"Jacob S. Gordon" <[email protected]> writes: > 1+, this makes sense to me. I also wonder about the other uses of ‘C-c > C-c’ when in column view, i.e. actions outside the headline like > > + renumbering a list, > + aligning a table or applying a formula, > + evaluating a code or dynamic block, > + updating a radio target.
You are right. > Personally I find it a bit surprising that it quits column view and > removes overlays in ‘org-sparse-tree’. I find this surprising too. I have seen this confuse users when showing them column view: they toggle column-view checkbox cells with `C-c C-c', but if they press the same key when point is not on such a cell, the whole column view disappears. I think column view should not special-case `org-ctrl-c-ctrl-c' at all. Column view is an overlay/editing view on top of an Org buffer, so it should not prevent the regular Org context dispatcher from handling plain-list checkboxes, tables, dynamic blocks, radio targets, and similar constructs. There is already a column-view-specific way to leave the view when point is on a column overlay: `q'. For the regular column-view key, I think `C-c C-x C-c' would be a better toggle: it currently starts column view, so it could also leave it when column view is active. That would let `C-c C-c' keep its usual Org meaning, while making the column-view command itself responsible for entering and leaving column view. The only downside I can see is that users who currently use `C-c C-x C-c' repeatedly to move column view between different local `COLUMNS' scopes might need one extra key press: first leave the current column view, then start it in the new location. But I think this is less surprising than making `C-c C-c' quit column view in contexts where Org normally has a concrete action. So I think the current patch is too narrow and should probably be withdrawn for now. It fixes the checkbox case, but the underlying issue is broader: column view currently takes over `C-c C-c' before the normal Org context dispatcher can handle other useful locations. I would prefer to rethink this as a key-binding/dispatch change rather than add another special case for list checkboxes only. Best, -- Slawomir Grochowski
