From owner-freebsd-questions Mon Apr 9 23:40:57 2001 Delivered-To: freebsd-questions@freebsd.org Received: from mail.freebsd-corp-net-guide.com (mail.freebsd-corp-net-guide.com [206.29.169.15]) by hub.freebsd.org (Postfix) with ESMTP id 373EF37B423 for ; Mon, 9 Apr 2001 23:40:46 -0700 (PDT) (envelope-from tedm@toybox.placo.com) Received: from tedm.placo.com (nat-rtr.freebsd-corp-net-guide.com [206.29.168.154]) by mail.freebsd-corp-net-guide.com (8.11.1/8.11.1) with SMTP id f3A6egk06196; Mon, 9 Apr 2001 23:40:42 -0700 (PDT) (envelope-from tedm@toybox.placo.com) From: "Ted Mittelstaedt" To: "Duke Normandin" <01031149@3web.net> Cc: Subject: RE: BSDi Acquired by Embedded Computing Firm Wind River Date: Mon, 9 Apr 2001 23:40:39 -0700 Message-ID: <000001c0c189$28163c00$1401a8c0@tedm.placo.com> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook 8.5, Build 4.71.2173.0 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3155.0 In-Reply-To: <20010409163545.B72259@mandy.rockingd.calgary.ab.ca> Importance: Normal Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG >-----Original Message----- >From: Duke Normandin [mailto:01031149@3web.net] >> > >> Many's the time I've seen people complain that FreeBSD didn't support >> their pet Wonkulating Gronkulator and that Linux did - and after you >> get into it with them, it always seems to come out in the discussion that >> their Wonkulating Gronkulator card costs $14, and the card that FreeBSD >> _does_ support that does the same thing, costs $25. (and is technically >> a better card) > >The way I read your parapraph I can only conclude that Linux et al *do* >support more $14 cards as an accomodation to the cheaps bastards out >there, than does FreeBSD. It goes w/o saying that FreeBSD *does* support >the same functionalities -- but only those of higher quality / or those >that the developers happen to favor. > Not always - for example the Realtek chipset-based $14 PCI network adapters are supported very well under FreeBSD - of course the card may not properly autoconfig to your network hub, but what do you expect from a cheap NIC chipset. However, I would say that I think if you were to do a survey and stack the OS's right up against each other, you would find that Linux _does_ support slightly more of the $14 cards than FreeBSD. Of course, you would need to subtract all video chipsets from this ranking - because video cards are supported by Xfree86, and thus the support is identical between FreeBSD and Linux. Also, the support for _old_ hardware is a political powder keg as well - since hobby folks often have very bad habits of picking up $2 cards from flea markets and saying "Wow, think of what I can build with _this!_" :-) Is it fair to castigate an OS because it doesen't support _old_ hardware? > >Are you telling me (us) that most private computer users, intent on >learning and *using* Unix, go out and buy the very best equipment that is >available? They don't simply go to Future Shop etc, and look for the >"best bang for their buck"? > Actually, yes, this is in effect what I'm saying. Not the very _best_ equipment (or we all would be running Alphas) but I think the private computer users running UNIX are less cost-conscious and more concerned with quality than most corporate buyers. On a scale of 1 to 10 with 10 being the best, I'd put most corporate buyers at a 3, and most home buyers at a 7. Of course, you have to get at them BEFORE the sale - far too many people go buy the hardware first then check the compatability listings later. Or they use old hardware they had from years earlier. > >If what I suggested above "more closely describes the actual fact than >most people would like", then I supposed I lucked-in and guessed >correctly. The facts then, seem to suggest a certain attitude, but like >to indicate below, there may be more issues involved. I can understand >completely though. You want to play with our little hobby OS? -- fine use >it and love as is! That's an attitude -- and perfectly justified. I guess >that *I* need to know this before making a decision, that's all. > I don't think that's an accurate description. I think it's more like "Quit telling me to write support in for your Winmodem you got for free out of a cereal box when you can solve your problem by putting it in a Windows box and spending $25 on a real modem" I've seen some very interesting reasoning from some users that are insisting that support be written in for their cards. The most common is "But, EVERYONE has one of these [insert Seagate ST01, SoundBlapper256, PCTel Winmodem, etc. here] cards and it's a _t_r_a_v_e_s_t_y_ that FreeBSD doesen't support it!!" (with a long whiney emphasis on the travesty part) It's like _come_on_, the developers aren't stupid, you know. They know perfectly well that you went to Fry's to buy a modem and right next to the $25 ones there was a $5 Winmodem with a $5 rebate sticker on it, and for a moment you succumbed to the lure of "the deal" Just because _everybody_ has one of these cards/devices isn't justification for publically castigating the FreeBSD community for not supporting it. This is what some people don't get - is that lots of times, the majority is _wrong_. 500 years ago the majority throught the world was flat but that didn't make it flat, and today just because millions of people are on AOL doesen't make it the best ISP in the world. >> >> But, most developers sift through the complaints and try to find >> commonalities. Say the driver is for a network card chipset, well the >> developer finds that 90% of the complaints are from 1 manufacturer's card >> that uses this chipset - well, the developer most likely will >obtain one of >> these cards and test against it and fix the driver. But, will >he then go on >> and fix the problems in the remaining 10% of complaints? Would >it be a good >> use of time for him to fix them? > >It depends! If 100% of the complaints represents only 10% of the possible >NICs commonly in use, then fixing 90% of the complaints isn't saying a >whole hell of a lot, is it? > Certainly! But, if only 5% of the people running FreeBSD are actually using the network chipset that commands 90% of the network chipset market, then you are indeed saying quite a lot! Every device driver in FreeBSD is buggy, no question about it. For that matter every device driver in every OS on the planet is buggy. But, those bugs are so subtle, or triggered in such rare occasions, or require additional variables, (like certain motherboards or combinations of other cards) or matter so little, that 99% of ordinary users are never going to see them even if they run the same version of FreeBSD for 100 years! In short, it's fruitless to work on fixing bugs that nobody, or a very small percentage, is complaining about. >> >> Of course the first step to solving this is to convince the end >users that >> they can use MANY DIFFERENT solutions to their problem - they aren't >> married to Microsoft's answer, nor to anyone else's answer. > >Exactly.... "they aren't married to Microsoft's answer,.."! It just seems >to me that if FreeBSD is interested in being "one" of the solutions for >the desktop-only/mostly crowd, then steps should be taken to accomodate >W/O compromising kernel integrity in any way. > Well, one big step is the fact that it's open source - if there's something that a desktop user doesen't like - well they always have the option of hiring a programmer to fix it. >> >> But, your missing something - the "UNIX Way" is to use the best >tool for the >> job, not to ram a hammer in there for all solutions. Therefore, >why use a >> substandard tool (ie: substandard computer hardware) to solve your >> problems? > >Now, this is a new one for me! I've never heard it said that the above >was a prerequisite for using Unix. As a matter of fact, in my visits to >the Open/Net/FreeBSD sites, nowhere have I read anything on using *only* >quality components. > No - but one follows the other, doesen't it? If you hire a professional roofer to roof your house, and the guy doesen't show up with a full set of ladders, a couple expensive nailguns and compressor, and the rest of the accoutrements, wouldn't you think there was something wrong if he goes up there and starts banging around with a finishing hammer? UNIX _is_ a step _up_ from Windows. If you want to work with it, it's because you want to work with a higher-performance OS, right? So, why would you want to work with garbage-grade components? It's not that good quality components is a prerequisite for UNIX, it's that people that want to work with UNIX want to do so to get better performance and reliability, and good quality components are a prerequisite for better performance and reliability, no matter what the operating system. >> anyone can use. But, what's the point of making the OS install so simple >> and easy - once the FreeBSD system is installed and reboots up to a login >> prompt, what are they supposed to do then? Sooner or later they have to >> start thinking, they might as well start thinking when they >start installing >> the OS!! >> >I think that we now are finally getting to the core issue of this >desktop/server FreeBSD/Linux deal. > >If FreeBSD exists to perform primarily* as a server platform - period -- >*then it should be marketed as such. Why? Because *all* peripheral >support is, by-and-large, developed for, tested on and limited to >server-grade quality stuff. It should then be plainly emphasized that >FreeBSD is *NOT* an appropriate OS for deployment on run-of-the-mill, >Microsoft-grade hardware. In the spirit of professional good-will, A >referal to some of the Linux sites could then be given. > It _is_ frequently emphasized that FreeBSD _is_ more demanding on the hardware. People post with problems all the time that are tracible to hardware faults, or hardware misconfigurations like overclocking. Your basically restating facts here. What _is_ the definition of low-quality computer components? Well, it's that they do NOT perform as well, and that they can also be flaky and unreliable. Understand also that unreliable means simply that - one person may find success with an unreliable component, another may not. Whereas a reliable component is one in which _the vast majority_ of people will find success with. Also, keep in mind that it's ALSO risky to attempt to use unreliable components with Linux too. While the incidence of "customer unsatisfaction" while using unreliable, low-grade hardware is likely less under Linux than under FreeBSD, when you or anyone _deliberately_ goes out and purchases components that are cheap and low-grade, your setting yourself up for failure under any UNIX or UNIX-like system. Your also greatly increasing the risk that you will fail even with operating systems like Windows 2K and NT4.0 You said it yourself: "Microsoft-grade hardware" If you purchase hardware that's specifically designed for use under Windows Millennium, which is a consumer-grade operating system, then why complain when it won't work with FreeBSD? It wasn't _designed_ for FreeBSD. By contrast, better quality components are designed to work with _all_ operating systems that may be in use. A hardware chipset designer may work for a company that has no intention of writing UNIX device drivers for their hardware, but the designer can do things that make it easy for _someone else_ to write drivers, OR they can take a bunch of cost-cutting shortcuts with the idea that the chipset will only run under Windows ME. Take, for example, the Adaptec 1520 card. This is a PIO-mode SCSI card - but early 1520's _had_ the jumpers for selecting DMA channels, even though NO driver that Adaptec ever wrote supported DMA on that card. The intent was that Adaptec knew on a DOS system that the DOS os was so low-performance an OS that DMA or PIO mode would make no difference in speed, but they knew that on a higher-performance OS that the DMA channels _could_ make a difference. While they wern't going to write the drivers themselves, they wern't going to hamstring anyone else who wanted to make the effort. >As well, it should be again plainly emphasized that FreeBSD is *NOT* an >appropriate OS for most folks wanting to migrate from a Windows/Mac >platform, as the level of computer literacy and competence required to >successfully install the OS, hardware peripherals, software, etc, far >exceeds that of the average Joe. Like you say above, the potential I don't agree - because you said "wanted to migrate" It's my experience that the average Joe User running Windows is perfectly happy with it, and does not WANT to migrate from it. But, if the "average Joe" feels constrained by Windows/MacOS, why then I would say that ALREADY he isn't an "average Joe" JUST BECAUSE HE FEELS CONSTRAINED! Even if he never seeks out a Linux or FreeBSD, the fact that he can recognize that Windows as a paradigm is very limiting, puts him in a class above the REAL "Joe User" The real Joe Users, DON'T feel constrained by Windows - the LOVE it. They _like_ everything being done for them, and they don't mind that their work output looks identical to everyone elses, because they really aren't individualistic with their computer. Most of the real Joe Users that I run into that are _complaining_ about Windows are NOT complaining because they don't like Windows - the paradigm. They have long ago bought-off on this. They are merely complaining because they don't like Windows - the implementation! They DON'T want the Windows to be replaced by KDE or Enlightenment - they just want "better Windows" whatever _that_ is. This is why, by the way, that OS/2 failed. (I have studied OS/2's failure off and on for years - I ran it for a long time as my only OS and I still have it on a partition) OS/2 failed because all that IBM ever aspired it to being is a "better Windows than Windows" What happened is that as long as OS/2 had a few things better than the real Windows, like better networking, all of the Windows users spent the time playing around with it. Microsoft recognized this and simply went into NT 4 and spent a lot of time correcting the problems (like piss-poor networking) that were making the Windows users go with OS/2, and once NT 4 was released with all these things fixed, all of the Windows users immediately dumped OS/2 and happily sucked down NT 4 without looking back. If IBM had done what Apple did with MacOS and basically told all of the OS/2 users that were screaming about stuff like Seamless Windows to go screw themselves, do it the OS/2 way or else, then OS/2 would still be alive today. Certainly it would be a niche OS - but it would never have been anything else other than a niche OS, and IBM would have built a revenue model around selling it as a niche OS. We might have seen things like OS/2 running in palmtops instead of Windows CE. After all, at OS/2's height, it had more installed base than MacOS, and yet today MacOS is alive, OS/2 is not. >Ted...I suspect that this thread is pissing off just a few folks out >there. So what! People can take some controversy, it won't hurt them! >The questions I asked were not meant to denegrate FreeBSD, the >developers, or their valiant efforts. I may have issues with some aspects >of FreeBSD, but my respect for the whole ball of wax has never waned. I'm >still in the throes of learning my way around Unix. I haven't yet made a >decision between a flavor of Linux or an incarnation of BSD. This >discussion has been enlightening and informative, but I think that it >would serve no further good to continue publically. Please email me >privately with any further comments/suggestions. Sorry, Duke, I don't operate this way - I've learned never to say things privately through e-mail and other trackable communication that I wouldn't say publically. I understand where your coming from though, and it's a more common "newbie" issue than a lot of old-timers understand, espically folks that got into FreeBSD from the commercial UNIX side, rather than the Windows side. This is why I'm spending time on this thread. Let me put it this way, your making 2 common mistakes. First, your looking at the FreeBSD "movement" or market, as a lot more black and white than it really is. FreeBSD is used on everything from 386SX25 routers serving 4 PC's to Pentium 3 2Ghz multiprocessing servers serving thousands. It is a very, very, very WIDE market. Yes, FreeBSD does have more problems with low-grade hardware than Linux does. But, that doesen't change the fact that lots of people still run it on low-grade hardware without problems. (or course, they probably are sitting on a stack of oddball components that they can use to play musical lets-try-this-card-in-that-machine-today games with) Secondly, your still thinking of FreeBSD as a defined, commercial market that operates under the same inputs and outputs as the regular commercial software market. It _doesen't_! Superficially, many of the happenings in FreeBSD, such as the Wind River<->BSDi merger, the BSDi purchase of Walnut Creek, and so on, yes these do appear virtually identical to operation of the commercial software market. But, I have come to the understanding that there are many, many, many subtle things that happen in the FreeBSD movement that do not occur as a result of market pressures. You cannot understand the FreeBSD movement without understanding how the _entire_ movement operates, and without understanding that many things that happen do so without apparent logical reason. I say the same thing about Linux too - you cannot _really_ understand the Linux movement without understanding the idealism behind the GNU License - while I may disagree with it, I have come to understand it myself, and I can tell you that GNU's philosophy is completely 180 degrees in opposition to BSD's philosophy. It really is a tribute to all of the individuals involved with both movements that there is so much friendliness between both camps and so much help that happens between them. Ted Mittelstaedt tedm@toybox.placo.com Author of: The FreeBSD Corporate Networker's Guide Book website: http://www.freebsd-corp-net-guide.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message