Microsoft Dynamics 365: Finance and Operations Developer Associate (MB-500)

B is correct.

Explanation: Azure DevOps disables the creation of TFVC repositories in new projects and organizations by default. You must confirm which version control system your project uses before you start development.

A is incorrect: Git is supported for X++ development, but TFVC is also supported and is the default option for finance and operations apps.

C is incorrect: Azure Repos is the recommended and supported version control technology for finance and operations apps.

D is incorrect: TFVC is the default option for finance and operations apps, but new Azure DevOps projects and organizations disable TFVC repository creation by default. You cannot assume that a TFVC repository is available.

B is correct.

Explanation: The evidence supports B because setting the AOS Authorization property to CreateReadUpdateDelete ensures that the AOS must authorize create, read, update, and delete operations on the table. Before it allows Read access, the AOS verifies that the user's role has permission to read the data.

A is incorrect: CreateUpdateDelete does not include Read. The AOS would not authorize read operations, so the requirement for reads is not met.

C is incorrect: Read covers only read operations. Create, update, and delete operations would not require AOS authorization.

D is incorrect: None is the default value and does not require AOS authorization for any operation on the table.

C is correct.

Explanation: UDE developer environments function more like Tier-2+ environments and run on Azure SQL. You must deploy and synchronize customizations before you can use them. Build the X++ code locally, deploy and synchronize the package to the cloud, and then run and debug the code in the same environment.

A is incorrect: Cloud-hosted environments are the traditional one-box development model. Provisioning one doesn't make your changes available in the UDE environment.

B is incorrect: A database refresh copies data. It doesn't deploy or synchronize your X++ customizations.

D is incorrect: The Get-DeploymentSettings.ps1 script applies to setting up on-premises environments through Lifecycle Services. It doesn't apply to UDE environments managed in the Power Platform admin center.

A is correct.

Explanation: Merge using Merge Tool shows both versions of the file side by side. You can keep or discard each change in either version, one at a time. Use it when you need changes from both versions.

B is incorrect: Take Source uses the file from source control and erases all your local changes.

C is incorrect: Keep Local keeps your local file, and you lose any changes that are present in source control.

D is incorrect: You cannot check in a changeset until there are no unresolved merge conflicts.

B and D are correct.

Explanation: The evidence states that you should keep your branching structure as simple as possible and expand the strategy only as needed. It also recommends starting with at least a three-branch strategy: one for development, one for testing, and one for releasing code to the production instance.

A is incorrect: Creating a complex structure for every developer and environment contradicts the principle of keeping the structure as simple as possible. The evidence also notes that branching strategies can change throughout the implementation.

C is incorrect: The evidence specifies that code must be deployed into a Tier 2 Standard and Acceptance Test environment and selected as a release candidate before it can be promoted to the production environment.

C is correct.

Explanation: Use a client plugin when the scenario requires form interaction, X++ business logic, or data modifications that require the standard UI. Define the action as a subclass of SysCopilotChatAction and use data contracts for inputs and outputs. Client plugin development requires the unified developer experience on version 10.0.40 or later.

A is incorrect: AI tools are low-code, platform-agnostic components for simple data queries and cross-platform scenarios. They have limited awareness of finance and operations form context or UI state.

B is incorrect: Power Automate automates business processes across services, but it is not one of the two Copilot extensibility approaches and does not provide Copilot form interaction in the finance and operations client.

D is incorrect: A Power BI report provides interactive visuals in a workspace. It cannot create a draft transfer order or respond to Copilot prompts.

A and C are correct.

Explanation: Multi-user testing requires Visual Studio Enterprise Edition in the development environment. It also requires a tier-2 or higher sandbox environment that uses the same application and platform update as the development environment.

B is incorrect: You must install the same certificates that you created for the development environment on all Application Object Servers (AOS) in the tier-2 or higher sandbox.

D is incorrect: Multi-user testing requires a tier-2 or higher sandbox, and the application and platform update must match the development environment.

A and C are correct.

Explanation: When several developers work on one project, perform a Get Latest process to merge new code from other developers. Then perform a local build to verify that your code does not disturb what is in the Azure DevOps repository. After both steps, check in your code by using Team Explorer.

B is incorrect: Scheduling production downtime happens after a deployable package is built and applied to at least one sandbox environment in the same project. It is not a step before check-in.

D is incorrect: A build pipeline in Azure DevOps creates the deployable package from checked-in code, and developers generally do not deploy their own code. A package must be applied to at least one sandbox environment before you can schedule a production update.

A is correct.

Explanation: For tier 1 environments, you must complete the Power Platform Integration configuration in the Advanced settings during deployment. You can't change it after the environment is created.

B is incorrect: Tier 1 environments don't allow this configuration to be changed after creation. If you wait until after deployment, you can't add it.

C is incorrect: The Azure connector links Lifecycle Services to an Azure subscription and region for creating cloud-hosted environments. It doesn't configure Power Platform integration.

D is incorrect: The Azure DevOps settings let Lifecycle Services create a build pipeline for a build and test environment. They don't configure Power Platform integration.

C is correct.

Explanation: Select the correct model when you create the project. The model is difficult to change after you add files to the project. You can edit the model dependencies list later as you add elements.

A is incorrect: You can edit the model dependencies list later, so you do not need to select every model now.

B is incorrect: The model is difficult to change after you start adding files, so set it correctly at the start.

D is incorrect: Clean removes previous build outputs. It does not select or change the model for a project.

D is correct.

Explanation: Use QueryBuildDataSource to add TrainerTable as a child of TrainingMaster, enable relations, and set the join mode to ExistsJoin. Then use QueryBuildRange objects to filter TrainingType on TrainingMaster and TrainerType on TrainerTable. This filters TrainingMaster records based on the related table.

A is incorrect: TrainerType is a field on TrainerTable, not on TrainingMaster. You must add TrainerTable as a related data source to filter on it.

B is incorrect: Separate root data sources don't establish the relationship between the tables, and a sort field orders records but doesn't filter them.

C is incorrect: addSortField only defines sort order. Filtering requires QueryBuildRange.

B and D are correct.

Explanation: Microsoft manages cloud deployments. With on-premises deployments, the customer assumes greater responsibility and extra costs. These costs include establishing and maintaining infrastructure and configuring high-availability disaster recovery solutions.

A is incorrect: For on-premises deployments, the customer sets up the sandbox environments. Microsoft doesn't set them up.

C is incorrect: The fully managed service by Microsoft applies to cloud deployments. It doesn't apply to locally hosted on-premises deployments.

A and C are correct.

Explanation: Use labels for all text on an SSRS report so that the text is automatically translated when the report runs for a country or region with a different language. This lets you maintain one report for all locations. Use groupings to organize data into a meaningful layout, such as a header and footer grouping and a table layout with columns and rows for transaction data.

B is incorrect: Labels remove the need to create and maintain an individual report for each language.

D is incorrect: Showing all data without groupings leads to a cluttered display. Groupings also let you share expressions for a data set.

D is correct.

Explanation: When Dynamic Fields is set to Yes on a data source, the system adds new fields automatically to the query field list. A query extension isn't needed for new fields. You still need to build before you use the query.

A is incorrect: A query extension can add fields, but it isn't needed when Dynamic Fields is set to Yes. It adds work without a benefit.

B is incorrect: The Clear value for Dynamic Fields results in a build error.

C is incorrect: A query extension can't remove existing data sources, fields, grouping, ordering, or having.

C is correct.

Explanation: The evidence states that SysGlobalObjectCache is server-shared caching across all user sessions on a specific server, used to store static values so the system does not need to reread or calculate them for each session.

A is incorrect: SysGlobalCache refers to a single client session cache that is not accessible in other system threads.

B is incorrect: A singleton class creates an instance for the user session, meaning it is not shared across all user sessions on the server.

D is incorrect: A disposable context class is used to transfer parameters without changing a method signature, not to share cached values across sessions.

C is correct.

Explanation: A tile opens a page through its linked menu item. When you set the Menu item type to Display and the Menu item name to FMSetup, the tile opens the Fleet Management Setup page. After you configure the tile, add a Tile button to SummaryTileSection in the form extension and set the Tile property to the new tile.

A is incorrect: Aggregate data entities provide data for KPIs. They do not set the page that a tile opens.

B is incorrect: A .rdl file contains SSRS report data, calculations, and layout. It is not used to link a workspace tile to a page.

D is incorrect: A Power BI report shows visualizations in a workspace. It does not open the Fleet Management Setup page through a Tile button.

B is correct.

Explanation: Use InMemoryRealTime when users need real-time data. This option uses Column store indexes of the SQL Server database.

A is incorrect: StagedEntityStore deploys the measure to the Entity store, and the data is refreshed periodically. This setting provides near real-time data, not real-time data.

C is incorrect: The Default aggregate property controls how the measure is calculated, such as sum, count, or average. It does not control whether the data is real-time.

D is incorrect: The Goal Type property defines how the KPI goal is determined. It does not control how the aggregate measurement retrieves data.

D is correct.

Explanation: To use a copy of the Production database on a VHD, first restore the database from Production to a Tier2 or above environment. Then export the database by using Microsoft supported processes.

A is incorrect: You can't restore Production directly to the VHD. The database must first go to a Tier2 or above environment and then be exported by using a supported process.

B is incorrect: Team Explorer connects Visual Studio to Azure DevOps for code that other developers created. It doesn't provide database data.

C is incorrect: Demo data supports general development testing. It doesn't contain the Production data that you need to troubleshoot this issue.

B is correct.

Explanation: You can't use an extension to create a unique index because that type of development is intrusive. For an extensible uniqueness constraint that spans multiple columns, consider creating a hash of the values of those columns instead.

A is incorrect: You can add indexes to a table extension, but you can't create a unique index by using an extension.

C is incorrect: Development for finance and operations apps is extension-based and doesn't allow customizing base code and elements.

D is incorrect: You can change the Extended Data Type property on an existing field only to an EDT that derives from the current EDT.

B is correct.

Explanation: The evidence supports B because a class extension adds methods and state to artifacts in other models without recompiling or overlayering those models. You define the extension in a new model that depends on the model of the original artifact. Because the base code isn't changed, your customizations remain upgradeable.

A is incorrect: Overlayering changes the original model. Class extensions are designed to avoid overlayering.

C is incorrect: Extension classes must be marked final, and the runtime system instantiates them. You can't derive from an extension class.

D is incorrect: Changing and recompiling the original model changes the base code. That approach doesn't keep the customization upgradeable.

A is correct.

Explanation: The Menu Item Name property links the Menu Item Button to a menu item. Select CustFreeInvoiceNew so the button opens the free text invoice page.

B is incorrect: The Caption property sets the displayed text of the Button Group. It doesn't define the page that opens.

C is incorrect: The Name property identifies the Action Pane Tab control. It doesn't link a button to a menu item.

D is incorrect: A data source provides tables and fields for the form. It doesn't create navigation from a button.

A is correct.

Explanation: The evidence supports A because Chain of Command (CoC) is the default and more robust approach for extending public or protected methods. CoC enforces the method signature at compile time and makes the execution order explicit through the next call.

B is incorrect: Pre-event and post-event handlers are subscribed indirectly through attributes, meaning some issues like underlying method signature changes are only detected at runtime.

C is incorrect: The wrapper method must have the exact same signature as the base method in a CoC extension.

D is incorrect: Wrapper methods in an extension class are required to always call next unconditionally, unless the method is explicitly marked as Replaceable.

B and D are correct.

Explanation: The evidence supports B and D because referencing only the models that your extension needs reduces compile times and dependency management overhead. Keeping customizations modular and reusable by breaking large extensions into smaller, independent modules is also a recommended best practice.

A is incorrect: Referencing models that you do not need adds to compile times and dependency management overhead.

C is incorrect: Models that are in the same package as the base objects are associated with overlayering, which causes longer compile cycles and larger package distributions.

B is correct.

Explanation: By decreasing the number for Bundle size, you produce more bundles and increase the ability to distribute the events to parallel threads.

A is incorrect: The maximum value for Processing threads is 4.

C is incorrect: By increasing the number for Bundle size, you produce fewer bundles and reduce the ability to distribute the events to parallel threads.

D is incorrect: Retry count is the number of times that the system will try again to send business events to an endpoint if an error occurs.

A is correct.

Explanation: The evidence supports A because you should start the extension class name with the name of the type that you augment, and end it with _Extension. You should insert your model prefix as an infix to distinguish the class from other types.

B is incorrect: Don't name an extension class <Element that is being extended>_Extension. The risk of conflicts with other models is too high.

C is incorrect: The extension class name should start with the name of the augmented type. You should use the prefix as an infix, not at the beginning.

D is incorrect: This format applies to extension elements, such as table extensions. An extension class name must end with _Extension.

A and C are correct.

Explanation: The evidence supports A and C because when you wrap a static method, the method in the extension class must include the static keyword. Additionally, if the base method has a default parameter value, you must declare the parameter in the wrapper without the default value.

B is incorrect: The wrapper must not repeat the default value. You must declare the parameter without it.

D is incorrect: The call to next must be in the first-level statements of the method body. You can't call next conditionally in an if statement unless the method is marked Replaceable.

B and D are correct.

Explanation: To use the Fin & Ops Apps (Dynamics 365) connector, the data must be in a data entity that is configured for public API. Verify that the IsPublic property is set to Yes on the data entity in Visual Studio. The connector is a premium connector, so the users who run the flow need a license that includes premium connectors.

A is incorrect: Templates are optional. You can also create a flow from My flows by selecting New.

C is incorrect: You can't use this connector to integrate with Dynamics 365 Finance + Operations (on-premises) environments.

A is correct.

Explanation: Synapse Link continuously exports data into a managed data lake in near-real-time. It supports large datasets and is ideal for data science workloads that require high granularity and historical depth.

B is incorrect: BYOD is refresh-based and not near-real-time. Its ML and AI readiness is limited, and it suits medium workload sizes.

C is incorrect: Aggregate measurements use a fixed aggregation grain that is defined at design time. They aren't suitable for historical trend analysis beyond their intended design.

D is incorrect: Data entities are a semantic and integration layer, not a datastore. They aren't optimized for high-frequency analytics when you query them directly.

D is correct.

Explanation: By default, OData returns only the data that belongs to the user's default company. To return data from all companies that the user has access to, add the ?cross-company=true query option, for example [Your organization's root URL]/data/FleetCustomers?cross-company=true.

A is incorrect: IsPublic controls whether the entity is exposed on the OData endpoint. The entity already returns data, so it is already exposed, and this setting doesn't change the company scope.

B is incorrect: The Metadata/PublicEntities URL returns metadata and actions for entities. It doesn't return entity records from other companies.

C is incorrect: A custom service adds development and maintenance effort for each company. The required behavior is available through a query option on the existing OData endpoint.

C is correct.

Explanation: The When a row is added, modified, or deleted trigger in the Microsoft Dataverse connector runs a flow when a create, update, or delete operation occurs on a finance and operations apps entity that is enabled as a virtual table in Dataverse. This trigger requires Microsoft Power Platform integration to be enabled, which this environment meets.

A is incorrect: An Instant (Button) flow runs only when a user manually selects a button, so it doesn't run for each data change.

B is incorrect: A scheduled flow runs at a specific time, not when each record changes.

D is incorrect: This action starts a Data management framework export project. It is not a trigger for record changes.

C is correct.

Explanation: To consume an external web service, you can create a wrapper class using a .NET programming language like C#. You set up a service reference to the SOAP web service in the class library, include the assembly under References in Application explorer, and call the wrapper class methods directly from X++.

A is incorrect: Writing a custom service under a service group exposes a new service on SOAP and JSON endpoints. It is not used to consume an external web service.

B is incorrect: ExportToPackage is used to initiate the export of a data package from finance and operations apps. It does not consume external web service methods.

D is incorrect: The Microsoft Entra applications page is used to maintain approved client IDs for inbound API calls, not to configure outbound calls to external SOAP web services.

A is correct.

Explanation: With a synchronous pattern like OData, success or failure responses are returned directly to the caller. The response contains the error message, and the caller is responsible for handling any errors.

B is incorrect: Polling for results is required for asynchronous batch data APIs, whereas synchronous OData calls return the result directly.

C is incorrect: Synchronous OData calls process immediately rather than in a batch, and the system does not push status notifications afterward.

D is incorrect: Resubmitting through a batch data API will only return an immediate response indicating if scheduling succeeded, not the actual validation error.

D is correct.

Explanation: The Data management framework package API is available for both cloud and on-premises deployments and is the only supported API for on-premises deployments.

A is incorrect: Recurring integrations are not supported for on-premises implementations.

B is incorrect: Recurring integrations are not supported for on-premises implementations.

C is incorrect: Recurring integrations are not supported for on-premises implementations.

A is correct.

Explanation: Power BI accesses public data entities in the finance and operations apps database. In Power BI Desktop, use Get Data, select Other, select OData Feed, enter the connection information, select the data, and then select Load.

B is incorrect: Publish in the Data Connector add-in saves your changes back to finance and operations apps. It doesn't load data into a Power BI Desktop report.

C is incorrect: The Fin & Ops Apps (Dynamics 365) connection in Power Apps connects an app to a data entity. It doesn't connect Power BI Desktop to the data.

D is incorrect: Power Automate connections support flow integrations. They don't provide the OData Feed connection that Power BI Desktop uses to load data entities.

D is correct.

Explanation: The volume is high, the frequency is periodic, and real-time data is not required. These conditions fit an asynchronous pattern. Use the batch data API for asynchronous integration scenarios. Determine the required entities, and make sure that data management is enabled for them.

A is incorrect: A custom service fits real-time lookups with low volume and unplanned frequency, such as 1,000 records per hour. A real-time call for each of 150,000 records per hour does not match that pattern.

B is incorrect: OData is recommended for synchronous scenarios. In a synchronous pattern, the caller is blocked until the recipient returns a response, which does not fit periodic high-volume loads.

C is incorrect: Consuming external web services lets finance and operations apps call external systems. In this scenario, the external system sends data into finance and operations apps.

C is correct.

Explanation: A data entity presents the customer tables as one de-normalized view. You can expose a data entity as an OData service for synchronous integration, including third-party mobile app integrations. The entity abstraction also isolates application code from changes to the physical tables between versions.

A is incorrect: Recurring integrations use the data management pipeline for asynchronous integration. They do not meet the real-time read and write requirement.

B is incorrect: A data package is a compressed file with a manifest and data files. Use it for file-based import, export, and ALM scenarios, not for real-time app access.

D is incorrect: Working directly with the normalized tables makes reads and writes tedious. It also exposes the app code to changes in the physical tables between versions.

B is correct.

Explanation: When you extend an entity by adding fields, or when the automatic mapping appears incorrect, regenerate the mapping. On the Mapping page, select Generate source mapping, and then confirm that you want to generate the mapping from scratch.

A is incorrect: Auto sequence determines the processing order of the entities in a project. It does not change field mapping.

C is incorrect: Auto-generated makes the system generate values for a field that is not in the source file. In this scenario, the fields need to be mapped from the source.

D is incorrect: Delete selection unmaps fields. It does not add mappings for new fields. Required fields, marked with a red asterisk, must be mapped to work with the entity.

B and D are correct.

Explanation: Start all task recordings on the main dashboard so that automation tools can run your tests. Each test case should also have at least one validation. Validate critical fields, such as order totals, that cover the impact of other fields such as unit price, quantity, discount, and tax.

A is incorrect: Starting in January 2026, you can only create, edit, or use task recordings in sandbox and production environments. Cloud-hosted environments no longer have this capability.

C is incorrect: Avoid printing a report in a test case. If a test case needs a report, select it on screen.

A is correct.

Explanation: Select a service account, such as a generic user account, as the Admin user. The Admin user is the only user who can sign in after initial deployment, and the only user who can sign in to a sandbox after a database refresh from production. Deploy production to the same datacenter where the sandbox environments are deployed and where UAT and performance testing were done.

B is incorrect: If you use a named user account and that user isn't available, you might not be able to access the environment.

C is incorrect: You can't change a production deployment request after you submit it. To correct a mistake, wait until the deployment finishes, and then log a support ticket to delete the environment so that you can redeploy it.

D is incorrect: Production should be deployed to the same datacenter as the sandbox environments where UAT and performance testing were done.

B and D are correct.

Explanation: The Enable triggers in target database option is disabled by default. When you enable it, export jobs enable SQL triggers in the target database. Use the trigger to determine when the sync from finance and operations apps is complete, and then load the downstream reporting systems.

A is incorrect: The staging tables hold data while the sync is in progress, so they change constantly. Reading directly from them causes inconsistent results.

C is incorrect: Recurring data jobs aren't supported for BYOD. To schedule a BYOD export, use the Export in batch option.

B is correct.

Explanation: Power Platform integration surfaces finance and operations app data in Microsoft Dataverse. Dual-write provides near real-time, bidirectional synchronization between finance and operations apps and Dataverse. Use this pattern when the data must also be available to Power Apps, Power Automate, or customer engagement apps.

A is incorrect: Recurring integrations exchange documents or files with third-party applications. This pattern is asynchronous and does not provide near real-time, bidirectional synchronization with Dataverse.

C is incorrect: The Excel Data Connector add-in interacts with Excel workbooks and OData services in a task pane. It is a user tool and does not synchronize data with Dataverse.

D is incorrect: BYOD exports entity data from finance and operations apps into an Azure SQL database. The export goes in one direction and does not make data available in Dataverse.

A is correct.

Explanation: Cross-company data sharing replicates reference, parameter, and group data between companies within an environment. Sharing payment terms and payment day definitions across legal entities is a supported example. Test and validate the policy in a sandbox environment before you apply it to production.

B is incorrect: Cross-company data sharing cannot be used for transactional data. Sharing transactional records for reporting or consolidations is not supported.

C is incorrect: Sharing across deployments is not supported. Data sharing applies between companies within one environment.

D is incorrect: Fields that reference financial dimensions, such as Ledger or Default dimension, cannot be shared across companies.

B is correct.

Explanation: The evidence supports B because Electronic reporting configures formats for incoming and outgoing electronic documents to meet the legal requirements of countries or regions. It targets business users. Because you configure formats instead of writing code, you can create and adjust formats faster when regulatory requirements change.

A is incorrect: SSRS is primarily used for business documents and high-volume printing, such as invoices and packing slips. New and modified SSRS reports require X++ development.

C is incorrect: Embedded analytics and native controls aggregate data to workspaces by using charts and grids. They don't manage electronic document formats for exchange with government bodies or banks.

D is incorrect: Financial Report Designer creates and customizes financial statements, such as balance sheets and income statements. It doesn't configure regulatory electronic document formats.

A is correct.

Explanation: Use a data source of the Barcode type to produce bar code pictures when you configure an ER format. QR Code is one of the supported two-dimensional bar code formats, and you can use the images on shipping labels.

B is incorrect: A default check number interval is used for interval validation of check numbers. It does not generate bar code images.

C is incorrect: Character validation shows a warning when a character appears in a check number. It does not create images on labels.

D is incorrect: A configuration provider identifies the owner of an ER configuration. It does not generate bar code images.

A and C are correct.

Explanation: The evidence supports A and C because a password lock adds security to one report component without securing the whole system. Other users always have read-only access to a locked component, and they can open, change, and save it under a new name. A user with the administrator role can always access and change a locked building block.

B is incorrect: A user in any role can lock a building block. You don't need the administrator role to apply a password.

D is incorrect: A lock doesn't hide the component. Other users keep read-only access, so they can still open the locked building block.

C is correct.

Explanation: The evidence supports C because when you add a dimension filter in Report options, the report displays only the data for the selected dimension value. This filtered report is a personalized view. Other users can't view it.

A is incorrect: The filter applies to the manager's personalized view only. Colleagues don't see the filtered data.

B is incorrect: Report options change how you view the report. They don't modify the row definition building block.

D is incorrect: The filter is applied when you select OK in Report options. You don't need to regenerate the report in Report designer.

B is correct.

Explanation: ER supports JSON output, is designed for business users who are familiar with Excel-based formulas, and lets you control and modify reports without a developer. Versioning helps you manage the definition lifecycle as requirements change.

A is incorrect: Developer customization is the approach that ER avoids, and it adds effort for each regulatory change.

C is incorrect: A configuration provider identifies the configuration owner. Creating a new provider does not update a format.

D is incorrect: A Barcode data source produces bar code pictures for documents and labels. It does not produce a JSON report.

B and D are correct.

Explanation: Set up the required document types in Document management before you configure ER. Then turn on Enable design mode on the General tab to enable the ER design experience in the current Finance instance.

A is incorrect: The default value of 0 uses the standard system timeout. A custom value only changes the duration of queries that the ER engine runs and is not required to enable design.

C is incorrect: Autofit row height only adjusts the height of Excel template rows that contain merged cells with wrapped text at runtime. It does not enable the design experience.

C is correct.

Explanation: When you base a custom format on another format, ER automatically upgrades the custom format when the original format is changed.

A is incorrect: Manual re-creation is not required because ER upgrades a derived custom format automatically.

B is incorrect: ER maintains electronic document formats through configuration, not code, so a developer extension is not needed.

D is incorrect: Dataverse is used to distribute formats to partners and customers. It does not apply changes from the original format to a custom format.

D is correct.

Explanation: The evidence supports D because when you request more than one report at a time, the result is a .ZIP file. You must download and uncompress the .ZIP file before you can use the Word documents.

A is incorrect: Multiple reports aren't downloaded as separate files. They're packaged in one .ZIP file.

B is incorrect: The reports aren't merged. Each report remains a separate Word document inside the .ZIP file.

C is incorrect: Word export is still available for multiple reports. The constraint is that you must uncompress the .ZIP file first.