Hello,
I've been using Brunch and noticed something with the brunch-setup script that I wanted to share, along with a temporary workaround that worked for me. I hope this feedback might be useful.
Observation
When I run the brunch-setup script and select kernel 6.12, it seems to only display one page of framework options. My understanding from looking at the script is that it should show two pages of options, similar to how it works for kernel 6.6.
I noticed this behavior across several different versions of Brunch (r134, r137, r139, r140). I also checked the script itself and could see that options_page2 is indeed defined for kernel 6.12, so the options should be available.
A Possible Theory
I'm not an expert, but my guess is that this might be related to how the shell environment handles variables.
I suspect that the options_page2 variable might already exist as an empty string in the shell environment before the case statement is executed. If that's true, the condition if [ ! -z "$options_page2" ], which checks if the second page of options should be displayed, might not work as expected.
A Temporary Workaround I Found
Based on this idea, I tried making a small change to the script by adding unset options_page2 right before the case statement.
My change (around line 190 of brunch-setup):
reset
unset options_page2
case "$selected_kernel" in
6.12)
#...
Result on My Machine
After adding that single line, the script started working as I had hoped. When I chose kernel 6.12, it correctly showed "PAGE 1/2" and then "PAGE 2/2". My device was then able to load the modules from the second page of options and has been working normally since.
I'm not sure if this affects other users or if it's just something specific to my setup, but I wanted to share my experience in case it helps.
Thank you for all your hard work on this project!
Edit:
I think the issue is more complicated than I initially thought. The solution above was just an experiment that luckily managed to solve the problem. The unset command seems to act merely as a "stabilizer" that resets the state.
The root cause is likely that the shell interpreter isn't behaving as expected (in fact, I've seen others in this issue report the same problem with kernel 6.6, not just 6.12 as in my case). Perhaps we'll just have to wait for a proper fix from the developer. xD
Hello,
I've been using Brunch and noticed something with the
brunch-setupscript that I wanted to share, along with a temporary workaround that worked for me. I hope this feedback might be useful.Observation
When I run the
brunch-setupscript and select kernel 6.12, it seems to only display one page of framework options. My understanding from looking at the script is that it should show two pages of options, similar to how it works for kernel 6.6.I noticed this behavior across several different versions of Brunch (r134, r137, r139, r140). I also checked the script itself and could see that
options_page2is indeed defined for kernel 6.12, so the options should be available.A Possible Theory
I'm not an expert, but my guess is that this might be related to how the shell environment handles variables.
I suspect that the
options_page2variable might already exist as an empty string in the shell environment before thecasestatement is executed. If that's true, the conditionif [ ! -z "$options_page2" ], which checks if the second page of options should be displayed, might not work as expected.A Temporary Workaround I Found
Based on this idea, I tried making a small change to the script by adding
unset options_page2right before thecasestatement.My change (around line 190 of
brunch-setup):Result on My Machine
After adding that single line, the script started working as I had hoped. When I chose kernel 6.12, it correctly showed "PAGE 1/2" and then "PAGE 2/2". My device was then able to load the modules from the second page of options and has been working normally since.
I'm not sure if this affects other users or if it's just something specific to my setup, but I wanted to share my experience in case it helps.
Thank you for all your hard work on this project!
Edit:
I think the issue is more complicated than I initially thought. The solution above was just an experiment that luckily managed to solve the problem. The unset command seems to act merely as a "stabilizer" that resets the state.
The root cause is likely that the shell interpreter isn't behaving as expected (in fact, I've seen others in this issue report the same problem with kernel 6.6, not just 6.12 as in my case). Perhaps we'll just have to wait for a proper fix from the developer. xD