• Added bash_preexec_imported for checking if bash-preexec is present. deprecated __bp_imported variable in favor of new public variable
• restore original $? in PROMPT_COMMAND for access to other functions using PROMPT_COMMAND still. $? is already available in precmd functions.

• Efficiency and speed improvements in preexec's critical path. Replaced command substitution with writing to variables in #112. Some initial bench marking showed at least ~2x speed improvement. Shout out to @dimo414 for the PR and contributions.

• Fixes for #97 to allow bash-preexec to be included at any point in bash profile
• Numerous other small improvements and tweaks

Bugfix release for #64 where setting IFS could break function invocation
Shout outs to @gnachman for submitting the issue and @dimo414 and @d630 for feedback.

Bugfix release for #61
Shout outs to @lguelorget for submitting the PR to fix this. 👍 💯 🙏

• Bugfix release fixing #58 and an issue causing crashes on installation.

• Preserve preexisting PROMPT_COMMAND as precmd function and DEBUG trap as a preexec function
• Preexec returns last error code from preexec functions
• Specify using builtin history in case users alias history. Fixes #55 and addresses https://gitlab.com/gnachman/iterm2/issues/6063
• Couple of test fixes.

Shout outs to @dimo414 and @jombooth for PRs in this cut 👍 🙏 🎉

• Thanks @gnachman for the continued usage and awesome work on iterm2

• Preserve return value of last function in preexec
• Remove trailing semicolon on PROMPT_COMMAND to allow appending
• Couple of test fixes

Shout outs to @brandonweeks for the multiple PRs that make up this cut. 👍 💯

• Stop clobbering $_ and make available for precmd functions.
  • Worked around bug that was always overwriting the $_ variable in bash to be __bp_preexec_invoke_exec
  • $_ is now available again in bash and in precmd functions.

• Disabling subshell support by default. It can be enabled by setting __bp_enable_subshells="true" It should be considered experimental as users have noticed issues with it closing terminals (i.e. #25)