I couldn't swipe out reservations to the implementation of FJPEM for a long time, then PGP finally appeared in front of me in the early summer of 1994. At that time, MIT tried to find a solution to make non-commercial PGP legal against the RSA patent, so called PGP 2.5. The release of PGP 2.6 settled down the relationship between Phil and RSADSI, yet PGP 2.6 based on RSAREF was under US export control.
Meet to 2.6ui based on 2.3a finally leaded me to be a PGP user. I cannot forget my first impression to PGP. It's a really well-designed program. In the summer of 1995, I took part in NDSS, so called ISOC Security Symposium in San Diego. In addition of the proceedings, PGP by Simson published from O'Reilly was delivered to early registered participants. It's a masterpiece of masterpieces. For these reasons, I have spent time to integrate PGP and MIME rather than to integrate PEM and MIME.
The idea that marks can represent encryption and signature was given by Mine when we wrote a paper for IPSJ together. It took a time a bit to understand its merit that marks could be canceled at any time.