This paper presents Gia, a new peer-to-peer (P2P) file-sharing system that improves the scalability of Gnutella-like systems. Gnutella's original design uses flooding to search for files, which is inefficient and not scalable. Gia addresses this by incorporating dynamic topology adaptation, active flow control, one-hop replication, and biased random walks. These mechanisms allow Gia to dynamically adjust to the heterogeneity of peer capacities and improve system performance significantly. Gia is designed to be decentralized and unstructured, similar to Gnutella and KaZaA, but with enhanced scalability. It retains the simplicity of an unstructured system while offering vastly improved scalability. The design of Gia builds on previous research, including the use of random walks instead of flooding, topology adaptation algorithms, and token-based flow control. The paper compares Gia to other P2P systems, including Distributed Hash Tables (DHTs), and argues that while DHTs offer better performance for exact-match queries, they are less effective for keyword searches and are more sensitive to node churn. Gia, on the other hand, is more robust in the face of transient nodes and supports general search facilities, which are important for P2P file sharing. The paper also discusses the performance of Gia through simulations, showing that it can handle much higher aggregate query rates than Gnutella and other systems. The results indicate that Gia's design, which combines dynamic topology adaptation, flow control, one-hop replication, and biased random walks, leads to a significant improvement in system capacity. The simulations show that Gia's performance is three to five orders of magnitude better than Gnutella and other attempts to improve it. The paper also discusses the impact of heterogeneity on system performance, showing that Gia's design is more effective when nodes have varying capacities. The results indicate that Gia's ability to adapt to node heterogeneity is a key factor in its improved performance. The paper concludes that Gia is a promising solution for scalable P2P file-sharing systems, offering a balance between simplicity and performance.This paper presents Gia, a new peer-to-peer (P2P) file-sharing system that improves the scalability of Gnutella-like systems. Gnutella's original design uses flooding to search for files, which is inefficient and not scalable. Gia addresses this by incorporating dynamic topology adaptation, active flow control, one-hop replication, and biased random walks. These mechanisms allow Gia to dynamically adjust to the heterogeneity of peer capacities and improve system performance significantly. Gia is designed to be decentralized and unstructured, similar to Gnutella and KaZaA, but with enhanced scalability. It retains the simplicity of an unstructured system while offering vastly improved scalability. The design of Gia builds on previous research, including the use of random walks instead of flooding, topology adaptation algorithms, and token-based flow control. The paper compares Gia to other P2P systems, including Distributed Hash Tables (DHTs), and argues that while DHTs offer better performance for exact-match queries, they are less effective for keyword searches and are more sensitive to node churn. Gia, on the other hand, is more robust in the face of transient nodes and supports general search facilities, which are important for P2P file sharing. The paper also discusses the performance of Gia through simulations, showing that it can handle much higher aggregate query rates than Gnutella and other systems. The results indicate that Gia's design, which combines dynamic topology adaptation, flow control, one-hop replication, and biased random walks, leads to a significant improvement in system capacity. The simulations show that Gia's performance is three to five orders of magnitude better than Gnutella and other attempts to improve it. The paper also discusses the impact of heterogeneity on system performance, showing that Gia's design is more effective when nodes have varying capacities. The results indicate that Gia's ability to adapt to node heterogeneity is a key factor in its improved performance. The paper concludes that Gia is a promising solution for scalable P2P file-sharing systems, offering a balance between simplicity and performance.