A dialogue system (DS), also known as a conversational agent (CA) or chatbot (CB), is a computer program designed to communicate with people using natural language in the form of text or speech. They can answer questions, give information, or help users do things like order food or find directions. Since the 1960s, DSs have made a huge progress. Early systems such as ELIZA were based on simple rule sets, whereas current models employ generative artificial intelligence (AI) to support complex multi-modal interactions. This paper gives a detailed review of their history. It also presents a clear taxonomy and classification of the main changes in DSs. It integrates under-explored areas such as transitional hybrid architectures, agent personality, generalization, multi-modality and domain adaptation. The paper also compares key developments and studies the role of open-source DSs in shaping the future of generative AI. The survey aims to differ from earlier studies by giving clear technical explanations, examples of DS, and new visual representations.