The European Legislation Identifier initiative (ELI) aims at bringing legislation into the global Web of data, to facilitate the access, sharing and interconnection of legal information. It proposes the creation of URI identifiers for legislation based on common components and the description of their metadata based on an ontology relying on FRBRoo; the ELI ontology includes in particular the description of the FRBR levels of abstraction, the needed date properties to describe legislation and links to relate legislative acts. Legislation metadata is thus viewed as a global graph of interconnected entities. While ELI tries to lower the entry barrier for legal publishers to disseminate structured metadata and currently counts 13 implementations, it is also facing challenges to progress towards its full potential: data quality, description of ELI datasets, alignment of thematic vocabularies or granular description of the text subdivisions. ELI has the potential to facilitate access to legal information by enabling unambiguous legal citations mark-up, giving legislation more visibility in major web search engines, describing early legislation drafts or facilitating the exchange of data between legal information systems. ELI is tightly connected to novel legal information system architectures, based on legal knowledge graphs; this style of architecture encourages legal publishers to move from a document-centric perspective towards a data-centric perspective, as exemplified by the Casemates in Luxembourg and the Cellar at the Office of Publications of the European Union.