We have some more information on this! One of @regehr's grad students

If that name sounds familiar to you, it's probably because his teapot is ubiquitous in computer graphics: graphics.cs.utah.edu/teapot/

UNIX News July 30 1975 : USENIX Association : Free Download, Borrow, and Streaming : Internet Archive

UNIX NewsVolume 1 Number 1, July 30, 1975

^{Internet Archive}

Some local stories about this tape:

attheu.utah.edu/science-techno…

ksltv.com/science-technology/u…

We're going to try to get it read in a couple of weeks

University of Utah team discovers rare computer relic

A research team at the University of Utah uncovered a rare piece of computing history.

^{Shelby Lofton (KSLTV.com)}

Here's the document release you were waiting for today!

The UNIX V4 tape!

archive.org/details/utah_unix_…

#retrocomputing

UNIX V4 tape from University of Utah (raw) : Computer History Museum : Free Download, Borrow, and Streaming : Internet Archive

UNIX V4 tape from the University of Utah, received by Martin Newell circa June 1974 around when he modeled the Utah Teapot.This is the raw analog waveform and...

^{Internet Archive}

do you know if V4 Unix would have included anything like a source tree, including the kernel?

because if so, I suddenly find myself curious as to how long it'll be until
A. someone gets it running on a VM (or even on real hardware!), and
B. gets Doom to run on it... assuming it doesn't already

@aud and so much of it `#include`s .c files

it was a different era

Sensitive content Clicca per aprire/chiudere

@aud yea, I think "boot" is a far-too-modern paradigm here.

I don't know much about it at all but my coarse understanding is that you'd read in however much of the tape needed reading, then enter `/etc/init` and now your system is running.

As evidence to support my hand-waving argument, `/usr/source/s1/init.c` includes `case reboot` that basically re-enters `/etc/init`

@aud and getting all the way down to brass tacks: that re-entry is done with `execl()`, which is implemented in assembly in `/usr/source/s4/execl.s`

Really not a lot of semantic distance between the user and "the operating system is just one program in a trenchcoat full of other programs"

@SnoopJ @aud Source for init in /usr/source/s1/init.c

maybe I will run this instead of systemd

@bk1e @aud love this comment from the scheduler:

```
/*
* no easy core,
* if this process is deserving,
* look around for
* oldest process in core
*/
```

weighing your process's soul to see if it is lighter than a feather

Sensitive content Clicca per aprire/chiudere

/*
 *      Copyright 1973 Bell Telephone Laboratories Inc
 */

<a class="hashtag" data-tag="include" href="https://donotsta.re/tag/include" rel="tag ugc">#[url=https://donotsta.re/tags/include]include[/url]</a> "../param.h"
<a class="hashtag" data-tag="include" href="https://donotsta.re/tag/include" rel="tag ugc">#[url=https://donotsta.re/tags/include]include[/url]</a> "../user.h"
<a class="hashtag" data-tag="include" href="https://donotsta.re/tag/include" rel="tag ugc">#[url=https://donotsta.re/tags/include]include[/url]</a> "../buf.h"
<a class="hashtag" data-tag="include" href="https://donotsta.re/tag/include" rel="tag ugc">#[url=https://donotsta.re/tags/include]include[/url]</a> "../conf.h"
<a class="hashtag" data-tag="include" href="https://donotsta.re/tag/include" rel="tag ugc">#[url=https://donotsta.re/tags/include]include[/url]</a> "../systm.h"
<a class="hashtag" data-tag="include" href="https://donotsta.re/tag/include" rel="tag ugc">#[url=https://donotsta.re/tags/include]include[/url]</a> "../proc.h"
<a class="hashtag" data-tag="include" href="https://donotsta.re/tag/include" rel="tag ugc">#[url=https://donotsta.re/tags/include]include[/url]</a> "../seg.h"

char    buffers[NBUF][514];
struct  buf     swbuf;

struct  { int int;};
<a class="hashtag" data-tag="define" href="https://donotsta.re/tag/define" rel="tag ugc">#[url=https://donotsta.re/tags/define]define[/url]</a> PS      0177776
int     tmtab;
int     raflag  1;

bread(dev, blkno)
{
        register struct buf *rbp;

        rbp = getblk(dev, blkno);
        if (rbp->b_flags&B_DONE)
                return(rbp);
        rbp->b_flags =| B_READ;
        rbp->b_wcount = -256;
        (*bdevsw[dev.d_major].d_strategy)(rbp);
        iowait(rbp);
        return(rbp);
}

@SnoopJ @aud check out this 🔥 /etc/rc:

```
rm -f /etc/mtab
chdir /tmp
rm -f [a-t]* [x-z]* v*
chdir /
/etc/update
```

@aud okay yes and if I had just slowed the heck down and read `/usr/sys/ken/main.c` a LITTLE more carefully I would see clearly that it is setting up `init` and then invoking the scheduler

welp

It's really cool how abstract-but-actually-very-concrete that is to me, someone who has functionally zero exposure to systems programming:

```
/*
* make init process
* enter scheduling loop
* with system process
*/

if(newproc()) {
expand(USIZE+1);
u.u_uisa[0] = USIZE;
u.u_uisd[0] = 6;
sureg();
copyout(icode, 0, 30);
return;
}
sched();
```

@aud thinking about this a bit more in the context of that `unix` binary, I'm probably wrong, but we're well beyond what I know about UNIX here.

But consulting V7 setup, it looks like `unix` indeed was the entrypoint. Maybe my mental model of `init` as the effective base of a process tree is also too modern!

wolfram.schneider.org/bsd/7thE…

It's been booted!

squoze.net/UNIX/v4/

Thread about the .tap format for those who want to dig in more:

github.com/larsbrinkhoff/emacs…

Repair Gosling emacs.tap from Universität Stuttgart by thaliaarchi · Pull Request #17 · larsbrinkhoff/emacs-history

Gosling Emacs file ftp.informatik.uni-stuttgart.de/pub/cm/dec/decus/emacs.tap is corrupt, as reported in #15. I repair two broken parts. Every 4096 bytes, the byte sequence "\0\x10\0\0\0\x10\0...

^GitHub