Unvanquished archæology: restoring 10 years old blog posts

The Unvanquished blog is now an archivist paradise.

Doing wizardry with the Internet Wayback Machine, I recovered the 43 articles from the old vBulletin blog we thought were lost forever, and some others for a total of 46 recovered news entries.

<granger> time machine? (illustrative picture CC BY-SA 3.0 Archivo FSP).

The loss

What happened is that the first Unvanquished website we had in March of 2012 was a vBulletin forum. and the blog was just a part of this forum. The team migrated to that website a bit in urge as all previous blog posts by kharnov from January of 2012 and the first Unvanquished release post published on last day of February of 2012 were deleted in the very immediate days following the release.

Six months later in October of 2012, we added a Joomla website to have a proper home page. That Joomla website became the new place for blog posts, but the old blog posts were not migrated to Joomla since the vBulletin forum was still used as a forums so they were easily available. It looks like when the vBulletin forum software was replaced with phpBB, the blog posts were not retrieved. Lately the Joomla website was turned into a WordPress website, and while the articles published on Joomla were imported into the WordPress software, the former vBulletin articles did not followed as they were never imported to Joomla to begin with.

We thought those blog posts were lost forever as everyone forgot what was their urls and without those urls we didn’t find them in the Internet Wayback Machine. Also it turned out many of them were only stored by the wayback machines with very complex urls with garbage GET parameters. The 19th blog post was even stored by the wayback machine with an url adding ?goto=prev parameter to the 20th blog post url, that says a lot about how those urls were unpredictable.

The recovery

I found trace of what was the rough shape of those urls so I started digging. I also recovered some articles thanks to finding old copies of the RSS feed.

Among those 43 articles,

  • 37 were fully recovered: title, date, author, full text with markup;
  • 4 were only partially recovered: title, date, author, first paragraph;
  • 2 are considered lost: title and date were recovered, author was discovered by doing math, content is unknown.

Some images may be missing in some fully recovered articles anyway. Many images were stored on imgur.com and either some of them disappeared, either imgur.com gave the unique id to another image so we now find totally unrelated things on the same url. But a lot of those images were properly recovered and re-uploaded there to not rely on a third-party service.

Some images that were also lost on our more-recent Joomla articles were recovered thanks to finding on the wayback machines some copies of the blog posts that were done on forum at the time (those images were probably stored in vBulletin, disappearing with it).

For the partially recovered articles we included the part of the text we had (first paragraph) and a comment saying the rest is lost.

For the lost articles a dummy article was created with proper title, date and author, and a comment saying the content was lost, so we at least have a proper history.

I also recovered the 2 blog posts kharnov published on the tremz.com website before the first release of Unvanquished. They were deleted with the very first Unvanquished release post right after the publication of such release post (this was not planned).

The release post of the very first alpha release of Unvanquished that was published on tremz.com was not recovered as it was so quickly deleted (it only survived 2 or 3 days, really!) it was never archived by the wayback machine. The date of the article is known as it is documented by a third-party website (the FreeGamer blog) and the title was recovered from a link on that third-party website that linked the article before it was deleted. A dummy article with the date and the title was created with a comment saying the content is lost.

The archive

Our blog post list is now complete with a proper entry for every blog post ever made, even if a few 10 years old ones are incomplete.

Among the articles retrieved from the Internet Wayback Machine, previously stored on the vbulletin forum, are the three very precious retrospective blog posts about Unvanquished history kharnov wrote in 2012, 10 years ago:

Maybe it’s time for a new retrospect article? There will be one. 😉️

A special history tag was created to look for history-related blog posts, this includes retrospective blog posts, but also review of work-in-progress efforts and blog posts sharing expectations for the future.

We now have a blog post entry for every release we made, though the first release post is a dummy, all others are now fully complete. Only some images may be missing but it’s rare.

Along the first release post the current website missed the release posts for the alpha 2, alpha 3, alpha 4, alpha 5, alpha 5.1, and alpha 6. They are now recovered and a special release tag was created to easily browse all our past releases. Next release post is in the pipe. 🤫️

This archæological effort was done while preparing a blog post for an important Unvanquished anniversary! If you like stories, stick around, you’ll get some!

One thought on “Unvanquished archæology: restoring 10 years old blog posts”

Comments are closed.