On Jul 9, 2014, at 11:54 AM, Jacques Nadeau <[email protected]> wrote:

> When there is disagreement, we need to work through it.

I agree - and here we are, working through it!

I’m sorry I made a commit after you’d expressed objections. That isn’t allowed 
in the Apache process, so I won’t do it again.

But I hope it’s clear that I was doing it in good faith. I’d discovered a bug, 
a test case, and a fix, so I checked it in. One extra test case passing is 
progress. I suspected that it would break your use of the code, so I started a 
dialog, trying to find out what exactly your problem was. I figured, heck, you 
could even use my test case as a starting point for yours.

I know you’d like to have the discussion about test cases in a separate thread, 
but they really are central to this issue. If I can’t change a piece of code 
because it breaks some code that I can’t see, it will slow the process down. If 
I’d not been proactive and had just committed this bug fix, you probably 
wouldn’t have noticed until a month or two down the line, and then we’d all be 
scratching our head. But if there had been a working test case, we’d not even 
be having this discussion. I’d not have made the commit, and in fact, I would 
not have made the botched merge that probably caused the whole thing in the 
first place.

Julian

Reply via email to