It's unclear, but possible. In 2BSD, 3BSD and 4BSD, there aren't very many copyrights, but they are all Regents copyrights on pascal, assembler, termlib and some plotting software. Well, there appeared to be two Bell Telephone Laboratories, Inc copyright in lex.c and sed.c. The vast majority of files had no copyright attached to them at all, even ones that were little changed from 32V or 7th edition. Same was true for 4.1BSD. 4.2BSD had rcs which was Copyright Walter F. Tichy and Ken Harrenstien, SRI International (mostly the former). fp which was Copyright Scott B. Baden, indent which was copyright Board of Trustees of the University of Illinois. sccstorcs copyright Kenneth L. Greer. and cpm which was copyright by Helge Skrivervik, UCB. and was one of the few files to have a permissions grant "Permission granted for use by UNIX* licencees." though many of these were for manual pages. And sccstorcs did have the permission * All rights reserved. No part of this software may be sold or distributed * in any form or by any means without the prior written permission of the * author. on it, through several releases (though it was removed before 4.3BSD-Reno). 4.3BSD added a bunch more copyrights of various people: Digital Equipment Corporation Tektronix Inc Advanced Computer Communications and likely others. Starting in 4.3BSD-Tahoe we see lots of * This code is derived from software contributed to Berkeley by * Excelan Inc. which have Regent copyrights, and sometimes the original contributor copyright. This was the time that we started also seeing the BSD license is a proto form. This continues in 4.3BSD-Reno to a greater degree. 4.4BSD continues this to a ridiculous degree. So, maybe it did happen, but I find no extant evidence of a copyright being removed and replaced by Berkeley. If anything, once files started being marked with a copyright notice, they seem to be retained over several releases and on the 2BSD series where the code was merged into. Now, it's not clear if all the code contributed by folks executed paperwork assigning the copyright to the Regents or not. But it looks like in many cases credit was given, at least in the time period starting with 2BSD. 1BSD lacks the word 'copyright' but kirk's archive has all the files in .a archives which are grepable. It didn't include any AT&T code. Warner On Wed, Sep 21, 2022 at 6:36 AM Rob Pike <robpike(a)gmail.com> wrote:

Yea, if you look at the license statements from the 80s and early 90s from USENET and the BBS scene, you'll see that evolution playout. If you subscribed to the 'trade rags' of the 80s, you'd find them talking about Copyright some, but not much about licensing, so there were a huge number of 'licenses' that we'd laugh at today as being totally insane / unenforceable / unclear / etc... There were also some presentations at DECUS in the mid 80s (and maybe earlier) on the topic as well as their distribution of tapes got large enough... There was also a pointer to Jeremy C Reed's http://www.bsdnewsletter.com/2013/12/Features186.html as well as his book on BSD history http://www.reedmedia.net/books/bsd-history/ (which I'd love to get a copy of). It wasn't clear if it had been published or not from the link I'd love to see John's handout... Warner On Wed, Sep 21, 2022 at 8:51 AM Rich Salz <rich.salz(a)gmail.com> wrote:

below in blue ... On Wed, Sep 21, 2022 at 9:47 AM Warner Losh <imp(a)bsdimp.com> wrote: The legal term '*copyright*' and historical term of '*provenance*.' I agree with Warner that I know of few if any cases where copyright was not maintained when it was in the code itself. And as he points out, please grep through the archives and I think that will be found to hold true. But I also think Rob rankle comment is fair. Joy and was noted for recognizing cool ideas and adding them into 'Berkeley UNIX. The line at the time was he took ideas and '*peed on them to make them smell like Berkeley*.' For example, 'Berkeley Joy Control' came from Kulp via Europe and MIT, the network stack famously started at BBN, and a lot of the support for limits and user controllers from Australia. Yes, the CSRG team did do a great deal of innovation as well as integration, but the line between the two was not always easy to see from the outside. And I think developers outside of UCB sometimes felt (to use Rob's words) 'rankled' for CSRG getting credit for some of innovation that really belonged to others, because the CSRG team was the distribution vehicle. ᐧ

On Wed, Sep 21, 2022 at 8:59 AM Clem Cole <clemc(a)ccc.com> wrote: That makes a lot of sense. When there was a name, it was preserved, but a huge amount of the sources had nothing at all in the source files to identify it. One big area of contribution was into the kernel where the options sometimes contained the name of where the code came from. In the 2BSD kernels we see eg TEXAS_AUTOBAUD, MENLO_OVLY, MENLO_KOV, MENLO_JCL, MPX_FILS, CGL_RTP and a bunch of UCB_ names. It's clear the non UCB ones came from elsewhere, but there's no info on where they came from (they are documented in setup.5 at least so I know what they are). So given the sparseness of the early marking for provenance, the coments make more sense and give a better timeframe to it. Warnerᐧ

On Wed, Sep 21, 2022 at 4:07 PM Bakul Shah <bakul(a)iitbombay.org> wrote: Indeed. One of the reasons that USL settled was because they released 32V without the proper copyright notices (whatever that means, but in this case there were none at all in the media) and the judge issued a preliminary ruling declaring that 32V didn't have copyright protection. They were scared of losing that and that motivated them to settle. Interestingly, in the settlement, USL specifically agrees not to pursue trade secret claims against anybody using 4.4BSD-lite, but copyrights aren't mentioned, outside of the restricted files getting an USL copyright added: "c. USL agrees that it shall take no action against any person who utilizes any methods and concepts in the restricted files which as of this date have become available to the general public by acts not attributable to the University, its employees or students...." and "i. USL agrees that it shall take no action base on the use or distribution by any person of material contained in the Unrestricted Files." also (for another message in the thread) "f. USL agrees that it shall affix the University Copyright Notice and the University Acknowledgement to the files in exhibit C.." Where the wording of those phrases is defined elsewhere in the settlement. And there's about 8 paragraphs as to exactly how AT&T will do this in both code and documentation... Warner

Sure. And of course there will always be interpretations of how much one idea salted another. But the seeds of the discomfort come from these various solutions and frankly how much the original authors were bought into/part of the integration process. On Wed, Sep 21, 2022 at 5:14 PM Theodore Ts'o <tytso(a)mit.edu> wrote: -- Sent from a handheld expect more typos than usual

Recall also that the AT&T code was protected by "trade secret" claims, so that and the Berne convention meant a copyright on each file wasn't needed. As the copyright moved to being used as a license, doing so became more important.

On Wed, Sep 21, 2022 at 5:50 PM Phil Budne <phil(a)ultimate.com> wrote: One of my favorite copyright notices was for /bin/true in System V. An empty file got turned into 7 lines of comments holding copyright boilerplate and an `#ident` line with an SCCS version number: progress! - Dan C.

I’ve thought of this example many times when the 9fans minimalists talk about empty-file `true` impls. Which of course would lead to the 3.8k reported issues of https://github.com/kelseyhightower/nocode -- ~j

Around 1977 I was working/volunteering/studying at the Dynamic Graphics Lab at the University of Toronto, where Unix ran on an 11/45 and we had a bunch of graphics hardware. Doing graphics on a PDP-11 was a challenge, but we managed. (For reference: later Dave Tennenhouse made a 256x256x8bit frame buffer, and that was the size of entire PDP-11 data address space.) Everyone was jealous of the C/A/T phototypesetter that Bell Labs Research used to print their documentation. One Friday evening I had the idea to use our stinky but effective Versatec plotter as an output device for nroff. In just a few hours - our libraries were already pretty good - I had something tolerable running. Tom Duff dropped by and helped make it faster by coding what we would now call the character blitter in assembler. Then Bill Reeves joined in, and Mike Tilson, and by the end of the weekend we had pretty good efficient output. (Still nroff; troff came later, mostly due to Bill I think, who did a lot of work on the character set.) It was grey and blotchy and smelly, but after a Xerox copy it looked pretty good for the time. Ron Baecker, who ran the lab and was the graduate advisor for everyone else - I was just an undergraduate physics student having fun - stopped by on Monday morning and was furious to see us all hammering on the code. Everyone was supposed to be working on their thesis and we had spent the weekend hacking. I was about to be in serious trouble for distracting the graduate students. But then he saw the output and completely changed his tune: "Can I use this to print out my new grant proposal?" For context, consider this: I used the system for my 4th year optics project report. The professor was furious with me for copying someone's work. He did not believe it possible to create output like that (and to be fair, it wasn't possible almost anywhere else). I had to take him to the lab and show him how I did it before he would let me pass the course. Until then, no one had seen a student capable of making text look good. The software went on the Toronto tape, with a top-of-file comment crediting me, Bill, Tom, and Mike. It emerged again from Berkeley with that comment replaced by the Regents' rankling rewrite. When I interviewed at Bell Labs, Dennis Ritchie saw on my resume that I claimed to have worked on the Versatec text output system. He asked why I had bothered, when Berkeley had already done it. "Because we wrote it first, and Berkeley took the credit," I said. Berkeley did tweak it, but honestly it was mostly our work. I didn't care so much about losing credit for the code, but the idea was 100% mine, and for a young punk the loss of credit was upsetting. Later Henry Spencer, another Toronto graduate, explained the story on Usenet. I don't know if he was believed, and through the 1980s it remained the "Berkeley typesetting software." It was all long ago, but seeing that "Regents" comment is, as we say today, triggering. But to be fair to Dennis, he believed me, and maybe that helped me get hired. -rob

On Thu, Sep 22, 2022 at 3:08 AM Andy Kosela <akosela(a)andykosela.com> wrote: Indeed. Ted has made this point frequently; Linux for example basically requires corporate sponsors to get new features into the kernel. Sure, some individual might come up with a great idea and implementation that'll make it in, but that's the exception rather than the norm. The flip side is that there's a lot of load-bearing infrastructure that is barely maintained, if at all. This xkcd seems perennially relevant: https://xkcd.com/2347/ I suppose the situation may be summed up as extremes at both ends. In any event, it's not great. - Dan C.