This reverts commit 3682d9bd5b.

This is required when moving the buttons_init() function to the
lcd_init() function

Since the class "MainMenu" was used within a static variable the
initialization of the object (constructor call) was done before Arduino
library startup. It always caused a crash when using AVRStudio with
JTAG debugger (caused from calling the LCD initialization / the lot of
I/O work / the stack used during this calls). By moving the LCD_INIT
out of the constructor and using an explicit call inside of Arduino
setup() implementation immediately fixed all problems and the JTAG
debugger runs fine.

"i" runs from 0 to 4 but "add_homeing" array size is 3 only. On the
other hand the calculated value gets overwritten by either one of the
if choice.

Since a removal or insert of the sd card takes a long time in firmware, just reinitializing the lcd does not hurt.
actually, it solves a problem with the UltiControlle

see https://github.com/ErikZalm/Marlin/issues/148

Changed SLOWDOWN. IF this does not work ok OLD_SLOWDOWN is the old algo.

Fixed small bug in arcs
M105 prints now actual_temp/target_temp

Moved the PID_dT in the Ki and Kd calculation from the configuration.h to temperature.cpp

Open Hardware board currently used by the group build of http://tvrrug.org.uk
The board can be found at http://solderpad.com/folknology/open-motion-controller/

Signed-off-by: Michel Pollet <buserror@gmail.com>

+ Removed most explicit pathnames, use the standard make "VPATH" to let
make find the files for itself.

+ Added a "hardware variant" variable that allows compiging Sanguino and
Gen7 as well as "generic" arduino.

+ Allows overriding the MOTHERBOARD define from the Makefile

+ Removed the 'preprocessor' bit that wasn't needed, now just "include" the
files that are needed, since it allows gcc to show the right file/line
when displaying error/warnings.

+ Uses gcc's own dependency generator to generate the .d files, and
and include these instead of self-patching the makefile

Signed-off-by: Michel Pollet <buserror@gmail.com>

One array was too short. This had nothing to do with long filenames, other than if they were 12 characters exactly, which could only happen if the extension and the text before were filled completely