Thanks for the explanation. I'd read that thread but thought the change was supposed to only modify the behavior when explicitly specifying the --debugger option. That also seems to be the gist of the original RedHat bugzilla request that prompted the thread: https://bugzilla.redhat.com/show_bug.cgi?id=1165793
> The --debugger option and -O extdebug do the same thing This might not be obvious.. I had assumed that --debugger does two things: load the debugger profile *and* turn on extended debugging features, from the description: Arrange for the debugger profile to be executed before the shell starts. Turns on extended debugging mode (see the description of the extdebug option to the shopt builtin below). Maybe the description of the extdebug option to the shopt builtin can be modified with something like If set as part of shell invocation, the debugger profile is executed before shell startup. If the profile cannot be executed, the option is not be enabled. On Tue, Apr 26, 2016 at 9:39 PM, Chet Ramey <chet.ra...@case.edu> wrote: > On 4/25/16 10:45 PM, Grisha Levit wrote: > > This seems new for 4.4. Having extdebug in BASHOPTS seems to change the > > invocation behavior to mimic that of having specified the --debugger > > option, which I don't think is usually the intention: > > It's always been the case. The --debugger option and -O extdebug (which > is the same as putting extdebug in BASHOPTS) all do the same thing and > try to start the debugger. Bash-4.3 was silent when it couldn't open the > debugger start file. Bash-4.4 warns the user when it tries to start > the debugger and can't open the start file. This change went in back in > November 2014 as the result of this discussion: > > http://lists.gnu.org/archive/html/bug-bash/2014-11/msg00137.html > > -- > ``The lyf so short, the craft so long to lerne.'' - Chaucer > ``Ars longa, vita brevis'' - Hippocrates > Chet Ramey, ITS, CWRU c...@case.edu > http://cnswww.cns.cwru.edu/~chet/ >