Thanks for the suggestions.
Some quick comments:
1) "Rows" being better than "Pages" (as said above) makes me wonder should default scrolling be "by rows" ? (I know I'm basing this question on one capture, but just thought I'd ask anyways).
Rows won't always work. For example it won't work in windows that require keypress simulated paging. But when they do, they can be much easier for SC to align, since scrolling by pages can leave little or no overlap to discover. That's why the "Test Window" procedure is basically trying to figure out if it can scroll by rows and how many rows at a time to scroll. The problem with setting to ByRows manually is that it can eat up a lot of memory unless you set a high enough "rows per scroll multiplier" (which is what Test Window tries to determine).
~ if horizontal scrolling is activated when there is no horizontal scrolling in the window, it doesnt work
~ if horizontal scrolling is *not* activated where there *is* horizontal scrolling in the window, it doesnt work.
not sure what you mean by "doesn't work". it should do fine capturing pure vertical content even if there is horizontal content and you disable it. if you try to capture horizontal content and there is none, it shouldn't cause any problems just some wasted time trying to horizontally scroll when there is no content. can you elaborate what didnt work?