I have a number of internal 8-bit SCSI-1 drives that I'd like to archive for historical purposes. These are all the old NeXT Cubes and NeXTStations (monochrome and color) that were used by id Software to create DOOM and Quake. I'd like to donate the machines somewhere they can do some good, but John Carmack had me promise more than a decade ago that I'd wipe the drives if I ever passed them along, and I'd hate to lose any priceless data that's on them.

In theory, I realize SCSI is backwards compatible, and so I may just be able to plug it into an appropriate cable & modern SCSI card...

But I suspect that it's not quite that simple, and further, it's not obvious what type of cable or adapter I would need and where to acquire one.

If I successfully do extract the data, I'll ping John and see if it might be ok to release the disk images to the public. It's his data, so it'll be his call. id has open-sourced much of their code, so there might not be anything useful here, but the geek in me would hate to lose it if there is. :)

  • John Romero said it was OK to just throw the drives in the garbage. Of course that was before the whole "Daikatana" thing.
    – joeqwerty
    Jan 17, 2013 at 19:42
  • There is no problem with connecting 8 bit wide SCSI drives to a wide (16 bit) hostadapter. And if you have any large SCSI drives using only 6 byte commands (basically SASI) then I am very interested in them to help resurrect an old computer. ( See tnodiana.com/node/12 . That mini moved to the TU/e and from there to the computer club. We still have it, but no drive and no network card)
    – Hennes
    Jan 17, 2013 at 19:50
  • 1
    A card with 50 pins (8 bit) connector can be found for as low 12 pound s on ebay. Just make sure you have a computer with a matching interface (e.g. PCI) and for ease of use that you get one with a 40 pins interface. As for termination: SCSI termination is done at both ends of the bus, pulling cables up to +TERMPWR with a 220 ohm resister and down to ground with a 330 ohm resister. Those old drives probably still have sets of removeable resistors on them.
    – Hennes
    Jan 17, 2013 at 20:00
  • 4
    Nextstep used its own minor variation of good old UFS so a Linux machine should be able to read it. Linux mount even has a ufstype=nextstep option.
    – user143690
    Jan 17, 2013 at 20:07
  • 5
    Even outside the scope of your question, you may consider contacting Jason Scott (not me) at textfiles.com, or @textfiles. He works for the Internet Archive and he's particularly keen on digital preservation.
    – jscott
    Jan 17, 2013 at 20:15

7 Answers 7


Old NeXT hardware. SWOON!

Getting a SCSI2 -> SCSI 1 adapter should be trivial. There were both internal and external varieties. Google is your friend. For some reason I thought the "newer" slabs should be SCSI2 but it's been a long time.

You still have to terminate the SCSI chain. Keep to addresses 0-7.

Further just Googling for "scsi-2 pci card" comes up with hits. An Adaptec card for instance should still have good working SCSI support on linux. Looks like one can be had for ~$20.

Here's the thing tho. How the heck are you going to read the file system using anything BUT NeXTSTEP? I suppose you might be able to boot an old NeXTSTEP/OpenStep Intel cd, and ftp/scp/rsync over the files you're looking to save after you've mounted the drive. The trick would probably be finding an Adaptec card which OpenStep supported.

Good luck!

  • 4
    Just cat or dd the whole disk to a file and worry about the format later? (Probably using loopback mounting later on).
    – Hennes
    Jan 17, 2013 at 20:02
  • 3
    Can be mounted in Linux as user143690 above already mentioned.
    – Tonny
    Jan 17, 2013 at 20:16
  • I believe the slabs have internal SCSI-1, external SCSI-2. But it's been a long time since I drooled over those spec sheets (and the engineering samples at NeXT HQ!) Jan 17, 2013 at 21:48
  • I was planning on using dd to create an exact copy and figuring out the rest later. IIRC (its been awhile), NeXT used a UFS filesystem similar to other BSD variants. I suspect Linux can just mount it natively, like it can other UFS variants, so I'd then be able to (with John's permission) pass around the disk images and anyone could mount them on a Linux box. Jan 17, 2013 at 21:52
  • The Adaptec 2904 worked for me in Linux a couple of years ago... Jan 19, 2013 at 23:38

I've got a NeXTStation slab, and I'm trying figure out why it doesn't boot. I removed the drive from the NeXTStation and installed it in old PC (IBM Z Pro) with an onboard UltraWide SCSI controller. I found an adapter which converts from the 80-pin UW-connector to the old style 50-pin connector used by SCSI-1 drives. Other obvious choices could be digging up an SCSI-1 controller. If it's a NeXTStation slab, the drive is probably already terminated with SIL-style resistors, so you don't need any on the SCSI cable.

I booted up the system and the drive was detected as /dev/sdc. Before doing anything else, I took a complete backup of the entire drive like this:

dd if=/dev/sdc of=backup.img bs=$[1024*1024*32]

It's a Seagate ST1280 drive, so the resulting file was around 238MB. After this, I manually added the ufs module to the kernel, although this may not be necessary:

modprobe ufs

I'm running Fedora 21, so the ufs module was part of the kernel-modules-extra package. After this, the disk can be mounted with this command:

mount /dev/sdc /mnt/nextstep -t ufs -o ufstype=nextstep

The UFS module installed from kernel-modules-extra only supports read only though, as seem with dmesg:

ufs: ufs was compiled with read-only support, can't be mounted as read-write

If you need to mount it read-write, you probably need to recompile the kernel and/or module.


Do the machines boot? It's quite easy to get them up on a modern network and tar/transfer the data via NFS...here's a picture of mine w/ my NAS/RAID mounted: http://hzsystems.com/scrap/nextnfsmount.jpg


You should indeed be able to plug them into any SCSI-1 or SCSI-2 device chain and then wipe them from there. However, if the machines boot, then what I'd do is to export a shared folder over NFS from a machine with enough space to hold the contents of the drives, then tar the complete contents of the drives to the shared folder, and then run dd to wipe them.

You might want to install gnu tar first, assuming that Carmack left you with gcc on the box, because the version of tar installed with NeXTStep has trouble with pathnames longer than either 255 or 1024 characters - I forget which.

I'd bet they boot. My own cube still does.


These can be read on a linux PC with a SCSI adapter. I would be willing to do it if you're close.

The best resource about these matters is the ClassicCmp mailing list (http://www.classiccmp.org/) - a very active community of very knowledgeable and technical people. Post to the list; I'm sure someone will be close enough to help.


The easiest solution is to read the drives on the original NeXT equipment. I have three NeXTStations color, and I would expect two of them to be fully operational, although I did not check that for many years. Please let me know how I can be of help - I am in AZ and CA.


As noted above, the easiest thing to do is dump them using an existing NeXT box. I have an external drive that I plug disks into, and dump the image using dd, cpio and tar, then FTP the images to my archive box. I do the same thing for MO/OD images (I have working drives). If you have a dual drive 50 pin cable, you can get by without the external drive.

Alternatively, if you can find an older pci adaptec 50 pin card, and plug it into a Linux box, you'll be able to dd the image in Linux, and later mount the disk using a loop back on Linux. In some ways this is better because dd on the NeXT doesn't dump the entire disk.

If you have trouble, I'd be willing to help or even dump the drives for you. I've done that for a bunch of people/companies.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .