minor comments, otherwise LGTM. https://codereview.appspot.com/11764044/diff/1/src/server/ct-rfc-server.cc File src/server/ct-rfc-server.cc (right): https://codereview.appspot.com/11764044/diff/1/src/server/ct-rfc-server.cc#ne... src/server/ct-rfc-server.cc:351: LOG(INFO) << "hash = " << b64hash << " size = " << qmap["tree_size"]; Is this leftover debugging or meant to be printed for every request? (In general it may make sense to print the incoming request path and body for all incoming requests) https://codereview.appspot.com/11764044/diff/1/src/test/run_log_server.sh File src/test/run_log_server.sh (right): https://codereview.appspot.com/11764044/diff/1/src/test/run_log_server.sh#new... src/test/run_log_server.sh:32: # echo "$HASH_DIR doesn't exist, creating" Simply remove this code? https://codereview.appspot.com/11764044/diff/1/src/test/run_log_server.sh#new... src/test/run_log_server.sh:40: echo "Starting CT server with trusted certs in $HASH_DIR" HASH_DIR seem to be completely gone from this file, so it should probably not be here as well. https://codereview.appspot.com/11764044/diff/1/src/util/util.h File src/util/util.h (right): https://codereview.appspot.com/11764044/diff/1/src/util/util.h#newcode41 src/util/util.h:41: Empty line at EOF.
https://codereview.appspot.com/11764044/diff/1/src/server/ct-rfc-server.cc File src/server/ct-rfc-server.cc (right): https://codereview.appspot.com/11764044/diff/1/src/server/ct-rfc-server.cc#ne... src/server/ct-rfc-server.cc:347: void GetProof(server::response &response, const uri::uri &uri) { response should really be a pointer here since it's an output.
Pushing... https://codereview.appspot.com/11764044/diff/1/src/server/ct-rfc-server.cc File src/server/ct-rfc-server.cc (right): https://codereview.appspot.com/11764044/diff/1/src/server/ct-rfc-server.cc#ne... src/server/ct-rfc-server.cc:347: void GetProof(server::response &response, const uri::uri &uri) { On 2013/07/24 14:56:29, Al Cutter wrote: > response should really be a pointer here since it's an output. I totally agree, however, cpp-netlib has it as a reference and so I am following the local idiom, even though I don't like it. https://codereview.appspot.com/11764044/diff/1/src/server/ct-rfc-server.cc#ne... src/server/ct-rfc-server.cc:351: LOG(INFO) << "hash = " << b64hash << " size = " << qmap["tree_size"]; On 2013/07/24 14:20:08, Eran wrote: > Is this leftover debugging or meant to be printed for every request? > (In general it may make sense to print the incoming request path and body for > all incoming requests) There's a VLOG to do that, I will remove this. https://codereview.appspot.com/11764044/diff/1/src/test/run_log_server.sh File src/test/run_log_server.sh (right): https://codereview.appspot.com/11764044/diff/1/src/test/run_log_server.sh#new... src/test/run_log_server.sh:32: # echo "$HASH_DIR doesn't exist, creating" On 2013/07/24 14:20:08, Eran wrote: > Simply remove this code? It's slightly tedious to work out, and we may well want to reinstate certificate hash directories, so my inclination was to leave it for reference. https://codereview.appspot.com/11764044/diff/1/src/test/run_log_server.sh#new... src/test/run_log_server.sh:40: echo "Starting CT server with trusted certs in $HASH_DIR" On 2013/07/24 14:20:08, Eran wrote: > HASH_DIR seem to be completely gone from this file, so it should probably not be > here as well. Heh, oops. I'll delete this. https://codereview.appspot.com/11764044/diff/1/src/util/util.h File src/util/util.h (right): https://codereview.appspot.com/11764044/diff/1/src/util/util.h#newcode41 src/util/util.h:41: On 2013/07/24 14:20:08, Eran wrote: > Empty line at EOF. Done.