We weren't seting the _DO_HADDOCK variables early enough.

I don't think it's necessary to build ghc-pkg that early.

Dependency problem was discovered by int-e.

I've also added some comments about what's going on.

Also a couple of other fixes and sanity checks along the way.

We now leave making installers to the Haskell Platform.

The file no longer exists

The build system thought that $(INSTALL_DYNLIBS) contained things
like "terminfo", but actually it contains things like
"libraries/terminfo".

In particular, this means that GHCi will use DLLs, rather than loading
object files itself.

We now put a handful of modules in a separate DLL.
For now the list is hand-written, but we could automate it in the
future.

Patch from Stephen Blackheath.

This is sometimes needed when cross-compiling, as some packages may be
built in stage 0 but not stage 1.

In order to make everything work out, this also removes the requirement
that the build-dirs are in dependency order

We now use the 'host' to determine whether stage0 builds terminfo,
and 'target' to determine whether stage1 does.

We don't build it if the platform is either Windows or ios.

This makes the build system a little simpler, and in particular
will make it easier to handle the changes needed for cross-compilation.

This means we don't define them multiple times

The hsc2hs, alex and happy options variables are now also
non-way-specific, as the files are shared between all ways.

From StephenBlackheath