> If you replaced the read-string with void, and the speed remained the same,
Sorry, I worded this terribly. What I meant to say is when I replaced read-string with void the performance across the two CHICKEN versions was similar; I to longer saw the large discrepancy. But you bring up a good test; read-string is also near instant from a local file for me so yes, it's more likely the difference is somewhere in http retrieval. Void is doing no actual reading from the socket and therefore also not exercising a key part of the call, so in using it I inadvertently cut out two factors instead of one. Thanks, Diego
