Rob Ricciricci
Dec 19, 2025, 7:18 PMThe attempt to read the UNIX V4 tape is underway!
Replies
- Rob RicciricciDec 19, 2025, 7:50 PM
There will be more video later (there's a TV news crew there) but here's a bit that Jon Duerig has sent from on-site
- Rob RicciricciDec 19, 2025, 8:02 PM
I'm told "there is data" but I honestly don't know what that means yet
- Rob RicciricciDec 19, 2025, 9:36 PM
Apparently the entire UNIX v4 tape was read successfully, what's going to need to happen next is to decode the signal from it, decipher the file formats, etc., I'll post updates as I get them
- Rob RicciricciDec 19, 2025, 10:22 PM
We have the binary image of the tape, link going up soon
- Rob RicciricciDec 19, 2025, 10:27 PM
So far some of our folks think they have found Hunt The Wumpus and the C code for a Snobol interpreter
- Rob RicciricciDec 19, 2025, 11:00 PM
- Rob RicciricciDec 19, 2025, 11:02 PM
- Rob RicciricciDec 19, 2025, 10:29 PM
Here's the document release you were waiting for today!
The UNIX V4 tape!
https://archive.org/details/utah_unix_v4_raw
Credits:
* Jay Lepreau for holding on to this tape
* Aleksander Maricq for finding it
* Jon Duerig for driving it to the Computer History Museum
* Thalia Archibald for doing a huge amount of research into the tape, its history, and file formats, and the upload
* Al Kossow for the tape-reading equipment and doing the actual read
* Len Shustek for the lab where the read was done and the software used to decode it - Rob RicciricciDec 19, 2025, 11:04 PM
Thread about the .tap format for those who want to dig in more:
gaytabasedysfun@treehouse.systemsDec 19, 2025, 7:20 PM@ricci i'm disappointed. i thought you were going to wave a magnet over it, speaking out 'one, zero, zero, zero, one...'
dominicbarometz@treehouse.systemsDec 19, 2025, 7:45 PM
Firehorseart Lives!srfirehorseart@ohai.socialDec 19, 2025, 8:00 PM
Apicultor 🐝apicultor@hachyderm.ioDec 24, 2025, 5:20 PM
Dec 19, 2025, 7:22 PM@ricci Good luck!🤞
John Francis 🇨🇦🦫🍁🫎johnefrancis@cosocial.caDec 19, 2025, 7:56 PM@ricci 😮 😵💫
Irenes (many)ireneista@irenes.spaceDec 19, 2025, 9:15 PM@ricci super exciting!!!!!!!
Dan PortsdanDec 19, 2025, 9:39 PM@ricci waiting for it to turn out to be an extremely low resolution video of Rick Astley
Jonathan Yujawnsy@mastodon.socialDec 19, 2025, 9:50 PM
David Douarddouardda@pouet.chapril.orgDec 19, 2025, 10:49 PM@ricci can't wait to have this in @swheritage archive!
Dec 20, 2025, 1:23 PM@douardda @ricci @swheritage LLMs can wait.
Dec 20, 2025, 5:00 AM@ricci I… I kinda want to write some SNOBOL
- Rob RicciricciDec 20, 2025, 5:03 AM
@mcc it's free and no one can stop you
- Dec 20, 2025, 5:09 AM
@ricci Right right it's just… there's something that feels more visceral about writing SNOBOL from a C implementation on a recovered lost unix tape. Like eating cheese out of the mummy tomb. I know that there's cheese around the corner but THIS cheese has a forbidden thrill
Dec 20, 2025, 5:03 AMI am *very* interested in the Snobol interpreter that’s going to be rad as hell.
PhosphenesPhosphenes@glasgow.socialDec 20, 2025, 2:30 PMBell Labs is going to demand you burn the tape for copyright violation. 😆
Dec 19, 2025, 10:31 PM@ricci@discuss.systems Aaaaahhhh, this is SUPER cool!!!!
Dianealienghic@timeloop.cafeDec 19, 2025, 10:34 PMIs the whole tape really only 14KiB of data?
Space Hobospacehobo@teh.entar.netDec 19, 2025, 10:56 PM@alienghic @ricci https://archive.org/download/utah_unix_v4_raw lists the component pieces including a 2.5MB "tap" file (which I assume is a structured decoding of the tape media) and a 1.6GB "tbin" file (which I assume is raw flux readings).
- Dianealienghic@timeloop.cafeDec 19, 2025, 11:07 PM
ah ok
kevinmehallkevinmehall@hachyderm.ioDec 19, 2025, 10:56 PM@alienghic @ricci If you click "Show All" there is a 2.5MB `analog.tap`. It has ASCII strings in it, so it must be a decoded image of the tape contents, not the analog waveform as the description and filename would suggest.
- Rob RicciricciDec 19, 2025, 10:57 PM
@kevinmehall @alienghic correct, unsure why the file is named that
GaryWiteWulf@cyberplace.socialDec 19, 2025, 10:59 PM@alienghic @ricci nope, 1.6GB
- Rob RicciricciDec 19, 2025, 11:05 PM
@WiteWulf @alienghic This is the raw analog readout from the tape, which was then digitized to make analog.tap
Dec 20, 2025, 3:20 AM@ricci @WiteWulf @alienghic
It's suspicious how the tbin timestamp is newer than the tap file then.Extrapolating from the rest, prior to digitization, the analog data was deskewed using peak statistics?
- Rob RicciricciDec 20, 2025, 3:21 AM
@aarbrk @WiteWulf @alienghic may just be the time that the upload to archive.org completed, the tbin was the last part uploaded
G. Gibsonmistergibson@turtleisland.socialDec 19, 2025, 10:41 PM@ricci Thanks for bringing this to my attention. Wow --> at long last.
ǝʌɐpdvshkn@treehouse.systemsDec 19, 2025, 10:49 PM@ricci incredible work! #unix_surrealism
Q.U.I.N.N.icedquinn@blob.catDec 20, 2025, 12:47 AM@ricci thought this was gonna be the tap test protocol- Irenes (many)ireneista@irenes.spaceDec 20, 2025, 1:07 AM
@ricci we made tap2raw work on analog.tap; for any of our followers wanting to poke at it, please enjoy our copy of analog.raw
- Irenes (many)ireneista@irenes.spaceDec 20, 2025, 1:08 AM
@ricci wait, sorry, file permission thing, give us a sec
- Irenes (many)ireneista@irenes.spaceDec 20, 2025, 1:11 AM
@ricci that should be https://dandelions.irenes.space/analog.raw
we haven't used these two file-hosting subdomains before and forgot which was which
- Irenes (many)ireneista@irenes.spaceDec 20, 2025, 1:12 AM
@ricci anyway the
filecommand calls it a "PDP-11 executable", so that argues strongly that we did it right - Rob RicciricciDec 20, 2025, 1:19 AM
- Irenes (many)ireneista@irenes.spaceDec 20, 2025, 1:20 AM
@ricci what a really fun thing for a friday night in December :3 we're so glad you were able to make this happen, and thrilled to be able to play with it
- Irenes (many)ireneista@irenes.spaceDec 20, 2025, 1:12 AM
@ricci we can't edit our toots but if you'd like us to make a new one you can reshare instead, let us know
- Rob RicciricciDec 20, 2025, 1:25 AM
@ireneista nah I'll quote boost the correct link
- Rob RicciricciDec 20, 2025, 1:26 AM
@ireneista Actually I can't, why don't you make one with the correct link and more descriptive text and I'll boost that
- Irenes (many)ireneista@irenes.spaceDec 20, 2025, 1:26 AM
@ricci thanks, cool
- Irenes (many)ireneista@irenes.spaceDec 20, 2025, 1:26 AM
@ricci thanks, cool
- Dec 20, 2025, 1:22 AM
@ireneista @ricci This link asks for a login?
(edit) But it seems that tap2raw from https://gitlab.com/dps8m/tap2raw works
- Irenes (many)ireneista@irenes.spaceDec 20, 2025, 1:23 AM
@me_ @ricci yes sorry for the trouble. https://dandelions.irenes.space/archive.raw
- Irenes (many)ireneista@irenes.spaceDec 20, 2025, 1:28 AM
- Dec 20, 2025, 2:28 AM
@ireneista @ricci No worries, thanks a lot for your work!
- Irenes (many)ireneista@irenes.spaceDec 20, 2025, 2:31 AM
- Irenes (many)ireneista@irenes.spaceDec 20, 2025, 1:30 AM
@ricci we made tap2raw work on analog.tap; for any of our followers wanting to poke at it, please enjoy our copy of analog.raw https://dandelions.irenes.space/analog.raw
the
filecommand detects it as "PDP-11 executable", so that gives us some confidence we did it right(we've verified this time that the link works. sorry for the trouble if you saw any of the several toots where, in our eagerness to share, we gave broken links. sigh)
- Irenes (many)ireneista@irenes.spaceDec 20, 2025, 1:36 AM
@ricci it is worth noting that you only need to do this conversion if your goal is to poke at the file format and try to decipher it (surely the bulk of it is a tar, and with sufficient inspection that ought to be possible)
if you're trying to use it with opensimh, the
.tapis already what you need. however, we haven't made that boot yet :) we'll report elsewhere if we do 
Eric CarrollEricCarroll@mastodon.acm.orgDec 20, 2025, 1:33 AM@ricci
Lol its rate limited right now.
Dec 20, 2025, 2:56 AM@ricci It's been booted!
Dec 20, 2025, 2:01 AM@ricci In theory, this is the format Plan 9’s tapfs is for reading. I’ll try it tonight.
- Dec 20, 2025, 3:07 AM
@ricci On a first pass, this doesn’t work; I get a bunch of “cksum error” (bypassing the checksum, predictably, doesn’t help). I wonder if there’s a header alignment issue.
I don’t entirely understand what tap2raw is doing, but the referenced analog.raw is readable, but just has a handful of short files with assembly instructions.
