04 Open-source ecosystem
How Project V, V2Fly, and Xray fit together
Projects, cores, clients, and node protocols are different layers. Before following a guide or reporting an issue, identify which layer it concerns. In particular, if an error occurs during startup, the name shown in the interface does not necessarily identify the core that is actually running.
- Project lineage
- Project V
- Core families
- V2Fly / Xray
- Graphical clients
- v2rayN / v2rayNG / v2flyNG
Quick reference →
-
Understanding modular proxy design through Project V
V2Ray originated with Project V. It separates inbound connections, outbound routing, and the rules that direct requests into distinct configuration components. An inbound accepts requests from a browser or another app and passes them to the core; routing selects an outbound based on conditions; the outbound handles direct connections, proxying, or other actions.
These three components also map to common symptoms. If the core receives no connection, check local proxy access. If a connection reaches the core but takes the wrong route, check rule order. If the request reaches the expected outbound but still fails, check server settings and network reachability. Identifying the layer is more reliable than guessing based only on a webpage that will not load.
-
V2Fly and Xray have different compatibility ranges
The V2Fly community continues the V2Ray core project. Xray branched from the same code lineage and has evolved independently; each is maintained by its own community. Their configuration concepts often overlap, but supported transports, security layers, and fields can differ. Using the same protocol does not guarantee that both cores accept the same complete set of settings.
If a node imports but the core will not start, first check which field the log identifies. Then determine whether the core does not support that option or the client omitted it when generating the configuration. Note the original core before switching. To compare results, keep the same node and network so you can tell which layer caused the difference.
-
Open protocol implementations do not make settings interchangeable
VMess, VLESS, Trojan, and Shadowsocks all have implementations in the open-source proxy ecosystem. A protocol defines how the two ends exchange data; the transport and security layer determine how that data is carried and protected. For example, the VLESS protocol name alone does not mean TLS is configured, and TLS is not the same thing as a transport.
When reviewing a shared node configuration, check the protocol first, then compare the transport and security layer with the corresponding fields and server settings. Copying a single setting from another guide can make two previously matching configurations inconsistent. Open protocols make implementations easier to compare and troubleshoot, but that does not mean a subscription or third-party service is part of the client project.
-
Three open-source clients for graphical configuration
v2rayN is a desktop graphical client for managing nodes, subscriptions, and routing, with a choice of supported core. v2rayNG uses Xray and provides mobile subscriptions and VPN connections; v2flyNG uses V2Fly and is suited to mobile configurations that need confirmed V2Fly compatibility. All three are open-source clients: their interfaces organize settings, while the underlying core establishes the connection.
Subscription services and clients are separate. Installing a client does not provide nodes, and a client opening successfully does not mean a subscription is still valid. Remote errors, expired nodes, and missing app permissions require checking the subscription source, node configuration, and device settings respectively. Identify which side produced the error before choosing the relevant part of a guide.
-
Clients, cores, and subscriptions update on different schedules
A client update may change the interface, system support, or packaging. A core release may add fields or fix connection behavior. Refreshing a subscription only syncs the node list provided by its source. A successful subscription update does not replace a software upgrade, and reinstalling the client will not restore an expired subscription.
Before upgrading, back up custom routing and configuration, and note the current core and connection mode that works. After upgrading, test the original node first, then restore optional settings one at a time. If a regression occurs, record when it started, the platform and architecture, selected core, reproducible steps, and redacted logs. That is more useful for diagnosis than a screenshot that only shows a failed connection.