Certified to suck.

[Kintaro]

Hey guys. Turns out, I had the awesome privilege to look in to Linux certification this past week. You know, like how when Novell says SUSE runs well on some laptops, then, well, it should? Well, turns out they have a funny definition of "well".

http://linuxhaters.blogspot.com/2008/08/certified-to-suck.html

[Refalm]

Wow, I didn't even know there was any 3D acceleration possible on Intel Graphics with Linux. I always suspected that if you're not on nVidia or ATI, you're fucked.

Dell seems to use the same videocard in their Linux laptops:
https://ecomm2.dell.com/dellstore/basket.aspx?c=us&cs=19&l=en&s=dhs&itemtype=CFG&oid=24726e90-c168-4494-bb3d-a6fa3113ed8b

I wonder if they spontaneously reboot with Compiz enabled as well.

Also, here's an article on why ACPI sucks in Linux:
http://www.osnews.com/story/17689/Bill-Gates-on-Making-ACPI-Not-Work-with-Linux/

At least that is something that can't be helped, but Intel Graphics? That sucks, any onboard nVidia or ATI videocard would be just as cheap.

[Kintaro]

While we don't know if he actually managed to do just that (creating problems to other OSes to work well with ACPI), but if he did, it is a good explanation why ACPI has been flaky on the majority of x86 computers with anything else other than Windows (the older, APM standard, seemed more compatible with alternative OSes).

Sounds like someone is making really lame excuses for a standard that was made with complete transparency. The reason ACPI sucks on Linux is because Linux doesn't follow published specifications on ACPI. Fuck, it pretends its Windows to sacrifice every good ACPI compliant motherboard for every bad one that only likes Windows. Why do you think FreeBSD never had these luser problems with ACPI? They followed the standard.

Just found this in the comments...

"3) Wireless connection was not tested due to a lack of an Intel Linux driver for the wireless adapter, will perform wireless stress test when driver becomes available."

Holy shit, wireless doesn't work on a notebook and it gets certified? WTF good is a notebook without wireless?

Could you imagine if Microsoft bulled this bullshit on you? It'd be a shit storm.

[Orethrius]

Sounds like someone is making really lame excuses for a standard that was made with complete transparency.

You got me there!  600 pages with not a simplified version in sight - I believe someone's invented "obscurity through absolute transparency"!  For the record, Microsoft is one of five partners including Intel (remember the "Wintel" days before AMD became any kind of force, because I sure do), HP, Toshiba (two historically major Windows resellers), and Phoenix (the only BIOS to date that has ever had the annoying tendency to hide drive specifications from the administrator).  Add in this little number, and I wonder why people might get the impression that something else is going on here.

The reason ACPI sucks on Linux is because Linux doesn't follow published specifications on ACPI.

Neither does BSD.  You should really keep up with Nate Lawson more often.

Fuck, it pretends its Windows to sacrifice every good ACPI compliant motherboard for every bad one that only likes Windows.

If you're referring to suspend_to_ram, that's a STANDARD that BSD still hasn't gotten to work.  If you're referring to X frontends, one must wonder whether you have HALF A FUCKING CLUE about how X works.  For the record, when implemented properly, it's not supposed to hook the system - hell, even THAT is left up to vendor-specific drivers.

Why do you think FreeBSD never had these luser problems with ACPI?

Obviously you're just misinformed on BSD's current ACPI support status.  Neither Hibernate nor Suspend has full support yet, so they're in much the same boat as Linux.

They followed the standard.

You're arguing that this can be fully implemented without running afoul of various dipshit hardware-specific decisions.  I know you can't be THAT stupid.

Holy shit, wireless doesn't work on a notebook and it gets certified? WTF good is a notebook without wireless?

*strangles Kintaro with homemade CAT-6* Sometimes, you can't recreate hard-wired performance - see recent 10-gigabit developments like FiOS for reference.  Oh, and for the record, Intel has THE BEST track record with open drivers.  They get the same slack that people afford ATi Omega drivers on Windows because they've earned it (that is, nobody has a gun to their heads).

Could you imagine if Microsoft bulled this bullshit on you? It'd be a shit storm.

Reference Windows ME, Windows XP Gold, and Windows Vista for proof.  There's a difference between "it's been delayed" and "not in your lifetime" and you know it.

[Kintaro]

FreeBSD will still actually boot on a lot of hardware Linux will not with broken ACPI.

Anyway, I don't know what you are on about after that because nobody shipped a wireless device without a windows driver, that got a certified PC. You go on about 98, ME, Vista, but fuck, they all had wireless drivers in their time? Just what the fuck are you on about. It had nothing to do with performance: NO DRIVER EXISTS TO TEST.

This is a clear issue of a dodgy certification, given to a laptop that doesn't support suspend, wifi, and hardware 3D which are pretty fucking important on laptops. I can't believe you defend bullshit on this level.

[Orethrius]

FreeBSD will still actually boot on a lot of hardware Linux will not with broken ACPI.

I have not TO DATE hit a system that doesn't load up.  Care to start issuing particulars, or are we going to throw GRUB/LILO misconfigs in with the mix too?

Anyway, I don't know what you are on about after that because nobody shipped a wireless device without a windows driver, that got a certified PC. You go on about 98, ME, Vista, but fuck, they all had wireless drivers in their time? Just what the fuck are you on about. It had nothing to do with performance: NO DRIVER EXISTS TO TEST.

Just how many of those Vista-certified systems could run Aero again?  Oh, and I never mentioned 98 - quite possibly because it NEVER supported wireless (unless somebody came up with a third-party application to handle that).  Anyway, you miss the central point, though I imagine that's your intent: NO DRIVER EXISTS YET != NO DRIVER EXISTS PERIOD.  Intel has a good track record with that, and I choose to trust them.

This is a clear issue of a dodgy certification, given to a laptop that doesn't support suspend, wifi, and hardware 3D which are pretty fucking important on laptops. I can't believe you defend bullshit on this level.

BSD DOESN'T SUPPORT SUSPEND.  Stop acting like it does, they're still having trouble at the CPU level.
WIFI WORKS THROUGH NDISWRAPPER.  No, it's NOT hard, people keep bitching because they don't want to have to deal with it.  Talk to the hardware vendors.
HARDWARE 3D IS SUPPORTED THROUGH HARDWARE VENDOR REPOSITORIES.  Again, not everyone has them, but ATi / nVidia / Intel most certainly do.  You expect me to give half a shit if Anlalifgani Tek doesn't render correctly?

Hey, Kintaro, what's the difference between "Vista Ready" and "Vista Capable" again?  How does Oxford define them? 

[Kintaro]

Well you just lost "The Year of the Linux Desktop" in this massive break from convention. Nobody wants to spend 10 hours getting ndiswrapper, and from someone it doesn't take ten hours for (me) it sure fucking crashes to point of being useless on half the drivers I try. Fuck ndiswrapper.

[Orethrius]

Wait, WHAT?!?!?  Oh yes, please goad me some more for a system that VISTA doesn't even properly support!  If we're going for outright honesty here, it's evidently not "The Year of the Vista Desktop" either.  I think we've all learned by now that you're an atypical user - I'd honestly be surprised if you bothered with lspci yet (yes, I know who I'm bashing, and I DON'T CARE anymore).  I'm personally looking into TuxOnIce and its source because of things I've heard about it from both sides.  If it's a legit package, that takes care of the suspend issues.  Nice work dodging the 3D support through hardware vendor drivers, though!  Next time you rag on ndiswrapper, you might dump your PCI bus and dmidecode for people who might be working on usable frontends.

[Refalm]

that you're an atypical user

Kintaro was just making a point here.
An operating system should be easy enough for my shallow teen cousin, as well and my grandmother.

This isn't the fault of Linux, but rather the hardware manufacturer and the lack of PR for Linux.
It's pretty funny. When Linux has crappy drivers for certain hardware, people tend to blame Linux, but no one blames Windows for your piece of hardware got supplied with crappy drivers.

[worker201]

Kintaro was just making a point here.
An operating system should be easy enough for my shallow teen cousin, as well and my grandmother.

Disagree.  Some operating systems should be easy enough for your cousin's grandmother.  Not all of them have to be, though.  As long as there are open document standards and open hardware standards, a multi-tiered system of abstraction layers or user-friendliness or support or a combination of all 3 is totally workable.  In fact, an open market kinda requires there to be computers for tinkerers, computers for steady workflows, and computers for grandparents.  It's the open document and hardware standards that are getting in the way of this ideal - not the ease of use.

[Kintaro]

Kintaro was just making a point here.
An operating system should be easy enough for my shallow teen cousin, as well and my grandmother.

Disagree.  Some operating systems should be easy enough for your cousin's grandmother.  Not all of them have to be, though.  As long as there are open document standards and open hardware standards, a multi-tiered system of abstraction layers or user-friendliness or support or a combination of all 3 is totally workable.  In fact, an open market kinda requires there to be computers for tinkerers, computers for steady workflows, and computers for grandparents.  It's the open document and hardware standards that are getting in the way of this ideal - not the ease of use.

No, obviously if you want Microsoft's new patented formats, you should pay for Microsoft Office. Your ease of tinkering doesn't defy the rights of others. It's like submitting closed, unwilling individuals to "open" towards medical experiments without their consent.

You can run Microsoft Office fine on Linux with crossover office. You greedy bastard!

[worker201]

Who said anything about "Microsoft's new patented formats"?  Certainly not me.  And medical experimentation is a ridiculous metaphor.  Try again.

[Kintaro]

Well good luck Sir, only I know you'll all resort to force eventually anyway when you realize the pipe-dream you live.

[Aloone_Jonez]

The idea that you can patent a file format is the most retarded ever because it means that one organisation can effectively own everyone's data.

[Kintaro]

The idea that you can patent a file format is the most retarded ever because it means that one organisation can effectively own everyone's data.

File > Save As > RTF

Was that hard to relinquish their ownership?

EDIT: Just looked it up and its not patented, freetards are just too dumb to reverse engineer it.