What is "the Outer Web"?

It's just a name we came up with to describe what's also referred to as the indieweb, blogosphere, small web, personal web, and old web. Some of these terms don't cover everything we're interested in, others are closely associated with a specific entity and we don't want to step on any toes.

As for the name, we like sci-fi and imagine the noncommercial web as galactic rim planets filled with scum, villainy, and Rust projects.

The Outer Web isn't a platform or product, it's just our way to describe the authentic part of the internet.

And "Navicomputer"?

Sticking with the sci-fi theme for a moment, Navicomputer connects the planets of the galactic rim. More concretely, you give Navicomputer a webpage and it will show you similar content. If you have a blog post about blue milkshake recipes, Navicomputer can link you to other posts about blue milkshake recipes.

So it's a blogstumbler?

Yes and no. Blogroulette sites link you to a random or featured post whereas Navicomputer provides links to similar content. But like blogstumblers, Navicomputer is a mechanism for exploring the web, it's just a more guided journey.

So it's a blogroll or webring?

Yes and no. Blogrolls and webrings typically connect web cliques on a site-to-site basis. This makes sense for topical discovery so long as people only blog about one or two subjects. Page-granularity linking is important because many web contributors produce content on a variety of subjects, like coding projects, cat pics, scuba dive logs, and blue milkshake recipes.

Blogrolls and webrings create neat loops through the galaxy of authentic web pages. With the Outer Web's Navicomputer we've created a traversable mesh of these same places, and no two neighbors need know each other.

So it's search?

Yes and no. Like search, Navicomputer links pages, rather than sites and is content-focused rather than clique-based. Unlike search, Navicomputer uses an entire webpage as input rather than a search string. From a user experience perspective, Navicomputer is meant to function more like a wiki rabbit hole than a query engine.

Also there are some great 'alternative' search engines out there already, we'd rather address a different use case.

Does it make the web better?

Yes and no. This tool doesn't help the web by itself. If no one uses Navicomputer the web will continue on its current trajectory, whatever that happens to be.

As discussed above, Navicomputer can be used to explore the web like blogstumblers and webrings, but with more granularity and guidance. This gives visibility to websites that might fall into the "Crawled - currently not indexed" pit of obscurity.

Providing a navigation engine for the Outer Web is nice and all, but Navicomputer has a use case that impacts the health of the internet in a more meaningful way. Alongside search-like results, Navicomputer spits out the an HTML snippet that can be added to your webpages. This snippet isn't a redirect to our site, it's static HTML containing the query results - title, description, and link. So, with the help of bloggers and personal site publishers, anyone can use Navicomputer without ever knowing it exists. Also worth mentioning:

• Navicomputer doesn't require user adoption to be useful; the results are the same if one person is using it or a million.
• The link snippets contain static HTML and thereby have no dependencies except the site on the other end of the link.

What counts as Outer Web?

It's a work in progress, but we look for three things:

1. Authenticity
2. Noncommerciality
3. Informational quality

Storefronts, AI-generated 'blogs', and affiliate referral farms never meet this criteria. On the other hand, good content certainly exists beyond the personal web. We might be largely disinterested in most of the microcorp.com, but if blog.microcorp.com has a post about porting freeski.exe to Microcorp's iZoon mp3 player, that could be kinda cool. On the other hand, CTO announcements about their latest data center box fail all three of the above tests.

It's murky:

• Personal sites are our primary focus and raison d'etre. They're mostly awesome, but can be painfully product-oriented or SEOed to the point of meaninglessness.
• Wikis are informative and authentic with light commercialization (associated more with the publishers/hosts than the contributors). Despite this, we've largely excluded wikis from our index since they're more about Q&A than web exploration.
• Most social media services are inaccessible from the open web, so they're out. Okay, that one wasn't murky.
• Journalism is authentic and informative in ways unmatched by the personal web (due to professionalism and access). We've indexed some media sources but are keeping our finger on the red button.
• Cottage and small business sites often provide standalone blogs that aren't laden with links to their product pages.

We're building the car as we drive it so at the moment we're more concerned about filter effectiveness than filter policy.

Okay you have my interest, what do I do?

If you don't have your own website, next time you read something interesting, paste it into Navicomputer and see if it gives you good reads.

If you run your own site, send your three favorite posts through Navicomputer. If the links look good, consider adding them to the footer of the original posts. Or create a standalone post about using Navicomputer to link the web. If the results are bad, ridicule us on your blog. Either way, it contributes to the health of the internet and will have Mark Facebook punching air.

What input can I provide?

Just paste in the plaintext of your post or page. Removing the window dressing like navigation and footers will slightly improve results quality. You can optionally populate the 'Keywords' field with space-delimited terms like 'golang or scuba diving'. URL fetch is coming once we have all the loose ends tied up.

How small or big can my input be?

There is a modest minimum input length, mainly because the algorithm is built to connect substantive content. If your post is "today I made breakfast", the matches could be almost anything.

We have an upper limit on input length so if your text gets truncated you've found it. If your HTML gets truncated it might not parse well, if your plaintext gets truncated the algorithm probably has enough to go on from what made it through.

My query results weren't great, what's wrong?

Good results typically require at least a few substantive paragraphs about specific topics. This can be supplemented but populating the keyword field.

There remains the possibility that we simply haven't seen much material on your subject matter. Blog and personal site content is skewed toward tech, many subjects exist almost entirely on social media platforms.

I got some links I wouldn't consider 'Outer Web'.

A report feature is in the queue. For the time being, please ignore any results that appear to have slipped past our filters. On the other hand, if your post matched a commercial link, does that say more about us or about your post? 🤔 (Definitely us.)

Speaking of undesirable links...

Yeah. At this point we don't have a content policy beyond the aforementioned authenticity/commercialism thing. That said, "ivermectin makes you fly" has negative informational value and we'll filter it when we find it.

You say this site is new and yet you have an FAQ, how can there be frequently asked questions?

Okay smartass, ever heard of a thing called convention?

Does the Outer Web link to me? I would like some exposure.

It might already! And if it doesn't, we'll be adding index requests soon. Just be sure your site is easy to index:

• It has a permissive robots file (or none at all if you live dangerously).
• It's not in a walled garden.
• It loads statically.

What happens if Outer Web goes offline? Will all my pages break?

Nope, not even a little. We provide static links that don't depend on us in any way. Of course, the site on the other end of a link might go offline.

Do you collect money from links?

Nope. There is no referral information in the urls we provide and we don't have promoted links. This site is entirely self-sponsored and not all that expensive tbqh.