Home > OpenBSD > Theo de Raadt Says Intel Only ‘Open’ for BusinessTheo de Raadt Says Intel Only ‘Open’ for Business Submitted by Johan Berg Thom Holwerda 2006-10-01 OpenBSD 76 CommentsTheo wrote a letter at large to the open source community urging us to take a stand. In the letter Theo provides information on how to contact Intel and helpful hints on how to explain benefits of open documentation to any vendor. About The Author Thom HolwerdaFollow me on Twitter @thomholwerda 76 Comments stare 2006-10-01 9:59 am EST http://ipw2100.sourceforge.net/http://ipw2200.sourceforge.net/http://ipw3945.sourceforge.net/ abraxas 2006-10-02 12:39 pm EST The problem that Theo is having is the distribution of the firmware. Without it the drivers are useless. Cloudy 2006-10-02 6:11 pm EST The problem that Theo is having is the distribution of the firmware. Without it the drivers are useless.No. The problem that he has is that he’s not willing to accept Intel’s terms for distributing the firmware. Intel is willing to distribute the firmware under conditions Theo is unwilling to accept. abraxas 2006-10-03 12:01 am EST No. The problem that he has is that he’s not willing to accept Intel’s terms for distributing the firmware. Intel is willing to distribute the firmware under conditions Theo is unwilling to accept.Exactly as I said. Theo’s problem is the distribution of the firmware. hhcv 2006-10-01 11:01 am EST It would be fantastic if we lived in a world where all hardware vendors realised it is actually in their interests for people to be able to use their hardware as they see fit.. But, some companies are too large, stubborn and short sighted to do so. The problem is, most of these companies will only react to the market, and the best way forward is to create a “market reaction.”I’d like to see a unix hardware compatibility index, where vendors, small OEMs and us diy-ers can support those companies that support us. If intel et al don’t want to support us, then there is no reason we need to support them – especially if there are alternatives.Open source projects allow hardware to do pretty amazing things – often better than their closed counterparts. 1st choice but second fiddle?Edited 2006-10-01 11:02 binarystar 2006-10-01 11:20 am EST AMD have been great supporters of open source … next time it comes to replace your desktop or work servers choose AMD hackmykack 2006-10-01 12:12 pm EST Hey binarystar,U just might have something there … I dont mean that I know if AMD is better at supporting opensource or not …. what I think is that too much time is spent on people like Theo or the Linux Kernel Devs etc. telling us … look he doesnt provide good support … boohoooo …Fine we use opensource coz we like it right … c’mon tell us what we SHOULD use then.Dont give us stupid news like X doesnt support our software.I know that everytime I look to build a new PC I have this dance with millions of forums trying to find out which mix of hardware will be supported perfectly with my choice of bsd or distro … none & I mean NONE of these guys have comprehensive HCLs on their websites …..Go figure ….Hence OpenLetters like Theos’ appearing from time to time has virtually NO effect on the BIG vendors.Kernel devs or BSD devs are in the best position to tell us what to buy …. they know perfectly well what hardware is supported properly.Its just that simple …. JATF 2006-10-01 12:26 pm EST Theo has recommended buying taiwainese stuff i believe Dirge 2006-10-01 12:36 pm EST Sure but what stuff is that? I agree, a complete HCL would be helpful.I did see these sites linked in the commentshttp://vendorwatch.org/index.php?title=Main_Pagehttp://www.fsf.org/campaigns/hardware.html hackmykack 2006-10-01 12:44 pm EST hey Dirge,took a look at those links man …thats exactly what I mean …Every release of the kernel or bsd should have a listing of new hardware supported perfectly … out of the box ….The devs are in the best position to tell us this …Then I as an opensource user can vote with my feet.Now wouldnt that be sooo much more useful … Ookaze 2006-10-02 3:17 pm EST Every release of the kernel or bsd should have a listing of new hardware supported perfectly … out of the box ….The devs are in the best position to tell us this …Then I as an opensource user can vote with my feet.Now wouldnt that be sooo much more useful …What you want is not possible and comes from the fact that you don’t know what you’re talking about.I mean, marketing is designed so that you don’t understand what you’re talking about.You see, there’s a difference between a product with some name, and the chips inside it, for which drivers are written.The kernel devs can know about those chips and write drivers for them. But they can’t know every name of every products that use these chips.And the IHV don’t want you to know these relations chips-product either, as they can make you pay more for a functionality that is just a config option to check in the driver (but not accessible to the consumer).Do you see the problem ? hackmykack 2006-10-01 12:39 pm EST hey JATF,That here is a perfect example of what I mean …What Taiwanese stuff exactly …Which model of which vendor with which firmware level, if applicable ….This is just the sort of thing I mean really ..C’mon kernel or bsd devs …. help us out some …. l3v1 2006-10-01 1:56 pm EST Well, not to protect them, but if you have some hw knowledge and have compiled some vanilla kernels a few dozen times, you will have a pretty fair knowledge about what is supported and what isn’t. With that I don’t want to tell you you’re ignorant – go figure – what I want to tell is there are lots of people out there who are not kernel developers and still could help in putting together a searchable HCL updated at every major kernel release. No need for concrete names and types, but at least for chipsets, firmwares, etc. Many less knowledgeables and beginners would find that fairly useful. Soulbender 2006-10-02 7:52 am EST “putting together a searchable HCL updated at every major kernel release”Except it wouldn’t work for products where the vendor changed chipset without changing the hardware version number in any way. As crazy and insane as that sounds it’s not uncommon practice. gpierce 2006-10-01 7:22 pm EST In my experience finding out exactly what hardware and chipsets are perfectly supported is not in itself difficult; however, it is not always apparent from the manufacturers website exactly what chipsets a wireless card may have. Branding constantly changes. An example are the Orinoco series of wireless cards. You would think they use the Orinoco drivers in the kernel, but not all of them do. It’s a pretty confusing situation in the wireless world. I rely on word of mouth but mostly on forums. SomeGuy 2006-10-03 12:20 am EST You mean a list like this one?http://www.openbsd.org/i386.html abraxas 2006-10-01 2:40 pm EST The best HCL for Linux is available on Redhat’s site.https://hardware.redhat.com/hwcert/index.cgiI used to use this list a lot but after configuring and compiling literally hundreds of kernels I am pretty aware of what is supported and what is not. mdoverkil 2006-10-01 10:06 pm EST Here is the OpenBSD HCL. Just pick which platform you are usinghttp://www.openbsd.org/plat.htmlhere is the i386 as an example:http://www.openbsd.org/i386.htmlFreeBSD HCLhttp://www.freebsd.org/releases/6.1R/hardware.htmlThose looks like a pretty complete lists to me =) hackmykack 2006-10-02 5:13 am EST Hey medoverkil,Thats the stuff ….I am not hardware expert or anthing but that is exactly what I meant …More Opensource Software projects should do that …Just state it explicitly ….An be sure to mention the level of hardware support … eg. works perfectly with all features etc.Also … if the vendor was helpful if providing docs etc.I know that many people dont care about hardware drivers being open or not but … then again … many people do … so that wont hurt tooo …When Hardware vendors see the trend of people naturally opting for “open” hardware over their “closed” counterparts … thats when its a done deal.Please note that I am not saying that linux does this better or bsd does that better … if u do choose to take it that way ..then I guess the human torch said it best …. FLAME ON ! … hehehe kaiwai 2006-10-01 2:17 pm EST Yeah, supports opensource so much that they can’t be bothered providing the necessary documentation for the ATI graphics cards that they now sell.There is a difference between wraping oneself in the opensource flag, and actually doing something that embraces the ethos of opensource and open documentation. MikeGA 2006-10-01 2:36 pm EST Hold your horses, the acquisition was only officially announced last month! happycamper 2006-10-01 11:59 am EST Don’t ever forget that you paid for their hardware; you don’t owe them anything. However, they owe you a working product.heck yeah, i want a working product.I tried Damien Bergamini drivers on my pentium m centrino based laptop, OpenBSD detected the intel centrino wireless network chipset except for the radio transmitter.Edited 2006-10-01 12:11 Adurbe 2006-10-01 12:22 pm EST their product does work, it up to the software to take advantage of what the hardware can do! dylansmrjones 2006-10-01 12:36 pm EST The software cannot utilize the hardware without specs… l3v1 2006-10-01 1:58 pm EST Yeah, like it would be so easy to use a blackbox hardware part. Real life often is different than a matlab system identification toolbox. lolownia 2006-10-01 12:02 pm EST help it spread: http://digg.com/linux_unix/Intel_Only_Open_for_Business . velko 2006-10-02 11:48 am EST Could somebody tell me why this post has been modded up? Is it particularly insightful or worth read? I have nothing against digg but I think that modding such comment up is very… strange. jessta 2006-10-01 12:32 pm EST This is one more reason the windows monopoly is horrible. By having such a large number of users on a single platform the hardware vendors can support only that single platform without lossing enough customers to care about the other platforms.I’m sure this will continue even with the increasing exceptance of GNULinux as a platform. Vendors are going to continue to release binary drivers that only work on x86 Linux, when they could release documentation of their hardware and have support for all platforms at no extra cost to them.Seems crazy not to do it. They must be crazy.– Jesse McNelis WorknMan 2006-10-02 5:41 am EST This is one more reason the windows monopoly is horrible. By having such a large number of users on a single platform the hardware vendors can support only that single platform without lossing enough customers to care about the other platforms.Well, it only sucks for those on the ‘other platforms’, but for the rest of us, it’s pretty sweet knowing that if a piece of hardware were released tomorrow, there’s a 99% chance it’ll work on our OS right out of the box.I’m sure this will continue even with the increasing exceptance of GNULinux as a platform. Vendors are going to continue to release binary drivers that only work on x86 Linux, when they could release documentation of their hardware and have support for all platforms at no extra cost to them. This is a good point. If I were going to sell a hardware device, I’d write a Windows (and maybe a Mac) driver and at the same time, offer enough technical specs so that the *nix crowd could fend for themselves. I mean, if there are dumbasses out there who are willing to work for free and write their own drivers, who am I to deny them of the privilege ? Ookaze 2006-10-02 3:48 pm EST Well, it only sucks for those on the ‘other platforms’, but for the rest of us, it’s pretty sweet knowing that if a piece of hardware were released tomorrow, there’s a 99% chance it’ll work on our OS right out of the boxThat’s assuming it will work on your OS out of the box because you have the driver. My experience show me that most of the time, the driver DOES NOT work well with the hardware. I would go as far as to say that the driver NEVER works well in Windows. So no, you know you will have a driver, you don’t know if it will work at all.Feel free to show me even 1 product which only had one driver, or no updated driver with fixes for critical flaws.My personal case is even worse : no hardware I own worked as well on Windows (98 and XP) as they do on Linux, some aren’t even supported on WinXP.Saying that it’s pretty sweet on your platform is a very short sightened view too.I mean, if there are dumbasses out there who are willing to work for free and write their own drivers, who am I to deny them of the privilege ? Are they really dumbasses ? I don’t think so. People who can write their own driver sure enough are not dumbasses, and most don’t do it just because they can.Most get a high benefit from doing it. deathshadow 2006-10-02 5:49 pm EST >> My personal case is even worse : no hardware I own>> worked as well on Windows (98 and XP) as they do on>> Linux, some aren’t even supported on WinXP.Guhugafugah?!?!? Now, I don’t mean to say my BS alarm is going off – but WHAT?!?Ok, I need details on this one, as that has to be THE MOST NONSENSICAL THING I’VE EVER HEARD. What are you doing, futzing around with ISA cards or something? Own nothing but SparcStations??? Own nothing but apple hardware?Trying to wrap my head around this one…Edited 2006-10-02 17:54 Ookaze 2006-10-02 6:49 pm EST That’s because you still live in your ideal world where Windows works every time without any problem, where every app works out of the box and as advertised, where all hardware have drivers.Well, I’ll just give you the hardware names out of the top of my hat : mustek 12000 SP driver (SCSI scanner), an old TV card (by a company bought by Hauppauge I think, can’t remember the name), serial gamepad adapter.And the scanner was always stuttering under Windows 9x, I thought it was normal. When I switched to Linux in 2001, it scanned flawlessly and smoothly.Strange that reality is nonsensical to you, though, you must be really young in computing. deathshadow 2006-10-02 7:33 pm EST Young in computing… BWAHAHAHA… oh man, you slay me.mustek 12000 SP driver (SCSI scanner) – ok, so a DECADE AND A HALF old 8 bit SCSI greyscale scanner you could replace with a USB device with better capabilities and faster scanrates for $40an old TV card (by a company bought by Hauppauge I think, can’t remember the name)– a product that likely also hasn’t been produced in years… Probably a Zoltrix, meaning again a $30-40 replacement for something with twice the capabilitiesserial gamepad adapter – Serial? Do you mean USB… no, you said adapter… lemme guess, the old hardwire a bunch of diodes to a PIC chip to run a NES/SMS controller routine? (I find it unlikely you have a Gravis Stinger – those never worked) Odd, PPJOY seems to work fine here… (but then, I’m running my SMS gamepad off the parellel port since I need the serial port for the Master System LCD shutter glasses) Not that said complaint can really be considered valid, since good linux games can be counted on one hand, unless of course you are playing two decade old arcade classics under MAME.Ok, so my initial guess was correct, you are trying to keep afloat a bunch of hardware from the ISA era… Usually I applaud the effort, but after THREE DECADES of doing this stuff, sometimes you’ve just got to throw it all in the trash – your complaint is akin to being upset your Trash-80 Model 1 won’t run that 20 gig hard drive. (you need a model 4 ‘cage array’ or model 4P for that)Unlike you, I don’t expect anything more than five years older than the last major revision of an OS to be supported in Windows… and I’ve found that my current hardware – Nforce4 mainboard, Ge7600GT, Audigy 2, Adaptec 29160, Acer 640S scanner… even my Brother printer and multiple monitor setups… Just work better / to their full capacity under XP – and /FAIL/ hard under linux. (on the other hand Windows isn’t allowed anywhere NEAR my Dual Xeon server – THAT runs Sarge) STILL have yet to get 3d accelleration under linux – and I’ve switched video cards 3 times, still no go even on distro’s that allegedly do it out of the box – much less working 5.1 audio, soundfont support… (I generally run 512 megs of SOUNDFONTS since I do music composition as well) doesn’t know what to do with my Thrustmaster cougar… etc, etc, etc.FIVE YEAR rule man – Three years is obsolete, five years is trash-heap time. Planned obsolescence and so forth… Of course that actually illustrates something I’ve always noticed about linux; You are more likely to find hardware support for decade old stuff than you are something you bought new last week… which is annoying since the more ‘useful’ desktops keep upping the system requirements to the point they run like sludge on anything less than the latest and greatest – BAD catch 22.Edited 2006-10-02 19:37 Ookaze 2006-10-03 5:27 am EST NO, my scanner was a color one, and was not 8 bit. It was 24 bit I think. Recently its power adaptor died so I bought another cheap USB scanner instead of buying another power adaptor.The TV card is NOT a Zoltrix. It was a very well known brand that was bought back by one of the two big TV card vendors right now (so I thought it was bought by Hauppauge).And I meant serial gamepad adapter for console gamepad (Playstation). And BTW on Linux, I have now a special USB gamepad adapter with a mat for pydance, and use it also to plug a gamepad to play ALL games, as, you see, I prefer using a gamepad instead of a joystick, they are also much more usable.And none of these devices are ISA, it’s one SCSI, one PCI and one serial. These were bought less than 7 years agon so you’re even more incorrect. I even have an old 8 GB SCSI disk that surely trash your latest SATA one in normal use. You seem to be the elitist type that buy back all its hardware every year because : it’s not supported anymore by your so called superior OS. FYI, my adaptec SCSI card was bought at the same time as the other hardware cited. It still works better in Linux than it was in Windows XP, where it caused a blue screen after installation (and the case is known to MS). The NForce4 on another PC mainboard works very well too. But I won’t continue, as your other facts are just wrong (like Ge7600GT not working with 3D acceleration on Linux, or Audigy2), and you’re obviously a troll. FYI, some people do professional music recording just like you on Linux.Like I said, 5 years sure enough is not obsolete.And what you simply miss, because you’re lost in your constant reinstalling or changing of OS, is that I setup my main PC in 2001 and installed Linux on it : it works flawlessly till now, I never changed the OS, and do not feel the urge to replace anything in it, except when sth dies. It still works as fast (no, even faster, with KDE and Gnome being faster) as before.And all my hardware (which is surely not a decade old) is still fully supported under Linux. Starting 2006, I made a PVR with my very same OS and MythTV, with a lot of hardware which is less than 5 or 3 years old (like SATA disks), for example.When I made my own custom Linux in 2001, Windows (yes, even XP) was still unable to burn a CD when doing tons of other thing, without a risk of losing the CD. They invented burnproof specifically for that !! This is just one of the thing that was vastly superior, driver wise. I cited the faster SCSI, the handling of unsupported hardware under XP, I’ve not cited the bluescreen of Windows XP with my Creative Live 5.1 card (which was dieing, but in Linux, never crashed the OS), the fact that when one of my MB died and was replaced, Windows refused to run or even be repaired with the new MB while Linux just ran like nothing happened, but there are others. deathshadow 2006-10-03 2:03 pm EST >> NO, my scanner was a color one, and was not 8 bit.>> It was 24 bit I think.Then it is NOT a mustek 12000 SP – The Mustek 12000 SP is a 12 bit grey 1200×600 optical SCSI 1 scanner designed for Window 3.1, NT 3.5 and MAC OS classic of that era – placing it WELL over a decade old.>> And none of these devices are ISA, it’s one SCSI,>> one PCI and one serial.The reason I mentioned ISA is it all SOUNDS like stuff of the ISA era, not necessarily ISA… It sounded like you were using stuff that was made when VLB was still standard equipment – now it just sounds like you don’t even know what your own hardware is.>> I even have an old 8 GB SCSI disk that surely>> trash your latest SATA one in normal useThat would depend on the drive and what you are doing with it… Since U160 and normal SATA end up roughly the same in performance (I doubt you have a 8 gig that supports U320), it’s probably a wash since most of the 8-10 gig drives, even the 10k RPM ones were 5-6ms on latency, while most SATA drives today are pushing close to 4ms latency (even if they lose a bit on seek time, they make up for it elsewhere)… of course, with those drives running SATA 3gb/s it probably starts to shift in SATA’s favor, and I bet my 10K RPM SATA150 WD Raptor trashes your old scsi drive rather handily – though 15K/U320 SCSI is still ‘where it’s at’ for speed. (which is why I have my windows swap, linux swap, autodesk hash and adobe scratchspaces all routed to a 15K RPM 36 gig Hitachi Ultrastar.>> The NForce4 on another PC mainboard works very well>> too. But I won’t continue, as your other facts are>> just wrong (like Ge7600GT not working with 3D>> acceleration on Linux, or Audigy2)WHEN I can get the mainboard to let the damned PCI express interface work right, letting the accellerated drivers load, I get ‘white space’ bleedthrough like frame limiting is off, and it invariably locks up X within moments of starting anything openGL based – and that’s after jumping through the endless hoops to even INSTALL the blasted thing… Sorry but if I wanted to spend two hours dicking on the command line to get something simple to work, I’d still be playing with Xenix on a Model 12.>> FYI, some people do professional>> music recording just like you on Linux.In yes… out? Not so much. >2 channel analog output is broken, scratchy, and reminiscent of the old VIA vs. SB Live ‘crackle’ – leading me to think it’s actually a conflict between the emu10k driver and ACPI. (since that was the cause of the problem with the live – you turn off ACPI in bios and problem solved)Smurf is pathetic compared to Vienna (which is usually what it’s compared to), there’s no midi comp program that actually lets you change soundfonts on the fly (you can only preset whatever happens to fit in memory) – Let’s face it, Cakewalk Sonar kicks ass, everything else is a chocolatte wannabe…>> was still unable to burn a CD when doing tons>> of other thingWin98 maybe…(though thinking on it… uhm, no…) 2k/XP had no such issues I was ever aware of – and at least back then I didn’t have to go to the command line every time I wanted to switch CD’s. (sorry, cheap shot)It’s like the old joke of the *nix user who claims years of uptime and swapping out Hard Disks, CPU’s and RAM without ever having to reboot… “Wait, didn’t you have to reboot last week?” – “Yeah, but that was just to attach my USB flash disk” (I say, I say, that’s a joke son…)Sorry, but your configuration seems highly unique (polite way of saying ‘out there’) and certainly your experiences run contrary to not only most everything I’ve ever heard about linux, but also my own experiences with it – which is why on my main PC it’s always been an ‘also ran’. Better hardware support as a rule is the LAST THING the majority of people think of when it comes to linux… unless of course you are building a server or are happy with early 1990’s level computing needs.Oh, and five years IS obsolete for one reason – you try getting 1:1 replacement parts at the three year mark, much less the five. The average life expectency of most electrical devices these devices is usually about 18,000 hours… aka just over 2 years continual operation… Most people don’t leave them running all the time, so naturally this works out to about 6 years on the outside before total failure… some devices will die sooner and need replacement, others won’t.Which is why if you tried to replace that… let’s see… 2001 so that’s a 1 to 1.5ghz class P3 or regular Athlon? 2ghz P4 or a XP 1500 if you sold your kids into slavery (go ahead folks, it’s ok to laugh)… If you bought Athlon, you’re at least lucky socket A held on like a rapacious swamp sow, but if that’s a socket 423, a replacement mainboard would run what? Does anybody even sell those new anymore?Edited 2006-10-03 14:17 WorknMan 2006-10-02 8:37 pm EST That’s assuming it will work on your OS out of the box because you have the driver. My experience show me that most of the time, the driver DOES NOT work well with the hardware. I would go as far as to say that the driver NEVER works well in Windows. So no, you know you will have a driver, you don’t know if it will work at all.Well, it depends. A bad driver is a bad driver, no matter what OS you’re using. But there are a lot of devices out there that work flawlessly. I generally tend to only purchase quality drivers, so if you spend $10 on a new video card, you’re pretty much gonna get what you pay for. I doubt the drivers would be any better in Linux is they bothered to write one.My personal case is even worse : no hardware I own worked as well on Windows (98 and XP) as they do on Linux, some aren’t even supported on WinXP. Ya know, I keep hearing how much better Linux works with hardware, but if that’s the case, why is it that so many people are on here bitching about it?Are they really dumbasses ? I don’t think so. People who can write their own driver sure enough are not dumbasses, and most don’t do it just because they can.Most get a high benefit from doing it.Well, even if I could write my own drivers, I don’t think I would, unless I were being paid to do so. Life is too short that kind of thing. macisaac 2006-10-01 12:47 pm EST hopefully this type of letter can dismiss the recent (in my opinion highly inflated) happy-feel-goodness folk were spreading about intel and open source friendliness due to them partially opening some of their graphics chips.intel is one of the most predatory and vicious corporations out there in the tech world (basically the microsoft of the hardware world in some ways, if not worse). an (even more than now) intel-dominated landscape is a very bad thing, we shouldn’t be encouraging it. now, if they actually want to change and clean up their act, fine. until then, bad press like this is well deserved. abraxas 2006-10-01 2:46 pm EST hopefully this type of letter can dismiss the recent (in my opinion highly inflated) happy-feel-goodness folk were spreading about intel and open source friendliness due to them partially opening some of their graphics chips. I will say from a Linux perspective that Intel has been very friendly towards OSS. They contribute to the kernel. They also have well supported in-kernel drivers unlike a lot of other companies. Nvidia has their own propietary drivers but have mostly been reverse engineered. Other companies don’t even have binary drivers or open specs.Could Intel be better? Yes, but as far as the big companies go they are one of the best. We should be more worried about GFX card companies and wireless cards from other vendors. deanna 2006-10-01 2:38 pm EST If this issue matters to you, please help get the word out by digging the story:http://digg.com/linux_unix/Intel_Only_Open_for_BusinessMore in-depth analysis on kerneltrap:http://kerneltrap.org/node/7184And the inevitable http://bsd.slashdot.org/article.pl?sid=06/10/01/0230234 Redeeman 2006-10-01 3:14 pm EST yes.. very very very open ipw3945 drivers… NOT. just as their supposedly open graphics drivers, except that they just arent open.edit:oh yes, and i almost forgot, their non-redistributable closed firmware for ipw2100 and ipw2200Edited 2006-10-01 15:16 ryan 2006-10-01 7:21 pm EST I’m replying to you right now via a ipw3945 chip on Ubuntu Dapper that installed without any intervention and runs perfectly both with the default tools and with NetworkManager. It may not fit your definition of open but it works great for me. deathshadow 2006-10-01 3:20 pm EST Sorry, but I know enough to get twitchy when people start throwing around words like ‘ecosystem’.There is this… “sense of entitlement” Open source advocates seem to be stuck on, where they seem to think the chip they bought on an entire board at $30 retail (which after packaging, distribution and taxes ends up at somewhere around $5 for the actual manufacturer) entitles them to somehow have access to the designers entire multi-billion dollar R&D budget to use it with something like linux – an application the developers did NOT spend money on creating it for in the first place.You can usually recognize this overzealous idealism and naivete when they start throwing around words like ‘rights’ or ‘ownership’ – without actually seeming to understand what those terms actually MEAN… combine this with the rampant anti-capitolist zeal statements like “only open for business” bring to the table, and the net result is that the majority of people who’ve actually got experience in business being tempted to tell said open source advocates to go back to their drum circles and tofu. twenex 2006-10-01 5:16 pm EST As a PC user, if I go into a store that sells hard disks, I can buy just about any hard disk available, whether from IBM, Western Digital, Fujitsu, or whomever.Software developers who develop closed-source software (CSS) are asking for special pleading. At the moment, unless you are lucky enough (a) to have hardware that has (usually reverse-engineered) non-Windows drivers, and (b) not to need software that only runs on Windows, and/or the advanced features of MS Office or other Windows-only applications, your only option is to buy a Windows-compatible operating system, and your only option in buying a Windows-compatible OS is to buy it from Microsoft.(I’m perfectly aware that ReactOS is a perfect alternative for some people – but how many people know about it? And is it usable for people whose primary operating system is a non-Windows OS already loaded on the disk?)This type of lock-in is brilliant for developers, but extremely detrimental for consumers – which is exactly why companies, from IBM to Microsoft to AT&T, have lost lawsuits and been bitch-slapped for using this tactic by what is arguably the most business-friendly national government in the world.It’s also why I refuse to use CSS except under duress.Edited 2006-10-01 17:18 Soulbender 2006-10-02 7:54 am EST How the hell did this get modded plus 4? ronaldst 2006-10-02 8:47 am EST Because it’s accurate. Ol’boy Theo is right there bashing away Intel because he didn’t get his way. Theo needs to grow up and so do the people cheering him.Apparently Intel deserves to look bad. But in the end, Theo comes out looking like a complete moron. And he’s built a reputation out of it. IanSVT 2006-10-02 2:47 pm EST Why does he need to grow up? He’s done more for open source than most “open source” companies have done. The guy actually codes too. He deserves to bash anyone who claims to be open but in reality, isn’t. I don’t care how abrasive he is. His abrasive personality has managed to do far more good for software than bad. Until the latter outweighs the former, he should be able to rip anyone standing in the way of progress. You can agree or disagree with him, but saying he looks like a complete moron looks like he bothers you on an emotional level and not technical. And I don’t think that’s relevent to this discussion.By the way, I’ve never touched openBSD. But any sys admin out there has probably used something he helped build or built himself. ronaldst 2006-10-02 5:15 pm EST @IanSVTWhy does he need to grow up?His abrasive attitude (his recklessness and lack of common sense) has cost him plenty of lost funds. PLUS you don’t make friends attacking people.Damn OSS fundies are thick headed. Get back into reality. IanSVT 2006-10-02 8:05 pm EST His abrasive attitude (his recklessness and lack of common sense) has cost him plenty of lost funds. PLUS you don’t make friends attacking people.Damn OSS fundies are thick headed. Get back into reality.Lost funds, friends? What’s next, is he going to be banned from a country club? What do either have to do with the technical merit of his argument(whether his is right or wrong)?How do you know I’m an OSS zealot in any way? Seems to me, he’s advocating open drivers. Let’s face it, hardware vendors make bad drivers. Most hardware vendors are offenders of this. But hey, you’re more than welcome to use craptastic driver sets hardware manufacturers write as an after thought to the hardware itself because you’re too thick to see that “open” drivers could actually improve the quality of those drivers. Don’t throw the notion at someone of being a zealot when you have proven yourself to be so completely one sided that you can’t see the benefits of OSS. They aren’t the be all end all. I can see that. However, you seem blind to the fact that there is worth there.And for de Raadt, he’s one voice in many. If he’s enough to turn intel off to the idea of completely open source drivers, then there’s something more wrong than him being an ass, which he can be.Edited 2006-10-02 20:06 Cloudy 2006-10-03 6:00 am EST What do either have to do with the technical merit of his argument(whether his is right or wrong)? The issue at hand isn’t about technical merits. It’s about the politics of corporations interacting with the open source community. Soulbender 2006-10-02 10:04 pm EST “His abrasive attitude (his recklessness and lack of common sense) has cost him plenty of lost funds.”His “abrasive attitude” has done more for OSS and software quality than any of your (or mine) comments on osners ever have.“PLUS you don’t make friends attacking people. ”I guess he’s not in it to make friends. It’s not a plushy, feel-good club.The people who needs to grow up are the ones who think that compromising your integrity and sucking up to big business in the name of “playing the game” is the right thing to do. At least Theo has enough backbone to stand up for his opinions and ideas. Cloudy 2006-10-03 6:06 am EST The people who needs to grow up are the ones who think that compromising your integrity and sucking up to big business in the name of “playing the game” is the right thing to do. At least Theo has enough backbone to stand up for his opinions and ideas.This isn’t an issue of integrity or backbone. If this was about Theo’s “integrity”, there would be no Intel version of OpenBSD; since OpenBSD won’t boot on an Intel platform without a BIOS, and BIOS code is distributed under even more restrictive licenses than firmware.This is an issue of knowing what fights are worth picking and which are not. The open source community needs to recognize that the cost to a company like Intel to provide the sort of support that Theo wants is far greater than the return to them from the small number of sales that would result.Let me turn the question around: How much are you willing to spend to make it worth Intel’s while to produce the sort of wifi chips that would satisfy the FCC’s certification process if they had open source implementations of the drivers?Theo calling people names is cheap theatrics. It gets the attention of the bloggers and it helps people feel warm and fuzzy about his “integrity”. But at the end of the day, they still don’t get what Theo is demanding, and Intel becomes just a little harder to work with. Soulbender 2006-10-03 8:12 am EST “How much are you willing to spend to make it worth Intel’s while to produce the sort of wifi chips that would satisfy the FCC’s certification process”Nothing, since the FCC certification means very little for people not living in the U.S.A.“But at the end of the day, they still don’t get what Theo is demanding, and Intel becomes just a little harder to work with.”Eh, what he is demanding is that they remove the completely pointless restrictions on the wifi firmware redistribution. Want to see theatrics? Read those restrictions, it’s theatrics at it’s best. Cloudy 2006-10-03 6:02 pm EST “How much are you willing to spend to make it worth Intel’s while to produce the sort of wifi chips that would satisfy the FCC’s certification process”Nothing, since the FCC certification means very little for people not living in the U.S.A.It matters indirectly, since Intel sells most of its product in the US. For them to ignore the FCC is to give up more than half the market for the product.“But at the end of the day, they still don’t get what Theo is demanding, and Intel becomes just a little harder to work with.”Eh, what he is demanding is that they remove the completely pointless restrictions on the wifi firmware redistribution. Want to see theatrics? Read those restrictions, it’s theatrics at it’s best.Theo’s been on this rant for over two years. Here’s what he says is “pointless” about the restrictions:Currently there is a contract that vendors must agree to, and a seperate contract that users must agree to. In their mind a vendor is someone who distributes. In our mind, anyone can be a vendor or a user. That is why we wanted no agreement, just let us do distribute,use, and redistribute these files. The vendor contract locks your users in.That’s right, folks, Theo’s sticking point is that he doesn’t want OpenBSD to be a ‘vendor’, although any court in the world would see them in that role.Now that is theatrics. Soulbender 2006-10-04 2:03 am EST “since Intel sells most of its product in the US. For them to ignore the FCC is to give up more than half the market for the product.”Eh, I seriously doubt that Intel sell most of it’s products in the U.S. Granted, it’s probably their biggest single market but I’m reasonably certain that Intel sell more gear to the rest of the world than they sell in the U.S.“Theo’s sticking point is that he doesn’t want OpenBSD to be a ‘vendor’, although any court in the world would see them in that role. ”No, he says anyone can be a vendor or a user of OpenBSD since anyone can redistribute it.Intel’s shrinkwrap license is downright silly and there’s no good reason why Intel cant allow free redistribution of their firmware when so many other vendors can. deathshadow 2006-10-02 3:50 pm EST >> How the hell did this get modded plus 4?Perhaps because some people agree with the viewpoint, and there’s nothing in it that should WARRANT it being modded down? Not one swear, and the closest thing to confrontational in it is the drum circle/peta comparison, which is a reasonable assessment by way of similie and certainly meant with a measure of jest.In fact, looking over that post I’d say that anyone modding it DOWN is in fact violating the rules here – of course that depends on where you draw the line on trolling – daring to post an opposing viewpoint is not trolling or confrontational, no matter how much you may disagree with it.It’s certainly better than the article itself, which BTW in most any forum I’ve dealt with has enough content in it to get someone a nice little instaban. You don’t call someone a big fat liar, much less an asshole (even if that was by comparison) then expect anyone in BUSINESS to pay you the least bit of attention…Lemme put this in language de Raadt fans can understand: His letter is enough for Intel to say **** OpenBSD and the open source it rode in on. deathshadow 2006-10-01 3:28 pm EST Those elements I pointed out, ‘Ecosystem’, ‘open for business’, calling them a fraud, and the general language of the entire letter means IMHO it will likely be taken about as seriously by Intel as a letter from PETA or Greenpeace.…and if you think anyone actually takes them seriously, you’re living in a dream world. deathshadow 2006-10-01 3:37 pm EST correction, the FBI and CIA do take them seriously, as fringe terrorist groups. deanna 2006-10-01 3:53 pm EST Yes yes, that’s the spirit. If you’re not pro-business, you’re pro-terrorism. It seems we’re both of us at the far end of our respective sides here. Arawn 2006-10-01 4:08 pm EST I bought a AMD Turion based one. With a ATi chipset (I know, but it’s a minor problem to me, I don’t need 3D in Linux), and a working Ralink WiFI. I just run Ubuntu, no probs.I actually did some research before buying, ended buying a MSI Megabook S270, and I hoped it came with the Ralink, and it did (it could have come with a InProcomm WiFi… 50/50 chance :/).I’m not the least repented with the choice I made. It has good battery capacity, good performance, and it works well with Linux. All I wanted.I agree with most opinions here, if vendors want their products to be bought, they need to pay attention to customer’s needs and give them support. So, I will buy what does work.Intel is the other edge of the Wintel sword. Cloudy 2006-10-01 4:13 pm EST So Theo calls an Intel representative a “big fat liar”, supposedly because of some slides. Said slides contain no lying at all.Then Theo *demands* “rights” that don’t exist.It’s great press, but as a negotiating strategy it’s right up there with a 2 year old throwing a temper tantrum – in the middle of a mosh pit.The way to convince Intel to change is to show them how its in their benefit to change. The way to convince Intel to stay the same is to throw temper tantrums and call their people childish names. yak8998 2006-10-01 6:07 pm EST A 2 year old throwing a temper tantrum – that seems to be par for the course when it comes to Theo doing anything. The guy makes some good points (usually), but he always comes across as an ass. They really need to give him the boot or at least get someone with more tact to handle PR… Janizary 2006-10-02 12:28 am EST Shut up, you know nothing about de Raadt, the position he holds or why, if you did you would not be making such a comment. If you cannot be bothered to understand the situation, why are you coming here bitching about it?De Raadt does not make these comments idly, they come to the public after frustration at dealing with the companies which spit in the face of open source, while pretending to aid it in order to give their image a nice, buzzworthy glow. This is the reaction to the last straw.De Raadt leads because he can, because he is smart enough, capable enough and because enough smart people like him and are willing to work with him that he is more important than some random pissant on a forum. His project produces cleaner, freer, more portable code than any other project in the world, his is the most secure networking operating system in existance – he creates things that matter and thus his opinion does.Who are you and what have you done?Edited 2006-10-02 00:31 AdamR01 2006-10-02 5:12 am EST “Who are you and what have you done?”One could say the same thing about you… Janizary 2006-10-02 5:59 am EST And yet I am not the ungrateful one wandering around shooting my mouth off about how childish the behviour of someone that is out there getting things done is, I am not calling someone that has been repeatedly recognised as one of the more important people in the free software world childish. yak8998 is, the fact that he thinks de Raadt should be, “given the boot,” illustrates how mentally inept he is, since de Raadt is OpenBSD.Edited 2006-10-02 06:01 yak8998 2006-10-02 6:26 pm EST Wow. Flame me over the internet. Look at the title of the thread and think about it a minute before you reply again. I never said he is a bad progammer or has not done good for BSD. I am just saying he is pretty awful at PR and needs to tone it down. Again, I said he makes some good points, but he could probably get more done if he wasn’t running around calling employees at Intel “morons.” But then again, I am probably wrong, because who am I and what have I ever done? Cloudy 2006-10-02 6:37 pm EST Shut up, you know nothing about de Raadt, the position he holds or why, if you did you would not be making such a comment. If you cannot be bothered to understand the situation, why are you coming here bitching about it?I know a great deal about Theo de Raad, the position heholds and why. I deal with the consequences of Theo’s destructive behavior every day.De Raadt does not make these comments idly, they come to the public after frustration at dealing with the companies which spit in the face of open source, while pretending to aid it in order to give their image a nice, buzzworthy glow. This is the reaction to the last straw.It is, perhaps, you, who should take time to understand both sides of the issue. Theo is burning bridges. This is not helpful to those of us who have managed to quietly get Intel to make what progress they have made towards opening their source.De Raadt leads because he can, because he is smart enough, capable enough and because enough smart people like him and are willing to work with him that he is more important than some random pissant on a forum.Theo isn’t much of a leader, more of a hothead. He was unable to work well with others in the BSD community and so split OpenBSD off, defocusing the community, and creating unnecessary work all around, rather than remaining and finding a way to accomplish his goals within the existing structure.His current rant is more of the same. He is actively undermining what little good will the open source community has with Intel because they wouldn’t do it his way.His project produces cleaner, freer, more portable code than any other project in the world, his is the most secure networking operating system in existance – he creates things that matter and thus his opinion does.Among the things Theo creates that matter are disharmony, dischord, and disgust. These negatives have done more harm than Theo’s contribution to the OpenBSD project has done good. On the whole, Theo’s contribution to date has been a small net negative.This latest tantrum is a classic example. If we are lucky, Intel will ignore it. if we are unlucky, Intel will respond, taking back some of the progress we’ve made in getting them to open sources and make them available.Who are you and what have you done? ,Someone who has worked with Intel to help them make steps towards open source and who prefers to be quietly anonymous; not someone who throws tantrums and burns bridges.But even if I were not, I would speak up against the excesses of Theo’s approach. Calling Intel spokespeople “big fat liars” does nothing but damage the open source community’s relationship with Intel. It makes it harder for those who are trying to fix the situation, not easier.Theo could have just as easily called for a letter writing campaign that was tactful and recognized that Intel is in the position to call the shots on this. Such a campaign might even have helped. Instead he opted to be openly antagonistic. This can only make things worse. ddpbsd 2006-10-02 3:03 pm EST Theo could have just as easily called for a letter writing campaign that was tactful and recognized that Intel is in the position to call the shots on this. Such a campaign might even have helped. Instead he opted to be openly antagonistic. This can only make things worse.Theo’s mentioned trying to be nice in the past, and it never seemed to work. Cloudy 2006-10-03 5:55 am EST Theo’s mentioned trying to be nice in the past, and it never seemed to work.Theo has the patience of a small child.I’ve tried being nice with Intel and it has worked.http://www3.intel.com/cd/software/opensource/asmo-na/eng/243377.htmlists many of the successes related to Intel and open source.Is it everything we want? No, not by a long shot. Is it better than what we’re likely to get as a result of Theo throwing public temper tantrums? Yes, by a long shot. Soulbender 2006-10-02 8:02 am EST “So Theo calls an Intel representative a “big fat liar”, supposedly because of some slides. Said slides contain no lying at all. ”No, but if you take them into the context if Intel behavior they do.Say one thing, do another. That has been called being a liar ever since I was a child and it still is.“Then Theo *demands* “rights” that don’t exist.”Perfectly reasonable demands *IF* a company actually wants to be OSS friendly/compatible. Cloudy 2006-10-02 6:09 pm EST “So Theo calls an Intel representative a “big fat liar”, supposedly because of some slides. Said slides contain no lying at all. “No, but if you take them into the context if Intel behavior they do.Say one thing, do another. That has been called being a liar ever since I was a child and it still is. Intel’s slides say “Share when it makes sense, leave closed when there’s a business reason.” Intel’s behavior is to share when it makes sense ad to leave closed when there’s a businsess reason.Where, precisely, is the lying in the slide? deanna 2006-10-02 6:38 pm EST Here is Intel’s purported stance toward Open Source as laid out in the slides:Enable the community to do as much as possible.Only keep internal those things that the community can not contribute toExample: Certification TestingIf you need to keep IP closed source (for example some whiz bang algorithm) document the hardware sufficiently that the community can provide their own.Treat the community as if they were a member of your internal team.Listen, and respond to, their input and feedback!!!Not one point in there is actually put into practice, and this is why we see accusations of lying. They look like lies to me. “Listen to, and respond to their input and feedback!!!” Where is Intel’s response to this debacle? Cloudy 2006-10-03 12:47 am EST Not one point in there is actually put into practice, and this is why we see accusations of lying. They look like lies to me. “Listen to, and respond to their input and feedback!!!” Where is Intel’s response to this debacle?You mean, I assume, “when will Intel respond with what we want to hear”, because they have responded. It was Intel’s response that Theo is ranting about, not their lack of response. Sphinx 2006-10-01 7:43 pm EST Money is the only voice big corporations can hear.Thanks for justifying my recent purchase of AMD hardware Theo, doing a great job, hang in there. TechGeek 2006-10-01 7:58 pm EST You know, Intel is one of the most linux friendly hardware makers out there right now. They may not open their drivers or firmware, but who says they have to? At least Intel hardware works with linux which is more than I can say about AMD stuff. I would love to buy AMD, they do make awesome hardware. But I will not buy a VIA board EVER again. Until AMD insists on quality for their chipsets like Intel, they will never take Intel’s marketshare. With Intel, it may not be the fastest (although core 2 duo rocks) but it does work like a champ. As for linux, we’ll see how linux friendly AMD is when they decide to open ATI’s drivers up. Or at least open the specifications so that open drivers can be written. Redeeman 2006-10-01 8:56 pm EST it doesent fit any definition on open, it simply isnt open, and they are even putting out lies about the openness, they claim the regulatory binary only daemon only is to ensure people doesent use other frequencies, yet that is not true, it does more than that.and even if it didnt, the daemon doesent protect shit, anyone that would be able to read the source and mess with it, can still do those exact things today. 1c3d0g 2006-10-02 1:59 pm EST I understand his position. With the focus of his distro so heavily based on security, I can see why he’d want access to the hardware specs. Someone should stand up for this, and he’s one of the few who has the balls to do it. Most people just talk, which doesn’t do anything.