> On Feb. 24, 2015, 12:48 a.m., Andrew Stitcher wrote:
> > There seems to be nothing wrong here, but it makes me ask what else changed 
> > to make this change necessary.
> > 
> > I'd be particularly concerned that the equivalent non-test code inside the 
> > broker and other executables didn't changing.
> 
> Steve Huston wrote:
>     1. I can check on the history of std::vector and validation checks 
> enabled in various Visual Studio versions if that would help.
>     2. The equivalent code in the actual broker uses the main(int argc, char 
> *argv[]) values directly so does not run into the std::vector check.

1. That's not really what I meant - the argv[] vector should never be 0 under 
usual circmstances as argv[0] is the exectable itself, why is the test 
infrastructure not honoring this - or when did it stop doing that.


- Andrew


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31330/#review73728
-----------------------------------------------------------


On Feb. 24, 2015, midnight, Steve Huston wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31330/
> -----------------------------------------------------------
> 
> (Updated Feb. 24, 2015, midnight)
> 
> 
> Review request for qpid and Andrew Stitcher.
> 
> 
> Bugs: QPID-6409
>     https://issues.apache.org/jira/browse/QPID-6409
> 
> 
> Repository: qpid
> 
> 
> Description
> -------
> 
> Patch to avoid taking addres and calling parse if there are no arguments.
> 
> 
> Diffs
> -----
> 
>   trunk/qpid/cpp/src/tests/BrokerFixture.h 1661591 
> 
> Diff: https://reviews.apache.org/r/31330/diff/
> 
> 
> Testing
> -------
> 
> Compile, run unit_test
> 
> 
> Thanks,
> 
> Steve Huston
> 
>

Reply via email to