"Surfaces into sessions" does not refer to surface files stored in directories. It refers to the internal structures that AFNI uses to organize datasets, surfaces, etc. Formerly, surface structs were attached to dataset structs. Now they are attached to session structs. This required modifying every place in the code that referred to the surface structs, since their parent structs were now different.

The purpose of this change was to make it easier to deal with multiple surfaces in a single session. Since the xyz coordinates of all datasets in a session are presumed to be aligned (for overlay purposes), it made sense to move the surfaces to the session struct. This made it possible to overlay surfaces on datasets that they weren't directly associated with -- e.g., a surface created from a set of MP-RAGEs then overlaid onto some other high-res anatomy. This example was the particular motivation, since there are people here doing very high res anatomical imaging (over small regions) and they wanted to overlay the surfaces from the MP-RAGE datasets onto their new-fangled images.

At present, AFNI does not read surface files from disk directly. It receives the surface geometry (node xyz triples and triangle ijk triples) via a socket connection with SUMA. Whether or not we will ever read surfaces directly into AFNI hasn't been decided.

Zorkon the Subsequent