a zip blog

Top-level Files of trunk
Login

Top-level Files of trunk

All File Ages Tree-View

Files in the top-level directory from the latest check-in of branch trunk

zb: a zip blog

This is my approach at creating a self-contained website. The whole site will be embedded as a Zip archive right in the server binary file, so it is probably the most portable website you'll ever have.

One of zb's unique features it that it will render HTML, Markdown and Org-Mode files, so if you prefer to write .org files instead of HTML, zb has you covered.

Motivation

I first saw the concept of single binary websites in Ted Unangst's blog where he complained that Go couldn't do this just as easily, because of its broken support for embedded Zip files. As I play with both Go and Rust every now and then, I thought "how hard could it be to do it in Rust?".

Quite surprisingly (for me), it wasn't really hard at all.

How do I use this?

You'll need a Zip file, containing your website. For the default start page, zb will search for one of the following files in the root path:

If you prefer to use a different file name, you can pass --defaultpage name.ext when starting the application.

Everything else in that Zip file can be any file of your choice. Relative references between them will work. (If they don't, please file a bug.)

First steps...

> fossil clone https://code.rosaelefanten.org/zb
> cd zb

> git clone https://github.com/dertuxmalwieder/zb
> cd zb

> cargo build --release

On Unix and Unix-like systems

Now, concatenate your zb binary with your Zip file. Assuming your Zip file is named index.zip:

% cat ./target/release/zb index.zip > ./zb

On Windows

Now, concatenate your zb.exe with your Zip file. Assuming your Zip file is named index.zip:

# PowerShell
> cmd /c copy /b .\target\release\zb.exe+.\index.zip .\zb.exe

# cmd
> copy /b .\target\release\zb.exe+.\index.zip .\zb.exe

Once done ...

Run the concatenated zb binary and your website will be delivered from port 8000. (You can change that: zb --port 8081 would make it run on port 8081 instead.) From now on, every time you want to update your website, just create a new Zip file and repeat the concatenation. (It should be easy to automatize that task.)

Automatic converted HTML routing

It is important to note that the file extensions for .org, .md and HTML files are optional, so the file stuff/demofile.md can be reached over localhost:8000/stuff/demofile as well.

The search order for all files is:

  1. /[path]
  2. /[path].md
  3. /[path].org
  4. /[path].htm
  5. /[path].html

How to contribute code

  1. Read and agree to the Code of Merit.
  2. Implicitly agree to the LICENSE. Nobody reads those. I don't either.
  3. Find out if anyone has filed a GitHub Issue or even sent a Pull Request yet. Act accordingly.
  4. Send me a patch, either via e-mail (git at tuxproject dot de) or as a GitHub Pull Request. Note that GitHub only provides a mirror, so you'd double my work if you choose the latter. :-)

If you do that well (and regularly) enough, I'll probably grant you commit access to the upstream Fossil repository.

Donations

Writing this software and keeping it available is eating some of the time which most people would spend with their friends. Naturally, I absolutely accept financial compensation.

Thank you.

Contact