Ruby Issue Tracking System

10/23/2009

07:05 AM Ruby Feature #2250: IO::for_fd() objects' finalization dangerously closes underlying fds

=begin
In comment 2 Nobuyoshi Nakada writes:
> This is a patch to add :autoclose option to IO.for_fd

Thanks! IIUC, your patch requires one to explicitly set
:autoclose => false to avoid this behavior. However, I'd
argue tha... pilcrow (Mike Pomraning)

10/22/2009

12:43 PM Ruby Feature #2250 (Rejected): IO::for_fd() objects' finalization dangerously closes underlying fds

=begin
1. Expected behavior:

An IO object created by IO::for_fd(a_socket_fd) should *not* call close(2) on its underlying fd upon finalization. The object did not allocate the fd, and so cannot safely nor politely close it.

... pilcrow (Mike Pomraning)