900 new features in WINDEV, WEBDEV and WINDEV Mobile 2026

148 , N e w f e a t u r e W D W B WM A MAJOR NEW FEATURE FOR HFSQL HFSQL 2026 introduces an incredible new feature: The ability to perform “semantic” searches, i.e. searches based on the meaning of text, rather than on the exact string. Semantic search consists in finding records not on the basis of exact keywords, but on the basis of the meaning or overall sense of the phrase being searched. This search examines text items or memos. Benefit of this new feature in version 2026: New, powerful search mode 149 N e w f e a t u r e W D W B WM SEARCH BASED ON THE MEANING OF WORDS AND PHRASES HFSQL 2026 makes it possible to search for generic terms in a database. For example, you can search for “men’s shoes“ in a product database instead of trying exact-match searches with "shoe", "shoes", "boots", "sneakers","sandals", etc. A new syntax is available for the FOR EACH statement: MAX_RESULT is int = 10 FOR EACH Product WITH SemanticSearch ( Description , "men's shoes" , MAX_RESULT ) DisplayProduct ( Product . ProductID ) END No need for third-party modules. Everything is integrated into the HFSQL engine. Benefit of this new feature in version 2026: A search mode that puts itself in your shoes 150 N e w f e a t u r e W D W B WM HFSQL WINDOWS AND LINUX The "semantic search" feature is only available in Client/Server mode. Users can launch a semantic search even from a mobile device, if the database used is in Client/Server mode. Note: The server hosting the database must be equipped with a modern graphics processing unit (GPU). For example, a 10-year-old server cannot host an HFSQL server for intended for semantic searches. Benefit of this new feature in version 2026: Semantic search available on all the platforms 151 N e w f e a t u r e W D W B WM TECHNOLOGY Semantic search is based on vector representations (embeddings) of texts mapped in a multidimensional space. Each document or text is encoded as a vector in this space. Search text is encoded in the same way. The engine measures the distance (or similarity) between the query vector and the indexed vectors. To this end, a new index format (.vex) has been added to HFSQL. The HFSQL engine handles this new type of index just like any other standard index. The closest texts or documents in this space are returned as relevant results. Benefit of this new feature in version 2026: Power at your fingertips 152 N e w f e a t u r e W D W B WM AN EMBEDDING MODEL IN HFSQL HFSQL integrates an embedding model and a specific index adapted to semantic search. No additional installations, no connection to external modules, no additional management. HFSQL does it all for you. Benefit of this new feature in version 2026: AI-powered search 153 N e w f e a t u r e W D W B WM RELEVANCE: NUMBER OF RESULTS (TOP K) AND SIMILARITY THRESHOLD Semantic queries must include 2 parameters: • maximum number of results, • minimum similarity threshold. The number of results is the number of records that match the search. The similarity threshold indicates whether the elements found should be more or less similar.The higher the threshold, the closer the meaning. Benefit of this new feature in version 2026: Two parameters, advanced search 154 N e w f e a t u r e W D W B WM CREATE A SEMANTIC INDEX Creating a semantic index is easy. Simply declare a semantic index in the analysis, and indicate the text items to be indexed. On existing databases, an “automatic modification” is run. Benefit of this new feature in version 2026: Total automation HFSQL SEMANTIC SEARCH: AI-POWERED INDEX Wh a t ' s n e w i n W I ND E V 2 0 2 6 WE B D E V 2 0 2 6 W I ND E V Mo b i l e 2 0 2 6 32