From owner-freebsd-hackers Thu Jul 18 05:55:55 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id FAA16284 for hackers-outgoing; Thu, 18 Jul 1996 05:55:55 -0700 (PDT) Received: from mole.mole.org (marmot.mole.org [204.216.57.191]) by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id FAA16279 for ; Thu, 18 Jul 1996 05:55:53 -0700 (PDT) Received: (from mail@localhost) by mole.mole.org (8.6.12/8.6.12) id MAA11084; Thu, 18 Jul 1996 12:54:27 GMT Received: from meerkat.mole.org(206.197.192.110) by mole.mole.org via smap (V1.3) id sma011082; Thu Jul 18 12:54:25 1996 Received: (from mrm@localhost) by meerkat.mole.org (8.6.11/8.6.9) id FAA27245; Thu, 18 Jul 1996 05:54:24 -0700 Date: Thu, 18 Jul 1996 05:54:24 -0700 From: "M.R.Murphy" Message-Id: <199607181254.FAA27245@meerkat.mole.org> To: dennis@etinc.com, terry@lambert.org Subject: Re: mitsumi CD-ROM Cc: hackers@FreeBSD.org, mrm@mole.mole.org Sender: owner-hackers@FreeBSD.org X-Loop: FreeBSD.org Precedence: bulk > > > Well, I hate to defend Terry when he's so off-base, but you hardly > > need the kit for FreeBSD, and it does add $25. or so. Agreed, The very good SCSI interface is, however, still nearly ten times as much as the somewhat crummy IDE interface, but the IDE interface works. With a 64MB P5/133, 2940, Quantum 4G, I get nearly 8MB/sec bulk single-disk throughput with 2.1.5. With a 16MB AMD586/133, VLB IDE, WD 1.08G, I get slightly over 3MB/sec with 2.1-stable and wdc0 flags 0x80ff80ff. The IDE uses a few percent of the CPU during this activity. The IDE is OK. What makes it so is that the CPU is fast enough, I think. The IDE is not as good as the SCSI, I also think we'd all agree on that. It does cost less. In some cases that's a reasonable consideration, in others it isn't. It astounds me that somebody can sell the IDE interface, in a box, shrinkwrapped, with documentation, shipping, marketing, and retail markup, for about $20. And they throw in cables, too. Did we mention how much good SCSI cables and external enclosures for the disk farm might cost? :-) :-) Accepting the adequate when the excellent is possible is difficult. > > Doesn't agree with Dennis :== off-base. 8-) 8-). :-) > > > There aren't enough on-board SCSIs to chose from to consider it, I have concerns about putting the kitchen sink on the MB. Then, again, I don't like the MB concept at all, preferring a passive backplane with alternate fastbus architecture. Oh well, it didn't happen. I'll make do. [...]> > > > The point is, if you have 50 machines that don't care much about disk speed, > > there certainly is a difference in cost. > > If you want to get down to brass tacks, you could argue that you don't > need disks for most of those machines. > Diskless, blech. Dataless, well, OK. Not quite a fair statement, since there are places for diskless machines, but I think paging over a network leaves some things to be desired. I've been watching the little red light on the machine on my desk as I type and revise this. It hasn't blinked very much. I'd just as soon have spent $200 for it not to blink much as to have spent $500 for it not to blink much. There's a place for IDE. There's a place for SCSI. -- Mike Murphy mrm@Mole.ORG +1 619 598 5874 Better is the enemy of Good