Mediawiki mirror/details

From Bjoern Hassler's website
Jump to: navigation, search
The "Mediawiki accessibility project"

General considerations - Mirroring - file size extension - email interface - W:Mediawiki OER export - [Edit]

This is a short writeup of a 'bespoke' mirror for wikieducator (would work the same for any mediawiki, such as wikipedia).

The original wikieducator is here: http://www.wikieducator.org/Main_Page

My local mirror is here: http://www.ict4e.net/mirror/wikieducator/index.php

The idea is that the mirror might be on the network boundary, rather than publicly on the internet.

1 Features[edit]

The mirror cache is filled automatically: Whatever pages are requested are then cached (but only updated if needed). If you ran a recursive wget on the whole mirror, it would cache the whole site, ready for offline use.

Really it does three things:

  • (1) It reduced bandwidth between LAN and the internet, as any cache would do.
  • (2) It works offline, i.e. when the internet connection goes, you can still browse whatever pages are cached (you can simulate that by using the 'go offline' button)
  • (3) As a side effect, the mirrored page is also much smaller than the original page, but a factor of 10 or so. So (as a side effect), the mirror also works as a site reformatter similar to loband (for that one particular site). Of course if you make the mirror look exactly like the original site, you'd need to put all the css back in etc.

2 Using the script on your laptop[edit]

Another use of the script would be to not install it on the LAN boundary, but to install it on your own computer. You could do that with a local 'LAMP' stack (such as WAMP or MAMP), which essentially means you can use the script as a local application, and you could thereby take wikieducator content with you.

3 Limitations[edit]

The script doesn't mirror images or attachments yet, because I've only spent a day or so on it so far, but it's easy to stick that in too. Also, it doesn't allow you to edit the original wiki 'offline' - that's a much harder problem. But the assumption is that more people want to read, so it makes reading really easy. The main bit of functionality compared to the original site that's missing the the site search.