November 30, 2019

1037 words 5 mins read

jstrieb/urlpages

jstrieb/urlpages

Create and view web pages stored entirely in the URL

repo name jstrieb/urlpages
repo link https://github.com/jstrieb/urlpages
homepage http://jstrieb.github.io/urlpages
language JavaScript
size (curr.) 1286 kB
stars (curr.) 1123
created 2019-06-30
license MIT License

URL Pages

jstrieb.github.io/urlpages

About

  • Create web pages in the simple, fast editor
  • Share code that others can edit and modify
  • Clone web pages with the bookmarklet (under active development)
  • “Publish” web pages instantaneously
  • Published links never stop working and ~cannot be taken down~ function as long as this site is trusted and extant
  • No dependencies
  • No signups
  • No tracking
  • No hosting
  • No cost
  • No commitment
  • Less than 200 total lines of clear, well-documented HTML, CSS, and JavaScript

Read the Hacker News Discussion here

How it works

As hinted by its name, URL Pages works by storing the entire contents of a web page in the URL.

Thus, as long as the URL exists, so does the page it points to. The rest of the URL Pages program is responsible for translating between web page code (HTML/CSS/JavaScript) and an “encoded” URL.

  • The main page takes encoded data from the URL, decodes it into regular web page format, and displays it to the user
  • The editor encodes user-created web page data as a link that can be shared
  • The bookmarklet takes a page that already exists and encodes it as a link that can be shared

When the main page is visited, the data is encoded in the URL using base 64 encoding via JavaScript’s atob and btoa functions in conjunction with its encodeURIComponent and decodeURIComponent functions. The encoded data is stored in the hash portion of the URL.

In the editor, data is similarly encoded, except that the HTML, CSS, and JavaScript portions are stored separately in one object that is converted to a JSON string before being base 64 encoded.

The obvious downside of URL Pages is that the links get very long very quickly. Luckily, some URL shorteners are able to accommodate fairly long URLs (shoutout to TinyUrl). In a strange way, this effectively means the link shortener is acting as the web host since it is responsible for storing the record of the web page’s data. For simple web pages (and even simple page hierarchies), URL Pages have proven reasonably easy and effective to use, however it quickly becomes infeasible to use for large sites or large embedded images.

Disclaimer

This just becomes a toy if I am the only one hosting a running version of this repository. If you believe it has real potential, clone it or fork your own version that addresses any non-fundamental problems you have with it, and host your own. The only way this actually becomes robust is if there is no single point of failure (i.e. my GitHub Pages)

Web pages in URLs are definitely not how things on the web were meant to be done, so don’t be surprised if trying to use URL Pages causes unexpected issues. For example, sharing these links may cause chat programs, email clients, and unsuspecting individuals to get confused, raise exceptions, or complain. Likewise, copy-pasting these links may take a long time, if it works at all. I’ve also noticed my browser running a little hotter while I’ve got 5MB links in the URL bar.

Furthermore, URL Pages is very much a proof of concept, and should not be relied upon for anything consequential.

Read the code and understand it before using so that you understand any associated risks. The codebase was written with readers in-mind. Since the codebase is intentionally short, it can be read and digested fairly quickly if you have prior experience with client-side web applications.

I originally conceived this as a simple, static CodePen clone, but I felt the “publishing” of pages as URLs was an interesting idea. So I decided to present that aspect of it front and center, even though it wasn’t really the point of the project at the beginning. About a year ago, I had a proof of concept version that I ended up using fairly frequently for sharing quick HTML/CSS/JavaScript experiments (never as a means of seriously publishing and sharing censorship-proof content). I found that if its use is limited to that case, it is actually very handy and robust!

Examples

The following examples were made and “published” using the provided code editor.

  • My personal website
    • Code in the code editor here
    • “Published” version here
  • Bookmarklet setup page
    • Code in the code editor here
    • “Published” version here
  • A page with embedded images (no external image host)
    • Code in the code editor here
    • “Published” page here

The following examples were cloned from existing pages using the bookmarklet.

  • My dad’s food blog here
  • The entire editor encoded in the URL here
  • This GitHub project page here
  • A cloned New York Times Article here

Bookmarklet

Currently, the bookmarklet is very much in-development (read: mostly doesn’t work). Feel free to try it anyway by visiting the link below and following the instructions, or pasting the code below into a bookmark:

  • Bookmarklet instruction page
  • javascript:window.open("http://jstrieb.github.io/urlpages/#" + btoa(encodeURIComponent(document.documentElement.outerHTML)), "_blank")

The bookmarklet enables some of the most interesting and promising opportunities for URL Pages. Namely: cloning pages for archival purposes, sharing restricted information to bypass censorship, bypassing paywalls, storing entire pages in bookmarks, etc.

Since its original creation, it has been forked many times. Please open an issue if you would like me to link back to a fork or mirror.

  • One particularly improved version is JSPen

Similar in some ways (though unrelated) to the following projects

TODO

  • Improve the bookmarklet – it’s mostly unusable as of right now
    • Fix relative vs absolute linking
    • Maybe try embedding images
    • Import all srced scripts directly
  • Test to make sure that everything actually works for other browsers, operating systems, and devices
  • Improve UI in general and editors beyond simple textarea (perhaps integrate Ace or CodeMirror)
  • Make the buttons better/more efficient (don’t update href on every key press)
  • Figure out and publish max URL sizes for various URL shorteners
  • Implement URL compression using Brotli for shorter URLs
  • Add option to “publish” pages using base65536 as suggested here
  • Allow users to sign/password protect published URLs
  • Upload examples of multi-page sites (tree hierarchy)
comments powered by Disqus