a rearchitected implementation of foldermark
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
10_File Types
20_Audio and Video
30_HTML Fragments
40_Downloads
components
css
fm3
images
lib
10_weasel-icon-512.png
20_test.markdown
README.md
favicon.gif
favicon.ico
favicon.png
index.html

README.md

foldermark-3

A rearchitected implementation of foldermark, a minimalist CMS that turns a directory structure of files in common formats together with a simple naming convention into a nicely structured website with readable urls.

In a nut

  • a foldermark site comprises a hierarchy of folders
  • each folder corresponds to a web page
  • the root folder is the home page
  • folder-names become page titles and urls (but you can order folders by prepending a nn_ which is omitted from the title/url)
  • contents of folders are assembled in order to become pages (and you can order them by prepending nn_ which is omitted from the title/url)
  • most of the file formats you care about are supported, including HTML (fragments), jpeg, gif, video, audio.
  • markdown files are supported (and encouraged)

History

The first version of foldermark created web-pages on the server-side using PHP and used a small amount of client-side javascript for interactivity.

The second version of foldermark created web-page data in json format on the server-side using nodejs and used a lot more client-side javascript to assemble the data into pages on the client.

This version uses a dumb PHP back-end to deliver indexes and all the "smarts" are in the client, which is built using b8r. The advantages are numerous -- the PHP implements a very simple protocol that can be re-implemented in other server-side languages with minimal effort, the indexes are highly cacheable (and could be statically generated).