Now It’s not easy to trace a +v 0v short on a board if it’s the main chip and if it is you have to remove it.
After the following there is no longer a short …
Let us just acknowledge that there are a LOT of pins on an MCP23017…
( don’t worry I have replacements…)
Wow 
Did it melt off?
I just was asking myself why @wyleu would want so many PCBs …
Now i’m starting to understand …
The best!
BTW, @C0d3man, a new competitor has entered the batterfield, and @wyleu’s tablecloth should be considered a hard rival!! OMG … and it’s soooo british!! 
This is the 12C over sockets machine’a encoders that I have been plugging and unplugging fairly viciously and this appears to be the result. The MCP23017 seems to fail with a pretty solid short circuit from +Vdc to 0V and it’s removal is ‘difficult’. I’ve used the old trick of levering from one end and continually melting the solder on the pins which I have used before ( I’ve even reused 16 pin CMOS IC’s after doing this in student times when getting hold of chips was a little more involved than amazon…)
IT’s the first time I’ve tried it with a 28 pin beastie and it’s also the first time I’ve seen inside a DIL package . . .
Not entirely sure the board is going to survive this process.
I have ripped whole pads/traces off with a desoldering pump, so I think you are doing pretty well 
Building analog synths in 741’s on veroboard taught me well!!
