Date: Wed, 22 Mar 2000 21:19:18 +0000 (GMT) From: Terry Lambert <tlambert@primenet.com> To: rsidd@physics.iisc.ernet.in (Rahul Siddharthan) Cc: tlambert@primenet.com (Terry Lambert), adsharma@sharmas.dhs.org (Arun Sharma), freebsd-chat@FreeBSD.ORG Subject: Re: On "intelligent people" and "dangers to BSD" Message-ID: <200003222119.OAA20201@usr07.primenet.com> In-Reply-To: <Pine.LNX.4.20.0003221411070.691-100000@theory8.physics.iisc.ernet.in> from "Rahul Siddharthan" at Mar 22, 2000 02:28:22 PM
next in thread | previous in thread | raw e-mail | index | archive | help
> > I think that if a professional software company were to do a > > formal code review on "fetchmail", the case-study basis for > > Raymond's paper, and publish the results, the entire Open Source > > movement might very well falter as "an obviously bad idea, in > > retrospect". The press is already starting to back up on the > > idea of Linux (right or wrong); see: > > > > <http://www.zdnet.com/zdnn/stories/news/0,4586,2470425,00.html>; > > That article isn't really backing up on the idea of Linux being > high-quality: it worries about fragmentation among distributions, > and "what happens if Linus gets hit by a truck" -- ie a possible > fork in the kernel itself. For the first question, the article > itself talks about the proposed linux standard base, and for the > second, I don't think any linuxers are worried: already Alan Cox > has as much respect as Linus himself, and there are others down > the line, so a code fork would look very unlikely even if Linus > did get hit by a truck. On the whole the article looked pretty > positive to me -- I've seen much worse, anyway. I think you are csting this in a more favorable light than perhaps a businessman not already a Linux advocate would cast it. The libc/glibc split, while claimed healed, and the problem with commercial software only being "supported" on a single distribution, are both looming, IMO. It also seems to me that Linux could take a page from my FABIO (Free Application Binary Interface Objective) proposal, and give vendors a way to disable "all that which is not cross distribution" by flipping a switch during porting, so that all distributions will be guaranteed to run a given commercial binary. This in itself is significantly less desirable than my initial proposal, which is that either the UnixWare or Solaris x86 ABI be taken as a base, and the same "disable everything which is not cross platform during porting" requirement, so that a single binary will run on all "FABIO compliant" (via self certification) platforms -- Level 1 compliance means that you can turn off the non-FABIO ABI components, and Level 2 (the default for the default binary compatability platform) means that you can't turn them off, so you shouldn't do your porting/developement there, but you can be guaranteed that you can run there afterward. As Linux distributions add "value add", as the article points out that they are doing, they are guilty of the same in-fighting that fragmented the UNIX market in the 80's and early 90's: they are attempting to differntiate themselves from the other distributions in their perceived ecological niche -- instead of looking at the big picture: "Oh, we _can't_ compete with Microsoft!" -- and in the process, they are screwing each other and ignoring the real threat to their market. And that threat is _not_ another UNIX. > > I think litigation is only going to be an issue when Open Source > > becomes more than a paper-dragon threat to commercial enterprises; > > It's good to note that you say "when" rather than "if". That's future perfect tense, based on "only going to be"; that's an implied conditional. I didn't mean to imply a belief in a manifest destiny for Open Source, so maybe it's not as "good" as you thought I meant. > The growth of linux in mindshare is pretty phenomenal, I increasingly > see ordinary people curious about it. This is a function of hype. It doesn't matter if it's deserved, or not, it's still hype. > Moreover, the open-source > people are widely seen as the "good guys", no longer confused > with the media image of "hackers" as vandals. Or the word "hackers", at all. The public still sees "hackers" as evil forces bent on stealing their credit card numbers, should they be so stupid as to buy something online. > BSD is not so well > known but it doesn't have to be: once the open source alternative > gets accepted, people will start getting to know BSD too. Not an issue, really. > I suspect that litigation on patent infringements, even if legally > possible, will become rather hard by that time, for PR reasons. > As one article some time back put it, it would be like suing > Santa Claus. Or suing God -- one of the higher rated Ally McBeal episodes, when it initially aired. I really don't see this as an issue. USL went after BSDI after their "1-800-IT'S-UNIX" phone number woke the lawyers up on a trademark issue, and they didn't go back to sleep, because that's the purpose of lawyers. USL went after UCB when BSDI hid behind the Net/2 distribution hoping UCB would shield them from USL trying to turn Trade Secret protection into some kind of Patent equivalent, when the law clearly states that once disclosed, the Trade Secret can never be reclaimed, and only the discloser can be sued, and then only for direct damages resulting from the disclosure (this whole area was well mapped out in case law, based on the steel industry and Andrew Carnegie). And then UCB caved, after Dennis Ritchie and other Bell Labs employees offered to testify on UCB's behalf, and MIT offered to bankroll them with money from their impressive patent portfolio. Law suits _will_ happen, and law suits _will_ get settled in dingily lit back rooms and in the offices of corporate VPs getting ragged on by their employees, as Mike DeFazio, VP of the Novell/USG was ragged on. Take it as a given that law suits are the result of a net positive coming out of a risk analysis equation, and that this will not change in the foreseeable future. > In any case, to protect the movement (and the industry as a > whole) those open-source people who are beginning to get into > influential positions in software companies should start lobbying > for reforms on software patents. Now is really the time, with the > statements by Tim O'Reilly and Jeff Bezos on the Amazon patents, > and Bezos himself calling for reforms. I'm sure if heavyweights > like IBM and HP can be persuaded to push for it it will be done. That's not going to happen. Like Spec. 1170, and the UNIX trademark branding process, and the POSIX certification branding process, and the Motif trademark branding process, and hundreds of others, a large patent portfolio is more like an admission ticket: it's not so much to support you, as it is to keep other people out, and if you can't do that (worst case scenario), you use it as trading cards to prevent other people keeping _you_ out. Right now, I think that's seen as a significant competitive advantage, and a reason for higher-than-would-otherwise-be market valuation -- and one of the businesses of publically held companies is to protet and increase shareholder value. Look at any patent-rich company's annual report, and you will see an explicit reference to their patent portfolio. > On that subject, here is a question I've often wondered about: > the FreeBSD distribution comes with the Unix compress command > included. Doesn't this infringe the Unisys patent on LZW? No. The command, unmodified, was permitted. The encoding is actually the only thing affected by the patent, in any case, since that is Unisys' (Terry Welch's, actually) contribution. The LZ (Lempel-Ziv) decoder will decode both, and is not patent protected. I would have to look it up, but I'm pretty sure that patent has expired, unless it was submerged (filed but not executed) prior to GIF images becoming common on the net. At the time, the enforcement was really more driven by CompuServe, attempting to combat ISP's for market share, than it was by Unisys -- though Unisys made the most of it to float their stock price a bit higher (it was in the toilet), with the press coverage. 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?200003222119.OAA20201>