Release Notes for Mindbreeze InSpire
Mindbreeze GmbH, A-4020 Linz, 2023.
All rights reserved. All hardware and software names used are brand names and/or trademarks of their respective manufacturers.
These documents are strictly confidential. The submission and presentation of these documents does not confer any rights to our software, our services and service outcomes, or any other protected rights. The dissemination, publication, or reproduction hereof is prohibited.
For ease of readability, gender differentiation has been waived. Corresponding terms and definitions apply within the meaning and intent of the equal treatment principle for both sexes.
Innovations and new features
Mindbreeze was awarded the silver certificate for web accessibility by the WACA initiative and the independent certification body TÜV Austria.
As a representative sample for the Mindbreeze InSpire cloud search web application, a use case in German and English language was selected for the certification audit. Mindbreeze meets the requirements of WCAG 2.1 - AA to a high degree and therefore received the silver certificate.
The following criteria must be met for the Web Accessibility Certificate (WACA) in Silver:
Website meets WCAG 2.1 - AA success criteria to the greatest extent possible.
All content is accessible to all users.
Basic functionality is fully accessible to all users.
Parts of the extended/optional functionality are more cumbersome for some users, but still accessible.
WACA is Austria's first and only seal of quality to represent web accessibility according to the international W3C guidelines. The certificate of the WACA initiative and the independent certification body TÜV Austria is intended to ensure accessibility for all people on the professionally tested website.
Link to the Certification
Mindbreeze InSpire on AWS Elastic Compute Cloud
Mindbreeze InSpire is now also available via Amazon Web Services. At Amazon Marketplace, both a Mindbreeze InSpire 1M instance as well as a Mindbreeze InSpire 10M instance are available as Amazon Machine Image.
For a subscription via Amazon Web Services, all you need is a valid Mindbreeze license.
Link to AWS Marketplace
Extended Relevance Model for Insight Apps
Mindbreeze InSpire now supports user defined document boosting rules where the boost factor itself can take the form of a property expression. Therefore, document boost factors can be computed given document metadata and even based on graph expressions using references.
This feature can be used to improve the relevance and accuracy of search results in Mindbreeze Insight Apps and enables completely new application possibilities.
As a concrete example we would like to include the priority (priority.value – 1 for cirtical up to 5 for planning) and the classified severity (severity.value – 1 for high up 3 for low) in the relevance of ServiceNow tickets. Both properties are more critical the lower the value they have. As new computed boost factors we can use the following expressions:
To boost priority:1 + (1 - priority\.value) div 4.0
To boost severity: 1 + (1 - severity\.value) div 2.0
Link to the Documentation
Object Recognition Based on InceptionV4
The new InceptionImageFilter plugin allows classifying images using Deep Learning. The plugin allows to collect, filter and search images and thumbnails in new ways.
For the object recognition the Inception V4 model is used, which is a very well-known and successful architecture for Convolutional Neural Networks related to image classification and object detection. In the filter-plugin configuration, you can further customize the filter for your application purpose.
Link to the Documentation
New and Advanced Connectors
The new Box connector enables indexing all content from a Box Cloud instance. This allows, for example, text documents, images, videos, notes, etc. to be indexed by Mindbreeze InSpire and used within Mindbreeze Insight Apps.
In conjunction with the new Box Principal Resolution Service, it is possible to download and resolve user and group permissions from your Box Cloud instance.
Link to the Documentation
Microsoft Graph Connector Supports Profile Endpoint
With the inclusion of new user profile endpoints, it is now possible to index advanced information about users/people from your Sharepoint Online, Office 365 and OneDrive cloud instances through the Microsoft Graph Connector. This enables new and improved search applications for a 360-degree view of projects, experts and skills in an organization.
The newly available profile-metadata includes:
Schools and education
Skills and expertise
Interests and hobbies
The additional metadata is automatically downloaded with the new version of the Microsoft Graph connector if the corresponding configuration option "Get Metadata From Profile", is enabled.
Link to the Documentation
Microsoft Graph Connector for Search-Based Organization Charts
Using new API endpoints in Microsoft Graph, it is now possible to index the responsible supervisor/manager of a user/person as a metadata reference using the Microsoft Graph connector. This metadata reference can be used to generate hierarchical organization diagrams in a Mindbreeze Search app.
In the following example, you can see such a generated organization chart view in a Mindbreeze InSpire Insight App:
Link to the Documentation
Revised Microsoft Dynamics (CRM) Connector
New optimizations in the Microsoft Dynamics (CRM) connector bring significant performance improvements for large data volumes. With over 1 000 000 simulated test documents, download speeds are now up to 10 times faster than before.
The flexible data structure in Microsoft Dynamics, together with Mindbreeze InSpire, offers a variety of different search applications.
Link to the Documentation
Insight Engine Query Performance Improvements
Software optimizations and adjustments to the CPU settings in the operating system, significantly improved the query performance of the Insight Engine. In our "General Web Index" benchmark, we were able to determine an average performance increase of 25%.
New Installer and Image Format
With 22.3, there is a new self-describing image format that is specifically designed to modularize the installation process.
The image format is verified and loaded by the InSpire System Loader. It is a repository consisting of container images alongside update instructions. Additional components, such as the host operating system update, are also part of the image.
The "InSpire System Manager" service is used for performing the update. If a more recent System Manager is part of this image, it is replaced by the Loader so that the new Manager can be used to perform the update. After the update, the System Loader is also updated, if it is part of the image.
Security relevant changes
- Security Patch: jQuery UI (CVE-2021-41182, CVE-2021-41183, CVE-2021-41184, CVE-2022-31160)
- Xalan Jars were removed from the product (CVE-2022-34169)
- Security Patch: Upgrade Python to 3.8
- Update Apache Commons Text (CVE-2022-42889)
- Update Batik to version 1.16 (CVE-2022-41704, CVE-2022-42890)
- Update OpenJDK to version '8u352' (CVE-2022-21626, CVE-2022-21624 und CVE-2022-21619)
- Fix for: Index may crash under load if boostings for non existing dynamic aggregated properties are used
- Include InceptionImageFilter in InSpire-Image
- Fix for: InSpire Update: chown of servicedata servicetempdata takes too long during update
- Fix for: Insight Apps using old adapted.css may render incorrectly
- Fix for: Alert filter and search tab creation problems
- Azure Principal Resolution Service use non-delta API to get all groups
- Fix for: Bestbets aren't shown correctly without KeyMatch
- C++ Servlets are instrumented in app.telemetry
- Fix for: Missing parameters for Client Service content fetch request instrumentation
- Fix for: Client Service Ressource Plugins do not consider the 'URL Base Path' setting
- Fix for: Collected Results: Result image is stored in persisted resources, but not used
- Conan Support for CentOS7
- Configurable sample text length for Insight Apps
- Fix for: CRM Crawler: mes:key duplicates with documents that have multiple entity types
- Customize the 'Manage Subscriptions/Alerts' Mindbreeze Dialog
- Fix for: Default-Language which is not matching the Browser Language
- Fix for: Documentum Connector: INFO Logging is not working
- Fix for: Excel Export doesn't work for hierarchical metadata (lists)
- FilterPlugin.InceptionImageFilter is available as ContentFilterService
- Fix for: Searches with labeled expression to fqcategory in combination with a NOT expressions yields to results from other fqcategories
- Fix for: Highcharts are not reset correctly
- Performance improvement of the Identity Conversion Replacement Transformer
- Faster start of subprocesses at syncdelta
- Improved Completeness of search documents if replace happens faster than inversion
- Fix for: Index Exception: Can't parse message of type "mindbreeze.common.Value" because it is missing required fields: kind
- Fix for: Index tmp-files are opened after unclean shutdown and lead to errors
- InSpire Update: Verifiy issuer of package signature and certificate on update
- Fix for: JIRA Principal Cache: Fetching users with latest API change
- Fix for: Jive Crawler: Thumbnailing cannot be disabled
- Fix for: Jive Crawler: Sitemap generator fails to generate full-sitemap
- Jive Sitemap Generator: Filter via administrator metadata instead of author metadata
- Fix for: LDAP PrincipalCache not using LDAPS when Connection Poolmanager is disabled
- Fix for: The search messages 'load more results', 'loading more results' are not shown without adapted.css
- Fix for: Logout not working in MMC
- Fix for: MaxTermsConstrainedQueryExprTransformer does not support EXPR_TERMS
- MESNode no longer runs as root
- Fix for: mesnode starts two crawler instances simultaneously if other processes cannot be stopped
- Fix for: MessageServlet instruments network requests incorrectly
- Microsoft Dynamics CRM Connector: ACLs can be disabled in the config
- Microsoft Dynamics CRM Connector: Improve paging performance
- Fix for: Microsoft Dynamics CRM: References to associated entities are wrong
- Fix for: Microsoft Dynamics CRM: StackOverFlow error and timeout setting not working
- Fix for: Microsoft File Connector: acl-dump-tool is not working
- Fix for: Microsoft File Connector: File handles in Microsoft Windows 2019 File Share remain open
- Microsoft Graph Connector: supervisor metadata is added
- Microsoft Teams Crawler: Improved delta crawling performance and stability
- Mindbreeze InSpire browser plugin available for Google Chrome and Microsoft Edge
- Fix for: Performance issues during profiling
- Fix for: Missing docid for (PDF) content fetch requests does not lead to fallback on key search
- Fix for: Multiple execution of the update action in MMC duplicates the cookie "mod_auth_openidc_state"
- Fix for: No search results when Client Service option "Enable Query Expansion for Similar Terms" is active
- Fix for: On mobile version of the InSight App, input elements are not focusable after a search
- Performance optimization for evaluation of character NGrams
- Fix for: Parsing error when using complex expressions with a string result
- Fix for: PDF-viewer-library does not render certain PDF documents
- PropertyExpr can be used for DocumentBoost
- Fix for: PropertyExprParser fails to parse with leading/trailing whitespace
- Fix for: Query termination cause timeout is not propagated to client service in certain cases
- Fix for: ReplacementTransformer does not update alternative indices on deletion
- Fix for: Bad search performance when using category-key boostings in indices with many deleted documents
- Fix for: Non-responsive search relevance editors and hidden scrollbars
- Set Linux performance settings to "Performance" instead of "Powersave"
- SharePoint Online Crawler: Allow indexing of lists and individual documents via CSV file
- SharePoint Online Crawler: New settings for completeness analysis
- SharePoint Online Crawler: Robustification for ACL indexing during delta crawl runs
- SharePoint Online Principal Resolution Service: Optimize container amounts in Cache
- SharePoint Online: Retry on HTTP 500 during search based site discovery
- Fix for: Sitecore crawler sometimes indexes wrong page title
- Fix for: Snapshot GUI produces multiple login windows during update
- Fix for: Special query leads to Internal Server Error on Client Service
- Fix for: SharePoint Online Principal Principal Resolving: Microsoft Graph groups inserted directly into site permissions are not resolved
- Fix for: SubQuery Event is missing label and description in app.telemetry
- Fix for: Switch client language breaks when closing preview dialog
- Limit number of task manager history entries
- Fix for: TERM_EXPR is not propagated as EXPR_TERMS which yields to complex OR expressions
- Fix for: Wrong total count in app.telemetry
- Release Box Connector
- WCAG21 certification for Mindbreeze InSpire
- Web Connector: Prefer newest Last-Modified date from HTML meta tags or HTTP response headers
- Fix for: Web Connector: Using OAuth 2 for downloading sitemaps fails
- Web connector OAuth 2 authentication is also working when refresh token is not available
- Fix for: Web Connector sends duplicate User-Agent header