Path: utzoo!utgpu!news-server.csri.toronto.edu!clyde.concordia.ca!uunet!ccicpg!conexch!amoeba2!rps2 From: rps2@amoeba2.UUCP (Rick Stevens) Newsgroups: comp.sys.amiga Subject: Re: 3000 autobooting a 1096N Message-ID: Date: 20 Aug 90 04:17:58 GMT References: <760.26C795C2@weyr.FIDONET.ORG> Lines: 41 In article <760.26C795C2@weyr.FIDONET.ORG> Randy.Coghill@f70.n140.z1.FIDONET.ORG (Randy Coghill) writes: >Dave what I would like is to find out if Commodore is going to fix the >problem (as far as I know there scsi routines are the only ones that >have this timing problem). While they do put these nice and fairly >expensive Quantums drives in the systems it is almost irrisponsible >(sp) for them to ignore the fact that some users would go with >different brands to suit there needs (for the same price as a Quantum >40 I got the Seagate 1096N which gives me double the storage capacity). > >The point is that Commodore seems to ignore most things and so >entirely as they please and the customer is not supposed to say >anything and live with it....Well I want a working product with the >hardware I have in my system. > >Randy First off, it's not a Commodore glitch, it's Seagate's. Seagate drives do not fully conform to the SCSI specification. If they did, the drives would show up as "not ready" on the bus after the specified bus delay time, and the 3000 would wait for them to go ready. The problem is that the Seagate drives don't show up on the bus AT ALL (meaning that, as far as the computer is concerned, they don't exist) until they load their firmware images from the drive platters themselves, which means that the drive has to spin up before the drives even show up as "not ready". CBM's SCSI routines go out and scan the bus. They see no drives, hence no autoboot. Now, CBM could put in a delay to wait longer before querying the bus, but that gets around a third-party manufacturer's problem. Sure, it'd be nice, but it's not necessary. This is not meant to flame anyone, CBM, Seagate, or Randy. It is just to point out what the problem actually is. ------------------------------------------------------------------------------- Rick Stevens, Small Scale Systems of Southern California sysop@xyclone.uucp | rps2@amoeba2.uucp | CIS: 75006,1355 | BIX: smallscale "I'm tellin' ya, Valiant! Da whole ting stinks like yesterday's diapers!" - Baby Herman -------------------------------------------------------------------------------