Jump to content

Talk:Research Unix

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

PDP-9?

[edit]

I deleted the PDP-9 from the list of systems. It wasn't one of the main development platforms. According to this note from Dennis Ritchie, it was only run on a PDP-9 for a few hours to play around. Since it says the PDP-9 belonged to a "different group", I assume this means it was in another part of Bell Labs other Dept 1127, and thus not strictly part of the Research UNIX project. Agarvin 21:44, 3 November 2006 (UTC)[reply]

Distribution terms

[edit]

I assume the code was copyrighted by AT&T, but on what terms was it published to other users? In particular, how did the code legally become part of BSD, which has a free/open source license? -- Beland (talk) 03:12, 4 February 2013 (UTC)[reply]

The article Berkeley Software Distribution covers that. All full BSD distributions required an AT&T source license. Only 4.4BSD-lite and lite2 were open sourced (although that term didn't yet exist), but in those versions all AT&T code was replaced. Except for a few bits, that is, which lead to the USL v. BSDi lawsuit. QVVERTYVS (hm?) 12:17, 6 October 2013 (UTC)[reply]

Proposed merge with Version 5 Unix

[edit]

IMHO, V5 Unix does not have sufficient notability to warrant its own article. The material in Version 5 Unix is already covered by Research Unix. (For this very reason, I merged my own former articles about V9 and V10 into Research Unix.) QVVERTYVS (hm?) 12:18, 10 February 2014 (UTC)[reply]

  • Support. I'd probably also throw in Version 8 Unix as well. Only Version 7, as the basis of the first commercial implementations, is really independently notable. 17:28, 9 April 2014 (UTC)
Done. I agree with V8, but before we go into big merge mode, note that actually V6 was the basis for the first commercial versions (and of the Lions book). QVVERTYVS (hm?) 18:46, 9 April 2014 (UTC)[reply]

Relationship between Research Unix, BSD, System III & V, and proprietary UNIX's like HP-UX, IBM AIX, and SunOS/Solaris.

[edit]

Given the complicated nature of how these different Unix versions/forks developed from the original AT&T Unix versions, maybe we should include a section here explaining how Research Unix fits in with and compares to the other version of Unix like BSD,System III and V, and proprietary UNIX's like HP-UX, IBM AIX, and SunOS/Solaris. If I understand the timeline correctly, BSD was based on early version of Research Unix while later version of Research Unix were influenced by BSD. System III & V were commercial releases of Unix AT&T sold once they where allowed to sell software commercially after settling the anti-trust case against them in the early 80's (which had begun in the 1970's). What could use some explaining in this article is how BSD version in the early 80's (prior the BSD Lite versions) were similar and different from Research Unix versions released in the 1980's (Other then the only partially released last two versions of Research Unix). Also, how did System III & V differ from Research Unix and did proprietary Unix's like HP-UX, IBM AIX, and SunOS/Solaris incorporate any Research Unix features in the 8th Edition or later? --2600:1700:56A0:4680:0:0:0:44 (talk) 22:20, 1 September 2021 (UTC)[reply]

v8 is based on 4.1c BSD?

[edit]

It appears there was no single official 4.1BSD release tape image. At least some V8 kits are based on 4.1_BSD_19810710-modified which may not have been released on 1981-09-01 as September 1, 1981 is the date actually recorded in /etc/motd as shipped in all subsequent 4.1BSD releases.... . The the article states that v8 is based on 4.1cBSD but is there any real evidence that it is?


Jamplevia (talk) 16:23, 21 December 2022 (UTC)[reply]

Inside 4.1_BSD_19810710-modified.tap is a stamp file that contains a timestamp for February 10, 1982 but there are files in the archive with modification times that are more recent than that. The most recent files have June 12, 1982 as the date of modification. At the least it shows how opaque the release process was. One of the issues is that just because some kits are using 4.1_BSD_19810710-modified as the base for v8 doesn't mean that it was used in the 1980s.
$ head stamp ; cat etc/motd
==> stamp <==
Wed Feb 10 17:10:24 PST 1982
Welcome to Berkeley Vax/UNIX (4.1bsd revised 1 Sept. 1981)
Yes, this comment contains original research, but I find it helps to know what you are looking for when you are searching for an RS.
---- Jamplevia (talk) 00:31, 22 December 2022 (UTC)[reply]
The CSRG CD-ROM, disk 1, has two versions of 4.1c BSD. Both of them have a ufs_nami.c with symbolic links. The 4.1_BSD_19810710-modified.tap has the old file name nami.c without symbolic links. The v8 tape (which overwrites files in 4.1_BSD_19810710-modified.tap) has nami.c with symbolic links. Therefore v8 is either based on an "old 4.1c BSD before nami.c was renamed ufs_nami.c but after symbolic links were added" or v8 is not based on 4.1c BSD, but an older version of BSD.
  • 4.1 BSD has the old file names (nami.c) and no symbolic link and no sockets.
  • v8 has the old file names and it does have symbolic links but no sockets.
  • 4.1c BSD has the new file names (ufs_nami.c) and symbolic links and sockets.
---- Jamplevia (talk) 01:34, 22 December 2022 (UTC)[reply]
The article states sockets replaced by Streams but, based on the kit, v8 was never built from a BSD with sockets and sockets were never added by Bell Labs, so they were not replaced.
---- Jamplevia (talk) 02:02, 22 December 2022 (UTC)[reply]