Possible differences between the VT330 and VT340 when using multiple aspect ratios

[j4james]

When testing Sixel images with multiple aspect ratios on a VT330, I noticed some interesting side effects.

1. If the aspect ratio changes partway through a row, the x offset seems to be set back to the start of the row.
2. Depending on what operations have been performed with the previous aspect ratio, the new aspect ratio will not necessarily apply to a DECGNL that follows it.

The script below demonstrates some of these effects.

multiple_aspect_ratio.sh

#!/bin/bash

# Test the effects of multiple aspect ratios in a Sixel image.

# The first set of tests demonstrate the effect of a new aspect ratio
# on the following graphic new line (DECGNL) commands.
# The second set of tests demonstrate the effect of a new aspect ratio
# on the x offset, when not already at the start of a line.

CSI=$'\e['			# Control Sequence Introducer 
DCS=$'\eP'			# Device Control String
ST=$'\e\\'			# String Terminator


echo -n ${CSI}'H'
echo -n ${CSI}'2J'
echo -n ${CSI}'?5l'

set_cursor_pos() {
  echo -n ${CSI}${1}';'${2}'H'
}

new_line_test() {
  local row=${1}
  local col=${2}
  local prefix=${3}
  local suffix=${4}

  set_cursor_pos ${row} ${col}
  echo -n ${DCS}'9;0q'
  echo -n '"10;1;1;1'	# Start with 10:1 aspect ratio
  echo -n '#1!60~'	# Render 60x60 block
  echo -n ${prefix}	# Output the prefix value
  echo -n '"5;1;1;1'	# Change aspect ratio to 5:1
  echo -n ${suffix}	# Output the suffix value
  echo -n '-'		# Move to new line
  echo -n '"1;1;1;1'	# Change aspect ratio to 1:1
  echo -n '#2!90`'	# Output marker lines to show how far we moved
  echo -n ${ST}
  echo -en '\b\b>'
}

auto_return_test() {
  local row=${1}
  local col=${2}
  local ar=${3}

  set_cursor_pos ${row} ${col}
  echo -n ${DCS}'9;0q'
  echo -n '"10;1;1;1'	# Start with 10:1 aspect ratio
  echo -n '#1!30~'	# Render 30x60 block
  echo -n '"'${ar}';1;1' # Change aspect ratio
  echo -n '#2!60`'	# Output marker lines to show the carriage return
  echo -n ${ST}
  echo -en '\b\b>'
}

new_line_test 4  5  ''  ''
new_line_test 4  18 '+' ''
new_line_test 4  31 '-' ''
new_line_test 4  44 '$' ''
new_line_test 4  57 ''  '$'
new_line_test 4  70 ''  '#3'
new_line_test 12 5  ''  '0'
new_line_test 12 18 ''  '+'
new_line_test 12 31 ''  '-'
new_line_test 12 44 ''  '?'
new_line_test 12 57 ''  '!1'
new_line_test 12 70 ''  '"'

auto_return_test 20 5  '5;1'
auto_return_test 20 18 '10;1'
auto_return_test 20 31 '5;0'
auto_return_test 20 44 ';'
auto_return_test 20 57 ';1'
auto_return_test 20 70 '0;1'

echo -n ${CSI}'H'

I'm not positive, but I suspect the VT340 doesn't work the same way, so I'm curious to see what output you get from this.

[j4james]

Forgot to add this here, but for comparison, this is essentially what it looks like on a VT330 (technically this is a capture from Windows console in VT330 mode, but it matches what I've seen on a real VT330).

Obviously the color will be different on the VT340, and note that the VT330 handles overwriting differently (the pixel output is ORed, so plotting color 1 on top of color 2 will change it to color 3), but what I'm interested in seeing is the positioning of the two horizontal lines in each test case.