We believe that one can have serious reservations as to whether heavy ion collisions (e.g. 100 GeV/n Au + 100 GeV/n Au) can lead to thermal and chemical equilibrium over large regions (particularly if it is assumed that this happens whenever QGP is produced at RHIC-that is, if it is produced). It is at present nor clear that the collision dynamics and times available will lead to this. An alternative scenario has been proposed by Van Hove where localized-in-rapidity bubbles of plasma may well be more probable, and may well occur at least some of the time, and some of the time mainly survive to the final state. We have developed a series of event generators to extend and describe these phenomena. A Van Hove-type (Van Hove L 1983 Z. Phys. C 21 93-8; 1984 Report CERN-TH 3924; 1987 Nucl. Phys. A 46) spherical bubble at eta = 0 is embedded in a reasonable event generator in qualitative agreement with Hijing etc (Wang X N and Gyulassy M 1991 Phys. Rev. D 44 3501; 1992 Phys. Rev D 45 844; 1994 Comput. Phys. Commun. 83 307). The plasma bubble hadronized at a temperature of 170 MeV according to the model developed by Koch et al (Koch P, Muller B and Rafelski J 1986 Phys. Rev. 142 167). The amount of available energy assumed in the bubble is selected as that in a small central circular cross section of radius approximate to 1.3 fm or approximate to 2.5 fm in 100 GeV/n Au + Au central events. The results predict (with the assumptions stated, possible) striking signals which may allow strong evidence for a QGP which cannot be explained by alternative conventional physics arguments, and thus may be crucial elements in establishing a QGP. We are also applying these techniques to investigating Kharzeev and Pisarski bubbles of metastable vacua with odd CP.