Cool, so I think the Java version is hovering around 700us, so its clear .NET may have some core performance advantages.
There is actually a lot of room to make this even faster, but not sure if its even worth going down that road. Things like removing the ngram parsing, better index structure, bailing out of bad patterns, etc, etc. Any word on updated data? I think we may be on our own with updating the device data. Someones got to do it. Is it just a matter of keep tabs on all the manufactures and upcoming releases? ________________________________ From: eberhard speer jr. <[email protected]> To: [email protected] Sent: Wednesday, July 3, 2013 3:23 AM Subject: DeviceMapClient Performance -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, made a few performance tweaks to the .Net version of DeviceMapClient (managed to cut out a few null-checks and a couple of ifs) average time per user-agent string is now consistently 0.10 milliseconds, accuracy around 96%. Nice ! esjr -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.8 (MingW32) Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iQEcBAEBAgAGBQJR09F8AAoJEOxywXcFLKYcM1kIAJyLlVarAH1zoivdHSIeaFWg l374J68ZP83FxfQRoVC8SiArPyMyJh1Wg/eApMF7x4MlBMUuvvZWSlvkehfypvyg jstrtWshzPgfnp0VQNtZ3jStNZhCwf9pr/S+MgRabso4VtSEhK9yPT6dEXjm8dho QCx7VI1UuiYMcyNiYw20EEaHio3Lx5N12GSh4hqmJxnsl1Cc4HKgzY739mu84RPu Izn4BDF0Q+t2UdVn1h8jK1S/wu6tSvV8Mw4rFX9KNRoJrSgMkrZkLMMoSooPrX+2 Ht7jnxvoNhI8tuPfLpiHjS8tM88AhRzIVeK7hWz4fWKiZ8N3uSQBhI1jUdLmwMQ= =hiIt -----END PGP SIGNATURE-----
