This is the final installment of the three-part blog series on identifiers. In the first part, we examined why identifiers are important in the professional media creation universe. In the second part (link), we discussed what well-designed identifiers look like and how the Ontology for Media Creation (OMC) leverages such identifiers for identifying and linking production information.
In this part, we present actionable practices that asset management systems, workflow and data platform designers as well as OMC data consumers should adopt to advance toward the 2030 Vision.
OMC Data Creators
The following general best practices apply to any participant responsible for creating OMC data. Depending on whether you operate an asset management system, a workflow system, a data platform, or a database-driven application, additional specific practices may also apply and are discussed in later sections in this article.
- Ensure Identifiers Exist: Every OMC entity must have at least one OMC identifier.
- Reuse Existing Identifiers: If the entity already has an OMC identifier, reuse that identifier if possible. If the entity already has an identifier outside of OMC, use that identifier along with a meaningful scope instead of creating a new identifier.
- Preserve Identifiers: Do not discard or lose any identifier assigned to an entity by any participant. Other participants might reference that identifier across different workflows, regardless of the origin of those identifiers.
- Create New Identifiers When Necessary: If no identifier exists or if organizational policies, operational requirements, or use cases prevent using an existing identifier, create a new one by following these guidelines:
- Determine the Appropriate Scope: Identify the most suitable scope for the entity so that OMC data consumers clearly understand the domain in which the identifier is used. Refer to Part 2 of this blog series for guidance on scope.
- Ensure Uniqueness and Persistence: Create a unique and persistent identifier as detailed in Part 2 of this blog series. Apply the principle of appropriate granularity as discussed in Part 2.
- Plan for Entity Access: Decide on the best method to enable access to the entity or its associated information. If identifier resolution systems are available, register the identifier with necessary information about the entity. Otherwise, ensure that all created identifiers and their corresponding entities can be listed, as different participants may request the associated entity using its identifier at various stages of production.
- Follow Character Set and Encoding Conventions: Adhere to the character sets and encoding rules described in Part 2 when generating identifier values and scopes.
- Validate Your Data: Use the OMC Validator to verify that your OMC-JSON data and identifiers meet all required practices.
OMC Data Consumers
Participants interact with OMC data at various stages of production, and the following guidelines apply when processing identifiers:
- Use Canonical Parsers: When extracting identifiers from OMC-JSON or OMC-RDF, use a canonical JSON or RDF parser to ensure special characters are correctly decoded. If identifiers are provided in other formats (for example, as part of a URL), employ the appropriate parser – such as a URL parser – to decode them.
- Preserve Identifiers: Once identifiers from other participants are properly decoded, do not alter them. This requires flexibility in scope, character set, length, encoding, and the number of identifiers received for each entity from expected OMC data sources. Furthermore, when sharing consumed data with downstream participants, do not discard or exclude any identifiers.
- Use Reliable Comparison Methods: Avoid visually comparing identifiers since some Unicode characters may appear similar but are distinct to computers. Always rely on string comparison utilities that support Unicode and compare identifier values separately from their scopes.
Asset Management Systems
Asset management systems, regardless of the production stage they serve, manage assets along with associated metadata. Practices stated below should be followed:
- Identifier Creation for New Assets: When storing new assets, create identifiers as described in the OMC Data Creators section.
- Preserve Identifiers: Sometimes the workflows that create assets will also store metadata about those assets in asset management systems. Asset management systems should follow the OMC Data Consumers section’s recommendations when processing OMC data and must store all additional identifiers provided with the entities in that data.
- Support Multiple Formats of Identifiers: Given that the policies and technical requirements at data origins may vary, asset management systems must be designed to handle multiple OMC identifiers per entity, even if these identifiers differ in length or character encoding or scope. If an identifier’s character set is unrecognized or unmanageable (which should not occur if all best practices are followed by the identifier creator), the asset management system should store it in base64 or hex and later return it in its original form upon decoding.
- Facilitate Data Retrieval: The system should enable asset and metadata retrieval based on OMC identifier. If the system generates its own identifier when an asset is stored, retrieval based on that identifier is sufficient; there is no obligation (although strongly encouraged) to support retrieval via identifiers created externally.
Workflows and Workflow Management
Workflows in media production can vary from conceptual development to pre-production, dailies, editorial, and post processes. Workflows consistently involve participants completing tasks that consume and produce assets. Since participants, tasks, and assets are OMC entities, they must all have assigned identifiers. The following practices should be adopted by workflow designers and workflow management solution architects.
- Identifier Assignment Across Entities: All relevant entities within a workflow – including participants, tasks, assets, and entities such as work units – should be assigned an OMC identifier.
- Preserve External and Internal Identifiers: Workflows may receive identifiers from external sources (e.g., the infrastructure supplying a user participant’s identifier) or generate them internally (e.g., for new assets produced during workflow execution). Regardless of the source, all identifiers must be preserved and included in the workflow’s output so they can proceed downstream where other workflow processes may need them.
- Return Valid OMC Data: Workflow management systems should encapsulate all relevant information as valid OMC data and return it to appropriate recipients.
OMC Data Platforms
The definition and scope of OMC Data Platform are still evolving. Conceptually, such platforms serve as centralized systems that allow different participants and workflows to store, locate, and retrieve OMC data. This centralization offers a coherent view of production status and likely delivers significant downstream benefits.
From an OMC identifier perspective, data platforms should follow these practices:
- Act as OMC Data Consumers: Since data platforms primarily consume OMC data, they should adhere to the guidelines outlined in the OMC Data Consumers section.
- Preserve Identifiers: Like asset management systems, these platforms must be designed to handle multiple OMC identifiers for each entity, accommodating variations in scope, character set, length, and encoding.
- Facilitate Data Retrieval: Like asset management systems, these platforms should support data retrieval by its OMC identifier, in addition to any other methods.
- Establish Connections Between Data Fragments: Data platforms may receive overlapping fragments of information pertaining to the same OMC entity. Although these fragments might initially appear unconnected, a common OMC identifier (identical in both scope and value) establishes the link. Platforms should record these connections, as and when discovered, and communicate to clients. Such connections reduce communication gaps, inefficiencies, and errors – a critical goal of the 2030 Vision. (A topic for another day.)
Conclusion
In a complex media production environment that has centralized systems and de-centralized systems in the mix, it is critical to understand how entity identifiers will be used across various applications and interactions. Knowing when to reuse existing identifiers, when to create new ones, how to assign scopes, and how to create and parse identifiers will make substantial difference for production teams.
In this three-part blog series on identifiers, we have discussed the motivation for using and creating well-designed identifiers, what makes them well-designed in the first place, and how the Ontology for Media Creation (OMC) uses them. Specifically in this third part, we provided actionable practices for various participants in the media creation world.
Implementing these practices will help reveal to participants the connections between information created and captured at different production stages, enabling systems and participants to quickly reference and locate entities using their OMC identifiers. If you want to discuss any of these blogs reach out to us to chat at omc@movielabs.com.
