Windows binary saga

[Dirk Hohndel]

In tonights installment of the saga...

Lots of progress.

I massaged MXE to go from MinGW 5.2.0 back to 4.9.2 (the compiler used in
the official Qt5.5 builds). That turned out to be a bit of a pain but hey,
I think I have it working.

Rebuilt all the libraries.

I can now create a Windows installer for Subsurface that actually mostly
works. Subsurface starts up, reads local files, etc.

I notice that even though this is using Qt5.5 it still identifies itself
as "Windows" not "Windows 10" on my brand new Windows 10 laptop.
Thiago, any idea?

The first time I tried to use cloud storage, Subsurface crashed. Oops.

I'll try to debug this tomorrow now that I have a Windows laptop with
working development environment including gdb - I need to do some tweaking
to my install paths but I think this will work just fine (as I had hoped).

When Subsurface crashed it left the local git cache in a damaged state.
That brought up a question and an idea:

a) where the heck is that cache on a Windows 10 machine and how do I get
there? Some searching with Explorer did show any obvious spots.
Lubomir, any suggestions?

b) if we detect that the local cache is inconsistent and unusable, should
we just delete it? I mean, even if it contained data that wasn't synced to
the server, if libgit2 can't read it I guess it's toast so instead of
giving a semi-useful error message it seems to be a much smarter idea to
just delete it and sync it back from the cloud, right?

Anyway, I'll stop here for today. Feel free to play with it - it's in
downloads/daily as subsurface-4.4.97-26-ge66f0895c68c.exe

I also notice that this installer is 12MB bigger than what we used to
have. Not quite sure why - more to investigate.

/D