Has anyone started this? I am willing to give it try, but I want to avoid duplication of effort. My main motivation is adding .vti support to Slicer, which was recently added to ITK via PR 6032.
Tagging people who might have started this: @hjmjohnson @blowekamp @lassoan @jamesobutler
I’ve been working around the edges with the goal of getting there. I’ve done BRAINSTools updates to prepare for this change. I’ve also looked into SlicerExecutionModel and think there are no issues there.
My focus has been on getting CTK updated to Qt6 before tackling the ITK v6 changes.
I believe the only hard part is cherry-picking 216f772 on top of current main. Applying the current style (gersemi, newer clang-format, etc) before cherry-picking should minimized number of conflicts that need to be resolved (e.g. sorting of list entries which happened between 5.4 and main). But then there are possible philosophical changes needed to that commit, which might be non-trivial. Slicer compiled (or near-compiled) with 5.4 with legacy turned off. It should compile with v6 and legacy on, with little to no changes.
I would also consider including this change: ENH: Add namespace to ITK autoload symbol by blowekamp · Pull Request #14 · Slicer/ITK · GitHub
The introduction of “ITK_LOAD_SYMBOL” in upstream ITK should also be considered.
Should we integrate that change into main? It seems completely backwards compatible.
I haven’t started any work related to this. So yes, it would be greatly appreciated 
@dzenanz if you work on this. The core ITK folks definitely are most familiar with any new deprecation handling and updating to new methods.