You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently, each atom is mapped to a URL based on its unique ID, as well as a configurable namespace. When the data is published (in RDF), the URL becomes the dereferenceable IRI of the topic.
However, these URLs are not very friendly-looking. E.g.
http://fortytwo.net/things/qAnrxm
Going forward, let's append the topic title to the URL, and instruct the Web server to redirect to the bare URL. With the proposed longer IDs, we might have:
It's optional, but applications will default to sharing and linking to the expanded URLs as opposed to the bare URLs. When you visit a page (and see HTML generated from the Markdown), you will have the full URL in your browser's navigation bar.
How IDs, IRIs, URLs and pages will work:
every topic has an ID. In the new model, a topic is nothing more than the ID, e.g. qAnrxm0gmzccMsYz for the concept of Komodo dragon. This is the identifier of the topic in a SmSn context.
for every ID, there is an associated IRI, e.g. smsn:qAnrxm0gmzccMsYz (IRI scheme or namespace to be determined; the example assumes a "smsn" scheme). This is the identifier of the topic in a Semantic Web context.
topics are described by pages, which have URLs based on the topic ID and a title. E.g. http://fortytwo.net/things/qAnrxm0gmzccMsYz/Komodo_dragon. As I have suggested, the title is not essential to the URL, and other titles are equivalent for the purpose of retrieving the page. This way, the title is free to change over time, while the ID remains constant. Page URLs are the identifier of a topic in a Web browsing context.
pages are contained in namespaces, e.g. http://fortytwo.net/things, which correspond to Git repositories or other SmSn datasets.
datasets are sets of pages. A dataset can be published in more than one namespace, but you can't have more than one dataset per namespace, nor more than one page for a given topic in a given namespace.
Currently, each atom is mapped to a URL based on its unique ID, as well as a configurable namespace. When the data is published (in RDF), the URL becomes the dereferenceable IRI of the topic.
However, these URLs are not very friendly-looking. E.g.
Going forward, let's append the topic title to the URL, and instruct the Web server to redirect to the bare URL. With the proposed longer IDs, we might have:
This URL would map to the same topic:
The URL is human-friendly and Linked Data friendly.
The text was updated successfully, but these errors were encountered: