Differents dev's of cdrecord

[asarch]

When I used to use Mandrake Linux 9.0 I used to burn with this command:

cdrecord dev=0,2,0 ...

Now that I am using Slackware Linux I burn with this command:

cdrecord dev=0,0,0 ...

The question is why this two different addresses and why do they change?

[bwkaz]

The SCSI device number that you give to cdrecord determines which SCSI device it uses to burn. Obviously on Mandrake, your CD writer drive was the third device on your first SCSI bus (it was also LUN 0, the first LUN). On Slackware, it's the first LUN of the first device on the first bus (0,0,0).

The address changes when your system changes the order that SCSI devices get recognized, or the number of other SCSI devices that exist.

[asarch]

Ah, that it means that it depends on how the distro mapped the SCSI device, isn't it?

By the way, what does lun means in

dev=scsibus,target,lun

because lun + a = luna, in spanish means moon? :D

[banzaikai]

Ahhh, memories. Ever own a PET or C64 with the IEEE-488 bus disk drive? It's kinda like that.

scsibus = which controller card/bus you're dealing with. If you have only one, then it's "0". If you start plugging in usb drives, they show up as 1, 2, 3, etc.

target = which device on that bus you're talking to, usually the ID number you have it set to. Most SCSI-I/II controllers handle 8 devices (0-7), with most having the controller card itself set to ID7. The better SCSI-III/W/UW can handle 16 devices (0-15), with ID15 as the controller. Now, since each device has it's own "smarts" on-board, the SCSI designers took advantage of this, and allowed the single on-board controller to extend the SCSI bus by adding:

Logical Unit Numbers (LUN) = the "sub-unit" device of the target address. The best way of looking at this is by thinking of a scanner:

The scanner is set to ID4, and sits on the only scsi card bus, so you'd see it at "0,4,0". Great. Now, you add a spiffy new Auto Document Feeder (ADF) to it, and the on-board scsi interface now needs to control it, So, it just keeps the scanner at LUN0, and assigns the ADF a LUN of 1. Now, any feed commands you need to send go to "0,4,1", and then you scan with 0,4,0. Feed with 0,4,1; scan with 0,4,0; ad nauseum. So, if your LUNs are actually sub-drives off a drive controller, you could theoretically have 49 drives connected (7 IDs x 7 LUNs) on SCSI-I/II, and 225 (15 x 15) on SCSI-III/W/UW. Cool.

Thing is, I've never seen this happen. In fact, most Amiga SCSI cards had the option of turning off the LUN inquiry to speed up the boot time (it only looks at LUN0, so it scans the bus for only seven devices instead of 49 - or 15 instead of 225). And, thus, the fond memory of my PET:

The IEEE-488 (HPIB) was set up very similar to SCSI. When using the devices (0-15!), you'd also specify the drive (Unit#) you're dealing with, too. This held over even with the single drives (the 1540/41). So, to initialize (reset) the second drive of an 8250 at ID8, we'd:

OPEN8,8,15,"I1:":CLOSE8

where the first 8 is the file number used by the PET/C64 to keep track of things, the second is the device, the third is the command channel (0-14 could be specified if you wanted to use the internal buffers for some weird purposes), and the command in quotes tells it to Initialize drive 1.

The IEEE-488 preceeded the SCSI bus by a couple of years, but the SCSI developers (Shugart and Fujitsu, who both made drives that used IEEE-488) borrowed the best ideas from it when they thought out SCSI. Neat.

banzai "OPEN8,8,15,"N0:BANZAI,BK":CLOSE8" kai

[EnigmaOne]

Originally posted by banzaikai
Ahhh, memories. Ever own a PET or C64 with the IEEE-488 bus disk drive? It's kinda like that.

Why do I suddenly feel as old as Don Lancaster?

[banzaikai]

Lancaster rocks! ( www.tinaja.com )

His "cookbooks" are must-haves for anyone delving into electronics. The only other names that come to mind along with his are Steve Ciarcia, Forrest Mims III, and Joseph Carr.

I even got my 64 to talk, thanks to a Byte article and an SPO256-AL2...

banzai "rosin core" kai

[EnigmaOne]

Originally posted by banzaikai
The only other names that come to mind along with his are Steve Ciarcia, Forrest Mims III, and Joseph Carr.
:D
Ciarcia, Mims and Carr were all in the old Radio Electronics/Electronics Now column mill circuit--those were the good old days!
I have to confess that I still have a few Mims and Carr books, along with that big, red Circuits Reference book by Graff, in the bedroom closet.

Originally posted by banzaikai
I even got my 64 to talk, thanks to a Byte article and an SPO256-AL2...

banzai "rosin core" kai
I have to check, but there may still be a set of those--in the original packaging--in my junk box...heh. Bet the data sheet is long-gone by now.

Now you have me racking my brain to remember the name of the guy who used to write the Antique Radio and Experimenters' Corner offerings in PE, back in the day.

I love the smell of Chemtronics in the morning!

(BTW, thanks for the link...I'm wget-ting the Hardware Hacker collection for posterity's sake.)
[Rats! None of his stuff from the 70's in in there.]