Update TODO

TODO.md

@@ -11,14 +11,12 @@
     * Save last results, check if difference is significant?
     * ProgressBar: `[{elapsed_precise}]` and eta_precise?
     * Test server over udp socket instead?
+    * Run multiple threads to test performance when contested?
 * Performance
     * cpu-target=native good?
     * mialloc good?
-    * https://docs.rs/zerocopy/0.3.0/zerocopy/index.html for requests and responses?
 * bittorrent_udp
     * ParseError enum maybe, with `Option<TransactionId>`
-    * Avoid allocating in conversion to bytes, send in a mutable buffer
-      instead, which is reused over requests
     * Avoid heap allocation in general if it can be avoided?
     * quickcheck tests for conversions
     * other unit tests?
@@ -29,4 +27,7 @@
   with AVX though)
 * `sendmmsg`: can't send to multiple socket addresses, so doesn't help
 * Config behind Arc in state: it is likely better to be able to pass it around
-  without state
+  without state
+* Use `bytes` crate for bittorrent_udp: seems to worsen performance somewhat
+* Zerocopy (https://docs.rs/zerocopy/0.3.0/zerocopy/index.html) for requests
+  and responses? Doesn't work on Vec etc