On Thu, 2013-10-10 at 17:43 -0700, Kevin Gillette wrote:
> I'm convinced that nobody *should* be relying on the output of %#v for
> any kind of exchange format, instead favoring one of the encoding
> packages for anything "real", yet it would still seem to fall under
> the Go 1 guarantee.
Agreed on both of those. The most likely place conflicts will arise is
in testing.