source software sucks.
Page 10 bump.
Bronibooru is The Round Stable's booru in case you didn't know.
whoever's doing the tumblrs, can you go over this one again https://the-ponyscope.tumblr.com/
one of the mods returned and released a bunch of incomplete art a while back
"Bronibooru - website dump" (inb4 loginwall):

Thanks to those who seeded this torrent (>>39672766) recently, for I guess IA blacklisted my IP address and I can't really peer with them anymore.
>And yes, for the 4 people who care, I have fixed deletion of pastes.
What types of sites would the each of the ponies make?
>Some users were going to share some files, but didn't yet
This also includes Not Twifag saying he might share the code to the Mega-like site he made.
(Related to the smutty.horse situation and this post >>39687342) UK champing at the bit to destroy end-to-end encryption, using "protect the children" and "child pornography" as a political excuse/cover:

Applejack would make a typical landing page website to advertise the apples sold by Sweet Apple Acres. Captcha: G8VMK
>UK champing at the bit to destroy end-to-end encryption,
I really wish I could leave this shitty country.
Newer videos not in TPA - "ACLP Films", which includes some popular videos:

Seems like I did not see any IA peer (USA) in one or two torrents recently.
Part of a YouTube comment on the longer video:
>I also have to wonder how governments manage to get mass electronic surveillance of all their citizens approved, via legislation - when ordinarily, if police for example, want to search a person's home for CSA/CSAM or anything else which might be illegal, they have to firstly obtain a search warrant. And in order to get that search warrant, they have to show reasonable cause & demonstrate that there is sufficient evidence to believe that such an illegal activity is taking place at that address. But with mass electronic surveillance of everyone - it's like assuming everyone is guilty until proven innocent - & even the innocent are surveilled, when the authorities have no evidence whatsoever that people might have done anything wrong or engaged in illegal activities. The government is now allowing the authorities to forcibly barge in through the front doors of our private lives via mass electronic surveillance - & they don't even need a search warrant, nor do they have any reasonable cause for that surveillance.
>And of course, as long as the technology is there - one has to assume that someone at some stage WILL put it to use for whatever purpose they desire - whether good or bad - & they won't always let us know, or ask our permission.
>All of which makes it even more important to adopt good strategies to ensure one's private life remains just that - "private" - if that is what one wants.
I'm confirmed for Mare Fair, me and Neighsayer plan on rerunning the pony archival panel for those who missed HarmonyCon.
What was wrong with it before?
Speaking of ACLP Films: https://www.youtube.com/post/UgkxuR1W7CzOhWA2xlFFI8M7ia2PBNfAWjgw
>I created a series called SFM Ponies: POV of Ponies on Life which deals with ponies giving messages to the viewers related to Mental Health and Motivation.
>At first, I thought it went well with dealing with all these problems like anxiety, depression, or PTSD. It took me a while to find out that these POV videos seem uncomfortable because of the camera pointing too close to the pony and the message that talks about the topic which has some criticism.
>I don't think I should continue to post any mental health topics on this channel. SFM Ponies: POV of Ponies on Life series will be ending soon on April 2023 with a final video.
>Some of the videos you'll see on this channel will be hidden.
>I'll be staying away from any Psychology and Mental Health topics on this channel because I didn't do much of my resources on how I can execute my videos.
Not in TPA - videos uploaded by RainbowDash72:
https://vid.puffyan.us/watch?v=83syJrJsLHI - "Rainbow Dash Sings Danger Zone"

In a previous thread, I think I read something like "you just can't delete pastes in PonePaste". Maybe there were some other quirks to it, hence why I asked "What was wrong with it before?"
>Page 9
I created a TAR file which includes a directory listing of IWIFTP:"/Pony/Art/Image archives/bronibooru.com/" in "bb_f.txt.gz" and a tarball index of "./bronibooru.com/s3.tar" in "s3.txt.gz". Said TAR file ("broni.tar") is in w3s:

$ tar -tvf s3.tar > s3.txt
tar: Ignoring unknown extended header keyword 'SCHILY.fflags'

"./bronibooru.com/s3/" is mostly an extracted version of "./bronibooru.com/s3.tar". Here are the differences:
Only in "s3.tar" (not in "./s3/"):
>-rw-r--r-- narylis/staff 104571 2012-02-05 05:51 s3/796a6d165e4ddc4a30d5dadeaa773d05.jpg
>-rw-r--r-- narylis/staff 168217 2017-10-31 10:09 s3/796d1e4ff10c245d73385ddfd63e87e0.png
>-rw-r--r-- narylis/staff 120 2021-03-10 10:01 s3/._.DS_Store
>-rw-r--r-- narylis/staff 6148 2021-03-10 10:01 s3/.DS_Store

Pic related is "./s3/preview@2x/796a6d165e4ddc4a30d5dadeaa773d05.jpg", which is probably smaller than "s3/796a6d165e4ddc4a30d5dadeaa773d05.jpg".
Sorta extracted here, but takes a long time to load or just doesn't load:
> https://ia601605.us.archive.org/view_archive.php?archive=/8/items/55a9eff4bf9f6b73a8b6fef5d18a3e/bronibooru/s3.tar&file=s3/796a6d165e4ddc4a30d5dadeaa773d05.jpg

Seems to be MD5 hashes: 796a6d165e4ddc4a30d5dadeaa773d05 and 796d1e4ff10c245d73385ddfd63e87e0 - maybe check "Search millions of images by MD5..." in the OP.

I think in pbooru.com you could search by MD5 hash, probably can't do that in the current Philomena boorus (and if you could the file might be changed so it would not work in that case).
Sorry for the delay. It was an issue with one of the IPFS instances using too much data and nuking itself. I tried setting the StorageMax to 1GB instead of the default 0GB. I'm not sure if that'll fix it indefinitely, but the daemon at least runs again, and it seems to work now.
Fuck yeah, can't wait!
check s3: >>39680066

Appreciate it, I will probably try to download/pin it now.

>setting the StorageMax to 1GB instead of the default 0GB
Didn't know the default was zero. I think if the amount of data exceeds StorageMax then garbage collection (GC) will delete unpinned data. I read something like GC can delete data being pinned but which is not done yet. Maybe setting "GCPeriod" to some large amount of time like "500h" (~21 days) will prevent this, or setting StorageMax to a large size like 99000GB. Anyways, I don't really know, guess you know more.
Seen in Amazon S3:

Non-static files - "AccessDenied":

Extracted via the following, filesizes matched:
>$ tar --extract --file="s3.tar" "s3/._.DS_Store"
-> in w3s: ipfs://bafybeibsxutketeldpvdqdclozv3swdpo4pk4zbwfad6la4xito5uuqtdi/._.DS_Store
>$ tar --extract --file="s3.tar" "s3/.DS_Store"
-> in w3s: ipfs://bafybeiawad4454cuemtkj7rdgem6rqk2smrmim76ywyyw5mxcw5eytnkea/.DS_Store
ds store files are mac resources files you negro
they're in there because I use macos
Those ds store files are only in the TAR file, not the extracted version. See >>39688998 . Said TAR file matches to the torrent >>39672766

Therefore, your system probably did not generate those files or maybe it saw them and moved them somewhere else. I don't know a lot about Mac. Why use Mac instead of Linux!?
File: 1336688596623.png (304 KB, 1366x1035)
304 KB
304 KB PNG
Adding peer 12D3...NSWd ( ponepaste.org/7515 ) seemed to help with download speed of ipfs://QmdMjH7EsHdd4gGgCUnssDWndf54rVXQANvaSZFnhp5Tnw/images/4chan-mlp :
>$ ipfs ls /ipfs/QmdMjH7EsHdd4gGgCUnssDWndf54rVXQANvaSZFnhp5Tnw
>QmXrFisyAvEjTUsEFzz1BEPxSS94WxXtgk413NqBjwZzXb - animation/
>QmU9rvTD52LHni5y5WCRcAXufRj5ckTWcmGjHEUzxh2etR - images/
>$ ipfs ls /ipfs/QmU9rvTD52LHni5y5WCRcAXufRj5ckTWcmGjHEUzxh2etR
>QmcG5DVwy2RsvYg3Ki2HhSCfLMjcuAZhRdY2HWHwyzPrMc - 4chan-mlp/
>QmNv4aBxVD7tJecoD1ETFAQPuriY4fKVecq5cBrZ4CEvtw - derpibooru/
>$ ipfs ls -s --size=false /ipfs/QmcG5DVwy2RsvYg3Ki2HhSCfLMjcuAZhRdY2HWHwyzPrMc
>QmU3w1QMrH3CivjKrQ9sgd6tkFWQPR4oP8RbkTvZF4hMNR desu.1-29p1/
>QmVw5NA1U6uR3isuqWdNmZRRHFeffZBLZ6bwKpe7mehnxC desu.1-29p3/
>QmfBai2XneZd6b6PFweuXuvHEGG5ke6u3Mbnc8gpvSF62S desu.340-459p1/
>QmTkD7QhTSFcVZXbAZoXJThBa1aNpa4GbAmgE1bbKzWX7M desu.340-459p3/
>QmPH5Y2vTAGYkbisRRNxC2hEZtdi9oxinfSubTjjUGLMXG desu.460-478p1/
>Error: canceled
>$ ipfs bootstrap add /ip4/
>added /ip4/
>$ ipfs get QmcG5DVwy2RsvYg3Ki2HhSCfLMjcuAZhRdY2HWHwyzPrMc
>Saving file(s) to QmcG5DVwy2RsvYg3Ki2HhSCfLMjcuAZhRdY2HWHwyzPrMc
> 48.93 MiB / 866.99 GiB [>------------------------------------------] 0.01% 72d19h42m17s

Or not ("72d19h42m17s"). I also ran the following commands and other stuff in ponepaste/7515:
>ipfs swarm peering add /ip4/
>ipfs swarm peering add /ip4/
>ipfs swarm peering add /ip4/

Even though I ran "ipfs bootstrap rm --all", it still connected to a bunch of computers, WebUI: "Discovered 704 peers". I thought the "ipfs swarm filters" set of commands could help filter out some address (so I only connect to whom I want), but I think it might have more to do with display than connections. "ipfs bootstrap rm --help": "--all bool - Remove all bootstrap peers. (Deprecated, use 'all' subcommand)." So it recommends that "ipfs bootstrap rm all" be used instead; maybe that would fix it. Also, desuarchive hash:
>$ md5sum "$file" | sed "s/ .*//g" | xxd -ps -r - | base64 - | sed "s/\//_/g" | sed "s/+/-/g" | sed "s/=//g"
Comparison: pbooru.com torrent (897.92 GiB) was added on "11/17/22 5:56 PM" and completed on "1/12/23 8:42 PM" = 56.12 days. 72 days is 1.28 times longer. 100 days is 1.77 times longer. Less informative than "get":
>$ ipfs pin add --progress QmcG5DVwy2RsvYg3Ki2HhSCfLMjcuAZhRdY2HWHwyzPrMc
>Fetched/Processed 236 nodes

Unless it messes things up or I get a faster Internet connection, I guess my ipfs daemon will be running for months. (So other links I posted in this general will be online too.) Reminder for other thing I was doing:
>$ find /mnt/e/dl/data/205c227/y.smutty.horse/y.smutty.horse_6718 -type f | grep "\.mp4$" | xargs -d "\n" sh -c 'for args do echo $args; mpv --fs "$args"; done' _
"ponut" is a form of a Finnish word. Captcha: 0MNAYX
>Page 9
No one really says ponut though, or ponu for that matter. People tend to just say pontikka or kilju.
Ponut is a plural form of ponu. "Ponu" is synonymous with another word which means "moonshine" or "homemade alcoholic beverage".
Yes, very good, you can read a list. Kukaan ei oikeasti käytä ponuttia kun puhutaan pontikasta, tai moonshinestä kuten sitä englanniksi kutsutaan. Joskus pontikkaa on saatettu kutsua kiljuksi, koska pontikka on laissa kielletty. Pontsukin on yleisempi termi mitä oon kuullut.
>some previously unleaked fla files that were circulating privately and being gatekept
File: mare stare.png (553 KB, 5000x5000)
553 KB
553 KB PNG
File: Cap Test.png (172 KB, 300x1316)
172 KB
172 KB PNG
I've been messing around with making captchas. What do you guys think of these?
>[4chan, system:] Couldn't get the captcha. Make sure your browser doesn't block cookies then click here.

What is it? Official?
>Bronibooru is The Round Stable's booru
I didn't know. I saw that some of the links in that booru linked to The Round Stable, so I thought there was some association.

More about bronibooru.com... Seems to be missing 196 pieces (torrent piece size is 256.0 KiB):

Checked in qBittorrent: "413,822 / 414,018 pieces". IA regarding "bronibooru/s3.tar": md5 4e82b78493c46a5165019c625da10cce , crc32 c197291f , sha1 218da9a73d0598b3899ef072be6cfd0d87de2d50 , size 108476611072 bytes, "filecount>A file count is unavailable for this archive". I could check that against iwiftp.yerf.org/iwiftp_sha1-md5_hashes_complete_2022-11-20.sqlite or generate my own hashes. I wish that qBittorrent could tell me the byte offsets of the missing pieces. Nevermind, just downloaded the remaining 49.0 MiB via BitTorrent. Still want to check iwiftp_sha1-md5_hashes_complete_2022-11-20.sqlite or something though. When I downloaded "s3.tar" from archive.org it was not all in one go: "wget -c ...", ctrl+z, and fg were used. I could have got bad/incomplete data due to not cleanly downloading it.

It might be neat if users could submit queries and get results from "iwiftp_sha1-md5_hashes*.sqlite" without having to download the entire file. Or it is better how it is, as it encourages the downloading of that file.

Update - generated my own (after +49.0 MiB fix, it matches so s3.tar in IA is fine):
>$ date; md5sum s3.tar; date
>Mon, 06 Mar 2023 09:13:13 +0000
>4e82b78493c46a5165019c625da10cce s3.tar
>Mon, 06 Mar 2023 10:37:07 +0000
This is interesting because it shows there was probably error(s) when it was downloaded via HTTPS - which was fixed by the checks from the torrent. A moral of the story is that when downloading large files over HTTP (especially if you pause and resume the download) check the download-finished file against a hash; it might not match.
Checking the vital signs, nice. It is possible that some of the video have already been deleted or "deleted".

This post is helpful in creating a better extracted version of Bronibooru. The attached image there seems to be Rarity from Pony Life. Ponies in "Pony Life" are like bobble-heads: some say (like me) that their heads are too big. Is her ear blue like that in that show?
>It is possible that some of the video have already been deleted or "deleted".
I did a quick check. Looks like no videos have been made private or deleted. Only unlisted so far.
Supposedly it's official though I don't know if there's a concrete way to verify that. Still, the naming convention for symbols seems correct.
The only thing I know is that this was available at least 3 years ago but not shared until now: https://desuarchive.org/mlp/thread/34819795/#34831110
File: bat_pony.png (3.7 MB, 1181x1633)
3.7 MB
3.7 MB PNG
>being this ill-willed
I looked at that image and thought "wow cute bat ponies"
File: 1648359028491.png (398 KB, 3246x3000)
398 KB
398 KB PNG
File: 1654182881567.png (186 KB, 585x681)
186 KB
186 KB PNG
>-List of alt boards https://ponepaste.org/8236
This guy said he would download as many desuarchive /mlp/ images as he could. If that anon is ITT, how is it going? In order to not put strain on desuarchive.org (which is asking for donations now), you could download heinessen-mlp-images.tar info@ >>39659894 + 931 gigabytes of 4chan /mlp/ images from /ipfs/QmcG5DVwy2RsvYg3Ki2HhSCfLMjcuAZhRdY2HWHwyzPrMc info@ >>39665604

If you download QmcG...PrMc via "ipfs get [CID]", how is the download speed? For me, it is slow, and it says it will take like 100 days to download (maybe I can change the settings and stuff to make it go faster). If you have a good or OK computer (by today's standards), it probably will not crash when trying to pin remote data QmcG...PrMc. I have an OK or less-than-OK computer which has crashed before in regards to this, plus it seems better to "ipfs get" it since there is better indication of download progress.

Further reason to download desuarchive /mlp/ full images + more info:
* "Desuarchive" seems to be acting more like a mirror than an archive due to mass deleting images from years ago >>39660396 >>39660410
** I guess they will keeping mass deleting images, next on the chopping block: 2017 images >>39669978 , 2018 image, 2019 images, ...
** More like "Desumirror"
* Ideas on decentralizing 4chan /gif/: >>>/t/1211438 >>>/t/1211439
Also: >>39669974 >>39672420
>Page 9
File: 1479069814473.png (91 KB, 346x452)
91 KB
"ipfs get ..." overwrites existing files (tested with zero-byte files with same paths), wish it had a no clobber option
Hope to find/make a thing to list all folders and files in a CID recursively. I made some progress on Linux or bash commands to do that.

Animation error from season 1
Is derpi's server having issues or did they quietly re-implement discouragement bans?
I haven't started yet, the scraping software I'm using has an issue with FoolFuuka and the dev has been busy. I already downloaded heinessen-mlp-images, and I did not know about that ipfs folder, I will probably download that soon. Once I've scraped all of desuarchive I will create a torrent of all the posts and images. It will all also be on ponybro.org once the Hayden frontend is ready. I will not delete any images unless I am served a valid takedown request, and if that happens the image md5 will be listed on a "transparency" page.
In other news I recently started archiving the posts and images from the following boards/websites:
>8ch: /pone/ /clop/ /mlp/
>endchan: /pone/ /mlp/
Once the site is up all this data will also be available to download.
>a centralized place for pony board archives
Very excited to see this.
Safety bump.
>Hope to find/make a thing to list all folders and files in a CID recursively
Something like this:
>for a1 in $(eval echo {1..$(ipfs ls $level0 | wc -l)})
>> level1=$(ipfs ls $level0 | tail +$a1 | head -n 1)
>> level1c=$(ipfs ls $level0 | tail +$a1 | head -n 1 | sed "s/ .*//g")
>> echo $level1
>> for a2 in $(eval echo {1..$(ipfs ls $level1c | wc -l)})
>> do
>>> level2=$(ipfs ls $level1c | tail +$a2 | head -n 1)
>>> level2c=$(ipfs ls $level1c | tail +$a2 | head -n 1 | sed "s/ .*//g")
>>> echo -n "|"; echo $level2
>>> for a3 in $(eval echo {1..$(ipfs ls $level2c | wc -l)})
>>> do
>>>> level3=$(ipfs ls $level2c | tail +$a3 | head -n 1)
>>>> level3c=$(ipfs ls $level2c | tail +$a3 | head -n 1 | sed "s/ .*//g")
>>>> echo -n "|-"; echo $level3
>>> done
>> done
This is like baby code that I have written a million times before. I wonder how to make it less crap and not brain-dead. Also this /mlp/ thing is taking a while to load (don't need to use -s or --size=false):
$ ipfs ls /ipfs/QmNW8RZvSUvpR2pMjJJHiFNY9gBnwkhTSiddgZ152xK4bp
How would I simplify this nested for loop into a function (that can go further levels):
while [ $a1 -lt $(ipfs ls $level0c | wc -l) ]
a1=$(( $a1 + 1 ))
level1=$(ipfs ls $level0c | tail +$a1 | head -n 1)
level1c=$(ipfs ls $level0c | tail +$a1 | head -n 1 | sed "s/ .*//g")
echo $level1
for a2 in $(eval echo {1..$(ipfs ls $level1c | wc -l)})
level2=$(ipfs ls $level1c | tail +$a2 | head -n 1)
level2c=$(ipfs ls $level1c | tail +$a2 | head -n 1 | sed "s/ .*//g")
echo -n "|"; echo $level2
for a3 in $(eval echo {1..$(ipfs ls $level2c | wc -l)})
level3=$(ipfs ls $level2c | tail +$a3 | head -n 1)
level3c=$(ipfs ls $level2c | tail +$a3 | head -n 1 | sed "s/ .*//g")
echo -n "|-"; echo $level3
for a4 in $(eval echo {1..$(ipfs ls $level3c | wc -l)})
level4=$(ipfs ls $level3c | tail +$a4 | head -n 1)
level4c=$(ipfs ls $level3c | tail +$a4 | head -n 1 | sed "s/ .*//g")
echo -n "|- "; echo $level4
for a5 in $(eval echo {1..$(ipfs ls $level4c | wc -l)})
level5=$(ipfs ls $level4c | tail +$a5 | head -n 1)
level5c=$(ipfs ls $level4c | tail +$a5 | head -n 1 | sed "s/ .*//g")
echo -n "|-- "; echo $level5

I realize I could use an array in some place, but that is not the issue. Bash has a thing to dynamically create variable names, but it doesn't work everywhere. Idea of dynamic names:
echo habc -> 123

I could probably make some eval thing to write the code to a certain level (amount of nested loops) then execute it, but that seems bad.
Thought this post might not go through

Tried posting that to >>/g/ daily programming thread at >>>/g/91960721 but it said it was spam.
>Once I've scraped all of desuarchive I will create a torrent of all the posts and images.
In that torrent could you exclude all the images that are already in "heinessen-mlp-images"? Unlike IPFS, files cannot be de-duplicated across torrents.
The pain of Bash.
>level0ae=($(echo -n $(ipfs ls $level0c | sed 's/^/"/g' | sed 's/$/"/g')))
>for x in "${level0ae[@]}"; do echo $x; done
That array should have 2 elements as specified by double quotes, instead it has 6 elements, interpreting quotes literally. Switching to single quotes didn't help.
Eval incarnation is seen below; it's better, but still bad. I tried hard with many different things to get max folder depth to be whatever, but failed:
>#run: ./thisscript.sh [CID here]
>for x in {1..5}
> eval1+='fast=$(ipfs ls $level'"$(expr $x - 1)"'c | xxd -ps -)
>> for a'"$(expr $x - 1)"' in $(eval echo {1..$(echo $fast | xxd -ps -r - | wc -l)})
>>> do
>>>> level'"$x"'=$(echo $fast | xxd -ps -r - | tail +$a'"$(expr $x - 1)"' | head -n 1)
>>>> level'"$x"'c=$(echo $fast | xxd -ps -r - | tail +$a'"$(expr $x - 1)"' | head -n 1 | sed "s/ .*//g")
>>>> echo -n '$ev'; echo $level'"$x"'
>> '
> ev+='-'
> evalend+='done; '
>eval "$eval1"
I can hardly believe it's already been a week since my last update. Actually a decent amount of stuff done. Also we've now got our first batch of performance improvements. Basically I finally made some testing tools and realized just how bad the situation was, with things starting to feel bogged down as early as 5,000 images. In general, I've been trying to avoid focusing too much on optimizations as all the optimizations don't really mean much if it's still an incomplete product. But the performance was really bad, so I did a few things. Namely fixing the search and creating an index. Now it's much more reasonable, still a ways to go before it'll be competitive, but it's manageable for the moment. On a test database of 5000 images, 10 random tags each, I saw an 2.5x performance increase. And beyond that, I would say it feels similar in performance to existing boorus up to about 500,000 images now (likely heavily dependent on your disk speed). Again, it's gonna need to be better than that, but not too bad I think for an entirely disk based database. In the future I intend to have an option to load search data into memory which I would imagine should bring some pretty substantial performance increases.

>Add template for user profile
Added basic skeleton for the user profile page.

>Link user profiles on images

>Show when comments are edited

>Add recently used filters quick select dropdown

>Add ability to set site name

>Add page titles (in the browser tab)

>Implement captcha system
This one took a while to do. Was actually pretty fun designing the captcha generation itself. They look exactly like you'd expect a captcha to look like. Less fun was creating the system to manage the captchas. The way the captcha system works is that there is a pool of potential captchas available that are gradually cycled over time. The cycle rate and cycle amount will be able to be set (though not yet as I haven't added the forms for it yet). The cycle rate will also determine the time the user will have to wait after reaching the max captchas allowed per time. This max captchas allowed to be requested in the time period will be determine by the cycle amount. This will ensure that a user cannot access the entire pool of captchas. There is also a system in place to ensure that the captcha issued for a form is the one the user submits. Currently the only page with a captcha is the registration form and only if captcha is enabled. There are three captcha difficulty levels to pick from. Currently, SunBooru will need to be restarted to properly apply changes to captcha settings. I will work to fix this in the future.

>Create task scheduler thread
Basically cron but for various booru functions. Currently the only thing it runs is the cycling of the captcha pool. In the future it'll also probably schedule database analysis and vacuuming.
>Fix featured image tags preview

>Fix filters not applied in the tag view

>Fix repeating message on mod tag category page

>Prevent empty tags and search operators from becoming tags
They would essentially become unsearchable tags otherwise.

>Lock hidden images from user editing

>Make search more robust
Search should not break anymore when you enter an invalid search string.

>Set the default filter on account creation

>Properly limit results in search
The previous arrangement was always going to be temporary, the only reason for it was I was still figuring out a good way of getting the total results count while using LIMIT and OFFSET in the SQL. On a big database it really slowed things down to a crawl and consumed large quantities of ram. Now that things are setup in a much more reasonable manner, this is no longer the case and search works much more efficiently now. Performance comparison before and after: Before on a single threaded test for a single tag, Sunbooru achieved 4.64 requests per second. After this change, the results went up to 7.32. This was measured on a database of 5000 images, each with 10 randomly generated tags.

>Fix the debug server starting worker threads twice

>Fix worker threads not getting started when running with a WSGI server
If you've tried to run Sunbooru on one before and wondered why nothing actually worked, this is why. Fixed now. Should work as intended under a WSGI server.

>Fix no trending images breaking home page
Was due to leftovers from old page count.
>Change flask-login to use alternate login token for remember me
Something recommended to do by flask-login. In the default setup, the user id was used in the generation of the remember me token. The issue with this is that the user id cannot be changed, and so for example if you were logged in on two computers with the remember me token, if you changed the password on one it would not logout the other. By using an randomized string for the alternate ID, we can just generate a new one whenever the password is changed. Doing so would log out other computers when changing the password. Will make more of a difference once I actually add a way to change the password.

>Restructure views into multiple files
Code is split up more sensibly now.

>Restructure forms into multiple files
Same as above.

>Simplify recent filters
Decided to not differentiate between recently used user and global filters in the recently used filter list. User and global filters are still of course separated on the main filters page.

>Minor styling cleanup on the filters page

>Consolidate database connection making

>Made some actual performance testing tools

>Create database index for the image_id in the tag map table
Indexes help speed up the database where they can be used to refine results. Before and after testing: Before on a single threaded test for a single tag, Sunbooru achieved 7.32 requests per second. After this change, the results went up to 11.33. This was measured on a database of 5000 images, each with 10 randomly generated tags. I will likely need to do more testing to see where more indexes could make a difference.
Have you considered using a search engine for search instead of a database?
Your posts motivated me to worked even more and make a better version of this software >>39703090 . This is working and could probably easily be put into a script with a for loop or something:

Contents of top-level folder
>$ ipfs ls QmU9rvTD52LHni5y5WCRcAXufRj5ckTWcmGjHEUzxh2etR > a.txt
>[exlude folders via adding " " to the end of the line]
Contents of first level folder
>$ cat a.txt | grep "/$" | xargs -d "\n" sh -c 'for args do h2=$(echo -n "$args" | xxd -ps - | tr -d \\n); h3=$(echo -n "$args" | sed "s/ .*//g"); h4=$(ipfs ls $h3 | sed "s/^/ /g" | xxd -ps - | tr -d \\n); h5="$h2"200a"$h4"; cat a.txt | xxd -ps - | tr -d \\n | sed "s/$h2/$h5/g" | xxd -ps -r -; done' _ | tr -s \\n > b.txt
Contents of second-level folders
>[alternate between "cat a.txt ..." and "cat b.txt ..." until everything is listed]
>$ cat a.txt | grep "^ \S.*/$" | tail -n +$n1 | head -n 1 | xargs -d "\n" sh -c 'for args do h2=$(echo -n "$args" | xxd -ps - | tr -d \\n); h3=$(echo -n "$args" | sed "s/^ //g" | sed "s/ .*//g"); h4=$(ipfs ls $h3 | sed "s/^/ \x20/g" | xxd -ps - | tr -d \\n); h5="$h2"200a"$h4"; cat a.txt | xxd -ps - | tr -d \\n | sed "s/$h2/$h5/g" | xxd -ps -r -; done' _ | tr -s \\n > b.txt
>$ cat b.txt | grep "^ \S.*/$" | tail -n +$n1 | head -n 1 | xargs -d "\n" sh -c 'for args do h2=$(echo -n "$args" | xxd -ps - | tr -d \\n); h3=$(echo -n "$args" | sed "s/^ //g" | sed "s/ .*//g"); h4=$(ipfs ls $h3 | sed "s/^/ \x20/g" | xxd -ps - | tr -d \\n); h5="$h2"200a"$h4"; cat b.txt | xxd -ps - | tr -d \\n | sed "s/$h2/$h5/g" | xxd -ps -r -; done' _ | tr -s \\n > a.txt
Contents of third-level folders and deeper
>not hard to implement
It probably will not have any errors as nothing happens if /\/$/ is not found.

These commands are helpful to pin folders a bit at a time for better indication of progress and stuff. Also because the "ipfs ls" set of commands has no recursive option to traverse all of the folders.
I'm sure the idea has probably crossed my mind at some point, though I admit I haven't given it much thought. It would mean another dependency and potentially a more complicated set up for people looking to run Sunbooru depending on the option chosen. I guess I'll see if my plans work out, if not I'll be looking into other options.

Glad to hear I've motivated you ^:)
just an update on mrponiator since it was talked about a while ago
playlist has been updated as of 3 days ago: https://www.youtube.com/playlist?list=PLZlT8cWAMLw7vSnU1803IdY2CUHcfAPI_ One additional video has been made private ("Don't Watch a Cartoon Called My Little Pony").
the youtube about section has been changed slightly. The tripcode is no longer there (!q1xtu0xiyY) but it does say "MLP:FiM is a good show and you should watch it.".

Probably old news that was covered before but will say again, eqd posts relating to mrponiator have been removed (eg. https://www.equestriadaily.com/2013/03/animation-derpy-story.html), probably because of this tweet: archive<dot>ph<slash>nIi2G Additionally, the twitter account no longer says "if you are reading this, get out" archive<dot>ph<slash>JqYSv

Not sure if there's anything else to save that hasn't been discussed previously.
partially recovered original sentence:
>!q1xtu0xiyY -- this code has been retired, because I forgot the phrase for it. I don't really visit the website it's intended for, either, ...
Although I suspect he still does given what happened last time
A few more things. He also deleted his comments asking for other videos with him in it to be deleted (eg. archive<dot>ph<slash>E14E2) A bit unexpected but at least those videos will stick around for now.

His new non-pony alias is MrFariator (https://desuarchive.org/mlp/thread/34179543/#q34181837), and that tumblr is also excluded from wayback machine. It's not just pony content. This guy really doesn't want any of his works to be found.
>tail -n +$n1
Should be "tail -n +1" instead. Or
>grep "^ \S.*/$" | tail -n +$n1 | head -n 1
should be
>grep "^ \S.*/$" | head -n 1

Contents of third-level folders, switch between:
>cat b.txt | grep "^\s \S.*/$" | tail -n +1 | head -n 1 | xargs -d "\n" sh -c 'for args do h2=$(echo -n "$args" | xxd -ps - | tr -d \\n); h3=$(echo -n "$args" | sed "s/^\s //g" | sed "s/ .*//g"); h4=$(ipfs ls $h3 | sed "s/^/ \x20\x20/g" | xxd -ps - | tr -d \\n); h5="$h2"200a"$h4"; cat b.txt | xxd -ps - | tr -d \\n | sed "s/$h2/$h5/g" | xxd -ps -r -; done' _ | tr -s \\n > a.txt
>cat a.txt | grep "^\s \S.*/$" | tail -n +1 | head -n 1 | xargs -d "\n" sh -c 'for args do h2=$(echo -n "$args" | xxd -ps - | tr -d \\n); h3=$(echo -n "$args" | sed "s/^\s //g" | sed "s/ .*//g"); h4=$(ipfs ls $h3 | sed "s/^/ \x20\x20/g" | xxd -ps - | tr -d \\n); h5="$h2"200a"$h4"; cat a.txt | xxd -ps - | tr -d \\n | sed "s/$h2/$h5/g" | xxd -ps -r -; done' _ | tr -s \\n > b.txt

>It probably will not have any errors as nothing happens if /\/$/ is not found.
If wipes a file (makes it zero bytes) if grep finds no matches, more reason to make a script so it doesn't go past the max number of matches. Or wrap it in 'if ! [ -z "$var" ]; then [do]; done'.

Wonder what the input text was that created that tripcode. Might be able to figure it out using tripcode explorer.
Don't think there is any working morty proxies, so I guess I will save those webpages to w3s or something for redundancy.

*[do]; fi
>He also deleted his comments asking for other videos with him in it to be deleted (eg. archive<dot>ph<slash>E14E2)
Comment was the following on video TUZsBSvppWM:
>mpoator [ https://www.youtube.com/channel/UCdJ2R1H-FwQL0r0z7iapwpQ ]
>1 month ago [before 3 Dec 2022 06:47:46 UTC -> "2022-11-03 06:47:46 Z"]
>To whomever it may concern, is it possible to have this video deleted? Thanks.

3 low metadata comments on video ID TUZsBSvppWM saved somewhere. In TPA as an MKV (didn't show up in hobune.stream):

Error with 4th level:
>_: 1: sed: Argument list too long

Some or all folders in directory "4chan-mlp", text file in w3s:

>Argument list too long
If it isn't universal maybe perl can handle larger text. Or I could use vim to input and output via files.
>MLP-related /fl/ thread:
>Dear anons, there is nothing sweeter in this world than mares.
>Crack open that horse puss, you know you want to.
>AI now makes it possible to somewhat easily create your own "My Little Pony: Friendship is Magic" content. The voices of the mane 6 can be made to say whatever via using machine learning-related software. The quality of the show does not have to be bad, like I heard the last two seasons of FIM were low quality.
Interesting image from the smutty.horse thread
I love how you guys are presumably concerned about archiving pony, but nobody itt ever said a word about how anonpone just vanished with hundreds of pony CYOAs it archived.
Is anonpone gone for good? Is there even an archive if it?
Nobody fucking cares. Youll rather archive a bunch of shitty barely pony related posts from some irrelevant altshitchan.
Oh, I wanted to do another grab of anonpone, I hope it comes back
What's the domain name?

>anonpone just vanished with hundreds of pony CYOAs it archived
More like "copied" or "mirrored"
That's how archives work, they copy things.
looks like another youtube channel has been hit by a crypto hacker, this time it's ILoveKimPossibleAlot: https://www.youtube.com/channel/UCCi_wGGQ4kcEH4Dd55bOKMw
All pony related videos should've been saved by now. Not sure about comments though.
"Archive" implies some sort of longevity and redundancy. www.anonpone.com died and there is apparently no backup, so it just "copied", "saved", or "mirrored" stuff to that website.
>Nobody fucking cares.
Or I/we were busy with other things. Everyone is in a goddamn competition to see how much stuff they can delete.
You mean like Desuarchive?
Seems like it wouldn't be that hard to rebuild it with desuarchive. I don't really see why they haven't tried, unless the source code got wiped too and it needs to be recreated from the ground up.
I had a dream, and in part of it Applejack said "I cast a net upon the world and my eyes return the catch." Poetic pony.
Copied here - b32eaa2a58756ea6c10936acb15710d312a6a01d :

Also in that dream: Sunburst had to write some thing for school/university that was at least ~3,400 words. He wanted to make it ~41,000 words instead. He had like a pallet-full of books next to him. 3,400 words ~= 6.8 pages, and 41,000 words ~= 82 pages.
2020-04, 2020-10

In the dream I think most of the ponies were humanized, but not like "Equestria Girls" with the weird non-realistic skin tones. They were actually/more white in the dream. I never watched EQG (yet?)
2023-03-11 04:43:16 GMT: https://www.anonpone.com/: "Website under maintenance [\n\n] Blame Wolvan"

Multiple days ago I had a dream that Fluttershy was underwater and was singing about how she wished it was her birthday everyday (more birthday partying maybe?)
Don't think this correctly shows his videos - I don't see any of them:
I see. I'll check both immutable IDs next time if one shows nothing.
>-Search millions of images by MD5 or SHA1 hashes, as of 2022-08-07 (>>38927537 >>38947506)

>[alive as of now]
Also here - loginwalled - "MD5 and SHA1 hashes of 28,956,490 images - pony and furry-related":

>Here's the final one:
>60093498 entries | 23433879552 byte (23GB)
>Index on file_hash_md5 for quick search
>[dead as of now, which is 2023-03-11]
Found it locally, now in ".../205c227/database_pony_furry.sqlite" = 21.8 GiB, and ".../data/205c227/database_pony_furry_in-progress.sqlite" = 10.8 GiB.
Info on files in folder "Pony", does not have hashes on files in the a "Furry" folder. Seemingly non-serious SQL database in there, can't count rows via sqlite3, but you can dump it to a text file; number of lines/rows in "database_pony_furry_in-progress.txt": <s>["wc -l" is reading the file now]</s> -> 31186604 ->
31,186,604 lines in there

* Info: https://desuarchive.org/mlp/thread/38904451/#38927698
* 60,084,657 rows ---> $ sqlite3 -csv -header database_pony_furry.sqlite [...] sqlite> select count(*) from local_files; [...] 60084657
* has info on non-binary files / non-image files / plain text files such as JSON files
* probably mostly has hashes on image files
* hashes on files in folders "Pony" and "Furry"
** however, this search returned no results, maybe I did it wrong: "sqlite> select * from local_files WHERE location_short LIKE 'Furry';"
*** <s>try with "Pony" instead for a baseline</s> no quick results from "... LIKE 'Pony'" - maybe sqlite uses a different thing compared to sql
File: giphy.gif (644 KB, 268x141)
644 KB
644 KB GIF
Typo, sorry. Correction
*Seemingly non-serious SQL database error in there
My isp died for a while yesterday, and with it, the scraper. Back now.
Here is a dumbed-down version for further clarity and stuff.

== Post >>39713041 ==

Shows the text of two of the referenced posts the OP: in the "Search millions of images by MD5 or SHA1 hashes..." bullet point. One of the two links in said two posts is dead, but I have a copy of the data.

== Post >>39713283 ==

About "database_pony_furry_in-progress.sqlite":
- There is a certain "Pony" folder which contains millions of images. This .sqlite file contains MD5 and SHA1 hashes of those images. In the same filesystem or directory tree, outside of "Pony", there is a "Furry" folder with many image files; said folder probably does have some MLP-related files. This .sqlite file does not have hashes of said furry images.
- This .sqlite file seemingly has a non-serious SQL database error. (IIRC, the error was caused by copying/moving the file before an SQL program finished writing to it.) The program sqlite3 can't count the total rows in it, but sqlite3 can dump it to a text file. I dumped the database to file "database_pony_furry_in-progress.txt". The number of lines/rows in that TXT file via "wc -l $file" is 31,186,604.

About "database_pony_furry.sqlite":
- sqlite command/query "select count(*) from local_files;" states that this .sqlite file has 60,084,657 rows. This database file has hashes on files in said Pony folder and the aforewritten Furry folder.
- As is the case with file "database_pony_furry_in-progress.sqlite", this file (database_pony_furry.sqlite) also contains hashes on non-binary files (which are non-image files or plain text files). It contains hashes on API-related JSON files.
- Probably most of the hashes in this .sqlite file are of image files.

38904451#38947506, 38904451#38927537, 38904451#38927698

=== Question on database_pony_furry.sqlite ===

Why did this search return no results, maybe I did it wrong: "sqlite> select * from local_files WHERE location_short LIKE 'Furry';". This .sqlite file should contain the text "Furry" in the rows of the location_short column. Replacing "LIKE 'Furry'" with "LIKE 'Pony'" did not show quick results so I must be doing it wrong or sqlite uses a different thing compared to sql.
File: AbsolutelyNoSense.png (1.24 MB, 1059x1023)
1.24 MB
1.24 MB PNG
Stop being autistic and use the proper one
is probably better or the "proper one" compared to

I kept database_pony_furry_in-progress.sqlite out of some interest and for certain use cases.

>doesn't make sense
You numskull
desu is down
select * from table WHERE column LIKE '%text/%';


Also down now:
2023-03-11 23:01:40 GMT - last seen online by me
2023-03-12 00:##:## GMT and now - offline

Wonder how this text is crossed out (<s>test</s>):
File: 1641210312443.png (30 KB, 128x128)
30 KB
*2023-03-12 00:00:15 GMT

ur too dumb to correctly shoot your brain and kill yourself
I put stuff on my archive in plain sight, and yet all of you fail at every step on the way. Truly amazing.
Not only local now - "MD5 and SHA1 hashes of 60,084,657 files (mostly images) - pony and furry-related":

Or if you mean the 70.5 GB file "iwiftp_sha1-md5_hashes_complete.sqlite" - maybe later.

**2023-03-12 00:00:15 GMT and now - offline
OK, so you probably were talking about iwiftp_sha1-md5_hashes_complete.sqlite.

It's fine to have smaller subsets of the data though, such as this: >>39714964

Also website-warc-image-hashes.sql >>39713041 is helpful since it has hashes of extracted files which I don't think you hashed.
it was scheduled maintenance, both are back up now
On page 10.
Pretty slow around these parts, huh?
Don't know how recent this is. Lots of videos are now public on Digibro's channel: https://www.youtube.com/channel/UCHhnf3RgHabfk5f2gUX6EVQ Fans were previously given a three week grace period to download the videos before they went private but it ended several days early.

The pony videos should have been archived by now, but it's worth checking again. This time there may be other pony videos that haven't been public for a very long time.
Forgot to check the livestreams. Looks like they returned around 12 days ago.
Found some MD5 mismatches thanks to "database_pony_furry.sqlite". Errors are from the tarball, or not.

filename , MD5 hash in IWIFTP
http://s3.amazonaws.com/bronibooru/7968149325a9af5521c6f8c83ea56bce.png , bb92b42f4f8d3dcb4edcb1a488c2be66
http://s3.amazonaws.com/bronibooru/79689fa8d28154b6fe28227550ccb0ef.jpg , d41d8cd98f00b204e9800998ecf8427e
http://s3.amazonaws.com/bronibooru/7969cb760aa62d8f7d6ffa5043ca651f.png , 3dc6e39fae1b99d4fc110c4ce227403d
http://s3.amazonaws.com/bronibooru/7969df4051bbecd44feede929c558676.jpg , bce8861ef13e7fa7c3f357dd988b3c7e

Check against:
ERROR - incomplete - https://iwiftp.yerf.org/Pony/Art/Image%20archives/bronibooru.com/s3/7968149325a9af5521c6f8c83ea56bce.png
ERROR - empty - https://iwiftp.yerf.org/Pony/Art/Image%20archives/bronibooru.com/s3/79689fa8d28154b6fe28227550ccb0ef.jpg
ERROR - incomplete - https://iwiftp.yerf.org/Pony/Art/Image%20archives/bronibooru.com/s3/7969cb760aa62d8f7d6ffa5043ca651f.png
ERROR - incomplete - https://iwiftp.yerf.org/Pony/Art/Image%20archives/bronibooru.com/s3/7969df4051bbecd44feede929c558676.jpg
Maybe I should extract those files from "s3.tar" and see if they have the same errors and hashes.
In other news, Endchan got overly jannied

Same problem in 7chan, also some textboards such as 4-ch.net
123,164 full images in bronibooru.com

>thanks to "database_pony_furry.sqlite"
Via this extracted section (in w3s):
I did know about "iwiftp_sha1-md5_hashes_complete.sqlite" since months ago, so I wasn't as unaware as you may have thought.

>Hashes of errored files all start with "796"
That definitely speaks to something. Also, glad that that AWS thing wasn't killed off otherwise those files might have been lost forever.
>Page 9
Wish this image was used as a foundation for this image:

If it was, subsequently derived images would have looked better.
~40 GB
I set up an rclone remote to show Google Drive folders which are shared with me. I shared that folder with myself, but command "rclone lsd remote1:" only shows older folders which were shared with me. I refreshed the JSON in here: "token" = "[JSON]", removed "token_url" = "https://oauth2.googleapis.com/token", removed "auth_url" = "https://accounts.google.com/o/oauth2/auth" -> didn't help. Re-add "client_secret"/"client_id" maybe or just re-do the whole thing. Recreated the remote = still only shows older shares. New ones might show up later (hours/days later)? Folder remove instantly removes it as seen in CLI and GUI, but not folder add. I removed then re-added this one folder, and that took a while. Maybe it takes longer with larger folders. Recently added folders sometimes show up in https://drive.google.com/drive/shared-with-me then later they don't. I love it when software is not easy to use.
"shared_with_me" and "starred_only" don't really work, but it looks like "root_folder_id" works.
Downloaded anonpone.com WARCs via
>rclone -P copy "remote1:Fan-made/www.anonpone.com" "www.anonpone.com"
"41.624 GBytes" in "5h56m16.9s", max download speed was about 24 "MBytes/s", avg. 1.994 megabytes/s. This might confirm something I was thinking about. Captcha: PKKKMJ.
Loaded "www.anonpone.com-2020-04-05-dc4c5c93-00000.warc.gz" into ReplayWeb.page. It looks like these forms of choose your own adventure (CYOA) stories, in /mlp/, are only linear and do not have branching storylines. Or I didn't check enough anonpone.com pages as some of them might branches. Some CYOA books, like this one I know about from like ten years ago, had branching storylines.

Anonpone image files have an MD5 hash in the filename, e.g.,:
>$ echo -n 497f52f7877eab13828e6654278b4fb7 | xxd -ps -r - | base64 - | sed "s/+/-/g" | sed "s/=//g"
This is helpful to find the image again in desuarchive:
-> https://desu-usergeneratedcontent.xyz/mlp/image/1465/25/1465251920936.gif
--> DEAD

Anonpone has some images which died in desu-usergeneratedcontent.xyz, such as "1465251920936.gif" / "497f52f7877eab13828e6654278b4fb7.gif", which is posted again here and possibly now restored in desu. Image saved out of ReplayWeb.page via www.anonpone.com-2020-04-05-dc4c5c93-00000.warc.gz > list of images > https://img.anonpone.com/image/49/497f52f7877eab13828e6654278b4fb7.gif > dev tools > Network > http://localhost:5471/w/id-97553da18d3e/20200406034732mp_/https://img.anonpone.com/image/49/497f52f7877eab13828e6654278b4fb7.gif > Preview > right click on preview > Save image as...
File: gay people make me sick.jpg (2.83 MB, 2192x1704)
2.83 MB
2.83 MB JPG
You have to be 13 or older to use this website
Don't blame me for Jewgle being broken.
I do blame you
Anyone have a link/archive to this guy's pony yt videos?


He unlisted them recently, music is still up on bandcamp.
I was only able to find the link to these two videos:


Just a though, has anyone backed up whatever random pony stuff turned up on ytmnd?
A few random links:
Also you might find links to other forgotten sites through that.
Wrong character tags:


currently the same as: https://www.youtube.com/channel/UCh1vaUr-WcLudHdlfFjEPpw/videos

Some of his videos in TPA: 9iV0oznhQjM 5aCzxo4qmwI cO3S7O_AVMs hLna6-iyrgA EpLKzje_Mn0 27XBU7OuFCY
File: 1415821673875.png (148 KB, 1143x677)
148 KB
148 KB PNG
It is in here I guess:

is 1420070401

pic related might be the newest image in heinessen-mlp-images: "1415821673875.png". Unix Epoch 1415821673 = 2014-11-12 19:47:53 UTC.
File: 1415502855033.png (563 KB, 1280x1440)
563 KB
563 KB PNG
Is there any backup/mirror/archive of gelbooru.com?

I see they deleted this upload of mine (not in wayback): https://gelbooru.com/index.php?page=post&s=view&id=4158354

No idea what it actually was. https://gelbooru.com/index.php?page=gmail&s=view&id=1092745&read=1&search=mod_notices
>Subject: User upload deleted. #4158354 09/26/21 9:01 PM
>From: Jerl
>To: [me]
>This is an automated message to let you know that an upload ( post #4158354 ) you have made was deleted. Don't get your panties in a bunch! You can still get the image restored at this link ( forum #484 ) if you believe it does not break the Terms of Service of the site. However, be sure that it does not break our terms or risk getting your access to the site restricted. Do not contact an administrator about this. You will be banned if you do. Use the forum.
>Deletion reason: ToS violation - screencap deleted by Jerl

MLP in Gelbooru.com - which runs on proprietary software:
pic unrelated
>I did not know about that ipfs folder, I will probably download that soon.
- It probably sourced images from archive.moe /mlp/ - 420 GB: https://archive.org/details/archive-moe-files-201510-mlp
- It might have pulled images from this too: https://archive.org/details/heinessen-mlp-images

Other important thing about "/ipfs/QmcG5DVwy2RsvYg3Ki2HhSCfLMjcuAZhRdY2HWHwyzPrMc - 4chan-mlp/": looks like it was created with the "raw blocks for leaf nodes" option on. Proof (also tested with two other images):
>$ ipfs cat QmUDm1Ma9JtcBBhmBokv68HP7AzjZwhiJpj8FjacXCM2ru > file.img
>$ ipfs add --raw-leaves file.img
>added QmUDm1Ma9JtcBBhmBokv68HP7AzjZwhiJpj8FjacXCM2ru file.img
Looks like it uses the standard 262144-byte chunk size. (I think web3.storage also uses raw leaves.) This is also good to know if trying to pin and dedup archive-moe-files-201510-mlp + heinessen-mlp-images with QmcG...PrMc.
Seems like YP does not have the "My Little Pony: Sister Switch Storytime" and "Sunny Starscout's Book of Adventure" books (among a lot of others). Is there an archive which contains these?
Things are going along, perhaps not too many stand out features this week but still a fair amount accomplished. Some fixes to captchas, I think they should be basically done at this point other than adding them to new forms in the future. I'm thinking I might try and get some more of the bigger additions done soon. On my mind is the forum, bans, and the database memory cache. I guess we'll see how fast I get to those. Currently working on the forum. Probably going to reuse a fair bit from the commenting system.

>Added forms for captcha settings
Available settings are enable/disable, pool size, maximum captcha age, cycle rate/max captchas per unit time, cycle frequency/unit time, serve captchas from memory and difficulty. Also available is customization on where captchas are needed. So you can have captchas enabled globally, but also disable them for certain things. Ares include registration, uploading, image editing, commenting, and the forum when it gets made. Captchas can also be made to be required even when logged in.

>Captchas now display the time they expire

>Forms that you would expect to have captchas now have captchas

>Add ability to add a custom spoiler image
Can upload any file you like, this is the image used to replace the thumbnails on spoilered images.

>Add list previews
Basically those search previews that you typically see on the user profile page. They also make an appearance in /lists. Wasn't quite able to fill all of them out yet, still need to add faved_by to search and a way for users to associate tags with their profiles. So those are still blank on the user page.
>Allow live updating of captcha settings
All captcha settings can now be changed and immediately applied without restart.

>Fix the captcha picker sometimes getting stuck
Was mainly an issue with small pools. Due to a typo, I forgot the = in +=.

>Fix taskQueue not initializing the schedule correctly
Was due to the schedule getting initialized at module import, moved schedule initialization into it's own function

>Fix taskQueue skipping all tasks if one was disabled
Had accidentally used break instead of continue in the loop processing schedule items.

>Fix captchas getting pruned too early.
I had forgot to convert minutes to seconds, and so captchas were getting pruned about 60x as fast as they should have been.

>Prevent captcha rate from being set larger then the captcha pool
Doing so would mean that the extra captchas generated would be immediately discarded upon generation. No point in that, so this prevents people from accidentally setting it up to do so.

>Do not increment tag count on hidden images
While I wouldn't expect hidden images to need much editing, on the off chance that they do we don't want it to throw off the count of tags as hidden images are excluded from the tag count.

>Ensure users can only submit one report per item

>Fix adding report category without specifying tier
The default is supposed to be tier 100 if it isn't specified. What was happening was that the category wasn't being created at all if the tier wasn't specified.

>Fix anonymous reporters being treated as user in report check
This bug was making it so only one anonymous reporter could report an item.

>Fix invalid filter cookie breaking site
The default gets selected if the requested filter does not exist.

>Change modForms to use redirect on form submit
Makes it so the browser won't ask to resubmit data when you refresh the page.

>Fix commenting not changing the comment count
>Convert captcha to template
A bit cleaner, allows for more html reuse.

>Merge display of reports for same item
Instead of having a separate entry in the list for each report on the same item, reports get merged into a single listing showing all the reporters and report reasons.

>Split modForm html files
While having the tabs to switch between sections without reloading the page was pretty cool, they did have some downsides. For one, every time you wanted to go to any of the mod settings, you would essentially have to load all the pages for that category instead of just the page you were after. Would have also made it much more difficult to keep track of pagination when I eventually get around to it.

>Centralize permissions
Will make it much easier in the future for me to adjust the default permission levels of various pages and tools. Also opens up the possibility of me making the permission levels user adjustable in the future.

>Split image edit forms into their own forms
Instead of reusing the same form for description, tags, source, each has their own form now. Mainly done just to be a bit more consistent with how I handle forms.

>Split routes into separate files
The Booru.py file was getting a bit long and it was getting a bit harder to find what you were looking for in there. So routes are now separated into different files by category. Makes things much easier to find.
>>Captchas now display the time they expire
Interesting, or better than what 4chan does.

I hope Sunbooru supports decentralized technologies.
>I hope Sunbooru supports decentralized technologies.
You know, one thing I was thinking of is having ipfs available as an option for storing images as a means of lowering the barrier to entry of running a booru. Of course, there likely several issues in the way of that. Poor native browser support (and the fact I haven't actually used ipfs yet). It was always the intent to have several options available for file storage, but at the moment none of that is really high priority. It'll be a while before I even think of doing anything more there.

The other thing I know I was thinking of earlier on was a means of sharing image/tag updates/favs/etc between instances. I'll admit I haven't really thought about that for a while. Still would be cool if I could think of a good system for that. Again, will likely be some time before I give that more thought. Once the usual booru functions you'd expect are more or less complete.
Can't you integrate activitypub? Isn't that a way to basically share data over instances?
I didn't really know about activitypub, which shows about how much research I've done into the matter so far. My thought was to do it off of a public changelog and match via file hash. I know I'll have to do some sort of changelog anyway for the mod log at some point. Basically, I'll see about figuring it out when I get there as things like that are still quite a ways out.
Is iwiftp's tor site not accessible anymore?
>go on derpi
>type sexy, solo, [-tags that remove non-ponies]
>it's just a bunch of semi-anthro ponies with incorrect anatomy standing on their hind legs
I wish the artists didn't have such an overlap with furfags, they can't appreciate the beauty of a pony.
Ponies are the best.
That's hot.
Twibooru has filter ID 6 for pony only. Normally you would see a bunch of anthro/non-pony from searching "paizuri"/"boobjob", but not really any of that here:
Might load:

I hope it is up or will be back up sometime this year/month. I will check later probably. Onionsite is possibly linked in the previous thread.
Asking for help tracking down some old mlp websites. Specifically, two that were mentioned in a conversation I’ve had with this anon here.
There’s supposedly an old website out there that documented fan beliefs about the show, both fanon and canon. He claimed that there’s also a big list of every pony related site out there somewhere.
If anyone knows what this anon is talking about and could possibly provide the links to the things mentioned, that would be awesome. It would be cool just to have them, or maybe have them linked in the op for this thread or the mlp general for the sake of accessibility and posterity.
>there’s also a big list of every pony related site out there somewhere
More specifically >>39731719
>if it doesn't exist already
If it does exist then it would've likely appeared in previous iterations of this thread. Otherwise it wouldn't hurt to start this list for future pony scraping. Start with whatever is already saved on IWIFTP.
410 MLP-related domain names (missing the lost safeponi.com) - text file in w3s:
>iwiftp onion
It works, I was using the old link instead of the new one. But sadly what I was looking for is not there anyways.
Did you see it here: >>39731905 - checking these?

>old link instead of the new one
So that was probably the shorter link ("deprecated v2 address") compared to the longer link.

What were you looking for?
Maybe you were looking for >>39727877
>Seems like YP does not have the "My Little Pony: Sister Switch Storytime" and "Sunny Starscout's Book of Adventure" books (among a lot of others). Is there an archive which contains these?

Check Library Genesis and https://archive.org/

Also, thinking about book titles in my local library = not there.
Anon, I don’t know what that is or how to use it.
Yes, I had the v2 one. I rarely use onion links, and from the timestamp of the shortcut I assumed it was the updated one.

>What were you looking for?
These : >>39732310 >>39732330
There was storytime threads for these (I ripped them) but usually because of 4chan's file size/dimension limitations these are not full quality (plus there are a few missing pages I assume for index/contents/legal/whatever, the actual "content" is there).
Thanks. Sorry, I don’t keep up with the internet privacy/freedom space. So I wasn’t aware of how to use the ipfs link.
>start this list
DMOZ or Curlie or might have created an index of such websites. Wikipedia info and example index:

Wikipedia on DMOZ (11/12/2016 article revision):
>DMOZ (from directory.mozilla.org, an earlier domain name) is a multilingual open-content directory of World Wide Web links. The site and community who maintain it are also known as the Open Directory Project (ODP). It is owned by AOL but constructed and maintained by a community of volunteer editors.
>DMOZ uses a hierarchical ontology scheme for organizing site listings. Listings on a similar topic are grouped into categories which can then include smaller categories.
This is great for making such a list of MLP websites and webpages:
>1 - 20 of 8,838

(DMOZ = dead, replaced with Curlie)
There should be more in these curated lists:
That shows stuff like "my little [some other keyword not pony]", so it has these matches: "my little -pony".

In quotes for exact text and "mlp" search:

Not giving Fandom Inc. my traffic:
I bet anime fags are seething because all of their boorus are shit+cucked.
Page 10 bump.
>Endchan overly jannied
Also Endchan posts are only accurate to the minute, not second, so another reason to dislike that imageboard (though I like their UTC time option).
Fix the OP, https://smutty.horse/ is still read-only.

Wonder what the max upload size of this is:
File: soyjak.download.png (71 KB, 1366x768)
71 KB
>Wonder what the max upload size of 'jak.download is.
Max filesize of https://'jak.download/ is 500 MB (pic related). TOS says they respect DMCA. I don't think the following Alex Jones video is really copyrighted. You can see the full movie (released in 2000) in one or more YouTube uploads and multiple archive.org items.

Tested (IIRC, I never watched this video):
>Title: "Dark Secrets Inside Bohemian Grove" (2000-10 video)
>English-language movie, AKA: "Dark Secrets: Inside Bohemian Grove"
>File: "/mnt/My_Passport/b/fc2a-3930/3/Documentaries/Alex Jones - Dark secrets inside Bohemian grove (1st) (english) (2000) --- 15-30.wmv"
>File source: probably some torrent from the 2000s or early 2010s.
>Filesize: 394,663,384 bytes
>SHA1 hash of file: f0ec18c355f1bc9852a6e574b3dc8c123c89f940
>SHA1 hash of file, more info: didn't see this in IA; should be the same downstream, didn't check yet, and unless I later post otherwise, it matches
>Access: 2020-09-24 03:45:59.269639300 +0000
>Modify: 2013-12-27 07:56:22.000000000 +0000
>Change: 2020-09-23 19:35:42.773295300 +0000
>IMDb: https://www.imdb.com/title/tt0379237/
>DL page: https://s'yjak.download/f.php?h=3E3JUrhZ
>DDL: https://s'yjak.download/f.php?h=3E3JUrhZ&d=1
Total images uploads as of now:
ponerpics.org - 6,643,260
twibooru.org - 2,911,597

Probably from 2009 torrent 2940a3207c0cd86a4a1afeb9b82f36a4fe791231
Why a difference of millions?
Ponerpics left a huge buffer empty so that they could just keep the same ID that's on derpibooru, Twibooru didn't.
File: zoomed_out.png (697 KB, 1365x1344)
697 KB
697 KB PNG
This image is from https://sirzi.deviantart.com/art/Difference-739424888 -> http://www.bronibooru.com/posts/113718 -> www.bronibooru.com-2020-04-07-08d56d35-00000.warc.gz -> Google Drive -> *00000.warc.gz -> replayweb.page -> www.bronibooru.com/posts/113718

Oh. I wish Ponerpics had some API info in one of the pages linked at the bottom of each page such as https://ponerpics.org/images/445556

All not what I was looking for (didn't work):

Should look like this (working):
Found it:
Such API links were not specified in Ponerpics, had to look elsewhere to find that.
PonerPics API is identical to Derpibooru, and Ponybooru. They all run Philomena. Just replace the domain name:
Useful to select a random ponebin paste:
File: file.png (6 KB, 543x182)
6 KB
Why is everything I try to download off of the pony archive corrupted?

Does anyone have fractilx's channel backed up? He privated all his pony videos last week.
>Page 9
Saw this horse advertisement above this thread
>- Image
>- Links to
>https://danbo org/gate [...] https%3A%2F%2Fdarkfantasies.co.uk%2Fproducts%2Ftwo-hole-horse-fantasy-stroker-1

If I were a rich man...
File: twibooru-1661477.jpg (153 KB, 1280x720)
153 KB
153 KB JPG
>>Argument list too long
>If it isn't universal maybe perl can handle larger text. Or I could use vim to input and output via files.
I used VimL to overcome this limitation:

>Why is everything I try to download off of the pony archive corrupted?
Give me the links to the corrupted theponyarchive.com files, including "fractilx - Chickenjack.mp3"
pretty much everything in here
This file, and maybe others, are composed entirely of binary zeros:
$ curl -Ls "https://theponyarchive.com/archive/ponymusicarchive/Pony%20Music%20Archive%2021.03%20(Raw%20Quality)/Artists/InklingBear%20(fractilx)/Chickenjack.mp3" | xxd -ps - | perl -pE "s/000000000000000000000000000000000000000000000000000000000000\n//g"

This could indicate that there was problem downloading torrent "Pony Music Archive 21.03 (Raw Quality)" or TPA didn't finish downloading that torrent yet.
2 other files also have the error "Entire file is binary zeros":
>$ TZ=UTC wget "https://theponyarchive.com/archive/ponymusicarchive/Pony%20Music%20Archive%2021.03%20(Raw%20Quality)/Artists/InklingBear%20(fractilx)/Flurry%20Heart.mp3"
>--2023-03-19 03:04:01-- https://theponyarchive.com/archive/ponymusicarchive/Pony%20Music%20Archive%2021.03%20(Raw%20Quality)/Artists/InklingBear%20(fractilx)/Flurry%20Heart.mp3
>Resolving theponyarchive.com (theponyarchive.com)...
>Connecting to theponyarchive.com (theponyarchive.com)||:443... connected.
>HTTP request sent, awaiting response... 200 OK
>Length: 14348473 (14M) [audio/mpeg]
>2023-03-19 03:04:04 (7.09 MB/s) - ‘Flurry Heart.mp3’ saved [14348473/14348473]
>$ exiftool "Flurry Heart.mp3"
>Error : Entire file is binary zeros
>$ TZ=UTC wget "https://theponyarchive.com/archive/ponymusicarchive/Pony%20Music%20Archive%2021.03%20(Raw%20Quality)/Artists/InklingBear%20(fractilx)/Lunar%20Ascension.mp3"
>--2023-03-19 03:04:29-- [...]
>2023-03-19 03:04:31 (4.62 MB/s) - ‘Lunar Ascension.mp3’ saved [4128210/4128210]
>$ exiftool "Lunar Ascension.mp3"
>Error : Entire file is binary zeros

Hopefully the files are not intentionally filled that way and it is a result of some mistake.
Download part of TPA:
>$ wget -U "Mozilla/5.0 (X11; U; Linux; en-US; rv: Gecko/20110929 Firefox/3.5.16" --recursive --level=0 --no-clobber --page-requisites --html-extension --no-parent --wait=1 --random-wait --span-hosts --domains=theponyarchive.com "https://theponyarchive.com/archive/ponymusicarchive/Pony%20Music%20Archive%2021.03%20(Raw%20Quality)/Artists/InklingBear%20(fractilx)/"

Could check other files under folder "Artists" or folder "ponymusicarchive", to see if they are also zeroed.
Didn't exclude subdomain new.theponyarchive.com
Looks like the copy on pony.fm is fine https://pony.fm/tracks/21837-chickenjack
It’s not an archived profile but created by the artist himself so you should probably get it before he remembers to remove that too.
Your first grave error was to use TPA for anything
--exclude-domain [list]
thank you
ARCHIVING GENERAL -> /pag/ - Pony Archival General #2:
Can someone please seed the My Little Pony 1986-1987 Disc 4 ISO? I've been stuck at 73% for almost a week. I will seed it afterwards whenever I can.
Will do. My share ratio on disc 4 is 8.65, so I quit seeding for a while. Share ratios on other G1 MLP ISOs = D1: 4.74; D2: 9.43; D3: 9.07.
currently equals https://www.youtube.com/channel/UC2bdrFW4uFo40vyfQWrtKUw/videos

pony.fm webpages with non-Latin script titles don't really work (text bytes above 0x7E): >>39742536
Fluttershy looks kinda anthro in the ad https://archive ph/shcXg
Used >>39740231 to list files in >>39691045 for hours. Maybe cloudflare-ipfs com (HTTP) is quicker to get directory listings and perhaps even the files themselves.
CF takes as long / times out / HTTP ERROR 524 with "untouched" folders. Plus, that .com site does not show the size of each file in bytes.
YouTube bloatware said themerefresh, so I don't know if the vids in sawt*'s 2nd channel were still up at this time:
Looks like Zippyshare's shutting down at the end of the month. Time to get saving ponystuffs. Again.
Do files on there only last for a maximum of one month? Does it have "permanent" files?
I honestly dont know anything about them. It's definitely temporary for anonymous uploads (deleted after 30 days of no activity), but I don't know if getting an account does anything for you.
All of the socials were purged within hours of each other. There should be some comments saved for the pony videos on this channel too (obviously wasn't going to spend time on that non-pony video 33 million views when all the videos were disappearing at any time). Will find my copy again.
Mon, 20 Mar 2023 02:43:50 GMT https://wiki.archiveteam.org/index.php?title=Zippyshare.com&action=submit
Mon, 20 Mar 2023 02:43:50 GMT https://wiki.archiveteam.org/index.php?title=Zippyshare.com&modqueued=1
archive ph has a dot in the submitted version:
>{{Infobox project
>| title = Zippyshare.com
>| image = <!-- . -->
>| URL = [https://www.zippyshare.com/ www.zippyshare.com]
>| description = Free File Hosting, "File Life: 30 days after no activity."
>| project_status = {{Endangered}}
>| archiving_status = {{Unknown}}
>| lead = [[User:Usernam]]
>'''Zippyshare.com''' is a temporary file sharing website. (It was "permanent" if uploads/file had enough activity.) In 2023-03-19, it was announced that it would '''shut down in 2023-03-31.''' In a post titled [https://blog.zippyshare.com/?p=1211 "Inofrmation about the closure of the project"] it was stated that the site would shut down due to being an old/old-style website, being less popular, being ad-blocked, and not being able to pay for stuff like electricity.
>www.zippyshare.com is [[List of websites excluded from the Wayback Machine|excluded from the Wayback Machine]], so https://archive ph/www.zippyshare.com is one source which can be used to discover files and accounts such as https://www.zippyshare.com/jillem (182 pages of uploads, all alive as of writing this).
>Zippyshare began in 2006. ZippyShare "outlived many similar websites, such as RapidShare, Hotfile, and Megaupload."[https://en.wikipedia.org/wiki/Zippyshare]
>[[Category:Partially Rescued Sites]]
Dead links here:

Also, /qa/ won: >>39735772
>not saving/archiving stuff in general
I can't really complain since I wasn't there for it and you were. "The hardest thing in history is being there."
What I mean is, after downloading everything top priority then you could have downloaded lower priority stuff.
Fair, especially since I wasn't planning to attend to the non-pony videos anyway. I also have a really bad habit of assuming things are going to be around in general. Speaking of which, ILoveKimPossibleAlot's youtube channel is back after being hacked by some crypto guy. Should get on to that later.

Anyway, I cannot stress how close this incident was. I saw the community posts being deleted in real time along with the other socials too so I couldn't think straight. Half an hour later, youtube was showing cached versions of the videos as they were being privated en masse.
Speaking of ILoveKimPossibleAlot, when I saw that posted above about all the videos being gone, it was sad to me because I think I saw some vids from that channel in the past which I liked. Glad it is back up.

>saving/archiving stuff in general
To preserve the memory of the "good old web", I am using grab-site to download zippyshare: start URLs via index archive.today/zippyshare.com and index archive.today/www.zippyshare.com
For each smaller independent website that shuts down, FAGMAN gains more power.
Is grab-site able to download contents in HTML/DOM shadow-root nodes?

Zippyshare DDLs have to be JS-computed (>>39745171), so this could be helpful:
>It's definitely temporary for anonymous uploads (deleted after 30 days of no activity), but I don't know if getting an account does anything for you.
has the text:
>26. Irrespective of Zippyshare.com entitlements provided for in the Terms and Conditions of Use concerning deletion of files, folders and other content, Zippyshare.com declares that it may delete files, folders and other content in the following instances:
> a. if more than 30 days have passed since the last download of a public file uploaded to the Site by a registered user,
> b. if more than 30 days have passed since the last download of a public file uploaded to the Site by a non-registered user,
> c. if more than 30 days have passed since the last download of a private file uploaded to the Site by a registered user,
> d. if more than 30 days have passed since the last download of a private file uploaded to the Site by a non-registered user.
> Zippyshare.com may also delete the user’s account and the files located therein if more than 90 days have passed since the user’s last login.

Also get links from https://archive.today/*.zippyshare.com
>Is grab-site able to download contents in HTML/DOM shadow-root nodes?
"It depends", basically
Now dead, why I don't bother making these threads
Screw it, here is the next ARCHIVING GENERAL:
Are you sure it's good to just spread things around a bunch? I think that's why those threads get bumped off the catalog.
I posted about archiving stuff in this general before and it annoyed at least one user. I also do not like to spread things out but I thought this general could be useful for:
- Not-so-summarized posts or posts about specific stuff
- (in the context of archiving) data analysis, developing code for whatever, learning about software
- Posts about websites where pony is not the main focus, like YouTube, DeviantArt, etc.
- Fun posts about cultural artifacts which were found to be interesting - "I like this [preserve it for future generations is implied]"

Maybe we should have a poll to see if all of the above types of posts are fine for /fsa/ or should be fractured and split into /pag/:
I think it's fine, whatever. I think whatever the anons were bothered with was more the style of your posts, and sometimes the frequency of them, rather the archival efforts themselves. A lot of those commandline commands, file hashes or other such things are pretty superfluous, perhaps spammy even, that you could easily just tuck in some pastebin somewhere.

Personally I have no interest in bumping two threads, when there's barely a few anons even posting in these threads regularly.
If a single thread can't get bump limit, what's the point to keep remaking it? Just keep it here
The amount of pastebin pastes or CIDs I would have to make for a small amount of text is the annoying thing, just like you said "I have no interest in bumping two threads". My time/effort is worth something. Archiving includes people who have an interest in specific stuff/information like hashes and CLI commands.

>when there's barely a few anons even posting in these threads regularly.
Another option is to renamed "Fan Site Alternative Thread" to "Fan Site Alternative Thread and Archiving General". Has a general ever been renamed before? Some of the opposition could be "Reeeee! I don't like change!"
Zippyshare does in fact have one or more pony files: >>39746305

Other thread, live, on zippy: >>>/g/>>92200188
Captcha: 4JAY4
First read it as "4AJ"
Brave Search: "mlp site:zippyshare.com" ->
>Name: Secret_Horse_Files_3_v36_unpacked.zip
>Size: 78.42 MB
>Uploaded: 07-09-2021 09:28
>Last download: 20-03-2023 17:52
File: pony_adult_female2.png (2.34 MB, 4096x4096)
2.34 MB
2.34 MB PNG
horse secrets revealed. pic related is an image in that zip
Billions must watch

[Advertise on 4chan]

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.