This tracks website development for Bedrock Linux
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.

Bedrock Linux Website

This is a repo of the Bedrock Linux website.

Most of it is in plain markdown, with a few minor tweaks. If you would like to submit a change to the website (such as a typo fix, a work around for a known bugs etc), it should be pretty easy to figure out what is going on and make such changes.

To generate the website from the markdown source, simply run the script. The only dependencies in addition to UNIX utilities (sh, awk, sed, grep, find, etc) is markdown and gnu awk ("gawk"). I might later drop the gnu-isms from awk if there is interest. If you don't have markdown in your repository (apparently Gentoo lacks it?), it's just a perl script you can download from here.

If you're interested in digging deeper, note that the website generation code was not necessarily designed to be portable to other projects. The atom.xml for example is hardcoded to refer to Bedrock Linux. If you want to press on anyways, the relevant information follows:

  • First line of a .md file should be Title: <title>, where <title> is what you would like the website title to be.
  • Second line of a .md file should be Nav: <nav>, where <nav> directs to the file which contains the nav bar that will be used. This will also be treated as markdown; however, all unordered lists will be given the id="nav"
  • To add an id, class, etc section within an opening tag, start the very beginning of the relevant section with a {...}, where the ... will be placed accordingly. Note the space after the closing brace. For example, ## {id="website_overhaul"} Website Overhaul will turn into <h2 id="website_overhaul">Website Overhaul</h2>
  • Since <ul> and <ol> are both immediately followed by another tag, if you would like to place an id, class, etc inside <ul>/<ol>, make the first list item just {...} (note lack of space at the end). e.g.: - {class="rcmd"}
  • Something within ~(...~) will be converted to <code class="changethis">.
  • Something within ~{...~} will be converted to <code class="keyword">.
  • Markdown tends to nest <code> directly inside of <pre>. The inner <code> tags are removed from this part.
  • The file header1 in the root of the markdown directory contains the beginning chunk of all of the websites. Baring a few things described below, this will be the same for every page.
    • The string TITLEGOESHERE in header1 will be replaced by the title set by Title: <title>
    • The string RELATIVEPATH will be replaced with the path to the root directory, ie, a bunch of ../../..'s as needed. This can be used to ensure everything is relative, which is useful in development.
  • Immediately following header1's content will be the nav as set in Nav: <nav>
  • Following the nav is header2's content.
  • Following header2's content is the actual markdown generated html from the markdown code.
  • Following the markdown code is the content of footer
  • Any files that don't end in .md, don't end in .nav, and aren't header1 header2 or footer are copied directly into the output section without change, as are directories. This includes .html files.
  • An atom.xml feed is generated from parsing news.html (which is generated from in the root of website. It expects news.html to be set up in a very specific manner:
    • Each item starts with <h2 id="X">TITLE</h2>, where X is used to get the hyperlink ( and TITLE indicates the title of the news item. This should be followed by a blank line.
    • Next should be an ISO8601 date without any time information, such as "1900-01-01", within <small> and <p> tags. This, too, should be followed by a blank line.
    • The next line(s) should be wrapped in a <p> tag. The first paragraph will be dumped into the summary; after that everything is ignored until the next title/hyperlink line.