From owner-freebsd-advocacy Thu May 6 9:57:38 1999 Delivered-To: freebsd-advocacy@freebsd.org Received: from b.mx.crl.com (bmx.crl.com [165.113.1.81]) by hub.freebsd.org (Postfix) with ESMTP id C3B8315A07 for ; Thu, 6 May 1999 09:57:35 -0700 (PDT) (envelope-from anarchy@crl.com) Received: from crl.crl.com (crl.com [165.113.1.12]) by b.mx.crl.com (8.8.7/) via SMTP id JAA04213 for ; Thu, 6 May 1999 09:57:35 -0700 (PDT) env-from (anarchy@crl.com) Date: Thu, 6 May 1999 09:57:35 -0700 (PDT) From: Ben Manes To: advocacy@freebsd.org Subject: [svlug] *sigh* (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-advocacy@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG might interest a few of you... ---------- Forwarded message ---------- Date: Thu, 06 May 1999 09:17:10 -0700 From: Derek J. Balling To: svlug@svlug.org Subject: [svlug] *sigh* OK, examples of why/where Linux FAILS in the real world. I work for a Large Web Company. (read the headers, do the math). We are a FreeBSD shop. We are installing Oracle. There is no Oracle for FreeBSD. Derek suggests Linux. Derek receives the usual sighs and groans in anticipation of the usual BSD<->Linux flame-war that occurs between myself and Jan (our security guy, who ALSO happens to be security-officer@freebsd.org so I'm told). Issue # 1: Get the DPT controller working. Motherboard has SCSI built onboard but nothing attached to it. In PCI0 is a spiffy DPT3334 dual-channel SCSI controller. Great, use the EATA-DMA driver and away we go. Except that for some reason Linux doesn't SEE the DPT controller unless we completely disable onboard NCR driver. Never mind that maybe our original intent was to boot off the NCR and let the DPT handle the RAID system for Oracle. Linux won't even touch the DPT until we tell the BIOS to pretend that the NCR doesn't exist. Issue #2: Choose your preference -- SMP or the ability to USE the box The person who did the install (before I got there) installed Red Hat 5.2... which is all well and good except that RH5.2 out of the box doesn't know squat about SMP. Considering all the "advances" between 5.2 and 6.0, silly me says "well, just upgrade to 6.0 for the SMP support and everything will be great." So now, after all is said and done, I have a choice -- reinstall 5.2 and not have SMP support, or have SMP support and a complete inability to compile basic simple tools like "ssh". No, I cannot just use the RPM for ssh. (is there even one? certainly not from RH) Is there ANY QA being done by RH before they release their code? I would think the inability to freakin compile would be something that they MIGHT consider checking? It's not just SSH from what I can tell by reading on Deja News but other minor things like BIND, Kerberos, etc. Any suggestions on how I should get this box operational TODAY using something other than NT would be greatly appreciated. But from my employer's perspective, Linux's second-chance-at-a-first-impression came and went (the first being years ago when we compared BSD/Linux and chose BSD) and I don't see it getting a third shot after this debacle. D -- echo "unsubscribe svlug" | mail majordomo@svlug.org ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ to unsubscribe see http://www.svlug.org/mdstuff/lists.shtml for posting guidelines. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-advocacy" in the body of the message