On Tue, Jul 20 2021, Norm Tovey-Walsh wrote:
> Yes, you’re right. I’m not sure what I was thinking. I guess because
> threading takes priority over the date order, and messages that aren’t
> threaded are basically independent, I just never noticed.
Ironically while I think it's logical that threads have independent
order, and it does work as you *think* it should (because the parent is
always shown as topmost), I got bitten by this sorting exception more
than one time when reading threads where all messages are new.
Speaking in broad terms, I wonder how the UI would feel if instead of
sorting in reverse and have these weird exceptions we simply *reverse*
the result list (just adjusting the threading symbols) before being
displayed. You would get a tree going up.
One thing I don't like when having normal (ascending) chronological
order is that mu4e seems to be really optimized to show you the first
results in the buffer, not the last ones. You'd need to jump to the last
item in the buffer as the first action after searching, for example.
Showing the search limit (query limited to XY entries) would need to be
placed on top to make sense. Nothing major IMHO, but it's a list of many
small things that would need to be adjusted.
So I definitely too feel the weirdness that Joseph is experiencing.
> The distinction I was making was between “j i” to view the inbox and “s
> someterm” to perform an explicit search. But on reflection, I think the
> difference is between my ears, not in the software.
No problem, I have brain farts all the time too :). For a minute I was
really wondering if there was something in mu4e I didn't know yet!