MultiProgress example does not draw correctly after atomic position

[djc]

When running cargo run --example multi-tree-ext on current main, the output sometimes becomes what could best be described as a jumbled mess.

Expected output:

[32/32]   ✔ the
[32/32]     ✔ quick
[32/32]       ✔ brown
[32/32]         ✔ fox
[32/32]           ✔ jumps
[32/32]             ✔ over
[32/32]               ✔ a
[32/32]                 ✔ lazy
[32/32]                   ✔ dog
████████████████████████████████████████  585/585

Actual output:

[ 5/32] ⠂ temp_1
[24/32]                   ⠉ dog
█░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░   28/585                               
[ 5/32] ⠂ temp_1
[ 0/32]                 ⠉ lazy
[24/32]                   ⠉ dog

Or

[32/32] ✔ the
[ 0/32] ⠉ temp_7
[ 1/32]     ⠙ quick
[32/32] ✔ the
[ 0/32] ⠉ temp_7
[ 1/32]     ⠙ quick
[ 4/32] ⠂ temp_6
[ 2/32]         ⠚ fox
[ 0/32] ⠉ temp_8
[ 1/32]           ⠙ jumps
[32/32] ✔ over
[ 8/32] ⠴ temp_5
[32/32] ✔ a
[32/32] ✔ lazy
[32/32] ✔ dog

This seems to have started with ea1f733. This of course substantially changes the timing so it's possible this is exposing some pre-existing race conditions.

[djc]

This turned out to actually be due to a panic from #405, masked by the fact that the MultiProgress drawing was overwriting the panic output. #414 contains a change (324f3dc) that will stop indicatif from drawing to the terminal while the thread is panicking, hopefully making this easier to diagnose in the future.