From owner-freebsd-advocacy Fri Mar 19 12:53:44 1999 Delivered-To: freebsd-advocacy@freebsd.org Received: from zippy.cdrom.com (zippy.cdrom.com [204.216.27.228]) by hub.freebsd.org (Postfix) with ESMTP id 4E23114BF5 for ; Fri, 19 Mar 1999 12:53:42 -0800 (PST) (envelope-from jkh@zippy.cdrom.com) Received: from zippy.cdrom.com (localhost [127.0.0.1]) by zippy.cdrom.com (8.9.3/8.9.3) with ESMTP id MAA52252; Fri, 19 Mar 1999 12:53:18 -0800 (PST) (envelope-from jkh@zippy.cdrom.com) To: Steve Price Cc: advocacy@FreeBSD.ORG Subject: Re: Linux emulation != FreeBSD sale (was Re: Netscape browser) In-reply-to: Your message of "Fri, 19 Mar 1999 08:34:42 CST." Date: Fri, 19 Mar 1999 12:53:18 -0800 Message-ID: <52250.921876798@zippy.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-advocacy@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG > Try as I might I can't get Netscape to record anywhere that > I'm downloading a copy of the Linux version only to run it on > a FreeBSD box. This is very bad! I download the Linux version. I'm afraid that this is something of a red-herring argument, albeit probably unintentionally. To explain why, let's examine the process which goes on behind the scenes in producing something like the Netscape directory server product and how you, the user, fit into all of this: 1. Vendor X (in this case Netscape) has product Y which they'd like to get to the maximum number of users, both for reasons of profit and to grow mindshare. They then do some research to determine which plaforms will give them the maximum bang-for-buck in this direction, since all development work is expen$ive. 2. Having gone and done various literature searches, vendor X determines that the major OS vendors are the commercial folks (Solaris, SCO, etc) and Linux, at least from the installed base perspective. FreeBSD doesn't quite show up here simply because our installed base is a lot more nebulous and we don't show up on the cover of Time regularly, so they're less inclined to settle for our nebulous numbers (vs Linux's almost equally nebulous numbers). 3. Engineering is given directions based on this survey work and, eventually, produce N binaries for these various platforms. 4. You, the user, are finally presented with this list of binaries on some FTP server. If it's there for Linux, you as a FreeBSD user can run it. If it's not, you can't and it doesn't matter (historically) how loudly you hell for a native version since you just didn't stack up in step #2 and no amount of yelling is going to immediately change that, if at all. 5. Given market feedback, Vendor X eventually goes back and does steps 2-4 again. This is where "emulation" hurts us, yes, but we'd also have never gotten past step 4 without it. Catch-22. Rather than take the emulation out and merely trade one evil for another, however, there are better alternatives. Pressuring vendor X to put "FreeBSD (with Linux emulation)" down as a survey item, for example, so that you can be a proper part of step #5 is one good approach. Another is to try and get vendor X to have their product do just a little bit of extra scratching around to determine whether it's running on a real Linux box or actually a FreeBSD box. This isn't so far-fetched since many vendors are having to do this kind of thing anyway just to figure out which *linux* flavor they're running on, assuming that such demographic/technical data is important to them. There's also a common fallacy here, which is that if we just but had a FreeBSD native version of product Y, life would be good and we'd be well on our way to establishing a good relationship with vendor X and the market it served. Our experience with Xi Graphics and the CDE product proved the exact opposite. XiG sells thousands of copies of CDE for Linux and, after significant pressure from ME (for which I feel more than a little responsible) to get a FreeBSD native version out there so we could reach parity with XiG's product line (and they do sell a lot of FreeBSD servers, so it seemed a reasonable thing to do at the time), they finally invested the time and money to do it. Tada! CDE for FreeBSD! Call marketing! Put it on the web pages! To make a long story short, they sold 3 whole copies of it and finally discontinued the product again. There just wasn't any interest, for some reason, and by demanding a native product I pretty much just wasted their time and money. That's not a good example of how you make friends with vendors or establish market credibility and, by shooting our bolt with CDE, we may well have a much steeper ramp to climb if we ever want to ask XiG for something similar in the future. It's sadly not at all inconceivable to imagine that we could well end up doing this AGAIN to some other ISV if we just blindly demanded FreeBSD native versions in knee-jerk fashion without actually checking to see if there's a *market* for them, and it's not enough to say that this is the ISV's problem and not ours since we're the ones trying to establish some credibility for ourselves - it IS our problem if the ISV goes away again in disgust. That, in a mere 5000 words, is why I'm now a pragmatist. :-) - Jordan To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-advocacy" in the body of the message