Well I *still* haven't built an XOR or XNOR gate, but I did learn a few important things which will result in yet another revision to many of my existing gate schematics in the very near future.

When I started this project I just used a bunch of diodes I had lying around from an old radio project. They are old-school germanium diodes rather than the ubiquitous silicon diodes in use nowadays. I built all of my circuits with them and all was well until I ran out. When i tried to use silicon diodes, I kept on getting wild oscillations all over the place. Eventually a helpful stackexchange answer set me straight. I needed to provide a pull-down resistor just after the diodes on any gate using them. This makes sense, but I'm still confused as to why the germanium diodes worked without the pull-down. Also, after much testing, I found the surprisingly large value of 470K ohms to work the best with both the germanium and the silicon diodes.

I also did more thorough tests of the LM393 and TLC3704 comparators to determine their usefulness as space-saving measures as I start to get into more complex circuits. Both are more limited than the LM319 as to which gates they can be used to produce, but they are both smaller packages so if I happen to need a few gates out of the list one or the other can make, I might do so rather than stick another 319 on the board.