Re: address@hidden: Eamcs 22.1 comint-scroll-show-maximum-output bug]

[Glenn Morris]

Richard Stallman wrote:

> I am sure it WAS needed.  If you search for bug reports in a few
> weeks before I made that fix, you will probably find the bug report
> which I fixed this way.

Can't find anything in pretest-bug, devel, or bugs. The archives of
pretest-bug only start a few weeks before you made this change.

> You mean, if the user sets comint-scroll-show-maximum-output,
> that should zero out scroll-conservatively?

Yes.

> Maybe that is correct, but I am not sure. You see,
> comint-scroll-show-maximum-output is only supposed to affect output.

But the default for scroll-conservatively is 0, and you saw no problem
in unconditionally changing it to 10000 in all comint-mode buffers for
the sake of comint-scroll-show-maximum-output.

> Perhaps the thing to do is bind scroll-conservatively in this code and
> force a redisplay there:
>
>                        (save-excursion
>                          (goto-char (point-max))
>                          (recenter (- -1 scroll-margin))))

Perhaps. I can't make it work for the case when
comint-scroll-show-maximum-output is nil and scroll-conservatively is
10000 in the buffer. Since I don't know what problem the setting of
s-c to 10000 was supposed to fix, it's impossible for me to test that
case. It all works fine for me with no setting of scroll-conservatively.

You should probably just fix this yourself as you see best. I've been
testing with:

M-x shell
bash 
while : ; do
echo 1
sleep 1
done