[phobos] I think we need to make an emergency release

[Don Clugston]

David suggested we need an emergency release for the array appender
bug, and I'm inclined to agree with him. My personal experience is
that it makes 2.048 unusable, and I've had to go back to 2.047. Here's
the bug:

Bug 4681 Appender access violation
This is a really horrible memory corruption regression, caused by svn
1813. Is there any reason we shouldn't just roll that back?

---
BTW, one compiler bug popped up in the last release, too. Fairly
obscure though. I created a patch:
4655 Regression(1.063, 2.048) goto to a try block ICEs

I also made a patch for this compiler bug from three releases back:
4302 Regression(1.061, 2.046) compiler errors using startsWith in CTFE

All other known compiler regressions were introduced in 2.041 or earlier.