pic1

Hugo is an open-source static website generator. It has many themes and many layouts ready to choose from. Content in Hugo consists of posts that you add using markdown format, and the posts are arranged in a logical directory tree structure. See the Hugo site here: Hugo Website.

I put together 3 Hugo sites, using the same set of post files. Have a quick look:

3 questions you probably have after looking at that:

  1. We have a github account?
  2. Holy Cow, how’d you do that?
  3. How do you keep documents private?

We have a github account? 

Yeah, github accounts are free. Pick a username and go. For the free deal, all your repositories are public.

For $99/year, repositories can be made private.

I have the password, I’ll give it to you if you want to host your repository here as well.

How’d you do that? 

While Github is known as a source code repository, it is also something else:

If the repository you are storing ALSO happens to be a Static Website, Github can serve up pages from the site.

Chew on that a good long time


It's completely unprecendented to have your code repository
ALSO serve up web-pages that happen to be stored there.

The key concept is that checking in your code and pushing
the resulting git branch to the server is EQUIVALENT to publishing
your website.

It's a new paradigm, and it's really incredible.  It makes
Documentation Management incredibly easy!

With proper configuration of the github repository, and proper layout of the data you are storing there, github can serve up the pages.

With Hugo installed locally on my linux workstation, here’s the process I use for creating that website:

  • edit content, simple .txt file
  • Run Hugo from within a shell script. This generates the entire static site.
  • use git to check-in changes, and push to Server
  • browse to site, wait a bit for changes to propagate

How do you keep documents private? 

1st of all, pay the $99 fee. That means the repositories are private, but the website pages are not. The key is knowing where the publicly browsable files are, and not putting private documents there.

The solution to pub/priv relies on a logical arrangement of the source code tree.

See the repositories at:

Each of those repo’s have 2 content directories, which are the Source Material of the documents.

  • content_pub       publicly browsable content.
  • content_priv       private content, not to be displayed on github.

My technique for public/private then is to generate 2 different sites, that are completely seperate:

  • One meant for public viewing on github
  • The 2nd one stored on github, But must be copied somewhere else to be displayed

To do this requires executing Hugo twice while using the proper Source and Destination flags.

For the publicly visible:

  hugo -c "./content_pub" -d "./docs"

To remain private:

  hugo -c "./content_priv" -d "./docs_priv"

The 2 output directories, docs and docs_priv, are both:

Standalone, Static, Hugo-Based Websites

docs gets displayed by github as default. docs_priv will have to be copied somewhere else to be displayed.

A few more notes 

See the file x.sh. It’s the shell script that automates the running of Hugo.

My platform is linux, there’d be just a bit of work to do getting this to run on Windows. x.sh might have to be turned into a .bat file. Haven’t done that yet. Hugo is supposed to run fine on Windows, but again, something I haven’t done yet.

Hugo should also be supported on Mac.

I’d be glad to help anyone get this up and running for themselves. Also, it’s not too difficult to set up different themes from gohugo.io.

JR