When Unix Landed - Computerphile
Vložit
- čas přidán 6. 06. 2024
- Professor Brailsford recalls the advent of Unix v7 and AT&T's licensing procedure.
/ computerphile
/ computer_phile
This video was filmed and edited by Sean Riley.
Computer Science at the University of Nottingham: bit.ly/nottscomputer
Computerphile is a sister project to Brady Haran's Numberphile. More at www.bradyharan.com
That's such a cliffhanger. Looking forward to the next one.
Google "reflections on trusting trust" if you want to spoil the surprise, I think that's what he's referring to.
Spoiler alert (do not unwarap if you don't want to be spoiled):
he will probably talk about the login backdoor that was not in the login program but hidden in the C compiler, which would add the special user/passwd check to the login binary when it was compiled from the source.
Intellectual Property is an oxymoron
I know, right?!
@@ChrisJones-rd4wb The term Industrial Property is more common amongst people who work in IP.
Really enjoyed this one!
Prof Brailsford... The David Attenborough of Computer Science history.
My thoughts exactly, the man is a national treasure
So much so.
Imagining Professor Brailsford in California in the the 1970s makes me smile.
i love how often professor brailsford sounds like a golden age jazz singer, recounting all the places he performed and the incidental legendary names that he encountered!
I absolutely love Professor Brailsford. He's like this perfect confluence of talent and humanity. I'm grateful for every one of these videos we get with him. Keep up the great work!
I think I could listen to Professor Brailsford for hours on end and it wouldn't get boring.
Always excited for a Professor Brailsford video!
I've been using UNIX for nearly 35 years, and even today I still think it rocks!
What unix system?
Can listen to Brailsford all day, so knowledgeable and entertaining at the same time. A true treasure!
I love to listening professor brailsford explanations and story tellings..
Thompson & Ritchie were NOT a beach band duo or crazy music producers, despite that luscious photograph.
No but I bet they were chill to hang out with if you were interested in computers. Ken probably still is, sadly Dennis is no longer among us
@@nicolaiveliki1409 Well Ken does R&D at Google now and is likely experimenting with a functional mock Skynet in some sort of lab.
@@nicolaiveliki1409 Yeah, ken's still pretty chill to hang with.
@@eideticex well, he was one of the people who worked on Go
They unleashed upon the world and unparalleled evil: K&R brace style.
11:45 "Actually, FORTRAN was the language of choice for the same reason that three-legged races are popular."
Now there's a quote. It's always nice to find academic writing with a sense of humor.
This made me pull down my old Bell System Technical Journal, 1978 Vol 57, #6, UNIX TIME-SHARING SYSTEM. The motto says it all, “Devoted to the Scientific and Engineering aspects of electrical communication”. I can only smile at the description of AT&T as a Phone Company.
Probably got the shirt when he was teaching in California
They give you a box of Hawaiian shirts with your computer science PhD. Pony tails are optional
@@matthewchunk3689 is that a stereotype? Because i fit it perfectly
@@minihjalte nah they actually give you the box, I got 2 😎
Always love to here from him. I was around for non of these events, but I like the feeling of being told a war story by your granpa. Computerphile always relights my passion for the field of computer science, thanks for that.
Professor Brailsford winging an American accent was golden!!!
I could listen to Professor Brailsford stories all day long...
this is probably the greatest cliffhanger of any computerphile episode :D
An upcoming video about "Reflections on Trusting Trust"?. I am seriously looking forward to it.
My father was an Architecture student at Oklahoma State 1980-84 and had to learn both Cobol and Fortran. That's an interesting possible connection! 🤔
Always love hearing more about the computer history. Especially when it is told by Prof. Brailsford.
There was no Internet, but you could get microfiches to put in your microfiche reader. At DEC we got a new set regularly. If you dropped the package, you were screwed.
When Litton was bought by NG in 2001, we sold off several of our buildings and in one of the trashcans was a complete copy of VMS 5.4 on microfiche. I couldn't let it go, it's still in my garage :)
@@VAXHeadroom Have you thought of letting Bitsavers make a scan of it?
@@lawrencedoliveiro9104 Noooo....
*sound of keys typing in a search engine"
When you dropped the package, you learned the hard way that fiche should have sequence numbers, just as punched cards should!
The Holy Trinity: Thompson, Ritchie and Kernighan.
There are also other unsung heroes from the UNIX room, such as Doug McIlroy, Rob Pike, Stephen Johnson, and many others. Everyone contributed in a way or the other, but without the PDP-7 and Ken's ideas, we wouldn't have got this wonderful piece of software
If anything, it should be Thompson, Ritchie, and McIlroy. Kernighan hardly did anything on Unix.
@@wiilillad To be precise, Thompson, Ritchie, Canaday and McIlroy worked on the operating system core (kernel, drivers, essential tools, compilers) and its design.
Kernighan wrote/co-wrote mainly languages (AWK, AMPL, ratfor) and tools (eqn, ditroff), other than taking part to the Linotron 202 hack later.
We remember him for his influential books (the K&R, _The Unix programming Evironment_, _The practice of programming_, ecc) and because he came with the name UNIX (probably as a pun on Multics).
@Mariano Bustos I can live with that.
Cambridge were allowed an IBM 370 "because they needed to work with US universities" - Algol68C was developed on it and used as one of the teaching languages, along with BCPL. The development team included Stephen Bourne who would go on to develop the Bourne shell on Unix.
This man's memory is absolutely incredible.
Ken and Dennis are on an absolutely different level.
Prof. Brailsford is a treasure.
It's immense historic talking about computer science by prof. Brailsford who have experienced the technological booming in the fist place.He is a dear person.
Another great video. I started work in the Data Processing Department of a UK bank in 1987. They had librarians whose sole job was to receive paper updates to manuals and file them in the correct manual. For example we might receive 5 pages for a PL/1, COBOL, CICS, IMS or DB/2 manual with instructions on what pages they new ones replaced. In those days we really did have to RTFM.
I recall a colleague, a former IBMer, mentioning how some of those pages went beyond the usual “THIS PAGE INTENTIONALLY LEFT BLANK” business. Instead, they said “DESTROY THIS PAGE”.
@@lawrencedoliveiro9104 Yes I remember those too. Of those 5 pages, page D-15 for example, might no longer be needed. So "DESTROY THIS PAGE" meant to take that page number out of your own binder and throw it away. We had a huge room with one whole wall made up of manual cupboards.
HBO called, they would like to buy your cliffhanger
To me it is cool that he lectured at CSUN where I graduated from even if it was 30 years before I went there.
Oh, he's a great storyteller 🙂 can't wait for future episodes
I'm so glad that we can get Professor Brailsford's amazing experiences recorded forever here on CZcams
Five users bogged down an 11/35? In the mid eighties, I was using mostly PDP 11/17s running RSTS/E. We often had 32 users logged in at once around the due date of our programming assignments, and it still felt like we had the whole machine to ourselves.
The VAX 11/780 running VMS was much more powerful, but always felt more sluggish.
Professor Brailsford needs a podcast
Lores and storytelling. We love you, Prof. B!
It's always nice to see and hear you, sir.
I've been hoping for a new Prof. Brailsford video AND tinkering around with old Unix lately, so this is perfect.
You guys should make a full two hour documentary with this quality instead of 'just' those small snippets here and there every couple months.
I love listening to this man tell stories from the past, even though I can't relate to most of the experience.
Absolutely riveting stuff, could listen to Brailsford all day, fascinating guy.
Oh man, is this the compiler thing?
Super cannot wait for next episode if so.
Or, if it's not, that doesn't really diminish my enthusiasm. Can't wait!
He was teaching at Berkeley when Cliff Stoll was there
When I arrived at UCL in 1984 I don't recall any ICL kit for undergraduate use, we were all on PDP-11s, and I think there were some VAXes, but definitely all running UNIX. At some point we got a Pyramid which ran System V and BSD simultaneously, that was very exciting.
Thank you Professor Brailsford! that was an amazing story and what a cliffhanger!
That's history of science class live. I love these videos and I'm not even into computer science. Thanks!
I am a simple man, I see professor Brailsford and I click (like).
Some day I hope you guys can edit up some of these into longer form documentary content, Professor Brailsford and the others are simply too amazing with their stories and histories
We had Unix V6 on a PDP 11/34 at Warwick in 1979 (and probably earlier). It was serving several of the professors and half a dozen students at the same time, all squished into 256k of RAM, with a 64k address space theoretically available to each and every program.
Oh the joy we had trying to add features to the kernel (for which we had the source, of course) and still keep it within that limit!
Finally moved up to V7 when we got a VAX 11/750 in 1981, and could breathe!
Love these right from the house's mouth kinda videos by the prof! His stories and story telling style is so so so good!
This bloke used to teach me about formal grammars. Always been useful to me.
I am looking forward to the next video. History of Computer Science is a very interesting topic!
RMS and Linus freed the world from the licensing circus.
Well, Prof. Brailsford just explained to you the first movement out of the "licensing circus". That very strict "free for Universities ONLY" agreement was the seed of BSD.
@@framegrace1 its really a shame BSD doesnt get more attention from non-corporate endeavors, i love the idea of BSD but for my purposes its practically unusable.
Only copyleft licences such as the GPL guarantee self-preserving software freedom
@@animeboitiddies6146 its license is its own enemy. corporations (ahem Apple, ahem) just take without giving back much value.
@@rationalityfirst i get the desire for the sort of license it is, but i think weve seen in the difference between the development of the two systems the pros and cons of each. both are vulnerable to different kinds of parasitism.
my college (in london) had a pdp-11/43 in the 70's & 80's. it ran rsts/e, though, not unix. they upgraded to a newer model and the students got 4,3 on it.
Unix changed my life, thank you Ken.
Thank you so much Computerphile.
Quickly! Someone hire this man to be life’s great narrator!
I am already so hype for the next video.
A week when @computerphile publishes a video with Prof. Brailsford is a very good week indeed.
Experience talking at it's best. Brailsford is an absolute genius.
Absolutely loved this history lesson.
I took classes under a couple professors at Oklahoma State who may have been there; one who worked on the Algol 68 compiler, and one who still swore by Fortran in the 1990s.
Sussex had PDP-11s and some Vaxen by the late 70s (11-780s, I think). I remember playing about with ADVENT* in 1979 on a VT100 terminal hooked up to their timeshare system. I had a lineprinter dump of the entire text segment knocking around somewhere until it got binned during an overzealous spring clean.
*Also Terry Winograd's SHRDLU, and Weizenbaum's ELIZA. Forty years later, the bug that bit me back then still has its jaws in me
These videos are very important, really. While much of this history is already documented, I think it’s so important to hear these stories from the people who were there. Future historians will thank you.
I cannot imagine what the process of physically bringing Unix over to the UK from the US was like then. I’m old enough to remember floppies, but it’s still hard to think about what type of media would have held an OS in that era…
You see now the impetus for RMS to start the GNU project, given the increasingly draconian restrictions on AT&T Unix licensing as time went on. They were bad enough to begin with, but after the breakup of the telephone monopoly, with AT&T now free to fully participate in the computer market, suddenly Unix became a whole lot more commercially valuable to them, and so they became less willing to let Universities and others have it, with full source code, at anything resembling a modest price.
But, BSD.
@@davidwuhrer6704 BSD postdates the start of AT&T license restrictions, and BSD being opening free of AT&T restriction postdates the completion of Linux with the GNU system running on it.
Hey, I was at OSU in 1978, and I used Algol-68 there!
*NIX is the only system that can do real work. Its amazingly easy to use and build a workflow.
I can swear, and I want the rest of the story! Such a beeswax these operating systems, a blessing and a curse.
Fantastic teaser at the end.
I'm so hype about this!!! IMO the true story of linux origins is often mis-told, and this is the perfect setup for what i believe may go on to tell the story correctly. really hope that this is part of the series.
What a marvellous storyteller he is
we love you, Professor Internet Grandpa
Twenty years later we were pushing our enterprise overlords to let Linux into the datacenter. They said no, so we snuck in through the backdoor. :)
@Gommash who are you replying to?
Documenting how and when the prosaic C first got its tendrils into UK universities and hence common computer parlance. Meanwhile zen-like Forth was eschewing the need for any kind of OS at all. Two philosophical paths: burgeoning complexity Vs irreducible simplicity...
Forth is beautiful. Unfortunately, it's not a language that's easy to use. This isn't a matter of 'getting used to RPN' either, it's far more fundamental. See, Forth demands a bottom-up approach to development that is extremely difficult for even very skilled programmers to get right. When it works, it's absolutely amazing. It's just that it works well so infrequently...
If that next video is what I think it is we are in for an absolute treat.......
Great story again, told by a great story teller! The shirt topped it of. I paused to get some 😎.
UNIX might have been flawed back then, but the system requirements sort of retro-actively applied to its development in the early days. The Kernel itself (and UNIX as a OS Family) were heavily restricted by clock cycle speed and memory access speed (read and write) that it was necessary at the time to "cut down" the amount of security related features that originally were planned, and further, error debugging code in the UNIX Kernel. In the original system, for any major crash (which pretty much was any crash), all it did was run the command "crash" and essentially "BLUE SCREENED". Just thought it would be interesting to note. This is why OS' built on top of the UNIX system (and the like) - in the modern day - add bits to the kernel or its own structure to "fill in" these gaps. Even then, some still exist and can be exploited. Despite this, in some regards, its light weight nature actually makes it faster in some regards to other systems like Windows.
"It's déjà vu all over again" Yogi Berra
"...then the status quo will return to what it was before" Murry Walker
Hadn't heard that Murray Walker one. Wonderful!
It' s new to me, too, but shouldn' t it be " The status quo ante will return to what it was before" ?
Well now I _really_ want to see the next one!
I was an undergrad in 1979, and teaching was done on an ICl with algol68. Punched cards and line printers. But if you asked nicely you could get to use the Harris Unix system with interactive terminals.
Ahhhh... the IBM 360/67 at Newcastle. I learned Cobol on it! (I'm 69, it was a long time ago).
I first hit Unix late 80's and it was total game changer. If it wasn't for X I reckon Unix would have taken over the desktop world as well. What a horror show that was/is
What do you mean by X?
@@hayden.A0Probably the X Window System, which unified and network enabled graphical user interfaces, maintaining the ability to run lower cost terminals with a central computer. As for the "horror", probably more FUD than experience.
Love this man
Hi Computerphile, I really enjoy your content, especially those of Pr. Brailsford. I couldn't find any episodes on shells. Would you be interested to produce an episode on the shell topic? I thought I had a clear idea on what that was until I learnt about the windows' shell that underlie the explorer.exe. It seems I can't get a clear abstraction about it (I know bash, zsh, dash, pwsh... but can't abstract a definition from). Kind regards.
Shell was a term from Multics, which was multi-layered - from the kernel, out via layers, to the software that interacted with the user - and by analogy with a nut, that outer layer was the 'shell'. In early days of Unix there was only a teletype as the user interface, no windows or mice. Users typed instructions into the command interpreter (shell) and saw results printed back. The idea that the command interpreter was just a regular program and not an integral part of the operating system was just one of the ideas so smart that at the time it seemed revolutionary. Initially, v6 Unix had just 'the shell' but the others with different command syntax started getting written and by the mid 80s there were several competing shells to choose from - csh with a c-style syntax for loops, Steve Bourne's bourne shell based loosely on Algol-68 and others. Hence linux now has 'bash' which is a development of Bourne's shell, the punning 'bourne again shell (bash)'.
Oxford got an ICL1900 installation with George and MOP
So they did! Many thanks for reminding me ....
the 2 vids shown at the end don't have clickable links over them...
Back when C was a high level language
lol
hahahah
At least the distinction was relatively obvious then. Is Rust a high level language? You can write everything from a BIOS to a highly threaded web server to a build script in (mostly) the same language safely. Go, Swift, Dart and others all support low-runtime high performance, but yet try to hide all the technical details and stay very flexible. JavaScript literally has script in the name, but is looking like it will soon be mostly used as a compiler target. Languages are getting pretty interesting!
I'm not sure if/how the bar has really moved for "high level" vs "low level" anymore, but even when I first learned the terms thirty years ago it struck me as a broken paradigm. Low level as the baseline of "it's essentially just human readable op codes" kinda makes sense, but only having one tier above that didn't really jive for me as a young teen hobbyist in the early '90s and has progressively made less sense as time has passed and complexity of relationships between languages has grown.
Low level means it is interpreted, or can be interpreted, step by step. Like assembler language or scripts.
High level means it can't. That there are concepts that require at least one full pass before it can run.
UNIX wasn't invented by AT&T or even Bell Labs; it was invented by Ken Thompson, who wrote it in three weeks while his wife was away, thinking he was going to get fired (they weren't allowed/given the budget to work on OS's). The guy's an f-ing legend and is as humble as pie - has so many great stories. He kept a pet alligator at his desk which once escaped which is how he began getting to know the staff.
When I first heard of UNIX I thought: Why do I care about the history of some boring company? It's a shame people focus so much on AT&T and its Bell Labs and forget to talk about the _people_ who were involved.
How did you know his name was, Frank J. Riffle Junior?
Well, he kept referring to himself in third person.
Great storytelling!!
Oh dear god they're gonna do Trusting Trust.
well, it's abstraction layers & trust all the way down....
Unlike the original audiences for that talk/paper, at least we can see it coming!!! What a mind-bender that must have been back in the day.
Never forget. Frank J. Riffle is a man of his word.
What is the following video where he talks about what security issues unix opened up? It's not in the description or other related videos or the title is not obvious continuation.
OK that's a cliff hanger! Hey Brailsford - you and Cant from NTU forged me into the monster I am. (Richard not can't)
More Professor B please! He is the f'ing bomb!
Has anyone watched a video of Professor Brailsford and thought "Well, that's 15 minutes of my life I'm not going to get back". No.... me neither.
I love this video
Has part 2 released?
I started watching this, went to do the dishes, came back and wondered if this man ever stops waffling incesantly… And then I released it was on autoplay and I'd gone through about 5 vids. Hehe.