On 6/3/14, 9:29 AM, Mike de Boer wrote:
Nope, you got me there - I generalised too easily. This statement is based on
personal experience, not science.
I'm not looking for science, necessarily. I'm looking for an
understanding of the problems we're trying to solve.
My basic issue is that from my perspective what I'm seeing is a request
that everyone change their workflow because the new setup is better
without a clear description of the problems that the new setup is trying
to solve. I'm sure there are problems that need solving here; I'm just
trying to get a handle on what they are.
To preserve backward compat, all the old function names of XPCShell assertions
do still work. I was planning to do the same with Mochi.
Sure. The big question is what we do moving forward for new tests.
The way I see it, a test harness has the user-facing API (what test
writers call) and it has a backend that tracks/reports the failures.
Improvements to the backend are generally a no-brainer in my book.
Changes to the user-facing API are something we should be more careful
with...
-Boris
_______________________________________________
dev-platform mailing list
dev-platform@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-platform
