Make the interaction between MultiProgress and other lines below it more robust #4737
Description
Follow up of #4560.
Now we know that the last line of MultiProgress lacks a newline, and instead it adds a bunch of whitespaces so that new contents are forced to be shown on the following line.
/ # RUSTUP_TERM_PROGRESS_WHEN=always rustup check^[[J^M$ ^[[1mstable-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[96mChecking...^[[0m ^[[1m^[[96m|^[[0m ^M^[[2K^[[1mstable-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[96mChecking...^[[0m ^[[1m^[[96m/^[[0m ^M^[[2K^[[1mstable-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[96mChecking...^[[0m ^[[1m^[[96m/^[[0m^M$ ^[[1mbeta-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[96mChecking...^[[0m ^[[1m^[[96m|^[[0m ^[[1A^M^[[2K^[[1B^M^[[2K^[[1A^[[1mstable-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[96mChecking...^[[0m ^[[1m^[[96m/^[[0m^M$ ^[[1mbeta-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[96mChecking...^[[0m ^[[1m^[[96m/^[[0m ^[[1A^M^[[2K^[[1B^M^[[2K^[[1A^[[1mstable-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[96mChecking...^[[0m ^[[1m^[[96m/^[[0m^M$ ^[[1mbeta-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[96mChecking...^[[0m ^[[1m^[[96m-^[[0m ^[[1A^M^[[2K^[[1B^M^[[2K^[[1A^[[1mstable-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[96mChecking...^[[0m ^[[1m^[[96m-^[[0m^M$ ^[[1mbeta-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[96mChecking...^[[0m ^[[1m^[[96m-^[[0m ^[[1A^M^[[2K^[[1B^M^[[2K^[[1A^[[1mstable-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[96mChecking...^[[0m ^[[1m^[[96m-^[[0m^M$ ^[[1mbeta-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[96mChecking...^[[0m ^[[1m^[[96m\^[[0m ^[[1A^M^[[2K^[[1B^M^[[2K^[[1A^[[1mstable-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[96mChecking...^[[0m ^[[1m^[[96m\^[[0m^M$ ^[[1mbeta-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[96mChecking...^[[0m ^[[1m^[[96m\^[[0m ^[[1A^M^[[2K^[[1B^M^[[2K^[[1A^[[1mstable-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[96mChecking...^[[0m ^[[1m^[[96m\^[[0m^M$ ^[[1mbeta-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[96mChecking...^[[0m ^[[1m^[[96m|^[[0m ^[[1A^M^[[2K^[[1B^M^[[2K^[[1A^[[1mstable-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[96mChecking...^[[0m ^[[1m^[[96m|^[[0m^M$ ^[[1mbeta-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[96mChecking...^[[0m ^[[1m^[[96m|^[[0m ^[[1A^M^[[2K^[[1B^M^[[2K^[[1A^[[1mstable-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[92mup to date^[[0m: 1.93.1 (01f6ddf75 2026-02-11)^M$ ^[[1mbeta-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[96mChecking...^[[0m ^[[1m^[[96m|^[[0m ^M^[[2K^[[1mbeta-x86_64-unknown-linux-musl - ^[[0m^[[1m^[[92mup to date^[[0m: 1.94.0-beta.5 (31a40f5c8 2026-02-20) ^[[1mrustup - ^[[0m^[[1m^[[33mUpdate available^[[0m : 1.29.0 -> 1.28.2^M$ / #
Clearly some terminal configurations are not very happy about this.
On the one hand, when the history is re-read under a different window width, the missing newline will be revealed easily.
On the other hand, in certain configs (which ones?) the line wrap is not visible at all, e.g. @epage's setup which is comprised of bash v5.3.9(1)-release inside of Zellij inside of Wezterm, running on Linux Desktop (but I wasn't able to reproduce on macOS, any DE/OS info?).