Scribd for UKOER resources

Overview: Scribd allows the sharing of documents, including short reports, posters, presentation slides, magazines, sheet music or full-length books. Typical use is for static text-and-image documents but spreadsheets are also handled. Metaphors used and functionality tend to be bibliographic: user accounts are described as "publishers", collections as "shelves", the metadata includes publishers, editors, ISBN etc.

Documents are viewable on the Scribd website and embeddable in webpages elsewhere in the iPaper format (which requires a Flash reader). Files on Scribd can be distributed either freely and openly, for fee, or privately. The service has been used by publishers to sell eBooks, but also (allegedly) for unauthorized distribution of copyright work.

Formats and Standards
For upload, Scribd supports Microsoft Office and Open Office formats for text documents (.doc, .docx, .odt, .sxw), presentations (.ppt, .pps, .pptx, .odp .sxi) and spreadsheets (..xls, .xslx, .ods, .sxc), as well as OpenDocument, Adobe PDF (so long as it is not "secured or password protected), PostScript, Rich Text Format and plain text.

On upload the files are converted to Scribd's own iPaper format, which can be displayed through Flash on the Scribd website or embedded into pages on blogs or other sites. Documents are also converted into PDF and plain text, and can be downloaded in the these or the original format.

API
Scribd has a REST-based API for uploading, editing the document metadata, deleting and searching documents. Data is returned in XML using a Scribd-specific schema. Upload maybe through multi-part POST of the document with the request or by reference to a URL for the document. There are client libraries in Ruby, PHP, C#/.Net, ActionScript 3.0, and Python.

Applications interacting with Scribd via the API must have an API key which identifies the developer. If there is no authentication in the application then the actions take place on the application developer's own account. There are API methods that allow users of an application to authenticate by remote login on Scribd or to share user credentials between the application and Scribd. Session keys are used to identify authenticated users. Authentication does not use OAuth.

There is also a Javascript API for the Flash iPaper viewer, allowing control of the player (full-screen or windowed, viewing pages in book mode, tiled, as slides) and of loading, navigating and enlarging documents.

Feeds
RSS2.0 feeds are auto-discoverable from most pages. Normally (e.g. from a user's profile page or from a document's canonical page) this will give a feed relating to scribd as a whole, however from a user's documents page the RSS feed gives a list of that user's documents.

From a document's canonical page there are auto-discoverable links to XML and JSON oEmbed information.

Other metadata
Metadata does not seem to be extracted from files (checked with PDF XMP).

As one might expect given Scribd's targeting of publishers as a market, there is fairly extensive provision (at least compared to other web2.0 services) for bibliographic metadata including ISBN, Author, Publisher, Editor, Series, Dewey Decimal class, edition/issue/version/part.

Collections and Grouping
Scribd partners may create a "shelf" to group similar works together, which might work well as a means of an OER project grouping resources by course.

Users may form "Groups" to discuss and share documents within Scribd, which might be useful in managing a distributed OER project or for studying OER resources.

Content export & embedding
Documents are displayed on Scribd, and can be embedded in pages on other sites, using Scribd's own iPaper format which is displayed using Flash. HTML and wordpress.com embed code snippets are provided to facilitate embedding in other sites.

Documents can be downloaded in their original format and in the PDF or plain text versions generated automatically by Scribd.

Visibility on Search Engines
Scribd documents are findable on search engines, they do seem to index phrases from within the text, however the visibility is no higher than for any other website.

Usage stats for resources
The number of "reads" the average user rating and comments are displayed on the document's page. This information does not seem to be available by RSS or API.

UKOER
Are any UKOER projects using the service?

Others

 * Project Gutenberg
 * Used by MIT Press (not OCW per se) for some free resources e.g. Opening Up Education

Notes and comments
Published resources can be updated: different versions can be uploaded to the same canonical page without affecting the usage data.