https://bugs.kde.org/show_bug.cgi?id=487296
Ivy Basseches <imb2...@columbia.edu> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |tz2...@columbia.edu
--- Comment #4 from Ivy Basseches <imb2...@columbia.edu> ---
That makes sense. I think you're right in that there doesn't seem to be a good
reason to isolate the "std" fds, as those are treated differently only by
convention. Like you mentioned, my proposed fix is not really solving the
issue. There's just not a great way of reliably determining if a std fd is open
or not, so I don't love the --track-fds=yes option as it stands.
If we're implementing your suggestion, there are certain potentially useful
cases that this option would now be ignoring, like inheriting a pipe, where one
might want to know if the child has closed its end of the pipe (as is good
practice). I think this is fine as long as the "all" option exists. The user
can make an appropriate decision about what they'd want to track.
I implemented your change along with a minor documentation change for the
--track-fds=yes/all options in m_main.c, will attach.
--
You are receiving this mail because:
You are watching all bug changes.
