Location fingerprinting is a technique for location sensing on 802.11 Wireless Local Area Networks (WLANs), using commodity WLAN cards and no additional hardware tags. Location fingerprinting is a two-phase process. First, a radio map of observed Signal Strength (SS) values from different locations are recorded during an offline calibration phase. Then, in real time, SS values observed at a users mobile device are compared to the radio map values using proximity-matching algorithms in order to infer current user locations. We present Locus, a software-only, platform-independent tool for location fingerprinting on 802.11 WLANs. Locus has an object-oriented design, and is implemented in Java with graphical display in Scalable Vector Graphics (SVG). While several proximity-matching algorithms have been proposed, very little research has evaluated their performance on existing wireless networks. Using Locus as a framework, we experimentally compared the performance of two proposed proximity-matching algorithms and also quantified the variance of observed SS values on five mobile devices. We find that in practice, due to issues such as access point occlusion from certain locations, in-building interference effects on signal strengths, calibration and signal strength detection difficulties on certain mobile platforms, the behavior of proximity-matching algorithms can be mobile platform and wireless network dependent, and can not always be generalized.