CSV reports may now include an unlimited number of entries, and the configurable maximum limit has been removed.

With this change, users initialize report generation using the UI, but if the report is not immediately available, V‑Spark continues to generate the report in the background. V‑Spark emails the requesting user when generation is complete and the report is ready to download. Reports generated with a relatively large number of entries may take a considerable amount of time to complete.

Previously, the number of entries in CSV reports was limited to a maximum defined with the export_filelist_limit system configuration option. This option has been removed.

Renamed and clarified expected values for the Servers field used when connecting a company entity to an automatic speech recognition (ASR) engine. The Servers field is now named ASR API URL. Examples in the field now show the format of the required URL, and UI text explains to use a semicolon (;) to separate multiple URLs.

Added the new restrict_application_editor system configuration file option (00base.cfg) to prohibit users with only View permissions from seeing application configurations. If restrict_application_editor = on, only users with edit permissions for the associated organization are able to open the application editor from the dashboard. This new configuration does not change any existing behavior related to accessing the application editor from the Settings menu.

Addressed cross-site scripting and other security vulnerabilities.

Resolved an issue with the transcript view's highlight navigation buttons ( and ) on the File Details page. The buttons now correctly cycle through multiple highlighted application matches. This resolves a regression first introduced with V‑Spark version 4.1.

Resolved a page routing issue that could sometimes cause certain valid URLs to be inaccessible and return a 404 error. Previously, attempting to access certain URLs resulted in a 404 error. These URLs now properly load.

Resolved an issue with NaN, +Infinity, and -Infinity values for custom metadata fields. These values are now converted and stored as strings.

Previously, when these values were submitted in JSON metadata, that JSON was accepted by the system but could result in records that could not be viewed on the File Details page. Additionally, those metadata values could not be changed with the API.

On the File Details page, hyphens are displayed as hash symbols (#), so -Infinity values appear as #Infinity.

Resolved an issue with exported CSV reports. Previously, generating a CSV report from the dashboard with All folders selected returned a report with incorrect values for folder and url column data. CSV report folder and url columns now contain the correct values when exporting all folders.

Removed end-user-facing references to copyright notes from the codebase.

Resolved an issue with sorting by custom metadata field in the Dashboard Files View. Previously, starting with version 4.3, some records appeared out of sort order when sorted by metadata fields. Records sorted by custom metadata fields are now in the correct order.

Improved audio ingestion processes:

1. Resolved an issue with Job Manager's zip file handling under certain race conditions. Previously, Job Manager sometimes checked for certain files before they were available on the target system, then reported the files not found. As a result, some requests were left unprocessed in a queue despite their overall validity. Files are now processed correctly under these conditions.

2. Improved invalid file and Elasticsearch error handling. Previously, Job Manager logged a WARNING-level message describing an error with Elasticsearch, but sometimes the problem was not related to Elasticsearch. Now, the Elasticsearch warning only appears for Elasticsearch errors, and the invalid file condition is properly handled.

3. Improved Job Manager data handling during V‑Cloud ASR transactions and added V‑Cloud requestID logging for improved traceability and support when using V‑Cloud ASR.

4. Improved error handling for V‑Cloud ASR transactions to resolve an occasional issue in which a temporary data retrieval error resulted in incomplete requests and so-called "orphaned" files in V‑Spark. Job Manager now correctly handles temporary errors when retrieving V‑Cloud results and retries the failed requests.

5. Improved Job Manager event logging. Previously, it was not possible to follow the full lifecycle of a particular job's data through transcoding and ASR processing. With this release, new and revised log messages associated with the changes listed above provide more and clearer status information linking V‑Spark requests and jobs to their associated audio and ASR processing.