Using dat-gateway to browse the web, which depends on hyperdrive-http, we've discovered a couple of path resolutions that websites expect servers to practice that hyperdrive-http does not. For example, see http://gateway.mauve.moe:3000/pfrazee.hashbase.io/ and click on a link on the first link and you'll end up at http://82kzddgmfbk9bf5wzx1jyu2cfeujj7n37718r5unh5pdxe0bubwg.gateway.mauve.moe:3000/blog/universal-publishing-and-the-webs-black-boxes

Now if you add HTML to the end, you'll see it does resolve. http://82kzddgmfbk9bf5wzx1jyu2cfeujj7n37718r5unh5pdxe0bubwg.gateway.mauve.moe:3000/blog/universal-publishing-and-the-webs-black-boxes.html

The other example is if you have a file at /foo/index.htmlt, if you try to go to /foo/ it will not give you /foo/index.html.

• What we accept with key
• Changes thing
• single archive URLs

You can put an archive at the root with the getArchive function by calling back with that archive if there is no key.

This is super helpful if you want to have one archive as the main archive at root. So we should document it! 😄

Also add tests.

For example, adding a few functions: .get(key).pipe(res), .file(key, filename).pipe(res)

var hyperdrivehttp = require('hyperdrive-http')
var express = require('express')

var router = express()
var drive = hyperdrivehttp(getArchive)

router.get('/hi/:key', function (req, res) {
  drive.get(req.params.key).pipe(res)
})

#19

Version 7.14.0 of hyperdrive just got published.

This version is covered by your current version range and after updating it in your project the build failed.

As hyperdrive is “only” a devDependency of this project it might not break production or downstream projects, but “only” your build or test tools – preventing new deploys or publishes.

I recommend you give this issue a high priority. I’m sure you can resolve this 💪

Your Greenkeeper Bot 🌴

be super cool to use https://github.com/mafintosh/directory-index-html

Some sites, like Fritter rely on a fallback page in order to have pretty URLs that don't correspond to files in the dat (Also for custom 404 messages).

Would a PR for supporting this be welcome? Might also tackle #43 while I'm at it.

This error should have a status code and message from callback, not just the error message. Need to document how errors should be handled in getArchive function.

hi,

I run node example.js and get:

`sodium.crypto_kdf_derive_from_key(seed, 1, context, secretKey)

Error: "key" must be crypto_kdf_KEYBYTES bytes long`