>> Kamesh Jayachandran wrote on Fri, Mar 11, 2011 at 18:12:21 +0530: >> > Unless there are no objection I will commit this patch post I >> > receive a r1075802 regression fix(and testcase) from either Daniel >> > Becroft or Arwin. >> >> Hold on, are you suggesting to commit the two thousand, six hundred and >> sixty-three line diff that started this thread?
>So what if it's long? I'd be much more worried if this diff was shorter. >It needs to make changes in lots of places. >It does a lot of things, including running every update during tests >with --dry-run, and checking whether the dry-run modified the working >copy, and making the test fail if it did. >And it checks if the dry-run output matches the output for normal update. >If all the tests pass with this, that's fine, isn't it? >Because how else would you do it? Not test it, just to make the diff shorter? Thanks a lot stefan for taking closer look at this patch. You voiced everything that I have to say about. With regards Kamesh Jayachandran
