The readme uses history so surely the author knows this?
function preexec_go() {
local command_line=$(history 1 | sed 's/^ *[0-9]* *//')
if [[ -n "$command_line" && "$command_line" != "preexec_go" ]]; then
# CHANGE THIS PATH to the location of your 'envoy' executable
(
~/Desktop/projects/envoy/envoy "$command_line" >/dev/null 2>&1 & ) disown
fi
}
trap 'preexec_go' DEBUG
I think the point isn't what the author knows or does not know but rather a question for the rest of us - what does this tool offer that bash history doesn't?
The list of features on the Github page doesn't provide an answer:
- Starts and stops on demand: Only logs commands when you explicitly turn it on.
- Saves to a custom file: You can specify the log file name.
- Cross-platform: Works on both Linux and macOS using bash or zsh.
- Minimal overhead: Runs in the background and has no noticeable impact on shell performance.
In charity, I think there is actually a product opportunity for improvements to the standard shell histfile.
I've often been frustrated by my history not being easily shared between concurrent terminals, difficulties in searching, and lack of other metadata such as timestamp, duration and exit code.
Although I suspect this repo was vibe-coded so far, I think there's a promising problem to solve here.
Not at my computer so I cannot share the sauce, but each mahcine I use has a shared bash eternal history - shared between every interactive bash instance on that machine, using nothing but bashisms.
I'm confused: doesn't bash already log all commands in it's history file?
The readme uses history so surely the author knows this?
I think the point isn't what the author knows or does not know but rather a question for the rest of us - what does this tool offer that bash history doesn't?
The list of features on the Github page doesn't provide an answer:
- Starts and stops on demand: Only logs commands when you explicitly turn it on.
- Saves to a custom file: You can specify the log file name.
- Cross-platform: Works on both Linux and macOS using bash or zsh.
- Minimal overhead: Runs in the background and has no noticeable impact on shell performance.
Bash largely covers all of those too.
...is this program just copying lines from shell history inti a different file?
How does this differ from Atuin? https://atuin.sh
Well this one looks like AI slop, so..
In charity, I think there is actually a product opportunity for improvements to the standard shell histfile.
I've often been frustrated by my history not being easily shared between concurrent terminals, difficulties in searching, and lack of other metadata such as timestamp, duration and exit code.
Although I suspect this repo was vibe-coded so far, I think there's a promising problem to solve here.
Not at my computer so I cannot share the sauce, but each mahcine I use has a shared bash eternal history - shared between every interactive bash instance on that machine, using nothing but bashisms.
If I think of it later, I'll add it....
Then you should look at https://atuin.sh/ although I am not certain about exit code.
Why do you suspect it was vibe coded? There are 2 substantive files that are each less than 100 lines...
Also the readme doesn't have the usual emojis for every bullet point.
The godoc on logging.go is in the "AI style" rather than in the style of a typical godoc.
Comments on every other line stating what the line does is very LLM