 My title, 32 years of metaphor. And why am I celebrating 32 years? Everybody else celebrate 30 years. What am I talking about 32 years worth? 32 and all the powers of 2 is like 2, 4, 8, 16, 32, 64. These are a big deal for computer scientists. The greatest day of my life was my 64th birthday. Really, we had a special my wife arranged a special surprise for me on my 64th birthday. And if anybody asked me what was the high point, we have to be that that day. Even the Beatles predicted this when they said when you're 64, you know, but but but I can't wait for metaphor to be 64. So I'm celebrating 32 for metaphor. Now, last last Wednesday, I was Luigi Cherbini's 256th birthday. And so I celebrated that by listening to more than an hour of his string quartet. So and so. So what was I doing 32 years ago today? Well, I was writing. I was trying to figure out how to advance to the next. Yeah, it's not doing anything. It seems to be dead. So this is called the intersection of art and technology. We have any representative Apple here tonight. So here is it going on. OK, so. So I was I was writing I was entering into my computer chapter two of this or this book, which was exactly this was September 20th. 1984. But earlier that year was it was the was the class that David referred to. It was the only time Stanford actually had a class to specifically devoted to font design and and Chuck Bigelow and Richard Southall. And I for 10 weeks, we taught a class about font design using Meta font. And at that time, Meta font was just being written workstations. The sunwork stations had had just been delivered. They had no operating system. They were using the first editor on them. Everything was crashing, but it was a really exciting class. And and so I'll tell you a little about this about this course first. Now, I also want to mention, though, before I before I do anything else that it's really appropriate to talk about this in San Francisco, because San Francisco is one of the only cities in the world that has a font boulevard. How many of you have ever been to font boulevard? And it's of course you look into the history and you say, why does San Francisco have a font boulevard? And the answer is that it's named after Father font, who was one of the early visitors to California in 1776. But anyway, I have to show you something that one of my students gave me a few years ago. Turns out that that the sign for font font boulevard was being retired and so it was so it was available for sale. And he found it. And so now I have it's about this big as part of place in my house. OK, so now this class that we taught. I don't know about three. There were three dozen students in the class and I can show you what one of the first exercises that we did. The students were supposed to design a border. So I had to had to design some character to use at the top of the border that would be that would be repeated. One for the left one for the right one for the bottom and then also four characters for the four corners. And they were supposed to fit together so that that would make a border and so we have about three dozen of these. And there's just one example. But you see second for the bottom is Dan Mills. He was one of the people that David said I started with Arnie all about work at the bottom. He was Stanford's printer of Bruce Leban went on to found several startup companies. In fact, everybody in the class turned out to have quite an interesting career during the 32 years. It'll have since then and last go Harville. I I'm not sure but her but her husband young Harville was one of the great pioneers of virtual reality. If you use there and so on. So and young was with our thing this class to. So at the end of this course, we had a day when we went to when we went to San Francisco and visited different places around the city. And we also made a pilgrimage to font street, as you can see. So so well, let's see. Second from the left is Dekron, who is here tonight. Richard Southall would have been he's a second from the right in the front row. Dave Segal, Lynn Ruggles. A lot of people that I'm not sure how many of them are connected to the audience tonight. But we also decided to make a condensed font. And of course, of course, we need an italic font as well. OK, so so so all of this story is told in a in a book that that that Dekron helped me publish called Digital Typography. And this this book has everything essentially that I know about topography that you know that my work on topography from from these these these days. And what David said, it was it was sort of a detour in my life or something. It was just how do I how do I say it? I'm so glad that to have had this experience, especially because of not only because of all the beautiful things that were made and that I got to see. But but because of the wonderful people that I met, somehow the graphic artists that I've that I've known have been about way up there. And it's the best people that I've had a chance to get to know during the year. I also should say, by the way, that this that the cover of this book was designed by Cope Compton, who happens to be here from Illinois today in the audience. OK, now this. Oh, yeah, this cover. The ends are by Matthew Carter variations on Microsoft. It did. But in the background, if you look closely, you'll see a font that's that's made entirely of ends. And that was an experimental font where we just wanted to have text. But we didn't want people to read it to read it. We were just experimenting with with line breaks and spacing. And and then we found out later that that if you if you were clever, you could act. You could actually read it, though, because some of the letters are wider than others and so it made an interesting code. OK, now I want to tell you how Meta font work because it was different from the other the other systems that people have been working on. For designing type at the time. But but it was something that seemed to me my goal was not to just copy what what somebody's design was. But actually to to teach a computer how to draw letters. So so you don't just say, you know, to take a photograph of this and trace these outlines and so on. I know you say somehow when you draw a letter, there's some meaning. You want to connect certain things to other things and that makes it look so. So anyway, I'm going to take you through the way Meta font drew the capital letter A. And this was five years before 1984. This was in 79 with when I did my prototype and experiments, which which led to this. So I start out and I know how wide I want the A to B because I I have been my model for this particular style of type. What was a monotype font called eight? It's called modern eight A and and I knew what the monotype had. I think it's 13 units in letter A or something like that. How many years there are. And then there's a little sidebar on there. So now so I tell the computer, OK, pick up a pen and draw a line, pick up a wide pen and draw a line. That goes from the from the top of the cap height down to the baseline. There's a cap right up there. Then you can see a place for the excite. And there's another line which is sort of where the bar cross on the comes through. And so go up, you know, just above the cap height and come down to the baseline with with your wide pen and starting to start about in the middle at the top. But come at the right, you want to be two units from the right end. Now the next step erase because because we're going to want to do something at the top and and and it got too black at the top. So I'm going to take an eraser and throw it away. The next step we prepared for that by taking a thin pen and going from the same point at the top, but down to the baseline and coming to units from the left hand side. And then the next step put in a bar line. Okay. So, so, so here at the at this at this I'll call it the E height of go from take take whatever pen I I told it to use. I guess it looks like the same pen we had for the left hand stroke. It says, you know, go from where the where where your second stroke was over to where your first stroke was at that at that level and join those together. So, you know, you see, I'm not tracing outlines, but I'm saying draw this according to a structure of a letter A. Next step draw little serif going out to up to the left of that bottom stroke. And what's next? Yes, we'll do the other half of this. Okay. So each of these was one line of code in the computer. This last one was done with calling a subroutine that makes service. And now watch closely the next step you might not see. Did you see any change? Go back. Okay. I took a little off on the bottom of the service there. That's cool. It makes it. You can charge another dollar for profanity. Okay. So then then we put in the serifs on the other side and and we have a capital letter A. All right. Now, that was that was all done assuming a certain width for the for the wide pen and for the narrow pen. But but but I can change that. And and so here's the same thing about making a bull face instead of instead of a normal A. And so you just for the bull face, say you just say that, well, these pens are actually the unit size could even be a little wider. What but they but but you make the pens bolder and and you adjust the other things accordingly. And so the same program draws bull facing. And so I'm saying the type designer tries to imagine not only not that that. You get that. The goal is to do one type face. But but you're imagining that that your boss is going to tell you each week to do another another version of that type face. So you might as well figure out in advance what you're going to do under all the all the things that he's going to do. So this is this is where the word metaphor comes from. This is a meta design. It means it covers more than one design. It's more abstract. The same thing for a typewriter style. Now, in a typewriter style, the the left stroke and the right hand stroke are both the same thickness. And and the you know, the serifs aren't so sharp. So on there, but the exact same program, but just change just change your preferences or whatever you want to call it. And you and and you get it. So this is this is way it was done in my first prototype system in 1959. And I I learned later that actually. This isn't this isn't really the best way to proceed for because the A will start to look a little too dark at the top where these two strokes coming down. And so in order to in order to do a little better, you can you can move the you can move the two strokes apart a little bit. And you can also open up the join away in order in order to do that. So so I learned these these refinements later, but they could all they could all be built in very simply to the instructions that I'm giving the computer on how to draw a letter. Now, now, starting in nineteen eighty, I I if that's when I first met Herman's up and he came to Stanford and here we are looking. This was I don't know maybe three or four days after we have met each other and and we were trying to design an F.T. Ligature to see if that would help if we had F.T. combination. And and so I'm sitting here with with with an old computer terminal that we had we had very limited equipment actually for displaying all these things in those days. Well, let me go back to the the problem I faced. I was trying to I was trying to make the fonts that would be used to typeset my book on art of computer programming. And the and my books had first been done in the in the sixties with with beautiful. Well, it was it was called the monotype eight, eight modern eight, eight fonts, which had which had many, many sorts to handle mathematics. And so at that time, all of my favorite textbooks, the ones that had had been most beautifully typeset, were using this monotype font and all the collection of mathematical things that it had with it. So that was my standard of excellence. And and I thought, OK, well, all I have to do is tell a computer how to how to do what that was. And then I didn't then I could then I could typeset my books because the there was a revolution in in in the in the industry where they essentially replaced all of the hot metal types with with at first with with photographic types. And then later with with computer types, but nobody cared about the way mathematics looked. And so and and so the proofs that I got from my publisher were incredibly bad. And I just couldn't stand writing a book that was going to look that way. And so I I found all it was needed was a small matter of computer programming. Just needed to do is write a computer program that would that would make the letter forms the way they should be. And then actually would it would be possible to do a decent book. And so I started looking at the history of it and here are some designs of the letter S that go back to the well, the fourteen hundreds and fifteen hundreds. And then the the the one at the lower right was designed for a Louis the fourteenth in in in France of the seventeen hundred maybe maybe seventeen thirty fourth. And so that's the way that they had used kind of a mathematical model for for letter form. But it didn't look you know, it didn't look like the letter S that I wanted to use in my books. And and so for a while, I thought maybe I should try to like, you know, just just rewrite my books so I didn't need the letter S. But then I'd have to leave Stanford. So so that wasn't an option. So I worked out for so I worked for a while and and and in fact, I went several days without sleep, not being able to figure out what letter S should really look like. And and I showed the I showed my my proofs to my wife in despair. And she she looked at it and said, Don, why don't you make it look like an S. Yeah, but what was it? But finally I figured out a nice a nice way to that I think the Greeks would have would have enjoyed the properties of of conic sections. And and I got so that I so so that I could make a fairly decent letter S. And then the question was what what slope should I use in the middle. So so so all of these S's come from a program that varies in one in one way. Just as to should it be pretty horizontal or should it cut at a sharp diagonal. And you know, and then I but but the main thing is I had to find some mathematics so that if I set that angle, then it would plan the rest of the curves and something that that would that wouldn't be too bad. So this was 1959 is when I'm just learning. But but it shows you some of the things that I got into when I when I faced the problems of telling a computer what what a letter looks like. Here's a here's a later example. This was, I think, you know, I presented this in 1983, which was which was a special conference of the of a type I. This is the Association Typographic International and Stanford had a wonderful week in the summer of 1983 where which was called digital typography, the computer in the hand. And we had all of the leading type people that Chuck had had had worked. Chuck people had worked very hard to plan this week. And it was it was kind of a climactic thing for the whole field in 1983. And and this is these slides are I gave as an example here I have here I have two. Two things vary in what in one case the it's the it's the slope in the middle. In the other case, it's the thickness in the middle because that's another question. How how thick should the SP have to do it? Here's an example. Another example that I gave designing this, the number six. How high should the ball be? So the same program will draw different variations. But you but you but you tell that you can you know, I can change the position of of one of the points. And then the other things try to adapt themselves in a decent way to the. Now, at that at that summer workshop, Summer Stone to develop the logo for for it that everybody was that everybody was wearing this t-shirt. I said I saved this t-shirt from 1983. It's just getting a little big but the but you can't see there are all the letters on there. But but but but he but all through that week we were seeing these five letters and and so I got I said, OK, in my talk, I should show how how how Meta font would approach a font of this kind. And instead of that modern a day font that I had before. So so here we have Summers a except that I that I abstracted it out and converted into a certain number of other points. But it still has many of the same properties. You know, you have this big stroke that goes from the top to the right hand corner and and here you know, I have an angle as to which as as to how the pen is going to do an amount of tapering that's going on in the pen. But you see there's a point number four there at the lower right. And I placed that that about a unit and a half from the right edge and and I placed the point two at the top of usually in the middle. And then I have these other things coming in the left. So so it's not it's not totally different from what we had before. And and with this program again, you have lots of parameters that got lots of knobs that you can turn if you want to change the. The characteristics of the so you can you can vary the amount of tapering and the thickness of the strokes and and so on. All coming out of the same program saying how to draw a letter A if you if you set your preferences for the for the different aspects of the thickness. And this is the program itself. You know, I don't expect you to read it. But anyway, you can you can see that that is there. And all of this is in in my book digital department. OK. Now, we were looking at the at the example of of the word metaphone in close to border designs. So the and in the title of. So when I use metaphone, I have a special font that I used for the word me to F. F. O. N. T. That's that's the logo. This font is called logo. And and it's a completely it's an incomplete font. It it only has it only has enough letters that it needs. And so it has the letter F. A. T. You can say the word fat and you can say the word Omen and you know, and that's enough to make metaphor. And and and we never did, you know, I don't know what a B looks like in this font or a C, although we did add a P and an S so that we could make metapost, which is another program. OK. But but but now this. This is an example of that. You know, that. Well, I say very simple to make all those letters because you see how short the programs are. Let's see what if I if I go up, can I take, for example, in here. So the program for the end is only one, two, three, four, five, six lines long. Well, the first line just says, you know, it's going to be an N that's going to be 15 years wide. And then it tells where they have formulas that say where points one, two, three and four and five are to be located. And for that, you says, but why one is is minus. So that says that the bottom of those things should be a little overshoot below the baseline. And the and the top should be a little overshoot above the above the height of the character. And then I say draw one, two to three. So that draws this and I draw four to five, which does that one. And that and the last line just says put labels on it one, two, three, four, five in case you want to see what in case you want to see the proof sheet. So here, here, here was an example of a font that also did their ligatures that bring the T and the A a little close together if if T is followed by a letter A. So that's an example of the code that that one writes for for create to create letters. So here's a here's a picture of some of our group in 1985. Richard South always the guy in the left. He died last year. He was wrote important books about typography and and he was he was kind of my main permanent. Besides Chuck Bigelow was my main permanent typographic advisor. Although we had many visitors to our project on our Samuel left at that time. He was maybe the world's only programmer over the age of 80. But he was famous for writing the first program to to play checkers and and and beat a human competitor. And and what they would use to all the other people. John hobby at the left at the right was the main wizard who worked out some beautiful mathematics by which metaphor chooses how to how to make very pleasant curves. And so so there were other people. These are the ones that happened to be there for the picture that day. One of the visitors we had was Herat Unger from Netherlands and his wife. If you go to if you go to Amsterdam, you'll you'll see examples of his typography in the airport. So honey, and in all the phone books and and so he's a wonderful type designer from Netherlands came to visit us for for three months. And and and I think it was eighty three I mentioned which year. But let's see. Yeah, you can say eighty five because it's on this slide. OK. But anyway, his wife, Marianne, also came and she talked about fashion style styles of furniture and and. Of pottery and and and other things not to do with type and and her husband talked about style in typography. And one of the points that that her art made was that a graphic style always lagged by 10 years from the from the other kind of styles. In other words, you know, the styles of furniture got to be something called streamlined. Well, then that there were types that sort of look streamlined 10 years later. And so it was a series of seven or eight lectures and and and they'd given the first six lectures and showing this theme in different aspects of the of the style of different kinds through the years of like, you know, the Netherlands. They had a period called they steal and then, you know, there's a view in steel and various other kinds of styles. And so it dawned on me on the morning there were there were scheduled to give at seven o'clock that night. They were scheduled to give the last lecture in the series. And I woke up in the morning and said, you know, if what he's saying is correct, then. There's going to be the next thing in type design is going to be something for a style that was prominent 10 years ago. So what was the style from from 1975? And the answer was punk. And so I said, OK, I'm going to design a font called punk that will be the just the right thing that the type world needs. And but I have to have it ready by seven o'clock that night. And so I sat down and and madly I don't know if I ever worked so hard. I mean, it's like the demo scene and anybody knows it where people work and do a lot of hacking. But I but I had only this this one day to do my font then. And and I decided that the thought would would consist of dots and and lines connecting them because because we had seen examples of of some patterns and furniture that was that were designed in this 75 that that had this characteristic. And so this is my first proof sheet before, you know, I hadn't debugged it at all. This is the first I saw of any of the of the letter and it looked pretty it looked pretty awful. But anyway, this was after I got it debug. And then we could make the entire character set that day. I only did the uppercase letters and I didn't make Lord make lowercase letter. Well, it's small caps anyway, so I use the same program for the lower cases as the uppercase. And I only had a few punctuation mark, but I was able to make a handout for that night and distributed at seven o'clock with a with a fund. That was definitely punk. And and so one of the things about it is that I might have found includes a random number generator so that every so that every time you generate the punk fund is different. You don't say go put a point here. You say put a point approximately here and then random number generator gives it a little jiggle and size where to put it. So so each time you you see punk, it's a little bit a little bit different. And and the programs for the for punk look like this. So again, they're they're they're fairly short. The first one's like P. P. says a punk point, which means start there and then jiggle. OK, so OK, so now it turned out there was some confirmation because of October of 1986. I was I was in the metro in Paris and I saw this sign at the bottom that someone in France had come up with that with a font that looked very much like like punk independently. So but but but also I at the Boston Art Museum, I found the pattern up in the upper right hand corner, which which Picasso had made in nineteen twenty four. So he was way ahead of all. OK, now the the highlight of of my work on on type design has series of fonts, family fonts called the computer modern typefaces, which which are the fonts that that are are now used in my books. And also I hate to say it, but maybe ninety five percent of all papers that are published by mathematicians and physicists these days. And and I have to have to confess that I haven't gotten tired of it yet in my my own books anyway. And I had lots of help on this over the years, of course, because of all the visitors that came. But but it's all summarized in in volume E of a series of five books, books are computers and typesetting. I discovered today that the San Francisco Public Library does not have this volume in its collection. I don't know if they do something about that. But but but the here's the here's the title page. You can see the computers and typesetting volume E at the top. And then then here it just shows the these letters. ABCD had been on the cover on the on the book jacket and also on the spine of the first of each volume. And you see that that that shows examples of type morphing from one set of parameters to another. So at the top we have my my standard modern font at the bottom. We have the type letter stuff on and then, you know, 10 steps to get from one to the other just a little change at a time. So these letters are we're done with trying to not only tell the computer how to make a pretty good letter, but to try to make make make a a top quality letter of for each of each of these letters. In fact, there's there's some 400 more than 400 different cliffs included based on what turned out to be 62 different parameters that that you could vary. So you have I talked about the you know, the center height, the body height that I said E high is really called the bar height. And so so those are some of the parameters that obviously we can change. Then then as we're drawing letters, we can we can talk about the different kinds of thickness of the pens. How how how thick is a stem? How thick is a curve? How much does it does a ball out to the bulb of flare out and so on. And on other fronts, we we can say how much of how sharp they are at corners and things like that. And and serifs have brackets that that we can vary how much how much it drops that how how how of our beat comes down, how far it extends. And you know, the the how much dish there is at the bottom of a serif. I showed you that before. At the right, we have corrections to avoid the V being filled in so much like I had talked about letter A. And a notch cut is how to open up to make it. So we have all of these all of these parameters. And then then we. So then I can make these, for example, these extreme examples of the letter capital A all with this program. And so this program, you know, it says pick up a certain pen, but each line of the program says how to position the pen. And, you know, one of them will say, you know, I want to go from the left at the left side of the right hand side of the where they intersect. And so and so there's a line there that says whatever you see. So it says choose whatever line works, whatever point works at the correct angle. And then it turned out that that's enough mathematically to to identify the correct the correct thing. So so anyway, this volume E contains the entire specification of how to draw all all of the letters that are used in all of my books, in particular, all these five books. So I don't know of anything else in history where the books describe exactly how they're printed, how they're made. Every letter in these books is described by the programs in these books. And in the other books, it tells gives the programs that that put the letters into position on the page. So it's a self describing book. OK, let's see. This was mine. It's right. I should be able to go to the next page. So so similarly, the you need you need a program for for a combination. If you're going to if you're going to sell any fonts in Denmark, where it's one of their basic letters or France and Tokyo. So so on the left hand page, you see the you see examples of the of the design with with different kinds of the parameters on the right hand side is the complete is the complete program. And and here are two of my main helpers that were advisors as as we were getting the bugs out of this design. So we looked at many, many proofs and I know and I would learn how to shave a little bit off of a stem here and there. So the optical effects. So that's that's permanent up in the middle and Matthew Carter and the right. Now, in conclusion, I wanted to show you some of the other things that we've been doing with that that that have been done. One of the one of the early projects done with visitor from China in the early eighties. And John hobby was to design. Nice subroutines to do the basic parts of Chinese characters. And so we have a way to make teardrop shapes and different kind of strokes and then quite adaptable. So that then this then can be put into Chinese funds that can be rendered in different stuff. One of the one of the fun things has been Christmas cards at the end of the year. This was was made by Hans Hagen in in Netherlands using Meta font. And the easy way just turned out that all of these candles are done by the same program using a random number generator to choose the color and and to choose the size of the of the candle and the and the flame. And so each time he generated it, he got another another picture like this. This was his Christmas meeting in year 2000. You sing have a good 21st century. Whoops. My wife and I. I got to show you our Christmas card for 1959. That's when I was first first working on metaphor. And and I got to use the computer equipment at Xerox Palo Alto Research Center. And at that time there was a new fangle thing called a color Xerox machine. And and so I thought, oh, wonderful. I'm going to have a Christmas card that says Merry Christmas. And it's going to be printed with this color Xerox machine. So so I will it turned out that the the size of the files that it took in order in order to send these letters to that machine were 10 times or 20 times more than the software had ever seen before. So so so I would get married to go and then and then it would then we'd send it to the next to the next step between it would break, it would crash. And so the patch that we get it through that one and the next one would break. And so. And so finally, I realized I was never going to get Merry Christmas done on the color Xerox machine until all the software had been rewritten at zero. So so I settled on the word Hark, which was which was just big enough that the machine could could handle that. And OK, so so that was 1959. And it showed the status of the of the funds that I had in the early days before I had met. Then down below with Christmas card 1981, where we used the Christmas story from here, I was combining font design with with typesetting. What have I got here? So so you can see that that I had a big that was able to use a bold font and switch with a light font. But then use my typesetting system also to arrange the the characters. So this is the entire Christmas story that that we all read to each other. OK, now then. Here's a here's a card I did 1985 where I made a font that that was intended for type setting Celtic knots. So so if you look, these little lines will cross over and under each other. And and so I until I could design this particular grid grid like thing to make a pattern based on Celtic knots. And and this this is not in my book to draw typography. This is in my book on fun and games, which is another great book that makes a great Christmas present. And in 1989, again, we use Meta font for a Christmas card. This time we had a font called Holly, which just made me Holly leaves and berries. But I also but then it's combined with another font that called Jen. My daughter's name is Jen. And I based this fun on her handwriting. And so it has lots of of special ligatures and things like this to be a fairly good imitation of of my daughter's handwriting. In recent years, this is this is a paper said to me from Italy a couple of years ago, Luigi. Garza, who is who is a quiz combined Meta font with with another programming language called Lua. So that you can do interactive you go between Meta font and a and a more traditional programming language interactively. And and so he's has been working on things like this. And my original title, this was this was some work that was actually done in the nineties in Poland. But but when you have the when you have a electronic form of characters that chose their structure, you can you can play interesting games with them. And here's another going while with with that. OK. All right. So so so these are some of the things that that they came out of this little project that I that that was born came alive 32 years ago. And as I say, it's been a nice part of my life. Let's see. I have a little demo that I could that I could try. But I but but maybe I'm running over time. So OK, so I might need I might need Frank's help, though. So yeah, so so so two weeks ago, a man named your little Meyer in Germany, sent me a program that he has just been developing for an interactive version of. Yeah, bring it up. That was with Safari. Wasn't it? Yeah. OK. So so so here he I just wanted to show you. Oops. Oh, I see. OK. So I mentioned John Hobby had worked on special formulas, making pleasant curves. And so if I draw another curve here, I connected to the wrong one. Well, let me move it down. But anyway, Metaphone has I have I think the world's best algorithm for figuring out if you have a bunch of points. What's what's the most pleasant curve that you that you can make without giving any particular extra extra help to it? And with his program, you can you can play around with and see what goes on. But but but John worked out some beautiful mathematics that made this happen. And then there's this program also also at the same time, it writes out a little metaphor program that corresponds to what you did. So now you can edit this program and you can say, oh, you know, pick up a different pen, draw the characters with with another. And and it has enough that you can actually save your letter forms and make a new font out of it. And there's all that behind the scenes. So so this is a is a program that that just right now in development shows that Metaphone is finding new new friends. The the the combination with Lua, by the way, also outputs to open type font as well. So so this is kind of a sampler of of of my life with 32 years of this. Thanks a lot. Anybody has questions raise your hand. OK, yeah, by the way, I should have mentioned the this yeah, this example in Poland. I should have mentioned that this is a jacovsky and roots go who did his work because that their names ought to be mentioned. OK, so lowercase a lowercase six. Hey, hey, why is that design in a way that's never written? Why is what lowercase a and design that's never written? I never hand write the lowercase a like it's tight. Yes, actually, I started writing that lowercase eight that way. Some of some of them so there's there's this font called Stone Informal, which uses the which uses the A that that we learn in first grade. And and it's kind of pleasant. But but but it's interesting that we don't that. Most of us don't we're never taught to write lowercase a the way it is in a book. Now, I also but when you asked your question at first, I was thinking you were going to say, why was why was the letter A so ugly in in nineteen nineteen fifties. Nineteen seventy nine when I had to mark the word because the A there was actually. Yeah, yeah, yeah, when I when I showed it to Chuck Bigelow, Hermans of Matthew Carter and and Richard Southall and and various other people, they all quickly corrected it. And so I learned how to completely change the the A so that, you know, my original what I thought before I had before I had had done any years of apprenticeship. What was was was quite different from the A that now is in computer. There's one question here. You know, I can repeat your question. OK, so the question is bringing artificial intelligence into this AI. In fact, this is this kind of interesting because because Doug Hofstetter, who's one of the he wrote a very influential paper, sort of saying that that he said the most fundamental problem of artificial intelligence is what is the letter A? And my answer to that, I mean, OK, you know, artificial intelligence AI. So the most fundamental question about AI is what is the letter A? And my answer was and the second most is what is the letter I? Some people get it. So but but but essentially he his article showed about 50 examples of the letter A. And, you know, in saying you in some way we we know that that some of these will not go with other letters B. Somehow, but, you know, from seeing the letter A, could you imagine the entire file? And he asks this question as an example of perception, just to point out how little how little we know how to automate things like this. So but my opinion is that, you know, we always get a lot of knowledge. If we try to if we try to teach a computer how to do something, then we realize how little we know. And then we so in the process of doing this, we're learning a whole whole lot more about whatever we're doing. On the other hand, I, you know, I since we learn more in the process, we're also getting ahead and and and and that that much further ahead than we were before. And so and so so far we the computer hasn't caught up. You know, we we we stay a leap ahead. But but but questions like this are always are always insightful to to improve our knowledge. I haven't had any questions. Hi, I'm curious if you wrote a brand new typesetting system to publish your book. And if Park couldn't even print your Christmas card, how did you manage to get your book printed? How did I manage to get my book printed? How did the publisher handle your files? Oh, so I had a very good relationship with with my publisher, actually. Addison Wesley was the only technical publisher that had its own in-house type people. You know, McGraw Hill and Elsevier and so on would would farm out their thing and outsource it. But Addison Wesley right next to the building was Hans Wohl's composition shop. And this this man dedicated his life to finding the right way to typeset technical books. Had had lots of monotype machines and really skilled people working for. But then the in the 70s, when all these all these hot metal machines were were retired and these other people who had done the typesetting. Had, you know, had nothing to do with it. Then Hans was at his, you know, was doing his best to try to find any place in the world that that could make my books look like they did before and and failing. But but but so so. Addison Wesley contributed, I don't know, $30,000 or something for for for us to buy the alpha type. Typesetter, which was one of the first digital typesetting machines. And and and so I that was the machine on which I made my edition of my book, which they which they printed that. So I would send them the output of that machine and they would photograph it the way they they used to from their from their other equipment and and and they took it the rest of the way. Hey, so I'm curious with your background in computer science and mathematics. How did you stand up? Sorry, yeah, I'm just curious. Like, how did how are you inspired to pursue this direction and creating a font format? And like, what was your inspiration? And or like, how did you become interested in calligraphy and type and become connected with Zapp and so first as an author, I had spent, you know, hundreds of hours proofreading books. And so I thought I knew about letters and then when I saw the only letters that were done with these infury equipment of the 70s, I, you know, I naturally had I don't know of anybody who would who would have reacted, you know, they could look to that and but I did. But but the aha moment was was when I was on on on a committee from Stanford to to redesign our our qualifying exams for our grad students. And so we were checking out the to change the reading list for the next year. And a new book had come out by an MIT professor that had been typeset in down in Southern California on a new on a new machine that was digital. So so we had at our at our lab, we had we had something called the Xerox graphics printer, which was it was about 150 dots per inch to 250 dots per inch, depending on where you were on the page. But it was it was very low resolution, but it was but it was digital. And it would make it would make something that looked a little bit like books. But it was it was it was crummy. So I never thought of of of doing it that way. However, when I went to Southern California to see the proofs of this new book on AI by Pat Winston, which has turned out to be it's in the third edition now or something like that. But anyway, the first proofs of Pat Winston's book were done for the first time with with this laser cut machine made at triple I'd in Southern California. They and and I couldn't tell the difference between that and metal type. And so I, you know, so I grew up in in in Wisconsin, where we have where where where it was illegal to eat margarine because because margarine, you know, wasn't butter. And and Wisconsin is the Wisconsin is the dairy state. And so and and and and and actually I thought margarine tasted awful too. And and and so it was something like that with type. You know, I would see crude digital type and looked awful to me. So I never thought about it. But when I saw this is this example of high resolution digital type done with a really precise machine, you know, I couldn't tell a difference. And so I said, this is like having butter instead of margarine, even though it's done with it's not done with metal. It done with totally with bits, zeros and ones. And I know how to do bits. I mean, if there's anybody in the world that's going to make that's going to be all I have to do is write write programs that that compute zeros and ones where I put one where I want ink and zero when I don't want ink. And so that's my life zeros and ones. And so I said, OK, I was planning to take my my sabbatical year in in Chile. I wanted to learn Spanish and and we had major major arrangements. But but within a week of my going to Southern California and seeing this book, casual all those plans and said, no, I can stay at Stanford. I'm going to I'm going to, you know, write these programs that that put the bits on the page. Well, I thought I could do it here. It took a bit longer, but but but that's how in general, I'd never been very good at estimating how difficult things were going to be. One last question. I think this will dovetail very well with the previous question. As you started on this journey, was there a point you thought, oh, my God, what have I gotten myself into as it proved to be more complicated than perhaps your first thought? Yeah, there was there was always a question of where to stop and how to make it and how to responsibly exit from from from this work to something that I knew I could do do well because I can't know you have all these all these artists who were so so so I worked out a strategy whereby I I could put the thing on a stable ground, but it would take five years of work. So so so and at the end, I had to make all kinds of changes in order to adapt to all the all the new, you know, unicode and things like this. At first, I was just designing without accents. But but I had to, but so many people were using my system that I had to had to find a way for it to become to work with all languages and not just English. So that did take five years. But but but then I worked out a way that that that it could become a stable fixed point on which other people could build and that I wouldn't have to be coming up with a new version every year. But that the one of the virtues of it was that it did not change and it was buildable. It was it was it was something that that that would stay hopefully 50 years from now. People will still be able to use the source files that they do now if they want to. OK, thanks very much.