OSNews

ACPI works fine.

It does NOT work fine. Eugenia has THREE laptops and on NEITHER of them does Linux's ACPI work fine. Hell, it's almost common knowledge that Linux's ACPI kind of SUCKS ASS.

The problem is hardware deviating from specs, making it harder to support it.

Oh really? Why, I never realized that... To the normal end user, it doesn't matter $%#Q$^ if the hardware deviates from the spec or not. If it doesn't work, it does NOT work. As simple as that. If Windows can support all those deviating specs, than why can't Linux?

You know perfectly well that Linux supports the major bulk of the x86 market...

Yup, I do, but that doesn't have anything to do with the fact that Linux's ACPI support for x86 sucks.

...and even several other platforms.

Windows XP is probably just as portable as Linux. In case you didn't know, Windows NT ran on x86, SPARC, PPC and Alpha. But that has nothing to do with Linux's ACPI either.

But then again, Eugenia. You like XP, so your thoughts on this hardly counts by any metric.

I do not like XP. Yet, I agree with Eugenia. Your point?

Thom, Thom. Why don't you read up before you yell at people?

If Windows can support all those deviating specs, than why can't Linux?

Here is the secret: Windows does NOT support all those deviating specs. It's the laptops that support windows broken implementation of ACPI.

Say after me. Hardware is designed for Windows ($VERSION), Windows is NOT designed for $HARDWARE. Buy hardware that is supported in Linux if you want working system. Don't whine because your "Desiged for Windows XP" system doesn't work with MacOS X or Linux..

Hell, it's almost common knowledge that Linux's ACPI kind of SUCKS ASS.

Everybody's ACPI support sucks ass. Your point is?

To the normal end user, it doesn't matter $%#Q$^ if the hardware deviates from the spec or not. If it doesn't work, it does NOT work. As simple as that.

In terms of getting a solution to an end user it bloody well does matter. ACPI is by no means a standard, and the only reason why Windows works is because hardware Microsoft and hardware manufacturers have hacked it into place. I can give you a list of laptops and even desktops where suspend to RAM and hibernate does not work with Windows, even after all these years and given the fact that Windows is the first thing these people should be testing.

The solution is clear - hardware manufacturers have to really help out and actually test their hardware with Linux, and the OSDL and Linux developers need to encourage them to actually create an actual standard that's predictable and that works.

From a Linux Journal article:

ACPI, which stands for advanced configuration and power interface, is a power management component eventually set to replace APM. The current state of ACPI as a whole is marred at best. Laptop manufacturers often deviate from following any type of standard when coding certain ACPI-related BIOSes and components. Rather than follow a standard, they rely on proprietary Windows drivers to keep things functioning. Unfortunately, in many cases, this leaves Linux users out in the dark.

http://www.linuxjournal.com/article/6891

This answers your question about why linux can't support all those deviating specs. If you don't have the specs, how can you support them? If the manufacturers alter the spec, and then provide a binary driver to one OS provider only, then it's pretty hard to support that spec isn't it.

Sometimes you make a complete arse of yourself on your own site.

It does NOT work fine. Eugenia has THREE laptops and on NEITHER of them does Linux's ACPI work fine. Hell, it's almost common knowledge that Linux's ACPI kind of SUCKS ASS.

I've gotta take the high-road on this one. You can't say Linux's ACPI implementation sucks, because it doesn't. Linux's implementation is based Intel's, and Intel wrote the fricking spec. What sucks is Microsoft's implementation. It just happens that most hardware has broken tables designed to be compatible with Microsoft's sucky implementation.

You can say "the user doesn't care who is at fault", and you'd be write, but at the same time, I'd argue it's a mistake to think that Linux should break its implemtnation to be compatible with Microsoft's. Standards are standards, and without standards we are nothing but savages. Compromising standards for compatibility is exactly what made web-browsers as ass-tastic as they are today.

Edited 2005-12-15 22:13

Oh really? Why, I never realized that... To the normal end user, it doesn't matter $%#Q$^ if the hardware deviates from the spec or not. If it doesn't work, it does NOT work. As simple as that. If Windows can support all those deviating specs, than why can't Linux?

Eugenia is no ordinary end user and she knows very well, that ACPI in linux works fine. The problem is that ACPI does NOT work fine on laptops.

She is badmouthing the wrong part in the problem. The problem is Microsoft raping standards once again and certain hardware vendors do the same.

That's why ACPI doesn't work well on laptops. Linux is doing fine, it's doing the right thing. The hardware vendors are doing wrong - and so is Microsoft.

Focus on the real problem, instead of blaming the innocent one.

BTW: Windows XP is not exactly portable. It might be, but we don't know since it only runs on x86 hardware. I do know it ran on a few platforms in earlier days. But that's quite some years ago by now.

Of course you agree with Eugenia. I'd expect no less. And I'd think less of you if you didn't give her backup.

linux's acpi support does not work fine for a lot of people... in fact, it's one of the reasons i'm using freebsd on my laptop. it just works. this is one thing linux really needs to work on. end users don't care if an operating system can run on a game console or pda or somebody's used gum that they stuck on the bottom of a chair somewhere. end users care that things work, and work well.

Like when you enter on a FreeBSD system : shutdown -h now

And instead of shutting down it write : You can now safely turn off your computer!

for me, that doesn't work either.