Date: Fri, 19 Sep 2003 03:00:40 -0700 From: Terry Lambert <tlambert2@mindspring.com> To: John Baldwin <jhb@FreeBSD.org> Cc: Nik Clayton <nik@freebsd.org> Subject: Re: Sorry about sysinstall. Message-ID: <3F6AD3C8.A9940709@mindspring.com> References: <XFMail.20030918110635.jhb@FreeBSD.org>
next in thread | previous in thread | raw e-mail | index | archive | help
John Baldwin wrote: > On 18-Sep-2003 Terry Lambert wrote: > > Nik Clayton wrote: > >> Then, if the customer has problems installing FreeBSd using your > >> installer, and comes to the project's mailing lists for help, they'll > >> be told to either contact your customer support, or use the project's > >> installer which you will have shipped as part of your product. [ ... ] > > Do you see the similarities between some hypothetical person > > with a third party FreeBSD installer that installs the same > > damn FreeBS plus the same damn packages, and having some well > > known FreeBSD problem biting them, being told to reinstall > > "using the project's installer", because *somehow*, "it *must* > > be the GUI version of sysintall that's biting them on the ass" > > when the battery monitor in KDE fails to work with their > > laptop? > > Try and find one instance on current, hackers, arch, etc. where > someone reported a bug and someone else asked them which installer > they used to install the box. There isn't one, because the project doesn't permit you to call something with a different installer by the name "FreeBSD". > > It's assinine to limit something because of a hypothetical > > situation that could be engineered against anyway, but even > > if it wasn't, will probably never occur. > > It's assinine to make bogus, unfounded statements about the > developers on the mailing lists and their responses to bug > reports. The existance of a separate installer which was not sysinstall in a distrubtion that was still called FreeBSD is why the situation in question is hypothetical. The statements were hypothetical, not unfounded, John. My point was that, even were the bar lowered from "must be byte for byte identical to FreeBSD Disc 1" to "post-install system must minimally contain byte-for byte identical commands and data as if a minimal FreeBSD install via sysinstall had been used", there's no loss of quality control on behalf of the project's management. My other example, which you snipped, was the FreeBSD "Live CD", which, since it's an already installed system on a CD that you are booting, gets around the FreeBSD restriction based on the distribution of uninstalled vs. pre-installed status. Why should the project care about anything but the post-install status of the system? If you *do* care about something more than the post-install state of the system, why aren't you pissed off about people who (1) ship "Live CD"-type products, or (2) ship machines with FreeBSD preinstalled on them? How would you feel about (3) HP/Compaq shipping a FreeBSD installation disk in their "Jump Start" format (i.e. using *their* "Jump Start" installer instead of FreeBSD's) with every HP/Compaq computer they sell from now on? Could they call it FreeBSD, even though it didn't use the FreeBSD installer? And if the FreeBSD Disc they shipped with the thing was "FreeBSD recovery Disc 1", instead of an identical copy of FreeBSD Disc 1? Is it a terrible thing that the Disc they ship came pre-configured correctly for the hardware it shipped with, instead of making the poor user futz around with XFree86 config files? To me, it doesn't matter *how* the FreeBSD got on the system, what matters is that it *be* FreeBSD, and not some mish-mash of something-not-FreeBSD and FreeBSD calling itself FreeBSD; *that* would cause problems. I really don't care if FreeBSD got installed because someone waved a magic wand over their computer. In fact, I'd prefer it: I'd buy a wand for myself, and visit every computer store in a 100 mile radius. 8-) 8-). -- Terry
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3F6AD3C8.A9940709>