Hi
   Whilst testing my 64 based project I noticed that a finger brought up to
   the quartz window caused the output leds connected (via 2003 darlintons)
   to ports C and D to behave oddly. This is not caused by the oft mentioned
   optical effect as the window was covered. Neither is it a decoupling
   problem as each pair of supply pins has 100n across them. The supply
   comes from a 78L05 and the timing is courtesy of a three terminal (4MHz)
   ceramic resonator.

I don't know if this is relevant, but recently I ran into this:
I have been building a PIC programmer, and I used a 7812 with
a voltage divisor  to boost the voltage slightly (to 14 V).
       _____            The voltage measured with a voltmeter
_______|    |________   looked fine, but when I looked at it
       |    |   |       with the scope, it was 14VDC plus a 10MHz
       --+---   <       oscillation with 1 Vpp amplitude. Kind of
         |      >       shook my confidence in 78xx parts. I suggest
         |      <       you look carefully for a parasitic oscillation
         |      >       (of course hooking scope probes could influence
         +------|       what you are looking for---welcome to the
                >       wonderful world of analog).
                <
                >       BTW, I ended up using LM317---works like a charm.
----------------+-----

                greetings

                        przemek klosowski (przemek@rrdstrad.nist.gov)
                        Reactor Division (bldg. 235), E111
                        National Institute of Standards and Technology
                        Gaithersburg, MD 20899,      USA

                        (301) 975 6249