Code review comment for lp:~unity-api-team/unity-scopes-shell/scope-harness

On Wed, 2015-03-18 at 10:17 +0000, James Henstridge wrote:
> Rodney: why are you marking this MP as needs-fixing if it is the other
> one that you have an issue with?
>
> Yes, size_t mangles to a different symbol on different architectures.
> But the branch fixing the symbols addresses that. This is hardly the
> only library that hits this 32-bit/64-bit difference.

This is the branch which introduces the size_t (and this library, and
the symbols file should really be part of this branch, not a separate
one). The correct way to fix this was to use an int and wrap the size_t
exposure out of our own API, using error checking to prevent negative
values or such.

We do have the same issue in other libraries, but that is not an excuse
to keep reintroducing the problem in new libraries we write.