Documentation preview for this PR (built with commit bfb292a; changes) is ready! 🎉
This preview will update shortly after each push to this PR.

There is some code duplication. Would it make any sense to have a function that would replace each instance like

if [ -f spkg-pipinst ]; then
    sage-logger -P spkg-pipinst "$SAGE_SUDO ./spkg-pipinst"
    if [ $? -ne 0 ]; then
        error_msg "Error running the pipinst script for $PKG_NAME."
        exit 1
    fi
fi

by a call to a function defined by something like

run_script() {
if [ -f $1 ]; then
    sage-logger -P $1 "$SAGE_SUDO ./$1"
    if [ $? -ne 0 ]; then
        error_msg "Error running the $1 script for $PKG_NAME."
        exit 1
    fi
fi}

?

(not tested)

I've made it less verbose in a different way, please take a look

It seems that some of spkg-*.log files are generated and some not. Which is which and why?

If any are still generated with this PR, that's a bug. Is this what you observed?

Can we test it here, with "CI Fix" tag? We can already see it.

I think there should be some explanation why you introduced the log files and why you think now they are not useful...

I don't have a better explanation than what I wrote in the ticket description: "I used sage-logger for prefixing output; generating the log/time files was a side effect, which has not proven to be useful"

You certainly added the logging feature expecting that the logs would be useful. Then now you simply say "which has not proven to be useful". I am curious what happened in the meanwhile...

I looked #37391 again. Right, you didn't add the logging feature. As you said, it is a side-effect of using sage-logger for prefixing. The generated log files are not useful, and hence here you add a feature to turn them off.

OK. I misunderstood the situation. Sorry.

No worries, these are all the right questions to ask!

Thank you!