RE: PATCH: isearch-yank-until-char

[Drew Adams]

> >> the use cases: when the goal character is a letter
> >> at all, one is either looking at that letter
> >> on the screen *or* the letter is some known
> >> syntactic delimiter and its case is therefore
> >> known as well even if the letter is off the screen
> >> right now.)
> > ...
> > There's no reason to have this command make an
> > assumption about whether its char-search should be
> > case-sensitive.  I don't think the either...or
> > assumption you made above is good for the command
> > to make.  Better to let users control whether to
> > search for the char case-sensitively.
> 
> When I think about the circumstances under which this
> command is actually used, which I tried to characterize
> above, I couldn't think of any in which case-insensitive
> matching would make sense
> ...
> Can you say which part of the assumptions I gave above
> seems wrong to you?

I guess my question is about the "circumstances
under which this command is actually used".  Why
are those two cases obviously the only ones?
 
I don't see why a user would necessarily either
(1) be looking at the target letter or (2) want to
grab text up to the first case-sensitive occurrence
of a letter.

A user might well want to grab text up to the next
occurrence of the letter `Q', regardless of letter
case, no?  That could be the case, for example, if
the text in the buffer is generally indifferent wrt
case.

And in such a situation I'd generally expect that
`case-fold-search' _would_ be an appropriate guide
for both (1) this grab-up-to search and (2) the
search afterward with the updated search string.

Perhaps I'm wrong, and I really have no axe to
grind (don't care) about this.  I just haven't (so
far) seen why the only reasonable scenarios are
the two you assume.

Can you give an example where you think case
insensitivity would be inappropriate for this, but
it would be appropriate for the search with the
updated search string?

(Not that I'm arguing for case-insensitivity here.
I just wonder why case-sensitivity is always TRT.)