Common Relations

Some relations are so commonly used, we suggest some standard names here:

Name Aliases Semantics (A .. relation .. B)

Name	Aliases	Semantics (A .. relation .. B)
`related`	`rel`, `is related`	Undirected link connecting A and B.
`same as`	`is same as`	A and B are referring to the same concept.
`is alias of`		Like `same as`, but with a clear main concept.<br/>This is like a sym-link.
`links to`	`link`, `see also`	Directed, untyped link from A to B. Default link type when four dots (`....`) are used.
`has tag`	`tag`	A has the tag B. Tags are not transitive.
`has type`	`type`, `is a`	A has the type B. Like rdf:type. Types are inherited via `has subtype`.
`has subtype`	`subtype`	The type A has a more specialised type B. Transitive relation. If it forms a cycle, all participants of the cycle are considered to be the same entity (`same as`).
`prefix`	–	When parsing a name like `A:foo` (starting with prefix an colon) the `A:` should be replaced by `B`. This is only needed for RDF processing. This processing is just a recommended way to interpret a ddot.it triple base. Not needed in non-RDF contexts.
`has content`	`content`	A has content B. In this case, B is usually a longer string. Maybe including line breaks (`<br>`). Maybe stated via https://ddot.it/block.

related

rel,
is related

Undirected link connecting A and B.

same as

is same as

A and B are referring to the same concept.

is alias of

Like same as, but with a clear main concept.<br/>This is like a sym-link.

links to

link,
see also

Directed, untyped link from A to B.
Default link type when four dots (....) are used.

has tag

tag

A has the tag B.
Tags are not transitive.

has type

type,
is a

A has the type B. Like rdf:type.
Types are inherited via has subtype.

has subtype

subtype

The type A has a more specialised type B.
Transitive relation.
If it forms a cycle, all participants of the cycle are considered to be the same entity (same as).

prefix

–

When parsing a name like A:foo (starting with prefix an colon) the A: should be replaced by B. This is only needed for RDF processing. This processing is just a recommended way to interpret a ddot.it triple base. Not needed in non-RDF contexts.

has content

content

A has content B. In this case, B is usually a longer string. Maybe including line breaks (<br>). Maybe stated via https://ddot.it/block.

Binaries: Images, Videos, Audio, 3D, …

How should a text-based format like ddot.it handle binaries? Looking at HTML, which is also a text-based format and the implied model of Markdown and AsciiDoc, we have several options:

Base64-encoded data: URIs
Links to local (relative links)
Links to remote (absolute links) files

Ddot.it can support all of these. However, data: URIs (and other inline approaches) don’t fit the lightweight human touch of ddot.it well. So we suggest using links, ideally relative one. That way, ddot.it files or files with ddot.it tripels int them can be packed together with the linked resources in a single zip archive.

Commands as Relations

The relation of a triple can also be a command. A parser lexes it as a command wherever it appears, but what a command means is decided downstream — an unknown or retired name simply keeps its text as the relation value.

label used to be such a command (a ..!!label.. Alice).
It is now an ordinary relation (a ..label.. Alice), defined in the Vocabulary Specification.