LWN.net Logo

Lindows CEO attacks Intel's Centrino Linux lockout (Register)

The Register looks into allegations made by Lindows.com CEO Michael Robertson. "Intel is going through a major internal struggle over desktop Linux, and the pro-Microsoft marketing droids are currently winning, according to Lindows.com CEO Michael Robertson. As evidence, Robertson puts forward the lack of Linux support for Centrino, the mysterious blocking of his company's request to participate in an Intel roadshow, and the last minute pullout of Intel speakers at his Desktop Linux Summit earlier this year."
(Log in to post comments)

and now for something completely different

Posted Mar 20, 2003 14:28 UTC (Thu) by coriordan (guest, #7544) [Link]

...a man with no credibility says that something that is not good for his
company is evil!

(news at 11:00)

and now for something completely different

Posted Mar 20, 2003 14:43 UTC (Thu) by havoc (guest, #2261) [Link]

But look at the bright side – because he's attached himself to Linux, he can lend his dis-credibility to the whole Linux community!

... oh, wait a minute.

and now for something completely different

Posted Mar 20, 2003 15:07 UTC (Thu) by proski (subscriber, #104) [Link]

If companies are going to make money selling computers (especially notebooks) with Linux preinstalled, having drivers for all hardware components is a must. No amount of disclaimers of implied warranty will make happy their customers, who expect to be buying fully functional systems.

That's a good thing. It doesn't really matter in this case if a certain CEO has credibility in certain circles or not. Lindows is in the hardware market, and it's a potential customer of Intel.

If a fool says the sun rises in the East, does that make it wrong?

Posted Mar 24, 2003 11:34 UTC (Mon) by Max.Hyre (subscriber, #1054) [Link]

Evaluate the case based on the information presented, rather than who the presenter is. In the case of Mr. Robertson, we just have to use a bit more care to determine that we're being presented facts. If they pan out, we should apply ourselves to fixing the problem, rather than berating the messenger.

anyone have details on the Centrino thing?

Posted Mar 20, 2003 22:08 UTC (Thu) by mattdm (subscriber, #18) [Link]

As the article explains, "Centrino" is a marketing term for the Pentium M processor bundled with their new "Pro/wireless 2100" mini-PCI 802.11b card. I presume that no special drivers are needed at all for the processor and that the wireless is the real issue. So, then, *is* it a real issue? Not providing a driver is not really a big deal -- most companies still don't do that, and it wasn't long ago when no one at all did. The important thing is: are they making the needed information available to developers?

anyone have details on the Centrino thing?

Posted Mar 21, 2003 0:21 UTC (Fri) by barrygould (guest, #4774) [Link]

It probably uses different Northbridge and maybe even Southbridge chipsets, so drivers would possibly be needed for IDE, PCI, ...

anyone have details on the Centrino thing?

Posted Mar 21, 2003 8:30 UTC (Fri) by mattdm (subscriber, #18) [Link]

Won't those likely be similar enough to just work? But even if they're not, the real question: is Intel really refusing to provide information, or are they just not writing drivers?

anyone have details on the Centrino thing?

Posted Apr 29, 2003 16:08 UTC (Tue) by devophl (guest, #10941) [Link]

I just recently purchased a Gateway Centrino Laptop with Windows XP professional on it. I partition magicked the 30 GB drive and opened up a 8 GB drive for Linux. I installed Red Hat 9.0 on the system. As far as I can tell, the only thing that doesn't work is the wireless card. GRUB, Red Hat, graphics drivers all work well. I believe this system is faster than my 1.8 Ghz P4 which is pretty impressive for a 1.3 Ghz laptop.

If anyone has any ideas on wireless support for Linux Centrinos, I'd be curious to hear it.

Lindows CEO attacks Intel's Centrino Linux lockout (Register)

Posted Mar 21, 2003 5:55 UTC (Fri) by burdicda (subscriber, #10272) [Link]

Comments 1 & 2
Troll Out !!!!!

"Branding Restrictions" Whoa !!!

That's kind of clever like "Intellectual Property" Eh ???

Probably came from the same source (Closed Source)

* Sometimes it takes alot of practice to take down a Goliath
with a little sling

*World Domination, It's only a matter of time, with or without Intel LOL

Antitrust issue

Posted Mar 21, 2003 7:34 UTC (Fri) by hazelsct (subscriber, #3659) [Link]

Interesting that most of the comments posted here have been anti-Lindows, given that lack of vendor support is a real problem for all of us.

But I see an opportunity here: why not support Robertson and turn this into an antitrust issue? He's the kind of guy with the guts (and money) to file a class-action lawsuit on behalf of the community (well, on behalf of himself, but which could benefit the community). We can certainly say that those hardware vendors which provide neither complete Linux drivers nor sufficient specs to write them are creating barriers to entry of a viable competitor (cite large and growing server market share, non-negligible desktop share) into the operating systems market which is dominated by a convicted monopolist -- and possibly force that monopolist to fund those vendors' Linux driver devlopment! Because of the "computer time" phenomenon, we can probably even force the simultaneous release of Windows and Linux drivers. While we're at it, if there are punitive damages involved, and Linux developers and users get some of those, well then all the better!

Our government may have sold out to Microsoft in the damages they required on appeal. But it's nonetheless possible to extract damages in related lawsuits based on their monopoly position, and more importantly, to extract drivers and specs. I say it's worth supporting Robertson if there's a chance that could happen.

Antitrust issue

Posted Mar 21, 2003 9:55 UTC (Fri) by mattdm (subscriber, #18) [Link]

I don't see this as helping. We'll just end up with a bunch of buggy binary-only drivers. We need *good* relations with the hardware companies -- not lawsuits.

Antitrust issue

Posted Mar 21, 2003 16:15 UTC (Fri) by hazelsct (subscriber, #3659) [Link]

And bad drivers will just shame the companies that write them, reflecting badly on the quality reputation of the company as a whole.

On the other hand, a company which newly provides binary drivers will get introduced to the community, and after a couple of API changes which break all their closed stuff, will learn the value of providing source too.

It's a process, we have to win them over, and not just think we can beat them all over the head with "you must provide source NOW!" IMHO, it would be a small net plus at first (I hate binary drivers too), but a big win long-term.

Antitrust issue

Posted Mar 23, 2003 4:28 UTC (Sun) by banshee_dk (guest, #10290) [Link]

But don't you see, some (many) of us are unable to make use of the binary, if the vendor chooces to make it NON-FREE! It would break our systems -- I mean: Then our systems would no longer be FREE systems!
But OK, if you are into that OpenSource shit...
But we NEED FREEDOM, not PRISON! And then I don't care, what you call the system, but if GNU/Linux is going to be NON-FREE, then I (and others) can't use it anymore...

I say: DON'T DO IT

Antitrust issue

Posted Apr 18, 2003 18:40 UTC (Fri) by ikg (guest, #10738) [Link]

I think your position is a bit too radical. Of course, free (as in speech) drivers would be the best solution, but I don't see the point in refusing to use binary-only drivers. If considering performance issues, binary-only or source-provided drivers are equivalent. If considering philosophical issues, binary-only software is bad, but as long as freedom of software is concerned, i.e. ability to modify software to suit to your specific needs, I think such a radical position is pointless, because you would rather never modify the drivers (well, most of the time). Of course if the drivers imply some security or privacy issues, then it is rather a company's interest to have free software drivers -- it's the company's advantage anyway, but here I mean the company's short-time-vision interest -- but for some "hardware support" drivers, I think one cannot be too demanding.

Extremism will not convince not-yet-convinced companies as easily as tolerance and openness (not meaning by that at all "open-source"). Be clever and adapt available software to suit your needs, I say. Show how wonderful things can be done with your free software and appeal non-free software-related companies to follow you. Besides, it's way easier to convince hardware manufacturers than software companies.

Lindows CEO attacks Intel's Centrino Linux lockout (Register)

Posted Mar 21, 2003 9:18 UTC (Fri) by strombrg (guest, #2178) [Link]

Why should Intel bother writing drivers when they can count on some geek with something to prove to write the drivers for them? To some extent it's the mark of a product's success if someone in the linux community writes the vendor's driver for them - at no cost to the vendor.

We need to stop buying hardware if the vendor doesn't give official linux support with it, unless there is no other choice, even paying a little more to get real support.

Until then, until there's a financial incentive to greedy vendors to support linux, they won't bother supporting linux.

Lindows CEO attacks Intel's Centrino Linux lockout (Register)

Posted Mar 21, 2003 10:15 UTC (Fri) by JoeBuck (subscriber, #2330) [Link]

For drivers to be written, developers need the specs. It is appropriate to ask the Linux developers to write the drivers if Intel will release appropriate specs that will enable device drivers to be written without heroic reverse-engineering effort. Linux will be hosed if another major hardware manufacturer follows Broadcom's example and refuses to release specs for an 802.11{a,b,g} chip. I don't know whether Intel plans to release such specs or not.

In any case, we want to have open-source, maintainable drivers, not closed-source binary-only drivers whacked out by some company engineer who may have an excellent understanding of his/her device, but is likely to have a weak understanding of the Linux kernel, driver writing etc. (An open-source version initially written by such an engineer, even if flawed, can be improved by others).

Lindows CEO attacks Intel's Centrino Linux lockout (Register)

Posted Mar 22, 2003 7:55 UTC (Sat) by truffels (guest, #10283) [Link]

The problem is that security issues more and more will me handled by hardware. This is also of strategic importance to the US. Linux is not good for the empire ;-)

Lindows CEO attacks Intel's Centrino Linux lockout (Register)

Posted Mar 21, 2003 10:26 UTC (Fri) by pizza (subscriber, #46) [Link]

Now how can some geek with something to prove write a driver, when Intel refuses to share any hardware/programming docs?

When it comes to wireless under Linux, you have three choices -- Intersil Prism2, Orinoco, or Cisco Aironet. Oh, the older Atmel USB widgets. These are now all obselete. And indeed, it's becoming increasingly difficult to working stuff, as vendors like Linksys and D-Link change chipsets at the drop of a hat -- without changing their model numbers so you're never quite sure what you're buying. Meanwhile, the second generation of chipsets is out, from Atheros, TI, Intel, SiS, ADMtek, Realtek, Cirrus Logic, and generally anyone with access to a chip foundry. Hell, even Intersil doesn't want to share the specs on their new 802.11g/a chipsets, despite all of the good things that happened the last time they did.

There has been *zero* interest from the chipset people in providing source; with a couple of players so far releasing very buggy binary-only drivers for one specific kernel and calling that "linux support."

And you know what? Welcome to the future of hardware support, where source is the enemy because it "exposes our valuable intellectual property". And since supporting Linux is nearly impossible if you don't release source code... they choose not to.

Lindows CEO attacks Intel's Centrino Linux lockout (Register)

Posted Mar 21, 2003 11:42 UTC (Fri) by Baylink (subscriber, #755) [Link]

> where source is the enemy because it "exposes our valuable intellectual property"

I believe you've mispronounced "because it exposes our violation of other people's patent rights".

Lindows CEO attacks Intel's Centrino Linux lockout (Register)

Posted Mar 21, 2003 15:35 UTC (Fri) by pizza (subscriber, #46) [Link]

*giggle* Very, very true. That's what they're really afraid of.

Lindows CEO attacks Intel's Centrino Linux lockout (Register)

Posted Mar 21, 2003 12:30 UTC (Fri) by mattdm (subscriber, #18) [Link]

Now how can some geek with something to prove write a driver, when Intel refuses to share any hardware/programming docs?

Are they refusing? Who's asked? What did Intel say back? I can't find any background on this via Google.

Lindows CEO attacks Intel's Centrino Linux lockout (Register)

Posted Mar 23, 2003 4:51 UTC (Sun) by banshee_dk (guest, #10290) [Link]

Very very true!
In all this we assume that what Lindows.com CEO Michael Robertson says is true, but has he ever really (I mean REALLY) tryed?


Who, said he had "no credibility"? ;-)

Lindows CEO attacks Intel's Centrino Linux lockout (Register)

Posted Apr 23, 2003 3:30 UTC (Wed) by tom (guest, #10836) [Link]

It amazes me ( a non developer with an interest in spreading the linux word) how we can expect anybody to move from w**dows to linux when we appear to be such a back biting bunch of ingrates.
I for one do not know if Intel will release the required information or not, or if anybody out there can write the drivers, but I am sure of one thing:
If the lindows ceo is right then slagging him off does only damage to all involved in linux in general.
If he is wrong then and only then can you be in a position to pass detrimental comments.

But, regardless of the outcome of this particular issue, isn't it about time that we all worked together?

Lindows CEO attacks Intel's Centrino Linux lockout (Register)

Posted May 11, 2003 23:14 UTC (Sun) by madmoc (guest, #11136) [Link]

ok im not quite sure as to the topic but if everyone wants to use wireless why not buy a router that runs linux on it.
if you have a PC that runs linux use linux not cisco or any other companys hardware and i know that there is linux routers out there. maybe not wireless but there is linux routers out there. hell cisco built there router OS on unix. pretty much the same thing as linux. if they want to play a game of we dont like you on the playground then move to the other side

maybe im wrong and if i am tell please.
madmoc ccna+

Copyright © 2003, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds