> On June 17, 2016, 10:07 p.m., Benjamin Mahler wrote: > > While it's nice to document that this is internally only doing a > > receive-side shutdown, I'm left wondering why. And now that it's > > documented, the restriction looks intentional but the intention is absent. > > It would be of great help to the reader of this code to include it! > > > > If we're hesistant to change it to `shutdown(int how)` to match the POSIX > > API and avoid confusion, it seems to be only because there is a potential > > reason for only allowing `SHUT_RD`. Let's get context from benh and joris > > to figure that out. If it was intentional, let's include the reason in your > > comment, otherwise let's allow the caller to specify `how`. > > Neil Conway wrote: > I agree that documenting the current situation (receiver-side shutdown) > is not ideal, but I think it is an improvement over the status quo. I created > a JIRA to track either (a) making this consistent with BSD API, or (b) > documenting why it is inconsistent: > https://issues.apache.org/jira/browse/MESOS-5658 -- I'd be happy to revisit > this once I've shipped some other higher-priority stuff. Sound good?
Having the TODO sounds good, as it signals the reader that this is not the intended end state. Last thing is the following was a bit confusing for me in that it seems to imply different behavior rather than just being more restrictive (`how` is always SHUT_RD). - Benjamin ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/46026/#review138325 ----------------------------------------------------------- On June 20, 2016, 12:35 p.m., Neil Conway wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/46026/ > ----------------------------------------------------------- > > (Updated June 20, 2016, 12:35 p.m.) > > > Review request for mesos, Benjamin Hindman, Benjamin Mahler, and Joris Van > Remoortere. > > > Repository: mesos > > > Description > ------- > > This is slightly confusing, because it doesn't match the > semantics of the shutdown(2) POSIX function. > > > Diffs > ----- > > 3rdparty/libprocess/include/process/socket.hpp > 4cb49680d1304899a4ee675ac07379e51d9c55b1 > > Diff: https://reviews.apache.org/r/46026/diff/ > > > Testing > ------- > > > Thanks, > > Neil Conway > >