Offline Versions

<- Back to OER Interoperability Scenarios

Offline Versions
The development of OERs (OCWs in particular) are based on the assumption that users have access to the Web, and in particular have broadband (or faster) access. Access to broadband access is not universal worldwide, quite a few areas of the world (especially in developing countries) do not have constant or near-constant access to the Web. The initial solution has been to create a static version of the website by direct copy or using a web crawler and send the content on hard disk or CD/DVD.

Examples:
 * MIT OCW has been working with a number of organizations to provide static snapshots of their site on hard drives to create local mirror sites around the world (e.g., [|iCampus OpenAfrica], [|MIT OCW in Ethiopia])
 * MIT OCW uses rsync (open source Unix protocol) to keep some of the mirror sites updated with new content via the Internet (content is normally batched)
 * OpenLearn has distributed DVD with standard's compliant packages of our content to be installed in various places, one in Africa, on in the UK Prisons (no internet access permitted for prisoners) and oddly for Scotland.
 * OpenLearn offers a "printable version" of content as a print CSS styled HTML document which is proving VERY popular.
 * USU OCW has distributed DVD and zip file versions of its site in Asia and Africa.
 * [FHSST] / Siyavula (no site yet - Shuttleworth Foundation project) - focus on being able to produce properly typeset material for printing, royalty-free material is more than an order of magnitude more cost-effective than current publishers alternative - allows cheap bulk printing and distribution to poor schools who have no computers at all
 * A WikiEducator / WikiMedia collaboration to produce a customized, typeset pdf document from a user selected collection of content articles in any Mediawiki installation. This is pretty sophisticated because it will fetch the high resolution images for printing. Phase 2 of this development will focus on producing an export in OpenDocument format to enable users to localise materials offline using any Word processor that supports OpenDoc formats. (I can demo a early prototype if you like. We're not advertising this too widely given the obvious issues around server loads on Wikipedia and challenges to the publishing industry ).
 * Open Office V3 has a Mediawiki export feature, which does a reasonable job of formatting for Mediawiki including basic tables. However images are not supported. This is proving valuable for WikiEducator users who do not have 24/7 connectivity.
 * Connexions collection printing tools produce PDF's from collections of modules (short or single topic educational entries), using LaTeX as the typesetting engine. We have also built an XSL-FO pipeline for transforming XML to PDF, but at the time the typesetting for Math was not as good as LaTeX. A beta version of the collection printing tools will be available in January.
 * OER Commons provided static UI, metadata, and K-12 resources from Connexions and others as a sample of content for use on OLPC laptop or local server or online.

Challenges:
 * Initial creation of, typically static, versions of content, the desire to retrieve a portion of the content, and keeping the content updated with newly published course materials. Opportunities to eliminate barrier to media-rich haves contrasted with stripped-down (text-only) have-nots.
 * (For print) - Allowing for proper typesetting can make both user interface and code more complicated
 * (For print) - Finding print solutions that support OERs using peer-collaboration production models
 * (For print) - Using LaTeX is complicated when trying to formatting PDF without author intervention.
 * (For print) - Again, using LaTeX, support for different character sets and flows requires special packages and LaTeX packages eventually interfere with each other. Finding local language experts to help develop would be desirable.
 * (For media) - Finding media solutions for low-bandwidth contexts (or, for that matter, visual learners/communities)
 * (For media) - Finding just a few "hardcopy" solutions that all OERs can support and use.
 * (From OpenLearn) Another challenge is in how to remove "old" materials e.g. when its gone out of date or because a rights issue has come up and something has to be pulled.  We're also concerned about the loss of tools around the content e.g. forums to help forum learner communities. Or if tools are provided the challenge is in synchronising activity for example merging distributed forum posts back to the centre when internet connections are available.  One of my colleagues is working on an offline Moodle which will do this sort of synchronising of content and user activity.  There seems to be a lot of buzz in the web development community about online/offline web apps - Adobe AIR, GoogleGears... I guess they would be enablers too.

Enablers:
 * Ability to export the entire site to run locally
 * The key enabler is XML - if the content starts out that way, it can be repurposed easily for multiple presentation formats. As we provide our content in RSS as well I'm seeing people make use of tools like XFruits http://www.xfruits.com/ to repurpose into their own preferred format.
 * (For print) LaTeX run server-side provides a free powerful typesetting tool
 * (For print) XSL-FO is a promising output formatting language with decent tools.
 * Mediawiki ==> pdf technology - should be available by the end of the year.
 * Dokuwiki ==> html & dokuwiki ==>pdf plugin's make static copies of the wiki content
 * ocw ==> wiki grammar (ocwinmotion now uses IMS Content Packages and RSS to retrieve content from MIT OCW and Educommons powered websites)
 * IMS Content Packaging
 * See eXe - a tool for offline learning design and content packaging - work ongoing to integrate with (e.g.) WikiEducator (and any SCORM/IMS compliant LMS or LO repository).
 * Talk to Derek Keats about "active dynamic mirroring" integrated in the Chisimba framework (AVOIR).
 * Possibly thinking about a subscription service so content providers do not have to update all the content, just that which is desired (also relates to Low Bandwidth/Mobile)

Scenario from Philipp Schmidt
We conducted a search for high-quality OER for secondary schools in South Africa (focusing on Maths and Science, but not excluding other subjects) and came up with a long list of resources. The objective is to make them available on an intranet server, because the schools we are targeting are connected to each other on a wireless LAN/WAN, but do not have Internet access. The solution we would like to build (but I am not sure we'll be able to overcome all the difficulties) looks like this: We store a list of all content in a wiki, where users (teachers and students) can comment on the materials. The wiki serves as a clearing house for meta information to enable users to find the content that would be useful for them. From the wiki we link to the individual resources which are stored as static html on the local server, or cached in a proxy server. The idea was to store as much of the materials as possible in a way that could easily be burned on a CD ROM for offline distribution. However, getting the materials into formats that could be hosted locally and put on a CD is proving extremely difficult. Copying and pasting the actual page content is a tremendous amount of work, and we are not sure which pages can be properly cached, and which pages could be downloaded (using the browsers save as webarchive function). Philipp Schmidt University of the Western Cape, South Africa United Nations University MERIT