Date: Wed, 8 Sep 1999 00:52:55 +0000 (GMT) From: Terry Lambert <tlambert@primenet.com> To: regnauld@ftf.net (Phil Regnauld) Cc: bright@wintelcom.net, davids@webmaster.com, brett@lariat.org, freebsd-chat@FreeBSD.ORG, walton@nordicrecords.com Subject: Re: Berkeley removes Advertising Clause Message-ID: <199909080052.RAA16284@usr06.primenet.com> In-Reply-To: <19990903111035.34383@ns.int.ftf.net> from "Phil Regnauld" at Sep 3, 99 11:10:35 am
next in thread | previous in thread | raw e-mail | index | archive | help
> > No need for sarcasm, the GPL makes it impossible for commercial companies > > to incorperate GPL'd code into a propriatary product. This kills the > > competative advantage the company has. > > There, you said it: "makes it impossible for commercial companies" -- > that assumes a certain definition of "commercial companies", and > to what extent it's "impossible" for them. In the traditional > approach of protecting a trade-secret, maybe. But I like > people to explain things instead of repeating them like lemmings. It has nothing to do with "trade secret protection", since if the code is published to other than a "select group", trade secret protection is lost. A "General Public License" does not qualify as restricting publication to a select group. The only possible recourse to the former trade secret holder is to attempt to recover damages from the act of disclosure itself. This is why BSDI hid behind UCB when USL came knocking. In point of fact, the danger lies in two areas: 1) Dilution of intellectual property rights, specifically implicit grants of license to use process patents when the patented process is embodied in software. It is my understanding that this is why IBM bought Whistle (FreeBSD based) instead of someone else (Linux based); the GPL would have been a deal-breaker. This is also why you should not expect Red Hat software to ever be acquired by anyone with any large amount of any kind of intellectual property: they will flourish or they will languish on their own, but they won't be bought out by a multibillion dollar corporation which wants to remain a multibillion dolar corporation. 2) Commercial companies expect to benefit from expenditures, to the point of recovering them, plus some profit margin. This is true of any company which meets I.R.S. requirments for definition as a "for profit" company. A "for profit" (subchapter C, or "C-Corp") must show a profit after two years of operation, or be disbanded. You could imagine other types of companies, which exist to spend money and never recover it, but generally relatives are a better means of losing money; at least you get birthday cards out of it. This means, in technical terms, that a company that has expended research and developement monies expects to be able to make the money back over the projected product lifetime. The product lifetime is defined (by companies which remain in business, anyway) as the amount of time that the product may be sold in the market place at a continued net profit. The total net profit is the amount of net profit recovered above and beyond the the initial investment, over the total product lifecycle. This means that if you sink 5 billion into a product, you must get 5 billion plus 5 billion times net profit margin out. If this isn't going to happen because someone can buy one of your widgets for a relatively small sum, and then demand your source code, then use that to produce their own widgets, you've effectively flushed your 5 billion down the sewer. Now it might be nice in theory to think that everyone will pool their R&D money, finance common projects, and reap a proportional share of the rewards... but this is reality, not some socialist fantasy-land. All it takes is one player who doesn't put into the R&D fund, and all of the players who did are screwed. This is why there is big money in industrial espionage (and why some companys are willing to engage in dire behaviour to deal with spies). > There was a very good interview (in one Login of L. Peter Deutsch about > his writing Ghostscript, and why he regretted using the GPL). I would be interested in obtaining a pointer to this interview, if you have it. > > Although many companies play the GPL game, this sort of symbiotic > > relationship would not be possible under the GPL and it makes for > > a much more attractive codebase. > > I don't know about "not possible" -- more difficult in > a very competitive envrionment, maybe. The relationship he is describing is one in which tactical information is disclosed, and strategic information is held back. Strategic information has a finite shelf-life, and eventaully degrades either into "tactical" or "don't care". This is why Whistle (my employer) funded the soft updates work in FreeBSD, with the understanding that the restrictive license would expire (in other words, soft updates would "turn tactical") after a certain amount of time. This particular symbiotic relationship would not have been possible under the GPL, since the code could not have been held back under another license, since it must be linked with the kernel, which, if GPL'ed, would render the whole GPL'ed. In this situation, FreeBSD would not have gotten soft updates at all, since it's only the fact that Whistle is able to amortize the costs of developement over a (generously short) lifecycle that enables Whistle to remain profitable. And therefore capable of funding such R&D efforts. Similar examples exist elsewhere. The Juniper Systems firewall code, and the resulting general availability of smtpd/smtpfwdd, is one. So is the Whistle-provided NetGraph code in FreeBSD, and the Frame Relay support it brings to FreeBSD with it. A license is a transaction: it pays people in value, and you must hope that (or work to educate them until) they are enlightened enough to see the difference between "tactical" and "strategic", and act accordingly. The Internet exists today because of the UCB license on the Net/1 and Net/2 code, which resulted in the ubiquity of TCP/IP. Can you provide one example of a monumental undertaking which would not exist without the GPL? Compilers don't count, since compilers existed before GCC; the entire output of the GNU project is merely a duplication of preexisting private (or in the case of Linux, public) efforts. The Internet was an unexpected consequence of the UCB license; yet consequence of that license, it is. Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-chat" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199909080052.RAA16284>