Make the power button work properly without ACPI
 

I cannot use ACPI with *BSD, because a BIOS bug in my laptop version (Aspire 3680-2633) causes resources for ethernet devices to be deallocated by the BIOS when ACPI is enabled. Unfortunately, I don't have APM support either.

This would be fine... Except for one tremendously stupid problem.

In Linux, when ACPI is disabled, the power button works properly. If you hit it, nothing happens. It only turns off the machine when you hold it down for five seconds.

Likewise in Windows: when ACPI is disabled, hitting the power button does nothing, unless you hold it down for five seconds.

In *BSD, though... When ACPI is disabled, the power button shuts down the machine the instant it is hit. In other words, if you accidentally brush the button... Poof. There goes your data. And this is unacceptable.

How can I force *BSD to do nothing when I press the power key, or any other ACPI related key?

(And really... I'm getting a bit fed up with all this. Windows XP is slow and very annoying to maintain, Linux is stupendously buggy, and Solaris doesn't support my hardware. The BSDs, by contrast, are generally pretty well behaved. And this one, utterly stupid bug stands between me and a usable BSD install. Help?)

Try upgrading your BIOS.

If that doesn't work, here's a quarter: go buy a decent PC.

And in the meantime, stop by a hobby store and buy some material you can epoxy to your PC so you don't accidentally brush the power button :hattip:

'Hope that helps .. PSM

PS:
What's all this nonsense about "Linux too buggy"?

Any BSD in particular, or do they all exhibit the same behavior?

And Paul, once you've experienced OpenBSD (that did it for me, at least) stability, any little hiccup in Linux (they definitely exist) make Linux seem buggy.

Hi, rocket357 -

I don't disagree with you that *any* OS can have "issues".

I just think it's terribly unfair to make such a broad, specious claim without stating any specifics, or giving any justification. Who knows - the problem might have been an application problem having nothing whatsoever to do with the kernel. Or it might have been some "niche" distro (like Oracle's self-serving answer to "This is Not RedHat"). Or it might have been some trivial annoyance the OP just characterized as a "bug".

IMHO .. PSM