Name Last Update
includes Loading commit data...
layouts Loading commit data...
plugins Loading commit data...
root Loading commit data...
static Loading commit data...
styles Loading commit data...
.gitignore Loading commit data...
LICENSE Loading commit data... Loading commit data...
base-config.yml Loading commit data... Loading commit data...
freecomment.php Loading commit data...
new-site Loading commit data...

This is the design for my Web site,

The contents of this repository are not intended to be used directly as a Jekyll site; instead, a new-site script is provided that creates a new Jekyll directory tree and clones this repository into _template inside that directory. To build the site, you use a GNU Makefile which compiles the stylesheets and runs Jekyll with the correct configuration files. (Two configuration files are used: one in this repository, which contains some settings necessary for this system to work, and one in your site's directory, which you can customize.)

In particular, you will notice that your site tree will contain no _layouts or _plugins directories. These are provided for you in the _template directory, and the Makefile makes sure Jekyll uses the correct paths for them.

When the site is built, the output will initially be placed in a ._build directory, and once the process is finished, _site will be deleted and ._build will be moved into its place. This is to ensure that the entire new version of the site goes live at once as opposed to each page separately. However, any .git, .gitmodules, or .gitignore files or directories contained in _site will first be moved into ._build; this way, you can still make _site be a separate Git repository (e.g. for hosting on GitHub Pages). To preserve other files, use the _postbuild script.


  • Setup
  • Front-matter parameters
  • Tags
  • Filters


  1. Install the dependencies:

    fedora$ sudo dnf install ruby ruby-devel rubygems nodejs npm make
    ubuntu$ sudo apt-get install ruby ruby-dev rubygems nodejs npm make
    any-distro$ sudo npm -g install stylus
    any-distro$ sudo gem install jekyll stylus pygments.rb

    PHP 5 is an optional dependency, necessary only for the provided redirect mechanism (_static/redirect.php).

  2. Download and make it executable.

  3. Inspect new-site.

  4. Run new-site <directory>, where <directory> is the directory in which you want your site to reside.

  5. Edit _config.yml (Jekyll configuration), _config.styl (stylesheet variables), _postbuild, and _redirects as desired. You can also remove _postbuild and _redirects if you don't need them.

  6. Add some content!

  7. Change into your site's directory and run make.

Front-matter parameters

There's a lot of these.



The page's title.



The URL, relative to the page itself, of an icon to use as the page's favicon and in navigation elements.

You may also use a Material Design icon without having to provide the image yourself. To do so, use material:<icon name> as the value for the icon property. Multiple words in a single icon name may be separated with underscores, hyphens, or spaces. This uses the Material Icons webfont, which uses ligatures to allow one to just type out the name of the desired icon instead of having to hunt down an image or manually specify a codepoint.

There is currently no support for using Material icons as the page's favicon with the webfont method; this currently only affects navigation elements. Favicon support may be implemented later by dynamically generating them in JavaScript using the Canvas element.



The page's description.



The URL, relative to the page itself, to use for the value of the Facebook og:image meta tag. This will be converted to an absolute URL based on the value of site.full-url or site.url. Protocol-relative URLs will then be converted based on the value of site.default-scheme.


(string) (default: same as page.title)

The title to use in the HTML <title> element.


(boolean) (default: False)

If False, HTML and newlines will be stripped from the title as used in the HTML title element.


(string) (default: same as page.title)

The title to show in the <h1> element at the top of the page.

(boolean) (default: True)

Controls whether any copyright statement is shown in the page's footer.

(boolean) (default: False)

Controls whether the recent posts list is shown on pages other than the home and blog pages.



Extra content to be inserted at the end of the <head> tag, but before page.stylus and/or page.css.



A Stylus stylesheet to be inserted in a <style> tag at the end of the <head> tag, but before page.css.



A CSS stylesheet to be inserted in a <style> tag at the end of the <head> tag.



Raw text to be inserted before any generated HTML, even before the doctype.



Raw text to be inserted after all generated HTML.


(boolean) (default: same as site.use-absoulte-root or False)

Controls whether absolute paths should be used for the site root path and the page's directory path (see the root and dir tags). This only applies when the page is being rendered as a standalone page. When being included in another page (e.g. a blog post listing or feed), the other page's setting will be used instead.



Controls how the page is displayed in navigation elements (the top navigation bar, subpage listings, etc.).


(boolean) (default: False (True if no layout))

Causes the page to not be shown in navigation areas.

(boolean) (default: True (False if no layout))

Causes the page to be shown in navigation areas.


(string) (default: same as page.url)

Used for sorting navigation elements.


(string) (default: same as page.title)

Overrides the title used in navigation elements. Different kinds of navigation areas may also support other parameters that override this one.

(string) (defaults: page.nav.title or page.title)

The title to use in menu-style navigation elements. Prefer to use page.nav.title instead.


(string) (defaults: page.nav.title or page.title)

The title to use in subpage listings. Prefer to use page.nav.title instead.


(string) (default: same as page.description)

Text to be shown when hovering over the menu item.


The window target to be used in the navigation item's <a> element.



The URL of the page to highlight instead of the page itself in the navigation area. This URL's parents will also be highlighted.


(boolean) (default: False)

Whether the page referenced in page.nav.highlight will be treated as the "current" page.


This is only a partial list of Liquid tags available.


Inserts the site's root URL. This may be relative to the current page or absolute, depending on the value of page.use-absolute-root on the page actually being generated or site.use-absolute-root.


Inserts the path to the current page's directory. This may be relative to the current page or absolute, depending on the value of page.use-absolute-root on the page actually being generated or site.use-absolute-root.


Inserts a literal enquiry character (ASCII value 0x05). This exists due to how the root and dir tags are internally implemented.


Inserts the copyright year(s) that apply to the site. This is based on the current year at the time of rendering and the value of site.copyright-start. The output will be of the format <start-year>&ndash;<end-year>, or just <end-year> if both are the same.


A block tag that resolves all references to the site's root URL using the settings that apply to the page or layout in which rootify is used. All uses of root, dir, or enquiry MUST be contained inside a rootify block, either directly or indirectly. The included layouts (except for none) will take care of this for you.


Usage: {% indent <level> %}...{% endindent %}

A block tag that indents its contents by level space(s). level may be a literal integer or a variable that contains either an integer or something that can be converted to an integer.

The contents of <pre>...</pre> sequences will not be modified except to replace newlines contained within them with the XML entity &#x0a;. Unfortunately, this is done with a regular expression.



Usage: {{ ... | gsub: <regex>, <replacement>[, <options>] }}

Converts input to a string and runs Ruby's String#gsub on the result. All arguments are to be supplied as string literals.

options should only contain the characters i, m, or x. These are the same as the same options in Ruby's Regexp class.

You should only use this filter with trusted input and trusted arguments.