> On Nov. 11, 2014, 7:49 p.m., Vinod Kone wrote: > > src/authentication/cram_md5/authenticatee.hpp, line 390 > > <https://reviews.apache.org/r/27493/diff/5/?file=756742#file756742line390> > > > > why the factory method?
AFAIK typed tests need a factory to function. We introduced such factory to the isolators as well when we switched over to typed tests for the same reason, if I recall correctly. > On Nov. 11, 2014, 7:49 p.m., Vinod Kone wrote: > > src/authentication/cram_md5/authenticatee.hpp, line 33 > > <https://reviews.apache.org/r/27493/diff/5/?file=756742#file756742line33> > > > > C headers should come first. If I did that without further changes, GCC would bomb out when including this file in an implementation that does not previously define standard C types (specifically size_t). To fix this, I will swap c-headers up again but prepend them with #include <cstddef> (which would be the C++ variant) or #include <stddef.h> (which would be the c variant). Actually, prepending it with cstddef would again result into this very issue you are commenting on, so stddef.h it is :) .... (loud thinking here). > On Nov. 11, 2014, 7:49 p.m., Vinod Kone wrote: > > src/authentication/cram_md5/authenticatee.hpp, line 58 > > <https://reviews.apache.org/r/27493/diff/5/?file=756742#file756742line58> > > > > virtual Try<Nothin> initialize( > > const process::UPID& clientPid, > > const Credential& credential); > > > > Also, why introduce initialize()? I wanted to differentiate between authenticate and a preparation to allow for a more flexible interface. However, I can just was well move all currently initialize specific code into authenticate to simplify -- guess simple is good :) . - Till ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/27493/#review60837 ----------------------------------------------------------- On Nov. 10, 2014, 1:50 p.m., Till Toenshoff wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/27493/ > ----------------------------------------------------------- > > (Updated Nov. 10, 2014, 1:50 p.m.) > > > Review request for mesos, Adam B, Kapil Arya, and Vinod Kone. > > > Bugs: MESOS-2001 > https://issues.apache.org/jira/browse/MESOS-2001 > > > Repository: mesos-git > > > Description > ------- > > see summary. > > > Diffs > ----- > > src/Makefile.am 443554a > src/authentication/authenticatee.hpp PRE-CREATION > src/authentication/cram_md5/authenticatee.hpp 3088a77 > src/examples/test_authentication_modules.cpp PRE-CREATION > src/examples/test_authenticator_module.cpp 4398d71 > src/module/authenticatee.hpp PRE-CREATION > src/module/manager.cpp 0d71e33 > > Diff: https://reviews.apache.org/r/27493/diff/ > > > Testing > ------- > > make check > > Note: All three CRAM-MD5 Authenticatee patches need to get applied before > running make check! > > > Thanks, > > Till Toenshoff > >
