More new features from @Josef_Prusa 's team - automatically correcting for sloppily built printers and not only lining up the bed with the rest of the machine, but also the axes to each other!
http://prusaprinters.org/first-printer-correct-geometry-axes/
Neat, but seems like an exceptionally complicated and specialized approach, compared to switching to a capacitive leveling sensor and building some kind of perpendicularity positive registration feature into the frame.
I have one MK2 on the test bench and it seems to be a very handy feature.
Sounds like too many things can go wrong. Just put a laser on the head and align it.
We’re too stuck on fixing everything in software, when proper hardware won’t have any issues.
IMHO lasers are not suitable for many users incl. schools, beginner’s kits. Prusa i3 MK2 is still close to its RepRap roots.
While I love what Prusa is doing and applaud them, it does sort of seem like solving a problem that mainly plauges the threaded rod i3 frame 
I also imagine that one of the main skews on an i3 (angle between Z and Y plane could not be compensated for this way.
@Nathan_Walkner My point is, fixing a fundamentally unreliable design (inductive probe hitting small PCB targets on a printer architecture that allows huge misalignment) with an elaborate algorithm in a printer-specific firmware fork is a BAD APPROACH compared to making the hardware more robust. A capacitive sensor would trigger anywhere on the print surface rather than needing to tag specific targets on the PCB heater underneath. (You can then measure the edge of the build plate if you still want to do skew correction.) And good hardware designs physically don’t allow Cartesian printers to be constructed so badly out of alignment that XY perpendicularity skew is a meaningful issue.
This is only useful for a tiny subset of people with a specific poorly-assembled printer. It’s too specialized to benefit the community at large. (Unlike something like mesh-leveling or delta auto-calibration which is sensor-agnostic and heatbed-agnostic.) So, in my opinion, it’s a one-off hack to make up for a hardware issue. That’s fine, really, lots of printer manufacturers use custom firmware to make up for hardware quirks, but it doesn’t really do much for anybody else.
To my understanding their fully built printers don’t have any issues, but when the kits are loosely built by end users which are not careful enough, things can go wrong. I like this automatic software fix.
All I’ve gotta say is: someone is actually adding useful features to 3d printing. Development and deployment of things that make 3d printing easier for people have all but stopped. I applaud @Josef_Prusa for development of this feature, which I’m sure will help lots of people get more accurate prints.
@ThantiK you must not be using RepRapFirmware or Repetier.
Excellent. This is what commercial machine tools do; geometry correction in software. The difference is, they require another instrument (a ball-bar, costing thousands) to measure the machine’s geometry.