Reconfigurable Intelligent Surfaces (RISs) have emerged as a pivotal technology for the Sixth-Generation (6G) communication system, showcasing the ability to configure wireless environment dynamically. Acknowledged as a breakthrough in enhancing network coverage, augmenting system capacity, and facilitating advanced applications such as Integrated Communication and Sensing (ISAC), RISs present a concrete approach to molding the future network evolution. The advancement of RIS technology necessitates a departure from idealistic assumptions and oversimplifications, compelling a progression towards models that more accurately reflect the physical attributes of hardware and the characteristics of propagation. In this paper, we delve into the practical constraints and limitations of current RIS design methodologies, conducting a comprehensive analysis based on the latest technological research advancements and product realizations. Our exploration is broad-ranging, encompassing the engineering challenges of single-point RISs, such as hardware impairments, intricacies of algorithm design, frequency spectrum-specific difficulties. A concentrated discourse is presented on novel near-field channel designs, the restrictions imposed by low-bit quantization, and the intricacies of amplitude-phase correlation constraints. This discussion aims to unearth the challenges, opportunities, and paradigmatic shifts induced by the practical deployment of RISs. The deployment challenges, networking dilemmas, simulation, and product evaluation is provided for RISs in large-scale networks from a broader system perspective. Furthermore, this paper highlights the critical need for accelerated efforts towards the commercialization of RISs. We explore the practical application revolution of RISs, encompassing engineering aspects and standardization processes. Our discussion aims to establish a foundational framework for introducing RISs into the market, acknowledging their significant potential as a game-changing technology in 6G communications.