> On Sept. 5, 2014, 5:01 p.m., Dominic Hamon wrote: > > 3rdparty/libprocess/src/encoder.hpp, line 124 > > <https://reviews.apache.org/r/24984/diff/2/?file=676526#file676526line124> > > > > there are a few other places in the code that work with > > Transfer-Encoding. check http.hpp and process.cpp. also http_tests.cpp > > checks for this header.
I took a look at all files you mentioned. In all cases "chunked" is used in "PIPE" mode: declared in http.hpp, implemented in process.cpp, tested in http_tests.cpp. - Nikita ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/24984/#review52470 ----------------------------------------------------------- On Sept. 5, 2014, 5 p.m., Nikita Vetoshkin wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/24984/ > ----------------------------------------------------------- > > (Updated Sept. 5, 2014, 5 p.m.) > > > Review request for mesos, Benjamin Hindman and Dominic Hamon. > > > Bugs: MESOS-1625 > https://issues.apache.org/jira/browse/MESOS-1625 > > > Repository: mesos-git > > > Description > ------- > > As mentioned in MESOS-1625 current HTTP Chunked Encoding implementation can > break clients (e.g. Tornado). As I see no need in Chunked-Encoding because we > know content size beforehand - I propose to use Content-Length. > > > Diffs > ----- > > 3rdparty/libprocess/src/encoder.hpp 9c5aa81 > > Diff: https://reviews.apache.org/r/24984/diff/ > > > Testing > ------- > > `make check` and manual testing with golang language native binding > > > Thanks, > > Nikita Vetoshkin > >
