Where Have We Been? Where Are We Now? Where Are We Going?

A conference summary necessarily glosses over or outright ignores a lot of interesting work. For that I apologize. However, I'm going to compound any such error by attempting to summarize our past, present and future even more briefly:

3 years ago

Only one algorithm was required and it was either RL or MEM, depending on who was talking.
Everyone thought that the PSF was knowable a priori.
As a community we had little experience with the deconvolution of HST data.

Now

We now know that from zero to many algorithms might be applicable.
We have seen that it pays to view the PSF an unknown to be determined.
As a community, we now have lots of experience of HST deconvolution.

3 years from now?

Our model of the telescope will be more complete: including SVPSF and a better noise model.
As our understanding of blind deconvolution improves, we will realize that it pays to view the aberrations (rather than the PSF) as unknowns to be determined.
We will have produced lots of documentation on deconvolution.
Post-servicing, the HST will have a better PSF. However, as a community, we'll still have a lot to do. The PSF for WFPCII is not perfect and varies across the field.

Finally, I should say that ST ScI has done a great job in pushing the cause of image restoration in the community via software, algorithms, user support, newsletters and via these meetings. Personally, I have enjoyed both these meetings tremendously, and learned a lot. I look forward to the next meeting in the series.