pull request

[Linus Torvalds]

On Mon, Aug 20, 2012 at 9:19 AM, Dirk Hohndel <dirk at hohndel.org> wrote:
>
> Urgs. That's even more ugly and inconsistent then. I can have the same
> set of dives selected, but depending on how I get there, summary items
> may or may not be selected?
>
> I /really/ don't like that, Linus.

Think it through. The *only* way to avoid that rule is by saying "if
any dive at all in a group is selected, then we always force the
group-entry to be selected". Because that has to be the rule for the
collapsed group.

So "consistency" is actually detrimental in that case. If you select
(by hand) a single dive, it would be *very* annoying imho to
force-select the dive group.

So now the "consistency" rule would need to be made more complex:
 - for collapsed entries, the group-entry is selected if *any* dive in
that group is selected
 - for expanded entries, the group-entry is selected if *all* dives in
that group are selected

You can make that consistent, and not overly annoying.

But that doesn't work well either. Why? Because it screws up the
"select by range" in the date view. You select a date view, and the
*last* group entry isn't selected. Ugh. Ugly. Visually very
inconsistent.

So the current choice gives the visually best solution without
unnecessary "select stuff behind the users back". And the one case you
complain about is the odd case when you switch between views, not a
*normal* case.

             Linus