From owner-freebsd-advocacy Sun Jul 9 11:30:12 2000 Delivered-To: freebsd-advocacy@freebsd.org Received: from lariat.org (lariat.org [12.23.109.2]) by hub.freebsd.org (Postfix) with ESMTP id 5B67937B66B; Sun, 9 Jul 2000 11:30:06 -0700 (PDT) (envelope-from brett@lariat.org) Received: from mustang.lariat.org (IDENT:ppp0.lariat.org@lariat.org [12.23.109.2]) by lariat.org (8.9.3/8.9.3) with ESMTP id MAA12179; Sun, 9 Jul 2000 12:29:58 -0600 (MDT) Message-Id: <4.3.2.7.2.20000709122318.04a05100@localhost> X-Sender: brett@localhost X-Mailer: QUALCOMM Windows Eudora Version 4.3.2 Date: Sun, 09 Jul 2000 12:29:53 -0600 To: "David Schwartz" From: Brett Glass Subject: RE: Emulation (Was: No port of Opera?) Cc: , In-Reply-To: References: <4.3.2.7.2.20000709004740.049f9740@localhost> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: owner-freebsd-advocacy@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG At 02:11 AM 7/9/2000, David Schwartz wrote: > While I think that this could potentially happen, I don't think it's >inevitable. A lot hinges on how vocal and visible the FreeBSD crowd is. The FreeBSD crowd isn't nearly vocal enough! In fact, certain people seem to advocate AGAINST advocacy. > Let's consider a case where it might matter. You have to make a few >assumptions, or it doesn't even matter: > > 1) The company has a Linux build. (Otherwise, it doesn't matter whether or >not FreeBSD has Linux emulation.) > > 2) They don't [yet] have a native FreeBSD build. (Otherwise, it doesn't >matter whether or not FreeBSD has Linux emulation.) > > 3) The Linux build works on FreeBSD under emulation. (Otherwise, it doesn't >matter whether or not FreeBSD has Linux emulation.) > > Now think about what happens given that these three things are true. Having >the Linux version being used on FreeBSD can only increase the number of >people talking about the product and FreeBSD. And telling the company, "Hey, it runs under emulation!" Which encourages it not to do a native port. > Now, two things are possible: > > 1) The non-native build works perfectly under FreeBSD. (In which case, >except for OS bigotry, there is no reason to have a FreeBSD-native version.) Not so. Users still won't be supported on a platform other than the one for which the binary is targeted. Also, you are making the dangerous assumption that the emulation is perfect and will always be so. This is not wise. The next version might be Linux-specific. > 2) A native build would work significantly better. (In which case, the only >reason for the company not to make a native build is if it doesn't justify >the improvement (in which case, they _shouldn't_ make one just to make the >FreeBSD crowd happy), or it does (in which case, if they're smart, they'll >make one.) Unless FreeBSD users refuse to use the Linux build under emulation, there will be no reason to "make the FreeBSD crowd happy." This is why the "portable ABI" is such a powerful concept. It turns the tables; the most advantageous route for the programmer is to write to the FreeBSD ABI and API for BOTH platforms. And since the Portable ABI would be BSD-licensed, they could distribute it with their code -- or even link it in -- without risk of GPL contamination. --Brett To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-advocacy" in the body of the message