From owner-freebsd-questions Sun Mar 25 11:29: 9 2001 Delivered-To: freebsd-questions@freebsd.org Received: from snipe.prod.itd.earthlink.net (snipe.prod.itd.earthlink.net [207.217.120.62]) by hub.freebsd.org (Postfix) with ESMTP id A740137B718 for ; Sun, 25 Mar 2001 11:29:00 -0800 (PST) (envelope-from lists@vivdev.com) Received: from [192.168.1.100] (dhcp043-10-151-24.nt01-c5.cpe.charter-ne.com [24.151.10.43]) by snipe.prod.itd.earthlink.net (EL-8_9_3_3/8.9.3) with ESMTP id LAA20072; Sun, 25 Mar 2001 11:28:50 -0800 (PST) X-Sender: lists@vivdev.com (Unverified) Message-Id: In-Reply-To: <004201c0b521$2192f040$1401a8c0@tedm.placo.com> References: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Sun, 25 Mar 2001 14:28:45 -0500 To: "Ted Mittelstaedt" From: lists Subject: RE: FreeBSD & GNU Cc: freebsd-questions@FreeBSD.ORG Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Ted - >>> >> >>OK, but how would anyone take control of the source in the case of FreeBSD? >>In the case if FreeBSD, the source is a bunch of volunteers? >> > >You just answered your own question - "as long as the source is a bunch of >volunteers" > >Who just announced recently they were synchronizing their _commercial_ >operating >system code to FreeBSD? Well, it's BSDi, who also owns Walnut Creek, the >major >FreeBSD distributor. In a few years, the source won't all be volunteers. > >Now, so far there's been no incidents of BSDi going to the FreeBSD project >and >saying "Don't make that change in the kernel there because if you do and we >make it then it will break all our commercial users" Hopefully that day >will never come, and BSDi will be wise enough to stay out of the Project in >these instances. But, >there's a lot of crossover of employees at BSDi and contributors to the >Project. >There are certainly going to be instances where this will create conflicts >of >interest on some technical decisions. While, at the current time I don't >feel that >any of the core FreeBSD Project members would be influenced by such >conflicts, >nobody knows what the future will bring, and it's always a risk to set up >these types of environments where there's a lot of corporate influence. > This answers my question exactly - the mechanism on the horizon is the BSDi/Walnut Creek influence over the project forcing decisions upstream instead of responding as technical improvements flow downstream from the project. With your caveat of if that happens. >>> >>>In the history of marketing, there's never been a single source supplier >>>that has lasted for more than a blink of an eye, just due to this issue. >> >>Maybe. How about Ma Bell? Didn't die due to lack of standardization. >>Died when the monopoly, which was underpinned by protected proprietary >>technology, was dismantled. >> > >Ma Bell was selling a product, dialtone, that for nearly 80 years was >very much like electricity in that it was pretty much unchanging. As such, >the market could tolerate the lack of innovation, since really there wasn't >anything to innovate. > >What set the seeds for Bell's destruction, ironically, was the invention of >the transistor by Bell Labs. With modern electronics, it became possible to >cheaply and rapidly make all sorts of new consumer telephone devices. The >first cracks appeared when people attempted to plug in these new devices and >were told by Bell that they wern't allowed to do it. The market revolted >and forced Ma Bell out of home ownership of the telephone. (that was a >significant >concession of their market, if you think about it) The transistor also >permitted modern, reliable and cheap phone switches which allowed the >competitive long distance carriers to take hold and once again the market >revolted when Ma Bell attempted >to block them, the result was the lawsuit. What you just laid out is the first salvo in the breaking of the monopoly over phone service. While it is ironic that the technology came from within the monopoly itself, the leverage point was in fact commercial/legal. Once the handset-nose of the camel was under the tent, the devolution of the monopoly was (in retrospect) inevitable. The distinction here, however, is that _dialtone_ never changed - backward compatability was maintained (I still have a rotary phone in the basement, I can still call you on it) despite the fact that the system itself made a fundamental technical transition from mechanical/analog to electronic/digital. So there was, in this case, no negative consequence to institutionalization/standards observance. In fact, the existence of an open standard allowed smaller and more innovative participants to determine the future, to the long-term detriment of the monopoly. This is what I suggest would happen if microsoft's source code were publicly available - a subset of IEEE-like standards would be developed, achieve a critical mass, and you would end up with a version history of that code like you have in UNIX (vis http://perso.wanadoo.fr/levenez/unix/ ) What that UNIX timeline demonstrates, btw, is just how organic the technical progress can be, incorporating as it has many innovations in markets, hardware, finance, and communications, when the commercial barriers which would otherwise constrain it are reduced or eliminated. [snip] > >>Also, in the space you are describing, shouldn't there be sun and hp >>machines? How do these offerings mix? The other thing is that the UNIX >>they are turning to now includes Linux and BSD, doesn't it? >> > [snip] >It's perfectly possible that >Open Source is such a fundamentally different means of software creation >and distribution that none of the older marketing rules apply, and as such >history can't repeat, because this hasn't been done in history before. In >fact, I even make that argument in my book. The older marketing and commercial rules will continue to apply, we can be certain. What varies is the complex of factors that make up the dynamical system that develops within those broad categories. As far as history repeating itself outright, unless someone can repeat literally what microsoft did - and that appears impossible due to many factors, not the least of which is that no one company (IBM) dominates this industry today (not even Microsoft beyond the intel desktop OS/Productivity suite software) - it won't. > >I'm lumping all UNIX together, both commercial and Open Source, because I'm >talking about general trends. I feel in my bones that the wind is changing >again and that corporate interests are much more open now to considering >UNIX >instead of blindly swallowing Windows. But, many will never be open to >Open Source, and will wish to continue to get those "commercial" systems, >even UNIX, and so Sun and HP and Compaq will have lots to do making products >for those folks. > My view is that the wind you are feeling is the steady trade winds of the mercantile forces that individuals and corporations always put at their collective back, as soon as the artificial blockades of monopolies are removed. As far as considering specifically to "Open Source", I think you are right on - many (in fact most) companies will prefer to sub-contract their development to specialists, and stick to whatever it is that brings in their business' revenue. How this _should_ develop is this: - The installed base will steadily move toward lower commitments of capital to expensive and over-capable hardware (PC's with bloatware), in favor of inexpensive, task oriented, appliance-like devices. - The value add will move from those who control sufficiently large segments of development capability and distribution to compete in _complete systems_ (e.g. microsoft), to those who can rapidly develop task-specific solutions that work immediately. - These two basic forces will drive buyers toward acceptance of inexpensive, unconstrained (in terms of hardware), robust, collaboration-supporting environments that don't need to be "upgraded" (in the commonly understood/microsoft use of the term) every couple of years, but instead - like dialtone - that doesn't change even as the technology in the background undergoes steady evolution - is steady and reliable year in and year out. As you point out above, there will be companies out there with vertically integrated products (Sun) that are right for the markets the seek to serve. [snips, edits] >>Btw, what is the status of vanilla UNIX? >As far as Vanilla UNIX goes, today there's only 2 kinds of UNIXes - those >blessed >by The Open Group (since they own the trademark to UNIX) as being Real UNIX, >and to get that you have to pay a big fee and implement a bunch of standards >that >nobody uses in your UNIX. The other are the non-UNIX UNIXes, like FreeBSD >and >Linux, which can't legally be branded UNIX yet implement 80% of what TOG >says a "Real UNIX" is supposed to have. > OK, but lumping UNIX together as you did earlier, is a big party - even mac os x. The question I meant to ask was answered when I looked at the history (linked above) - UNIX Time Sharing System v10 October 1989 appears to have been the end of the line for the progenitor system. >>Is NT a dominant player in that market? Seemed late to the party, not >>better, and frequently to disappoint it's customers. >> > >NT/2K is _the_ dominent player in mid-level servers today with Linux a close >second. > Alright, but because it is superior? I bet you'd find many of your peers (of which I do not claim to be one) that would oppose that point of view, if anyone is taking it. How much did commercial clout contribute to the market penetration of NT? >> >>Returning to software, aren't the variant forms of commercial linux supply >>examples of a business model that uses the standardized software and >>hardware to it's advantage in pursuing a very different business model? > >The penetration statistics of the various "brands" of Linux today are >one of the more hotly debated arguments. I don't have an answer because >I don't know if todays Linux market is that of one single monopolistic >dominant player (ie: Red Hat) with a lot of smaller ones, or all of the >brands have equal penetration. > But don't combine my point with this penetration analysis, which has more to do with installed base and other commercial realities of microsoft business practices than my argument. My point here is simply that red hat and caldera are not basing their business on control of the source, as microsoft has. They are trying to profit based on customization and service, which microsoft is decidedly not. [snip] >> >>Microsoft's two main sources of revenue to this day are windows and office. >>Without the ability to make products that are un-knockoffable due to >>proprietary information that is not shared with third party (competitive) >>developers, they would have _no means_ of enforcing their agreements with >>Dell and others. >> > >Part of this argument is the "DR-DOS" argument and it's been proven to be >fallacious. No. Nothing I am saying is predicated on this. My point here is three-fold: 1. If the source were opened, developing a viable windows alternative would be _possible_, and once it was possible, there would shortly be alternatives. Microsoft, in order to quash the upstarts, would simply lower the price, in an effort to retain market share, accepting a less but still highly profitable business. 2. Without the advantage of developing the Office suite to a secret API, Microsoft loses it's control of that source of revenue. Fast. 3. In the absence of the leverage that comes with controlling the means of production (and I am specifically drawing the historical paralell with industrial monopolists of a hundred years ago), Microsoft loses it's power over the hardware manufacturers. People who don't wish to worship in the Cathedral are free to form their own sects, and can go off and even found a secular society (to mash a bunch of historical metaphors and a few hundred years of cultural evolution together in a single sentence). :-) In short, if you remove the barriers to competition, competition will occur, until there is no profit in creating an alternative. At _that_ point, standardization can and usually will take place on an open platform since it benefits all participants and penalizes none excessively. This evolution can even be anticipated and attempts can be made to establish it for communal benefit (e.g. bluetooth). [snips and edits] >a competitive DOS to MS-DOS and price it cheaper and split the DOS market. >It never worked because DR-DOS didn't offer any increased functionality over >regular MS-DOS that was significant enough to induce people to switch. The DR-DOS experience as you've laid it out supports my view, in that DR-DOS _wasn't_ an alternative other than in name. > >Extrapolate this to be 2 competitve versions of Windows - the runner-up >would >probably fail for the same reasons. > But that is not my position. >You are right about one thing - the competitive edge to the Office suite, >though. >Clearly, withholding details that would help Office's competitors certainly >does indicate that Microsoft is actually treating both the OS and the >Application as a >single product. This means that out of the desktop PC software market, the >actual coverage of the single windows/Office product is far larger. That >weakens their argument that they are not a monopoly. Yup. Anyway, it isn't market penetration that makes them a monopoly. It is their business practices, and it is for the latter that they have been successfully prosecuted. But again, my question goes to the inevitability of stagnation given an open source, rather than the particular merits of the microsoft case. With the contribution of the history of UNIX site, I am rather optimistic that the genetic makeup, if you will, of a well-attended open source project bodes well for it's evolution and adaptation over time. thanks, chris lists@vivdev.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message