About the restoration of alt-webring

The original alt-webring

Launched in April 2003 by Andrew McPhee, the stated purpose of alt-webring was to combat dissatisfaction with other webring systems that the creator felt had become either too big and overly complicated or were abandoned and unsupported.

The site was a passion project, with no intent to become commercialized. It also intended to "be around for a long time". This recreation effort is in the same spirit.

You can view an archived version of the original alt-webring's about page here.

What were webrings?

Webrings were a website discovery method popular from the mid-1990s through the early 2000s. They worked as circular networks of sites linked around a common theme, like fan pages, hobby sites, personal blogs. Each member displayed a small navigation bar with links to move forward or backward through the ring, jump to a random site, or visit the central hub. Before search engines became reliable, this was one of the best ways to find quality content on niche topics.

The system worked because it was human-curated. Site owners joined rings relevant to their content, and administrators often reviewed submissions before accepting them. For independent creators, it was a practical way to get traffic and connect with like-minded people. Webrings faded as search engines improved and social media absorbed most personal web activity, though they've seen a small revival among people interested in alternatives to algorithm-driven discovery.

There is an excellent write-up of webring history at brisray.com.

🐰 Bunny: DNS, CDN, storage, and more.

This website is hosted and served by Bunny, a provider of DNS, CDN, and web storage solutions. If you happen to be in the market for such services and would like to help offset the cost of hosting and serving this site, use this affiliate link when you're ready to sign up for Bunny.

Occasionally Asked Questions

The directory page is missing a webring, a website within a ring, or a link to a webring or a website is missing.

Yes, this is a partial recreation from content archived by the Wayback Machine. Not all rings, sites, or destination webpages were archived.

For sites in the directory that are still online today, why do you link to the Wayback machine archive and not the current site?

This is an intentional choice based on two factors:

1) The directory is intended to be a time capsule. The link to each site's archive is generated based on the most recent snapshot timestamp of the redirection to that site.

2) The web continues to evolve. Sites that are still online today may not be tomorrow, and I'd like to avoid contributing to further link rot.

I acknowledge the irony of fighting link rot by directing everything to one single point of failure.

Were LLMs and/or Generative AI used in the creation of this website?

You betcha. A lot of the work involved in this project was creating and running scripts to extract and transform very old data. This is an excellent use-case for LLMs and Generative AI. I would not have taken on this project without the barrier of entry being lowered by these tools.

What tools were used when making this site?

A special shout out to WaybackPack and Beautiful Soup, two tools that were very helpful to this effort.

I'm curious about the traffic stats for a site like this?

Same. This site uses Cabin Analytics, a privacy-preserving solution. Here is a public stats page.

How can I support you for this work?

This is very kind of you, but is unnecessary. Please consider donating to the Internet Archive, which runs the Wayback Machine service instead.

How can I contact you?

The following inbox is loosely monitored with frequent delays: email@alt-webring.com

Approach to this Recreation

Note: I am not a professional archivist / preservationist. This is a summary of the methods I used for this effort, but these may not be best practice in any number of ways.

Step 0)

Purchase the domain name alt-webring.com, and set up a basic landing page with analytics. For all hobby projects, it is important and customary to pay too much for the domain name.

Step 1)

Inspect traffic logs and investigate the Wayback Machine for older versions of the site to determine the best way to extract data.

Step 2)

Get stuck and give up for a bit as life gets busy. "A bit" is a universally accepted non-standard unit of time that, in this case, is approx 2 and a half years. Whoops.

Step 3)

Several years later during a holiday break, decide to take a fresh pass at the challenge.

Step 4)

Start by downloading a copy of every archived snapshot of the main.pl page using waybackpack. Then use Beautiful Soup to extract the list of ring names and metadata. Upload this data to a Google Sheet and de-duplicate the content based on the most recent timestamp. Remove obvious spam content.

Step 5)

Proceed to begin downloading a copy of every archived snapshot of the list.pl?ringid= pages using waybackpack. As before, use Beautiful Soup to extract the list of ring names and metadata. Upload this data to a Google Sheet and de-duplicate the content based on the most recent timestamp.

Step 6)

Begin prodding the goto.pl link associated with each site to extract the destination URL, which was dynamically served by alt-webring. Complete two passes, one to provide the Wayback Machine's indexed URL, and the second to provide the destination URL.

Step 7)

Merge the dataset and complete a few manual checks and spot-fixes for process errors.

Step 8)

Take the prior custom website design and vibe-code improvements. Modernize the code, change the layout, and add quality of life features.

Step 9)

Review the original URLs that remain in-the-wild, along with analytics stats, to create Edge Rules at Bunny to audit inbound traffic and provide a suggested quick jump if the redirecting ring can be detected.

Step 10)

Programmatically generate the html code for the directory using the Google Sheets dataset.

Step 11)

Add finishing touches (ex. an about page) and launch.