This confusingly doesn't always work

  The trick was that when a line that was exactly filled was followed by an
  empty line, we'd ECH when the cursor is virtually on the last char of the
  wrapped line. That was wrong. For a scenario like:

  |ABCDEF|
  |      |

  We'd incorrectly render that as ["ABCDEF", "^[[K", "\r\n"]. That ECH in the
  middle there would erase the 'F', because the cursor was virtually still on
  the 'F' (because it had deferred wrapping to the next char).

  So, when we're about to ECH following a wrapped line, reset the _wrappedRow
  first, to make sure we correctly \r\n first.

…hich are blocked on #4213

# Conflicts:
#	src/buffer/out/textBuffer.cpp
#	src/buffer/out/textBuffer.hpp
#	src/host/screenInfo.cpp

(cherry picked from commit c21f740)

…conpty-wrapping-003

…BufferLine interface

…ly, esp conhost v wt v gnome-terminal. Remove ambiguity - just hardcore move the cursor in this scenario.

…avior quite right

…inal behavior quite right"

This reverts commit 0a98cce.

…-actually-did-it-this-time

…low-buffer-on-resize

…/migrie/f/reflow-buffer-on-resize

# Conflicts:
#	src/renderer/vt/XtermEngine.cpp
#	src/renderer/vt/vtrenderer.hpp

…-buffer-on-resize

…-buffer-on-resize

…-buffer-on-resize

# Conflicts:
#	src/cascadia/UnitTests_TerminalCore/ConptyRoundtripTests.cpp

# Conflicts:
#	src/cascadia/UnitTests_TerminalCore/ConptyRoundtripTests.cpp
#	src/cascadia/UnitTests_TerminalCore/TerminalBufferTests.cpp
#	src/renderer/base/renderer.cpp
#	src/renderer/vt/paint.cpp

…-buffer-on-resize

…-buffer-on-resize

…-buffer-on-resize

  one was weird: after the reflow, the conpty would know that the cursor was in
  a different position than the last time it wrote text, so it would emit a
  sequence to position the cursor. However, with quirky resize, the terminal's
  cursor actually _was_ in the right position. So when conpty would emit a \n to
  move the cursor one line down, that would cause the Terminal to be out of
  sync.

  The second: we were calculating the new position of the top row of the buffer
  using the first cell of the row, not the last. This meant that if the top row
  wrapped in this new resize, we'd place the top of the new viewport one row
  higher than conpty would

… top to calculate our new position

  The trick with this case was that when there were multiple wrapped lines
  inside the viewport, the lastLine of the buffer would be farther down than we
  actually wanted. We'd still want the position of the top row in the buffer
  here, because the lines in the buffer could have _only_ de-wrapped, taking up
  less lines than before.

  This is however, _not_ right. When you execute `ll`, then increase the height,
  then decrease the width, we'll snape the bottom of the viewport to the bottom
  of the output text, as opposed to pinning to the top of the text.

  This fixes the bug from before. This breaks when we simultaneously change
  width and height. Scenario is powershell, `type rwr-temp.txt`. Increase font size. Viewport is not at the actual bottom of the buffer anymore -.-

  I think we need to use the scrollback value unless we're decreasing the
  height of the buffer as well

  If the terminal height is getting smaller, always use the lastline to figure
  out where the new bottom line should be oh no that's not going to work is it.
  what happens with dy<0, dx>0?

  * dy<0, dx<0:
    - The Terminal is getting shorter, buffer less wide. Use the end of the text
      to find where we should place the viewport, since lines might have newly
      wrapped
  * dy<0, dx>0:
    - The Terminal is getting shorter, buffer wider. ?? Use the old top to find the
      new position. Lines migh de-wrap, and we'll want to stay stuck to the
      topmost line of the old buffer. ??
  * dy>0, dx<0:
    - The Terminal is getting taller, buffer less wide.
  * dy>0, dx>0:
    - The Terminal is getting taller, buffer wider.  Use the old top to find the
      new position. Lines migh de-wrap, and we'll want to stay stuck to the
      topmost line of the old buffer.

…-buffer-on-resize

…committing and pushing

…-buffer-on-resize

# Conflicts:
#	src/cascadia/UnitTests_TerminalCore/ConptyRoundtripTests.cpp