back in 1969 a couple of computer scientists here at Bell Labs started to develop some programs they needed for their own use what Ken Thompson and Dennis Ritchie started developing then has evolved into the UNIX operating system which by now is widely used not only in the Bell System but in other places around the world in the next few minutes we're going to tell you about a few of the problems that UNIX is useful for and we're going to talk some about both operating systems in general and the UNIX operating system in particular to show
you how UNIX helped people solve problems if you look particularly at Bell Laboratories but at high technology industry in general there is a tremendous need for better techniques and tools to make our work more productive we just simply cannot like all the software that needs to be written in fact a good analogy is that a number of years ago when telephone calls were switched not by computers but by human beings operators people predicted that sooner or later the entire population of the United States would have to be telephone operators to switch all of the calls
that needed to be switched we have the same problem in programming productivity today now keeping large amounts of software working and keeping it working in the face of change simply takes a lot of people the bell system we have many millions of lines of software we have thousands of many computers we have hundreds of very large computers and there is a lot of software that needs to be kept working there's all different sorts of software there are is the switching software that switches your telephone calls there is administrative software that administers the entire process and
there is for instance the billing software how many systems do you know give you a bill for something that costs as little as 10 or 20 cents there's a lot of software so we need lots of techniques and tools to make this more productive the programs which tell a computer how to switch a telephone call or computer payroll or play an electronic game are called application software there's another type of software the programs which control the resources of the computer itself the memory of the computer the processing elements and the terminals connected to the computer
this latter type of software is called an operating system another way of saying what an operating system is is that it's a collection of programs which make the intricate hardware of the computer see more simple and comprehensible from the point of view of an applications programmer so that the applications programmers can create software more easy we are trying to make computing as simple as possible in the late 1960s Dennis Ritchie and I realized that the then-current operating systems were much too complex we attempted to reverse this trend by building a small simple operating system on
a mini computer what we wanted to preserve was not just a good programming environment in which to do programming but a system around which a community could form fellowship we knew from experience that the essence of communal computing as supplied by remote access time sharing systems is not just a type programs into a terminal instead of a keypunch but to encourage close communication UNIX started out as a two-man effort and by now it's used all over Bell Labs we have close to 20,000 computer terminals in this company roughly one per employee and most of them
are used for communicating with UNIX systems one of the main reasons that UNIX is popular around here is because it provides graceful facilities for decomposing complex computing tasks into simple subtasks the UNIX operating system is basically made up of three parts the kernel or operating system proper is the part that manages the control of the Machine and supervises scheduling of the various user programs the shell or which is the name we give to the command interpreter looks after the communication between the user and the system itself the third part which is actually the largest is
the various utility programs which perform specific tasks like editing a file or sorting a bunch of numbers or making a plot in other words all the other programs that are not provided directly as part of the operating system kernel one of the things about UNIX is the ability that we have to create complicated programs by building them out of simpler programs rather than writing programs from scratch we can often construct them just by gluing together existing programs almost like building blocks let me show you an example of how that would work by writing a program
without actually writing a program to find spelling mistakes one of the things that UNIX has been used for it for a long time is document preparation helping create letters and papers and books and so on so let me write this spelling mistake program I'm not going to do it for a whole document because we don't have one but let me print a single sentence the sentence comes from a paper that John Massey and I wrote and if you look at it carefully you can see that they're actually a couple of spelling mistakes in it so
suppose that we wanted to use a computer to find those spelling mistakes instead of doing it by hand basically what we want to do is to take each of the words in that sentence and compare them against dictionary to compare it against the dictionary which is sort of one word for line it's certainly going to be easiest to convert this into one word per line so let me run a program called make words program that already exists we'll take its input from sentence and we'll dump the output into some thing that I will simply call
words words is a file just like sentence let's look at that and there it is all of the words one word per line now if I want to compare those against a dictionary I've made some progress but it's clearly not done yet because for example my dictionary doesn't have the word Bell or the word Laboratories with a capital letter at the beginning it only has them in lowercase and so if I do a blind search I won't get the right answer so the next thing I want to do is to convert words and the lowercase
and so let me just run a program called lowercase on words and dump the output in this case into something that'll call LC words inventing file names as I go here and let's look at that one and see how we did and sure enough bail and laboratories are in lower case so we're getting warmer we're almost at the point where we could compare these words one at a time against the dictionary the dictionary is in alphabetical order and it will be a lot more efficient if we only go one pass through my list of words
and one pass through the dictionary rather than sort of poking randomly at the dictionary so let me sort my list of words so I'll say sort kalsi words and I'll put that one into sordid words and let's see how that worked and sure enough they're sorted into alphabetical order we're sorting it shows us something else notice that the word systems appears twice there and that seems kind of silly and now in a single sentence it doesn't really matter but in the big document it'd be very nice to get rid of all of the duplicates so
let me run another program which exists on UNIX called unique and I will say unique from the sorted words and to a unique words and let's see what we got here sure enough the word systems the duplicated word has gone away and now we're almost at the point what we want to do finally is compare the list of words that I have here with the dictionary and print the words that are in my list that are not in the dictionary and we'll do that with a program called mismatch where I simply say mismatch and then
we'll go from the unique words and the dictionary those are my two sources of input and mismatch will simply print all of the words which are in unique words that are not in the dictionary and this one takes a little longer because the dictionary is actually fair size but sure enough we got laboratories and we got provide which are the two spelling mistakes and that tells you that this is actually quite a reasonable way to look for spelling mistakes unfortunately we've got a couple of other things we got the word time-sharing and we got the
word unix and this tells you why it isn't a perfect way to look for spelling mistakes time sharing is a fine example of jargon it's a word that means something to everybody in computing and probably means nothing at all the people who aren't in the computer business and UNIX is an example of a proper name and is not likely to be in the dictionary ever so what do we do when we get this list of words what we do is that we fix the spelling mistakes in the sentence and that's done and we take the
words that weren't spelling mistakes but that we're useful jargon like time sharing or useful proper names like UNIX and we use them to improve our dictionary so not only are we finding the spelling mistakes but we're also able to do the job better by taking that output and enhancing the dictionary that we use to find further spelling mistakes and so the next time we do it we'll never see those words again and so I think that what we've seen here is the the way that programs get developed on UNIX very often is not by somebody
sitting down and saying I'm going to do this complicated job like printing like creating a program to find spelling mistakes what I'm going to do instead is to try to break that big job up into little pieces to see if I can do a part here a part there and a part somewhere else and then stick them together to get the job done and then the other thing that sort of comes with UNIX is that they're already very large collection of useful programs little tools kind of building blocks that mean that you don't have to
build your own building blocks you can use the ones that are already there and I think this is one of the reasons why UNIX is a very productive system for most of us because the set of building blocks is already there and you can glue things together very very quickly in the way that I've shown you here and that way get your job done in a hurry and that in a sense is what productivity is one of the strengths of the UNIX system is its file system let me give you an oversimplified example UNIX file
system is like a file cabinet in a file cabinet there are folders within folders there are letters sheets of paper and on these letters and sheets of paper are words and characters similarly in the UNIX file system there are directories that have tags on them and within directories you can find other directories or files and this allows you to be able to retrieve a file very simply through this kind of indexing system files in the UNIX system are form atlas that means that they are simply a string of characters or bytes any format imposed on
that data is done by the programmer rather than by the operating system this makes programming very easy for for programmers for let me contrast other operating systems where at file creation time a programmer must specify what information will be in the file how big the file will be etcetera and once that is done the programmer is locked into it having locked into that kind of file then later it becomes impossible to put another kind of data in that file the analogy is to trying to put a legal sized document into a letter size file it
simply won't fit the UNIX system you don't have this problem because data in files is form atlas and data can be easily moved back and forth between files form atlas files in which the data consists solely of a stream of bytes it's uninterpreted by the operating system simplify life because it means that any program can process any file when you combine this with the concept of pipelining your stream processing it makes UNIX and extremely powerful programming tool I think the notion of pipelining is the fundamental contribution of Unix the notion that you can take a
bunch of programs two or more programs and stick them together end to end so that the data simply flows from the one on the left to the one on the right and the system itself looks after all the connections all of the synchronization making sure that the data goes from the one into the other the programs themselves don't know anything about the connection as far as they're concerned they're just talking to the terminal what Brian showed you earlier was that rather than write a special program to find spelling mistakes it's possible with Unix to cobble
together already existing programs and end up with a program that basically finds spelling errors now he did that by writing all by using one program on his sentence and then putting that output into a temporary filing and running various programs on the temporary files it's possible using UNIX stream processing or pipelining to eliminate the use of temporary files and let the system put the data from one program into the next those are called pipelines let me show you how that works with Brian's same example I'll use the same programs in the same text file so
I first run on make words on my text file and I run that through lower case I run that through sort and I run that through unique and I run that through mismatched and out of the end of the pipeline comes the spelling errors now it's also possible to sensing you get tired of typing the all the commands on one line to put the sequence of commands in a file and run only that file the system will then interpret what's inside that file is what you want to run so I have this same sequence and
if I called check and I can run check on my text and again I'll get it get out my spelling errors now let me show you another example I have a desk calculator program I'm going to run the output of the desk calculator through program called number which turns the numbers into English and then I'm going to run that through a program called speak which turns the written English into spoken English and we now have a talking calculator so if I say on 8-3 for example it tells me five now I can put this in
a file put this pipeline in a file name talk calc and get the same thing now we'll raise 2 to the 100th power 100 million 267 octillion 650 septillion 660 Liang 228 quintillion 229 quadrillion 401 trillion 496 billion 700 3 million two hundred five thousand 376 now person using this program doesn't need to know that it's that it's more than one program running and all of these things make our life much easier you don't have to keep typing the same thing over and over again and that's important for several reasons the first one is that
I as a single programmer can can up a set of commands that it would have had to have typed but now I can get it by just typing something fairly short I issue one command and what's really done is quite a few commands the second thing is because it looks the same as the existing commands on the system it's very easy for other people to use we've seen a great deal of Darwinism in the development and spread of eunuchs initially people in the computing Science Research Center broke programs and commands that we'll use were useful
for their own needs but a number of other people discovered that they were useful for their needs as well and as a consequence these programs and spread rapidly throughout the laboratories what this all means is that there is more leverage for everybody that this is sort of a continuation of progress in the computing field in which we have gone to higher and higher level languages and what has happened with UNIX is that over the years we've developed an increasing family this family of useful programs has gotten bigger and bigger and bigger and of course people
who have used it for a long time have a whole bunch of members of this family at their fingertips so that they can really get things done very quickly and I think it's that that really makes UNIX as productive as it is people have used the UNIX utilities to build all sorts of valuable applications and all of these are now part of the UNIX family of programs some of these applications turn out to have usefulness that goes way beyond the bail system for example the programs which we call the writers workbench were originally designed for
the use of our own people here but by now are used by people in a lot of other places as well in the bail system we produce enormous amounts of information for craftspeople that tell them how to install maintain and repair equipment this documentation is much more important these days than it used to be 25 years ago most of our equipment was mechanical a person who was trying to install a repair piece of equipment could frequently figure out how to use it just by looking at it these days however our equipment is much more complex
it's made up of computers and computer programs called software to debug or fix this kind of equipment you almost always need the documentation this documentation is usually written by the developers after all they know the most about the equipment but they weren't necessarily trained in how to write well so to help people like this and improve the bail system documentation we've developed some writing aids called the writer's workbench to begin I'll show you the output of a proofreading program called proofer as you can see here the file has only four lines in it or three
sentences probably by just looking at it quickly you can tell that it has several errors in it if it were a longer paper however you might have trouble figuring out what those errors are so let me show you what our proofreading program would do with this text the first information that comes up is a spelling check this program is basically a spelling checker that's similar to the one that Brian Carnahan demonstrated earlier in this film one of the advantages of UNIX is that we were able to take this program and/or one fairly similar to it
and build it into a proofreading package so you see that it found the two misspelled words here it then goes on and looks for simple kinds of punctuation errors we've gone however beyond simple proofreading to try and analyze the style of people's written text as well remember they frequently don't have a professional editor to look at their writing and there's lots of things about writing that are important especially in training documentation but people frequently don't know for instance you want to have the smallest number of passive verbs that you can and still have the text
read fluidly because the passive verbs make text difficult to read so we have a program that analyzes many of the different features of the syntactic style of a paper I've run it here on a memo of mine here it tells you what the readability score is it tells you that it's 11th grade it tells you that that's a reference to 11 years of schooling that you have to have to read this text and tells you that that's a good score it then goes on and describes the variations and says that you've had a good distribution
of simple compound complex and compound-complex sentences following the short version it tells you that 57 words is pretty long for a sentence and that really you probably ought to change it and it goes beyond that and tells you how you could find what that sentence is and all the sentences in your text that are over some lengths are UNIX operating system helped us a lot to develop these programs quickly for instance there were many modules like the spelling checker and the diction program that were already existed and we're already debugged so we didn't have to
start from scratch computing is going to be more and more it or woven with people's lives as the years go by so computer technology is going to have to evolve to be easier for people to use UNIX is certainly not the end of the road in this regard but I think it's a good step along the way you you you
