From owner-freebsd-chat Wed Sep 8 1:22:33 1999 Delivered-To: freebsd-chat@freebsd.org Received: from lariat.lariat.org (lariat.lariat.org [206.100.185.2]) by hub.freebsd.org (Postfix) with ESMTP id C20A315A7D for ; Wed, 8 Sep 1999 01:22:23 -0700 (PDT) (envelope-from brett@lariat.org) Received: from mustang (IDENT:ppp0.lariat.org@lariat.lariat.org [206.100.185.2]) by lariat.lariat.org (8.9.3/8.9.3) with ESMTP id CAA11404; Wed, 8 Sep 1999 02:19:33 -0600 (MDT) Message-Id: <4.2.0.58.19990907215517.047c9880@localhost> X-Sender: brett@localhost X-Mailer: QUALCOMM Windows Eudora Pro Version 4.2.0.58 Date: Wed, 08 Sep 1999 02:17:30 -0600 To: "Jordan K. Hubbard" , Rod Taylor , mmeola@uswest.com, freebsd-chat@FreeBSD.ORG From: Brett Glass Subject: My proposals for the future [Long] In-Reply-To: <7838.936165770@localhost> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: owner-freebsd-chat@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Rod Taylor writes: >I'm mildly interested in what you consider these [problems with the >promotion of FreeBSD] to be. I've elaborated on some of these in earlier messages. To cover them with a rather broad brush, they include: 1. An exponentially growing gap between the sizes of the BSD and Linux user bases, coupled with slowly declining market share; 2. A lack of the sort of aggressive, contagious advocacy that continues to fuel the spread of Linux; 3. A lack of the sort of venture capital and well-financed commercial promotion now being expended on Linux; 4. A lack of interest on the part of large application developers (due primarily to item 1 above); 5. An "elitist" attitude which discourages even very talented potential contributors to the code base; 6. A shortage of contributors to handle known problems in the code base (due to items 1 and 5 above); and 7. A failure to recognize the value of a large and growing user base. >Can we fix these problems? Are they problems, or choices >with long term payoffs rather than immediate. They are problems. There is no long-term benefit to falling behind Linux, either technically or in market share. > Does FreeBSD want a larger userbase, afterall... With the good, >comes the bad. A larger user base helps everyone involved with FreeBSD to better realize his or her goals. For example, suppose you are a developer of FreeBSD. You may be doing it because you like doing good work, or because you want people to benefit from using your code, or because you take pride in a job well done, or because the availability of FreeBSD helps you to perform other tasks or makes your business more successful. You may also wish to make money supporting or extending FreeBSD. In all of these cases, a larger user base furthers your goal. If you have a sense of pride and accomplishment when you write good code that benefits others, you will know that you have benefited more people if the user base is larger. If you use FreeBSD in your business, a larger user base means more contributions and improvements and hence a BETTER tool for the task. What's more, it means a larger range of available commercial applications for your use. (Many are now available for Linux but not for FreeBSD; they only work under emulation, if at all.) If you want to make a living supporting or extending FreeBSD for clients, or writing about it, a larger user base means more demand for what you do. Some people on this list have sneered, somewhat contemptuously, at the "hordes" of newbies that flock to a popular OS. But this is elitism at its worst. We were all "newbies" once (as I was when I first encountered UNIX in 1977), and everyone deserves to benefit from a good thing. Jordan Hubbard writes: >...I don't see what purpose would >be served by trying to do a "Michael Jackson", only to become rapidly >overwhelmed by an influx of screaming fans (though ours scream for >different and less pleasant reasons) and resulting in us seeing core >and all our principal developers retiring to various fenced retreats >with only monkeys and small children for company. Matt Meola writes: >To specifically target the masses >for wholescale conversion to FreeBSD is to degrade the overall quality >of the userbase. The above comments seem to echo Aesop's fable of the fox and the grapes: "I don't WANT all those users; they're no good anyway." Fact is, the majority of Linux users are not "screaming fans" but rather satisfied customers. While they may not recognize the harm they may do by supporting a GPLed product, the fact that they've recognized that a collaboratively developed UNIX-like OS is a valuable asset is certainly a point in their favor. If you don't want 'em, can I have 'em? ;-) Jordan Hubbard writes: >Sometimes it's >possible to grow TOO fast, and let me just remind you that where >"Linux" gets to spread the load across a whole bunch of different >distributions and literally thousands of developers, many of whom are >paid for their full-time work by the likes of Red Hat, Caldera, SUSE >and TurboLinux, we have only one FreeBSD and one group of people to >hold the line. Should FreeBSD gain user base, the number of contributors will likewise grow. There will be room for more specialization in the core team because there will be more skilled members. Linus Torvalds does not seem to have become any less productive in his improvements or refinements to the Linux kernel since Linux grabbed the spotlight, so there's no reason to fear that FreeBSD will have such problems. As for having "one FreeBSD" -- there's no reason why there can't be more than one distribution. (In fact, there are now. There are Cheap Bytes and PicoBSD as well as Walnut Creek's multiple FreeBSD packages.) However, it's likely that distributions of FreeBSD will have lots more in common than the many distributions of Linux, which is a good thing. Rod Taylor writes: >Rather than defending your yet to be shared thoughts. Please discuss, with me >if noone else, just what we can do. FreeBSD needs a strongly marketed, strongly evangelized, heavily supported distribution that surpasses anything now available in the Linux world technically and can catch up with Linux's growth curve. As I've mentioned, I've been approached by some folks who know of my previous work (most of which has been done under contract; you'd recognize some of the clients' names) and asked if I'd be willing to take one of the BSDs -- FreeBSD or OpenBSD, most likely -- and do a strong, commercial (but open source) BSD UNIX. This leads into some of Jordan Hubbard's comments. I wrote that I was concerned that Walnut Creek -- which, since it employs Jordan and runs the main distribution site for FreeBSD -- has a very strong grip on the development of FreeBSD, and might do things to discourage competition with is own sales of FreeBSD disks. Jordan writes: >I'm also not sure if I even understand Brett's concerns about "Walnut >Creek CDROM taking FreeBSD private" since that's nothing we've ever >shown the slightest inclination to do and to even worry about such a >thing happening strikes me as nothing short of paranoid. If one is >frequently chased by little green aliens then one has every right to >worry about them, but if one has never even SEEN a little green alien >then it's probably rational to assume that it shouldn't rank high on >one's list of personal worries. I've seen this particular "little green man" before -- it's called "All's fair in love, war, and business." I don't subscribe to that attitude, but many businesspeople do. >It has always been Walnut Creek >CDROM's wish that FreeBSD be as open as possible, we've never made any >bones about that fact, and we've even sneakily tried sending sample >copies of our CDs to known pirates in Asia in hopes that they'd copy >the heck out of the disks and spread them (and the FreeBSD cause) far >and wide. Since that doesn't seem to have worked (the pirates are >picky and only seem to go for the high-dollar value products), Bob and >I are going to Japan, Hong Kong and China next month to try the more >direct route. We want FreeBSD to cover the earth and we'll do >whatever we can to make that happen. While all of this may be true, intentions can change. And -- again -- Walnut Creek does have much more control of the development of FreeBSD than, say, Red Hat has of the development of Linux. So, it's something the investors are justifiably concerned about. Remember, since Jordan works for Walnut Creek, the company can literally claim ownership of what he does, even over his protests. We need strong assurances not just from Jordan but from Walnut Creek as a company that this will not happen. Jordan further writes: >I also take issue with Brett's assertion that he's somehow >single-handedly responsible for FreeBSD's recent exposure in the Linux >market since that only a grave disservice to the time and energy that >I and many others have invested in showing up at Linux conferences, >participating in open source panels, going to the O'Reilly show, doing >television interviews, writing articles for electronic and paper >publications, etc etc. In fact, I can't recall seeing anything from >Brett since the article in "Smart Reseller" and that was more than 8 >months ago. I'm happy and even a little proud to say that my own >recent efforts at doing PR have been far more timely and effective >than anything I've seen recently from Brett, and I've every reason to >be - I've worked damn hard at it lately! My remarks in earlier messages were not intended, in any way, to belittle the contributions of others, but merely to point out that I've done a lot -- much of it behind the scenes -- to make sure that BSD UNIX is mentioned in the same breath as Linux whenever possible. Some of what I've done is in print; some is in e-mail to fellow writers; some is in public forums. But I've noticed a STRONG trend since I started doing this. The exposure seems to build on itself, and journalists who mention BSD along with Linux once tend to continue to do it from then on. And it spreads from colleague to colleague, too. I see the words of my messages paraphrased or even duplicated in the new material. So, I think I deserve a lot more credit than Jordan gives me for effective networking and consciousness-raising in journalistic circles. Jordan also seems to take issue with my notion that it is important to point out the deficiencies of Linux, especially its licensing. He writes: >Even though it may generate what look like short-term gains to turn >around and slam your competition, I think the long-term effects are >only negative and there is always a large contingent of folks who >*don't* react well to this kind of "advertising" and will, >furthermore, never forget it if you do it so much as ONCE. When it >comes to mud-slinging, memories are long and I think we've all worked >too hard to have a reputation for more maturity and level-headedness >than this. Sure, there is probably always some collection of angry >teens out there who won't flock to your cause if you're not burning >something or getting a gang together to go kick the other side's >butts, but that kind of "action" is not our cup of tea and I prefer to >leave the knife-fighting for the West Side Story fans and those who >feel that they're not achieving something if they haven't kicked a few >puppies before breakfast. To characterize the sort of advocacy I'd like to do as "kicking puppies" is, of course, to miss the point. Good marketing involves a push as well as a pull; people need to realize that there are deficiencies in what they are already using before they will consider alternatives. ("You mean that ALL operating systems don't crash as often as Windows? I thought that was just part of the way computers worked.") What's more, the idea that a problem exists is more likely to rouse people to action than the idea that all's right with the world. The notion that there's a problem -- or, worse yet, a looming crisis -- is more contagious, or memetically fit, than the idea that things are fine. Aaron Lynch, in his book "Thought Contagion," writes: "Thought contagions spread fastest via *proselytic* transmission.... The conviction 'My country is dangerously low on weapons' illustrates proselytic advantage. The idea strikes fear in its hosts for both their own and their compatriots' lives. That fear drives them to persuade others of military weakness to build pressure for doing something about it. So the believe, through the side effect of fear, triggers proselytizing. Meanwhile, alternative opinions such as 'My country has enough weaponry' promote a sense of security and less urgency about changing others' minds." This principle accounts for the success of Christianity, as well as other religions -- such as Islam -- which preach that one is in imminent danger unless one takes action to achieve redemption and salvation. Does this mean that we should "bash" Linux or Windows at every opportunity? No -- however, the existence of a lunatic fringe, as with religion, is a sign that advocacy is successful. (Anything that successful stimulates emotions and creates strong attachments will cause a small number of people to go too far; it's just inevitable.) Fortunately, the downsides of having such a lunatic fringe are smaller for an OS than for a religion. The advocates of an OS are unlikely to try, for example, to take over the government, burn crosses on lawns, or stone you if you drive a car down the wrong street on the Sabbath. The Linux Loonies may be loud, but they make news and attract needed attention -- witness the "Windows Refund Day" events. And people are -- for the most part -- tuning them out if they find them obnoxious and using Linux anyway, as evidenced by Linux's continued growth curve. Nonetheless, effective advocacy of any product -- from soap to cars to long distance telephone service -- often involves convincing the user that he or she will be at a disadvantage, or will lose something -- perhaps time, money, or productivity -- by sticking with what he or she uses now. An operating system that does not do this is not memetically fit -- and, in keeping with the basic theory of replicators -- will be driven from its niche in the software ecology. It may not die entirely -- heck, there are folks still running all sorts of obsolete OSes -- but will not achieve the benefits of a widely used, actively developed, and well-supported OS. Jordan, however, questions whether a BSD UNIX distribution could ever catch up with Linux. In somewhat derogatory terms, he writes: >You can't really be so >egotistical as to asssume that you, single-handedly (or even you and >50 other guys with a few million dollars worth of capital), can >compete with Red Hat and its 2 billion dollar war market cap? Why not? Red Hat is competing with Microsoft, which recently surpassed the market cap of GE to achieve the largest market valuation of any company on Earth. >And >that's just ONE of groups you'd have to defeat to push Linux from its >perch. The folks at SUSE and TurboLinux might not be Red Hat but they >have their own millions, and I can assure you that they wouldn't take >any serious attempt to unseat them lying down. Yet still you come >here with your little war plans and your pointy stick and tell us that >we can all charge the enemy's line of assembled M1 tanks if we're all >just pure of heart and make sure to smear the magic chicken blood on >our chests which wards off bullets. It's got to be better than woad. [CF the song "Woad Warrior" (http://www.sff.net/people/Julia.West/songs/actslyrc.htm#woad) by Zander Nyrond (http://www.cix.co.uk/~zander/index.htm)]. >Uh huh. Go for it, Brett, and >we'll all be behind you in the follow-up attack, I promise. :-) Or would Walnut Creek -- and you, as an employee thereof -- shoot me in the back? Seriously: may I have your word, and Walnut Creek's, that you won't sabotage what I do if I manage to get these investors to cough up some money? If you do this, I could announce some positive results as soon as FreeBSDCon. >I'm still waiting for some concrete proposals from you which explain >just how exactly we're going to compete with Linux where it's >currently strongest without compromising our own principles for being >the technically sound, well thought-out solution. I've made quite a few, but they all boil down to this: good code and good memes. This is what makes a successful software product. But code quality is less important than good memes -- as is shown by the dominance of Windows, NT, and Linux over BeOS, OS/2, and FreeBSD. With better memes, FreeBSD could be successful. And there is nothing about success -- which I strongly sense from your messages that you fear -- that would compromise the technical soundness of FreeBSD. --Brett Glass To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-chat" in the body of the message