TUHS

tuhs@tuhs.org

50 participants
6506 discussions

Unix filesystem semantics history - files with holes

by Erik E. Fair

When did the Unix filesystem add the semantics for "files with holes" (large, sparse files)? Just an idle question that was sparked by a conversation I had today. Erik <fair(a)clock.org>

6 years, 1 month

Re: [TUHS] Paper discussing Unix boot process?

by Pat Barron

The more I think about this, the more I'm sure I'm barking up the wrong tree... >From bits and pieces I've been able to recall, the thing I am looking for was not about Unix - it was about TOPS-20. It was a timeline of the system bootstrap activities from power-on to the point where users could log in. I still don't remember where I found it originally, but at least now I'm pretty sure I've been looking in all the wrong places... I believe it originated at CMU, but I don't know for sure that that's where I originally located it. The actual problem I'm trying to solve is, at this point in my professional career, I'm starting to interact with a lot of people (even experienced software developers) who just have no clue of what has to happen to get a computer from the point of "power-on" to the point where they can actually use it to do things. This makes me sad... So, I'm looking for something that I can point these people to that could clue them in... I think the whole bootstrap process is useful to understand for a lot of reasons, partly because it makes you think about all the little fiddly details that have to be attended to to make the computer do what you want - when I was first learning about this, I remember being particularly fascinated by what had to happen to prepare for that moment at which you turn on the MMU, to make sure that the system continues executing in a place you expect it to, in the right processor mode. I know most people that I interact with are using Linux or Windows on Intel-architecture machines, but the boot process for Unix on the PDP-10 or VAX (or even TOPS-20 on the PDP-10) I thought would be a much simpler thing to understand. Though maybe that's the wrong thought process, maybe I should just find something related to Linux that is comparable (even though I think it's more complicated). While searching, I also came across a decent presentation by a friend of mine who teaches at CMU, and discusses hardware that people probably actually work with right now, but I think it would be best consumed along with the actual lecture that it goes with. http://www.cs.cmu.edu/~410-f08/lectures/L20_Bootstrap.pdf Maybe I'll find what I was originally looking for at some point, but after spinning on this for most of the day, I don't think it's related to Unix... --Pat.

6 years, 1 month

Re: [TUHS] TUHS Digest, Vol 41, Issue 2

by Doug McIlroy

here

6 years, 1 month

Re: [TUHS] List ping

by jnc＠mercury.lcs.mit.edu

This thread brings to mind a wonderful saying which I just saw in another forum: "A wise man speaks when he feels he has something to say - a fool speaks when he feels he has to say something." And to reply in advance to the 'but I did't realize so many other people had sent in replies' - try scanning your emailbox before replying to messages to a large list. S/N is to be hallowed. Noel

6 years, 1 month

Re: [TUHS] Does anybody have this book and can look for something for me?

by Pat Barron

This book can apparently be "borrowed" from the Internet Archive. I'm not sure how they do that, haven't tried to, I just see it says you need to log in to borrow the book for 14 days. https://archive.org/details/unixprimerplus00wait --Pat.

6 years, 1 month

Does anybody have this book and can look for something for me? - solved

by Jon Steinhart

Thanks to Pat, I looked at the book at the Internet Archive and it is there.

6 years, 1 month

Does anybody have this book and can look for something for me?

by Jon Steinhart

OK, something that's not a ping :-) I'm trying to track down the author of a cartoon that I'd like to use in my book so that I can try to get permission. Last one that I need! It's a cartoon that I only have on paper and don't know where it came from. It has two frames, then and now, the first with a bunch of cavemen grunting awk, grep, mkdir, yank, the second with a bunch of people sitting at computers uttering the same. I recently stumbled upon something that said that these were in a book called "UNIX Primer PLUS". Anybody have a copy of that? If so, can you please check to see if that's the original or whether they got it from somewhere else? Thanks, Jon

6 years, 1 month

List ping

by Warren Toomey

Just checking you are all still out there :-) Cheers, Warren

6 years, 1 month

Re: [TUHS] List ping

by Nemo Nusquam

On 04/10/19 09:59, arnold(a)skeeve.com wrote: > Nemo <cym224(a)gmail.com> wrote: >> On 10/04/2019, Warren Toomey <wkt(a)tuhs.org> wrote: >>> Just checking you are all still out there :-) Cheers, Warren >> Well, this is not "Forever September"? #6-) I just finished reading a >> fascinating article on Inferno and was most amused by the comment in >> Rob Pike's biblio note at the end. N. > So, please share article link and comment with the list? Thanks, Arnold Apologies -- I found it here: https://ieeexplore.ieee.org/document/6772868/ Bell Labs Tech. J., Vol. 2, Iss. 1, 1997 (or here https://onlinelibrary.wiley.com/doi/pdf/10.1002/bltj.2028 but there must an open version available by now). Pike wrote: "He has never written a program that uses cursor addressing." N.

6 years, 1 month

Retaining file-modified timestamps

by jnc＠mercury.lcs.mit.edu

So, a while back I mentioned that I'd done tweaked versions of 'cp', 'mv', 'chmod' etc for V6 which retained the original modified date of a file (when the actual contents were not changed). I had some requests for those versions, which I have finally got around to checking and uploading (along with 'mvall', which for some reason V6 didn't have). I've added them to a couple of my V6 pages: http://www.chiappa.net/~jnc/tech/V6Unix.html#mvall http://www.chiappa.net/~jnc/tech/ImprovingV6.html#FileWrite Note (per the page) that the latter group all require the smdate() system call, which was commented out in 'vanilla' V6 (because using it confused the backup system); the page gives instructions on how to turn it back on. Noel

6 years, 1 month

Re: [TUHS] a possible source for 4.1BSD tapes

by Richard Tobin

> "taperead" in http://github.com/brouhaha/tapeutils can extract files > from a tape image. The format is very simple: a 32-bit little-endian record length, followed by that many bytes, followed by the length again for integrity checking. A record length of zero is a file mark. -- Richard -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.

6 years, 2 months

a possible source for 4.1BSD tapes

by Mike Haertel

I noticed that the TUHS archive does not include a 4.1BSD distribution. Also, while poking around the net, I've found a number of purported tape images of 4.1BSD dated 7/10/1981 that look to me to a little sketchy, since most contain files dated well into 1982. So it appears to me that 4.1BSD is semi-lost. While googling all this, I discovered that the School of Computer Science and Statistics at Trinity College Dublin has an online archive catalog which lists a couple of 4.1BSD distribution tapes in the "John Gabriel Byrne Computer Science Collection". https://scss.tcd.ie/SCSSTreasuresCatalog/ Perhaps someone from TUHS who lives near Dublin could investigate and see if images can be made of these tapes?

6 years, 2 months

Another old DMR and Brian Redman story

by ron＠ronnatalie.com

Years ago just before one of the USENIX meetings in Atlanta Dennis made some joke comment that nobody had ever asked for a plaster cast of his genitals. A bunch of us thought it would be fun at the conference to hand out genital casting kits to Dennis and certain others of note. We ran down to a local art supply store and bought some plaster and portioned out into zone ziplock bags We added some paper cups to use for molds and wooded sticks to mix with. We needed a release agent. Vaseline would work, but I couldn't figure out how we'd get small portions (I couldn't use the ziplock bag idea practically). Fortunately, there was a little gift shop in the hotel lobby and they had these travel size jars. Perfect. Now the interesting thing was that concurrent with USENIX was the Southern Baptist Conference meetings (this led to some odd events at local restaurants). Anyhow, I walk up to the cashier and plop down ten jars of Vaseline. She looks at me and says, "I guess y'all aren't with the Baptists." Oddly, most recipients took the gift in good spirit but Redman had a fit for some reason. Babette suggested perhaps we made the kit too large for him.

6 years, 2 months

Happy birthday, NetBSD!

by Dave Horsfall

According to my (possibly inaccurate) notes: NetBSD checked in 1993 Message-ID: <alpine.NEB.2.20.1803211456560.25928(a)t1.m.reedmedia.net> Revision 1.1, Sun Mar 21 09:45:37 1993 UTC (25 years ago) by cgd http://cvsweb.netbsd.org/bsdweb.cgi/src/sbin/init/init.c?rev=1.1&content-ty… "Today is commonly considered the birthday of NetBSD. As far as I know, it is the oldest continuously-maintained complete open source operating system. (It predates Slackware Linux, FreeBSD, and Debian Linux by some months.)" -- Dave

6 years, 2 months

Re: [TUHS] Bell Labs data center in 1969/70

by Doug McIlroy

Strangely I have no recollection of a Bell Labs branch further west than Denver. What did they do in Oakland? Doug

6 years, 2 months

Early DNS software

by Witold Krecicki

Hello, I'm looking for old DNS software - servers, clients, libraries. The oldest one I've found is BIND 4.3 from 4.3BSD (and it works as a resolver on 2019 Internet), but there were earlier ones - http://www.donelan.com/dnstimeline.html says that UCB released first BIND in 1985, something was running on earlier servers. Any help would be appreciated. Witold Krecicki

6 years, 2 months

In memoriam: John Backus

by Dave Horsfall

We lost computer pioneer John Backus on this day in 2007; amongst other things he gave us FORTRAN (yuck!) and BNF, which is ironic, really, because FORTRAN has no syntax to speak of. -- Dave

6 years, 2 months

Early DNS software

by Witold Krecicki

6 years, 2 months

Re: [TUHS] [patch] do not strip mdoc macros

by Doug McIlroy

>> But sed, awk, perl, python, ... lex and parse once into an AST or >> bytecode, removing the recurring cost of comments, etc. that impact >> groff. So I don't think it's an even comparison. > > Of course it's a valid comparison. Which sed or awk or shell script is > distributed in a stripped/compressed form? Do they store their AST > somewhere, so as to avoid recompilation? They do not. Just as > with groff, every parse starts anew. Comments inside of a macro definition get scanned each time it's called. This justifies the first paragraph above. In the wild, almost all comments occur outside macro definitions. This justifies the second. Thus comments are harmless in practice. Doug

6 years, 2 months

Bell Labs data center in 1969/70.

by Dan Cross

TUHS related due to the BTL connection. This came across a mailing list at work today (as a, "hey, this is cool!" thing; nothing work related) and I thought some on this list might be interested. http://www.larryluckham.com/1969%20&%2070%20-%20Bell%20Labs/album/index.html - Dan C.

6 years, 2 months

P J Plauger in popular culture

by Tony Finch

An amusing Unix-related snippet from a science fiction site: https://www.tor.com/2019/03/12/more-please-authors-we-wish-would-publish-mo… > Back when the world was young and a ten megabyte hard drive required a team of six sturdy workers to move, P. J. Plauger quite reliably delivered to the world a story or so per year—memorable tales like “Wet Blanket” and “Child of All Ages,” stories that won him a Campbell for Best New Writer and a Hugo nomination for Best Short Story. Tragedy struck when he was enticed away from science fiction by the seedy world of Unix, which offered its arcane practitioners unnecessary luxuries like indoor living, food, and even health care. (The rest of the page is not relevant to this list) Tony. -- f.anthony.n.finch <dot(a)dotat.at> http://dotat.at

6 years, 2 months

Failing Memory of an Algol Based System from years ago

by Rudi Blom

>A bit off topic (sorry) but wondering about that PDF conversion. This >may be a dumb question but did you ever try the PDF conversion in >calibre ( https://calibre-ebook.com )? To answer my own question. Yes, that was a dumb question. I completely over looked the 'image' in Nelson's post >The result is a >searchable document, with a single page image per PDF page, rather >than the mixed bitmap scan of 1-up and 2-up pages in the original PDF >file. Calibre's PDF to whatever conversion doesn't do anything worthwhile with images.

6 years, 2 months

Re: [TUHS] Failing Memory of an Algol Based System from years ago

by Nelson H. F. Beebe

The posting of the link http://bitsavers.org/pdf/regnecentralen/RC_4000_Reference_Manual_Jun69.pdf brought back old memories. When I moved to Aarhus University in Denmark in December 1973, I found that the Department of Chemistry where I worked had a Regnecentralen RC-4000 minicomputer with paper tape input that was used to manage the department's accounting. It had a dedicated operator who kept in running nicely until at least after I left in 1977. It had too little physical memory to be considered practical for the quantum chemistry work that I was then engaged in, and may not even have had a Fortran compiler; instead, we used the campus CDC 6400 for our research. In memory of the RC-4000, and Per Brinch Hansen's (13 Nov 1938--31-Jul-2007) many contributions to the literature of computer science, programming language design, and parallel computing, I prepared an enhancement of its manual, available here: http://www.math.utah.edu/~beebe/RC-4000/ The Web page that comes up gives a directory of the available files, and documents the steps needed to produce them. The result is a searchable document, with a single page image per PDF page, rather than the mixed bitmap scan of 1-up and 2-up pages in the original PDF file. Despite my 40+ year engagement in the TeX community, I only recently learned via the texhax mailing of some PDFTeX internals that allow one to construct a file that can retypeset n-up files into 1-up format. I therefore make this posting in the hope that someone else might be encouraged to tackle similar document improvements in the bitsavers archives. Although it takes a bit of experimentation, with the exception of the OCR conversion, the entire operation can be done with free software on pretty much any modern computing platform, thanks to the portability of the needed software. ------------------------------------------------------------------------------- - Nelson H. F. Beebe Tel: +1 801 581 5254 - - University of Utah FAX: +1 801 581 4148 - - Department of Mathematics, 110 LCB Internet e-mail: beebe(a)math.utah.edu - - 155 S 1400 E RM 233 beebe(a)acm.org beebe(a)computer.org - - Salt Lake City, UT 84112-0090, USA URL: http://www.math.utah.edu/~beebe/ - -------------------------------------------------------------------------------

6 years, 2 months

Failing Memory of an Algol Based System from years ago

by Rudi Blom

>Date: Sat, 9 Mar 2019 15:29:11 -0700 >From: "Nelson H. F. Beebe" <beebe(a)math.utah.edu> >To: The Eunuchs Hysterical Society <tuhs(a)tuhs.org> >Subject: Re: [TUHS] Failing Memory of an Algol Based System from years ago >Message-ID: <CMM.0.95.0.1552170551.beebe(a)gamma.math.utah.edu> ... snip ... > http://www.math.utah.edu/~beebe/RC-4000/ > >The Web page that comes up gives a directory of the available files, >and documents the steps needed to produce them. The result is a >searchable document, with a single page image per PDF page, rather >than the mixed bitmap scan of 1-up and 2-up pages in the original PDF >file. >Despite my 40+ year engagement in the TeX community, I only recently >learned via the texhax mailing of some PDFTeX internals that allow one >to construct a file that can retypeset n-up files into 1-up format. > >I therefore make this posting in the hope that someone else might be >encouraged to tackle similar document improvements in the bitsavers. >archives. Although it takes a bit of experimentation, with the >exception of the OCR conversion, the entire operation can be done with >free software on pretty much any modern computing platform, thanks to >the portability of the needed software. A bit off topic (sorry) but wondering about that PDF conversion. This may be a dumb question but did you ever try the PDF conversion in calibre ( https://calibre-ebook.com )? I like the PDF to htmlz conversion even if mostly the result still needs (a lot of) extra work. Cheers, uncle rubl

6 years, 2 months

Re: [TUHS] Failing Memory of an Algol Based System from years ago

by Doug McIlroy

Clem, I think the "Algol machine" you have in mind is the RC-2000 (not quite sure of the designation--could look it up in the attic if it matters) designed by Per Brinch Hansen for Regencentralen (again, the name may not be quite right). The manual used Algol as a hardware description language. The instruction set was not unusual. It has come up before in TUHS. I have the manual if you need more info. Doug

6 years, 2 months

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

TUHS