On 02/04/2013 03:53 PM, Eric Blake wrote:
On 02/01/2013 09:41 PM, Stefan Berger wrote:
> On 02/01/2013 09:16 PM, Eric Blake wrote:
> I definitely prefer the ARRAY_CARDINALITY() instead of npos in my
> version of the test. Beyond that, it looks like your test copied and
> pasted so that the set vs. clear tests were independent, while mine
> interleaved the two to traverse over the same set. Coverage wise,
> either one of our tests gets just as much coverage of the added code;
> at least, I didn't see anything added by your test8() that wasn't
> present in my revised test4().
> Maybe we can merge the two for a joint effort :-) -- leaving test 4
> untouched.
Anyone else have an opinion?
ACK.