The text discusses the evolution of OLAP systems and the communication protocol between a standalone semantic layer and data consumption layers, specifically focusing on SQL as a potential solution. Historically, MDX was used for OLAP servers, but it has largely been replaced by DAX within Microsoft's ecosystem. However, outside of Microsoft, SQL remains a dominant protocol due to its widespread adoption and maturity in the database world. Despite being well-suited for this purpose, SQL faces a significant challenge: its bottom-up evaluation flow is incompatible with multidimensional primitives like measures, which are crucial for analytical queries. To address this issue, rewriting queries to align with bottom-up evaluation is necessary, but it requires careful transformation to avoid incorrect aggregation issues. The text concludes that rewriting queries is not new and can be efficiently achieved using E-Graph theory, as demonstrated by Cube's open-source semantic layer engine.