Oldfriend Archive

Updated 2024-04-07

This is a searchable archive of old (late 2000s to early 2010s) 4chan posts. Currently it hosts the "Archive Ten Billion" dump of 2005-2008 era 4chan threads, a complete scrape of the textboards, most of the text content from chanarchive.org, another dump of mid-2011 posts, and about 400 2003 posts from 4chan.net. Among these are the very earliest posts for many boards: /fa/, /fit/, /hc/, /jp/, /n/ (the Transportation reset), /r9k/, /tg/, /toy/, /trv/, /x/. It includes posts from almost all* 4chan boards of the era - about 173 million posts/11.2 million threads in all, ranging from 2003 to 2014, but mostly from mid-2006 to November 2008 (see the stats page). Most of this content, as of this time, is not hosted elsewhere.

*missing /f/ (flash), /j/ (the sekrit club) and /yg/.

There are no NSFW images hosted here, but plenty of NSFW language.

For a different way to browse some of the contents of this archive, check out 4museum (unaffiliated, est. 2019).

Reports / Contact

For personal information takedown requests, valid links to illegal content, or bugged post data, use the report function. To report other problems or for other inquiries email admin[@]sage.moe.

Q&A

What's the point of this site?
This site exists to be an easily-accessible repository of late 2000s - early 2010s internet culture. The era before Facebook/Twitter took over. I hope that it will help people researching internet history (such as the first SCPs), nostalgic oldfriends, and newfriends wanting to RP as oldfriends.

Who is this "Four Chan"?
For some time in the late 2000s (the main period covered by this archive), 4chan was a nexus of internet culture. In that period, it generated countless memes and other original content which propagated out to the rest of the internet.

It was also one of several major gathering points for anonymous online agitators (i.e. the internet hate machine), who would organize raids (concerted trolling) against various targets. In 2007, this activity had earned them enough notoriety to be recognized as "hackers on steroids" in an infamous Fox News broadcast. By 2008 with Project Chanology, Anonymous had exploded into real life and normies everywhere lived in fear of masked hackers blowing up their computers.

Of course, the vast majority of the content was just inane bullshit, and that is reflected here.

Where did you get this data?
I sourced most the data from the Archive Ten Billion dump Internet Archive item, which was uploaded by Jason Scott in 2018 after originally acquiring the data in 2009. The original scraper was not identified (it probably wasn't Jason). The textboards content also came from an anon on archive.org, the chanarchive.org and 4chan.net content was recovered in HTML form from waybackmachine scrapes and the 2011 dump was hosted on a filesharing site.

The original IA upload was borked in a few ways, and the data needed non-trivial repair work before it could be fully migrated to a searchable online archive format. See History for more.

Who originally scraped the Archive Ten Billion data?
The archive of database files in the Ten Billion IA item is named chanarchive.tar.gz, which might lead someone to assume that the dump comes from 4chanarchive/chanarchive, the best-known archive from this early in 4chan's history (4chanarchive ran from 2006 to 2013). But this is a coincidence.

The data might originate from a site called Rapidsearch, which was an index of links to uploads on file-sharing sites like Rapidshare. This would better explain why the first posts in the dump come from 4chan's /r/ (where file-sharing links would frequently be shared), why the dump contains threads from other sites besides 4chan (which match sites covered by Rapidsearch), and why the scrape coverage is comprehensive instead of selective like 4chanarchive (which served under 20,000 saved threads at its peak).

Whoever it was though, thanks for sharing.

How can I trust the data here?
I compared sample posts/threads to the GETs encyclopedia, the 2008 /a/ overlap with Desuarchive, and 4chanarchive threads on waybackmachine (e.g. compare the first thread here to this). Most of the posts here can't be found anywhere else, but those that I've been able to test match other sources.

Where are the images?
For the Archive Ten Billion posts, image data minus a binary "post has image" indicator is missing in the source. This includes the thumbnail/full-size image files, original filenames, file sizes and md5 hashes. A placeholder image is used here (with dummy metadata).

For the early 2010s posts, most image metadata was kept in the source, but most of the images themselves were not scraped. The placeholder thumbnails for these posts will often appear distorted, since the original aspect ratios have been maintained.

Were any posts removed from the source archive?
About 10k posts with odd characteristics had to be removed from the Ten Billion dump to get FoolFuuka working. Quite a few of the textboard dump and chanarchive.org posts were malformed in one way or another and also had to be discarded.

What other data is missing from the posts?

See the section for each source archive.

• Posts were only timestamped to the minute in the Archive Ten Billion source, except for on /b/ for some reason. This means that posts ordered by timestamp are often disordered by number, since many will share the same timestamp. The same is true for some of the chanarchive.org (>2008) posts.
• Capcodes are missing from the source.
• Some ban messages seem to be missing from the source. Others are preserved.
• All HTML tags have been stripped (some were stripped in the source).
• Oekaki post data (drawing duration etc.) has been stripped.

What's with the numbered boards (/n1-2/, /sp1-2/, /r9k1-2/. /hc1-2/, /con1-4/)?
During the period covered by this archive, some of the boards were reset (cleared, post number reset to 1). This means that in some cases, post ranges in a board would overlap with those of an early iteration of the board. To prevent collisions with thread links and avoid confusing the software, affected boards have been split up into numbered sections.

• /n/ switched from 'News' to 'Transportation' on 2008-02-19, so posts/threads from both of these iterations of /n/ are in the archive. Before switching to 'News' in 2006, /n/ was 'Nature' in 2005 and 'Trains' in 2004 (neither covered in this archive). The post count was reset when /n/ transitioned to Transportation.
• /sp/ was killed in 2006 and was brought back with a post number reset on the same day as /n/ became Transportation, 2008-02-19.
• /r9k/ was killed in early 2011 and brought back in late 2011
• /hc/ was killed in 2009 and brought back in 2011
• /con/ was reset 3 times during this period between different cons.

What time zone does the archive use/what's with the timestamps?
Ignore the displayed date; this will be 4-5 hours off the correct time for your time zone. Treat the "4chan Time" date in the alt-text (mouse over the date) as correct UTC (regardless of your local time zone, and with the exception of some annual 1-hour periods on New York DST start dates, where 3am-3:59am happens twice). In most newer archives, "4chan Time" is ET/New York time, not UTC. Credit to this guy for noticing the discrepancy.

I might adjust the "4chan Time" to ET in the future, which would fix the displayed (localized) date as well. This would involve doing some ugly things in the database. The "4chan Time" issue affects other 4chan archives in various ways and has roots in the fact that early scrapers did not have access to a UTC field in the HTML.

Could I get a dump of the data?
You can download an dump of the posts database including the Ten Billion, textboard and chanarchive.org content here. You can download a repaired version of the "Ten Billion" data here.

Source data

Archive Ten Billion (2006-2008 full archive)

This is the 90% complete 2006-2008 dump which makes up the majority of the archive. Tons of memes and other original content make their first appearances here.

2014 textboard progscrape scrape (2004-2014 full textboard archive)

This is a mostly-complete scrape of the 4chan text boards made in 2014, before their closure in 2015. The most notable content would probably be the programming discussion on /prog/ (companion board to /g/ and by far the largest board). It also goes back just a little further than the oldest posts in the Ten Billion dump (to December 2004).

chanarchive.org waybackmachine scrape (2006-2013 partial archive)

waybackmachine scraped almost all the text content of chanarchive.org (formerly 4chanarchive.org) before it closed in 2013. These threads were added to the archive via a voting system, so only threads considered particularly noteworthy at the time made it in. The content is notable for bridging a period of time not covered by other archives, containing a lot of "nostalgic" OC, and being a "best of" collection (for what it's worth).

k47 5M post dump (mid-2011 ~full archive)

This is a mass scrape of ~5 million posts from ~all boards from a short interval in mid-2011. It comes from a blogger by the name of k47.

4chan.net 2003 waybackmachine crawls

This is a very small collection of posts taken from all 4chan.net board pages captured on waybackmachine in early October 2003, starting within 3 days of the site's launch.