One of the biggest problems (IMHO) with OER is discovery of existing resources.

Domain Name System

DNS SRV resource records are used to discover endpoints for various services. Described in: RFC6335

Popular examples are SIP, LDAP, Minecraft and XMPP. This mechanism would mostly enable crawlers to discover them, since most users are not fluent in DNS. It is also unclear how the service is structured, if there where such an endpoint. As of today, several parties develop and maintain their own repositories.

Embedding Metadata in Websites

Enables Webcrawlers and their associated search engine providers to reason about website content. There is also the effort of the LRMI to provide schemes for educational courses. With version 3.2, there is also schemes for educational content such as Course (or at least in pending, whatever that means).

RDFa has several limitations and it’s use is not exactly encouraged.

JSON-Linked Data (JSON-LD) usage is encouraged by most search engine providers. My problem with JSON-LD is, that it must be embedded by an inline script tad, practically duplicating most information already provided in the website.

Also, see: Most of that information is five years old, with most mentioned projects already disbanded. Which, sadly, seems to be the case for all projects regarding discovery and metadata (e.g. MozCC and it’s successor OpenAttribute).


OpenBadges might help spread the existence of LMS and associated content, as they convey information about where they came from. Expected impact seems marginal, though.

Seems to be the only project still relevant in Mozilla’s educational activities. At least their user community seems to be active in recent days.

Standard development was moved to IMS Global, which will host an event on 2017-08-17 at the University of Michigan. Current draft is v2.0 and will be discussed there.

Content Sharing

General Media Metadata

XMP seems to be the de-facto standard for attaching metadata to various Multimedia Content

Moodle already has support for files in external repositories, although citations for (CC) licensed media leaves all the work to users. Example: Merlot Integration is integradet by default, as well as Wikimedia

Dublin Core

Learning Object Metadata

RDF/XML (IEEE 1484.12.3 and IEEE 1484.12.4 respectively)

Sharable Content Object Reference Model

Interoperability format between LMS, sends calls through client side JavaScript, insecure by design.

xAPI (aka. SCORM v2)

Improves on many limitations of SCORM, like cheating participants, but still seems to be very complex. Centered around a key-value storage (Learning Record Store), aggregating Statements (Actor Verb Object) via RESTful services ( Communication is transfered in JSON.

Still unsure about xAPI is meant to deliver actual content or is just there for reporting.


it seems that, like most broad standards, both SCORM and xAPI are difficult to use and hard to implement. xAPI states that Tapestry is one of the great and fruitful projects that is using xAPI. As many projects that implement SCORM, Tapestry already terminated their services.

Although the containers improve sharing content seamlessly, few people seem to actually use thr format for their content. This is also reflected in the Moodle FAQ for SCORM.

Used at:

MIT crosslinks


Librarians and Libraries

Since librarians have most of these problems solved already, there are severel things that I learned when I spoke with them.

These are just keywords, that need expansion.

Librarian Lingo

OPAC: Online Publich Access Catalogue. Letztendlich die öffentliche Schnittstelle über die Bibliotheksinformationen abgerufen werden können. List of next gen library catalogs

RDA: Resource Description and Access.

kein plan.


Globale, zentrale Bestandsliste aller teilnehmenden Bibliotheken. Bibliothek der Uni Ulm gibt ihre Bestandsliste an die Landeszentrale (Name vergessen), diese wiederum trägt dann bei Worldcat ein.

Verschlagwortung und Kategorisierung

Bibliographic Ontology Specification

Gemeinsame Normdatei

Liste der fachlichen Nachschlagewerke (299 Seiten) aus: Abeitshilfen zur GND


Normalerweise im MARC21 Binärformat, aber auch als MARC21-XML verfügbar. DNB stellt auch Testdaten zur Verfügung. MARC21 Standard Documentation: und

Es existieren Python2/3 Bibliotheken zum Arbeiten mit dem Format, diese weisen darauf hin, dass MARC21 ein Format ist, das nicht heuteigen Standards entspricht.

Beispielsweise ist ein Kritikpunkt, dass MARC21-records alle Informationen über ein Objekt enthalten, was dazu führt, dass die Informationen redundant in den Records verwaltet werden. Das Wiki, in dem die Kritik gesammelt wurde (letzter sinnvoller Beitrag ist vom 2011-11-17) hat einige Verweise zu den Designproblemen von MARC21. Auch die Anstrengungen der futurelib scheinen im Sand verlaufen zu sein.

Meines Erachtens ist, aus Sicht der Archivierung, die Redundanz hilfreich: so können die Records alleinstehend verwaltet werden, ohne Hinzunahme zusätzlicher Referenzen. Nachteilig ist bei dieser Form der Datenhaltung ggf. die fehlende Eindeutigkeit der Autoren, diese Datenstrukturen müssen zur Laufzeit reproduziert werden (keine Aussage, sondern eine Einschätzung).

Citation and Identifiers

Persistent Identifier (PID)

mix between ID and location. Several different standards exist

Document Object Identifier


Most research and resources should be accessible through a DOI. All documents with an associated DOI must be accessible through e.g. (where 10.1126/science.169.3946.635 is the DOI). For more information about DOI Content Types:

Built upon the Handle System Infrastructure, EIDR is using the DOI System, also.

From crossref:

the South Park movie , “Bigger, Longer & Uncut” has a DOI: a) So does the pornographic movie, “Young Sex Crazed Nurses”: b) And the following DOI points to a fake article on a “Google-Based Alien Detector”: c)


Open Researcher and Contributor ID

Uniform Resource Name

Archival Resource Key

DOI vs. ARK:

name to thing

Persistent Uniform Resource Locator

Object Identifier

Code and Data

How to create citeable github releases, through Zenodo: a service funded by CERN and horizon2020. Zenodo also provides citeable data storage up to 50 GB per file. When selecting a research data hoster the regisrty of research data repositories can be used to discover an appropriate provider.

Linked Data

The Web is not a glorified USB Stick