Dependency Graph Editor

---

A dependency graph says which actions include, imply, equal or contradict which other actions. It is what the reasoner reasons with, so changing the graph changes the conflicts it reports. The editor lets you keep graphs of your own, so a check can follow the rules you hold rather than only the curated ones.

To read the curated graph instead, open the Dependency Graph Viewer. How reasoning works walks through what a graph is used for.

What you can do with it

• Start from something. Copy a core graph and change only the statements you disagree with, begin with an empty table, or upload a Turtle file you already have. A file you download from a graph here can be uploaded again, and it gives you the same statements back.
• Write the statements in a table. One row is an action, one of the four relations, and a second action. Both actions are picked from the DALICC vocabulary, so every statement you write is one the reasoner can use. If a statement cannot hold, the editor says so before anything is saved.
• Keep every version. Every save is stored, nothing is overwritten and nothing is deleted, so a conclusion drawn from a graph last month can still be checked against the graph as it was. Publishing gives the graph a public address and a version number of its own.
• Share it. Give somebody else the right to change the statements, or only to read them, and pass the graph on to a new owner. Somebody without an account yet can be invited to it.
• Reason with it. Your graphs are offered in the License Composer, in the License Compatibility Checker and through the API, drafts included, so you can try a graph out before you publish it.
• Offer it to everybody. You can submit a graph that others should reason with for promotion to a core graph. Nothing is deleted and no address stops working: the statements are copied into a curated graph that points back at yours.

---