[Insight-developers] Notes from Friday May 18 TCon

[Will Schroeder]

Here's a brief summary of discussion items:

* CMake will be separated from the Insight repository. Users will have to install CMake separately. This is to shield the Insight developers from the rapid pace of CMake development which has lately disrupted builds.

* Iterators are to be made bi-directional. The API to support this functionality needs to be specified. (Josh Cates)

* A proposal was made to rename the iterator classes. Iterator connotes a linear traversal, and Insight has special iterators that do not act this way (e.g., some contain an iterator within the iterator for region walking). Proposed names include Walker, Strider, Scanner, etc. The consortium is to think about this and get back together with names.

* Filters that process PhysicalImage are not setting the output Spacing and Origin from the input. This needs to be added in the superclass. (Jim Miller)

* Copyright location is in Insight/Copyright.txt. To create a new file, paste this into the file, with the possible addition of CVS keywords. (Any source code header can be used for CVS qualifiers; CVS will replace the values with the correct values.)

* Extensive discussion on the meaning of life, origin, and the location of voxels. You had to be there.

* Paul H. suggested tagging clean compiles. Jim Miller suggested moving tags. Bill Hoffman is sending an email addressing this: using moving tags, the user could do a single "cvs update -r LAST_KNOWN_STABLE_BUILD" and from then on just do a "cvs update" and always get a working version (because of the sticky tag). (See the rest of Bill's email.)

* Grouping in doxygen and documentation. Luis has had some success with this mechanism, as well as adding some general documentation (independent of the source code...reference material) into the doxygen web pages. The general feeling that this approach could be used to create on-line reference material; the user-oriented documentation would still be crafted better with LaTeX. There was some concern about duplication of effort.

Will