There appear to be a diversity of star formation properties between apparently similar galaxy types. This appears to correlate with neither the metallicity of a galaxy nor its gas content. In this sample there is no direct evidence (MSTOs) for any discrete strong bursts of star formation. Comparing the results of the four dI galaxies in the HST sample of Skillman: Sextans A, Pegasus DIG, Leo A, & GR 8 a trend of higher sfr per area with larger MHI / LB was found (Dohm-Palmer et al. 1998). Although Pegasus and Leo A have clearly had higher global star formation rates in the past, whether this took place over several Gyr or in a short discrete burst cannot be discerned with these data. Leo A is the best nearby candidate for a primeaval galaxy. If there was an epoch during which galaxies had frequent bursts it was beyond z = 0.1. Deeper observations of more galaxies will push this limit back in redshift, and make the connection to the faint blue galaxy population more convincing. The giant galaxies (M 31, the Galaxy, M 33) have large old populations, and must contain the bulk of star formation in the LG since they contain something like 90% of the stellar mass, but this does not mean that they have always dominated in luminosity as completely as they do now. This is why it is important to look for detailed evidence of bursting behaviour in the most common (by number) galaxies in the LG - the late type irregular galaxies - by looking at the fossil record of the old/intermediate age resolved stellar populations.
Acknowledgments
I gratefully acknowledge the organisers for a generous grant, and I also thank my collaborators on the ``HST dwarf galaxy consortium'', namely, Evan Skillman, Robbie Dohm-Palmer, Jay Gallagher, Andrew Cole, Mario Mateo, Abi Saha & John Hoessel who have been involved in all of the projects I have reported on here.