>> Welcome to CS50 this is the start of week two and I actually got an email just this weekend from one of our own TFs who had been here at last week's lectures when we talked about some of these more arcane topics, namely one called typecasting which we'll look at in more technical detail today, but his email started as follows. Hi David, you introduced the concept of typecast, recall call this as converting a char to an int or an int to a char, last week, and exemplified the importance of understanding the types of operators. In case you want to insist on this point one day, here's a video of an event that changed my life. I was 14 years old watching this on live TV so what you're about to see is a video of ARIAN 5, Flight 501. This was a European rocket, there were no humans on board this rocket but at the risk of pulling up YouTube in class again, let me go ahead and show you exactly what this teaching fellow is referring to. 

[ background noise ] So there weren't, to be clear, any humans on board but the explanation for this explosion, which wasn't all that many years ago, was this. Because of a different flight path, a data conversion from a 64 bit floating point value to a 16 bit signed integer value caused a hardware exception. More specifically an arithmetic overflow as the floating point number had a value too large to be represented by a 16 bit signed integer. As it was an unmanned flight there were no victims but the breakup caused the loss of four cluster mission space craft, resulting in a loss of more than US $370 million because of one line of code. So if you find yourselves getting hung up on some of the esoteric nature of some of the lower level syntax realize that these are just building blocks. They're doing far more interesting things and certainly when you're actually coding real applications far more important applications do arise. So I asked this teaching fellow if he could elaborate on, you know, why was this actually changed your life? And he responded, I think that at first I was fascinated by space and I wanted to engineer rockets and satellites. To me computers and programs were amazing but a toy. While the conquest of space was serious stuff. Progressively however as I was studying engineering and sciences in high school and then college the memory of this event helped me realize that software engineering and computer science were much more noble then I initially thought. I think the reasons were three. One, software is useful enough so that some of it was in this rocket sent into space. Two, it is critical enough that a malfunction result in the rocket exploding and three, writing software is actually such a craft that even these rocket engineers from all around Europe could get it wrong. It made software engineering and computer science a whole lot more valuable to my eyes. He wasn't the only one so we've linked to the Wikipedia article that discusses the real in depth history of this particular story but one or part of the article concludes as follows. Flight 501's high profile disaster brought the high risk associated with complex computing systems to the attention of the general public, politicians and executives resulting in an increased support for research in ensuring the reliability of safety critical systems so we appreciate, if nothing else, as you dive into your own problem sets. One this week and into future problem sets that even these small bugs absolutely can recur in the real world and if you think about some of the P-Set problems you're working on now, your code might be this long or this long, a few dozen lines, maybe 100 at some point very soon, but if you're finding yourself having a bug with this line of code, imagine what happens when you literally have their implement rocket ships and operating system which are millions of lines of code long. So thankfully there are tools and there are other languages. And one of the goals of this course is to introduce you to progressively better and better approaches to design, to style, to correctness and so by semesters end it won't necessarily feel like you're struggling with these if conditions and loops so much, you'll actually be thinking the much bigger picture and that will happen quite soon. What some of you might have seen as you came in today is a little something called ASCIImation. This is free time to the extreme. This is someone having taken this idea of ASCII using ASCII characters to actually represent, not just pictures but an actual movie so this is his interpretation of the original Star Wars movie. I bring this up only because in problems that won for problems that won, well you dabble a little bit in this idea of ASCII art but realize that as the semester progresses we'll very soon have more expressive graphical capabilities. I thought, and speaking of, what you'll find if you haven't poked around recently on the course's website is you. So we took all of your answers to Problem Set 0 where one of the demographic questions, where is your home town, recall we didn't ask for your home address, just your city, state and country And what we did was, we used an off-the-shelf API. Application Programming interfaces, it's called. The students last year in fact used this for one of the tail end problem sets in the course and we embedded a Google map in the courses website and then we proceeded to represent each and one of you with a little marker. The projector in here is making this a little darker than it actually is on a real computer but if we zoom in down here we might see that Dapo is from Lagos, Nigeria. And if we click on any number of our other folks here, is he here now, you will see where each and every one of the 500 of you are from. So with that said, where are we going? So we're going to lunch this Friday. Not all of us but one of the traditions we've started in this course because its fairly large, is, on a semi-weekly basis to get to know some of the teaching fellows, the course assistance, myself and so we have CS50 lunches. Either in one of the dining halls or somewhere in the square and we generally do this by RSVP first come first serve and its really just an opportunity for 20,30 40 of us to head somewhere and grab a bite together so that the form is not on line yet but starting tonight if you are free this Friday at 1:15ish p.m. which is otherwise when lecture would be if we still had it on Fridays. Go to cs50.net/rsvp and think of this as just an opportunity to chat us up, ask about the course, talk about whatever, and its sort of very low key and edible office hours. So we'll do that on occasion if you are interested. So problems Set 1 is now out. if you haven't dived in yet that's totally fine. I take great thought, I have great fondness for those who do dive in at 7:01 PM on Friday nights much like myself back in the day. But just a couple of comments on this. So last week got very arc and quickly. For those of you without prior programming backgrounds all of a sudden we had this blinking prompt and a black and white terminal and I was banging out code fairly rapidly and was fairly out of context most likely for all of you. so realize, if you walked out of last week thinking, wow, scratch just got harder, well realize once you actually sit down with this week's pdf and walk yourself through the check boxes, well you find that it actually all very quite accessible and you get familiar with the new interface soon, we'll see different interfaces over time and so let me make and emphatic point here too. Especially as you do this P-Set, if you are finding yourself a bit overwhelmed or if you're just feeling like there's a disconnect between what you heard in lecture and what your expected to do, talk to us. Talk to me, drop me a not, attend office hours, there are any number of resources and at the end of the day, please, if you're on that point thinking, damn, I really should not have signed that study card with this course on it, at least have a conversation with me. I would love to be your barrier between exiting the course altogether and just kind of sneaking away thinking this isn't for you. we can make it work for you. so what is this P-Set all about? well there's essentially three noteworthy parts in each of the addition and in the standard addition you are introduced to the candy machine that is free in the lounge in Maxwell Dworkin which you are welcome to start frequenting. We'll occasionally hold some office hours there and hang out but the challenge here is really just to get your hands dirty with Linux and with C. And what you'll find in, even though the problem is relatively simple in concept, we challenge you to implement a program that allows you, the human, to guess how many Skittles are in a candy machine. You simulate this candy machine using some pseudo-random numbers and we walk you through the process of generating pseudo-random numbers in a computers memory and then ultimately you, the human, gets to interact with your own program and for those of you who have prior backgrounds you'll realize this is really just a warm up exercise if nothing else to get you acquainted with the new environment, the cloud, with the new language, C. but then do the problems to become a little more sophisticated. so for instance, the next question asks you to take on the role of a cashier. So it turns out that in the United States and in Europe and in other countries the currency systems are designed in such a way that very naively can a cashier at a store, if they have 20s and 10s and 5s or the similarly labeled quarters, nickels and dimes, if they're all in sorted order that cashier can very mindlessly hand people change by just starting with the biggest denomination in the drawer and working his or her way on down, handing the person the amount of change they're due and this is an algorithms that's just generally thought of as greedy. In other words if you literally take the biggest bill possible, a 20 but if you owe them say 30 buck you obviously don't want to take another 20. you then take a slightly smaller bite but a big one, a ten. Do you ultimately hand that human being the fewest number of bills or the fewest number of coins. So what's nice is when greedy algorithms which are generally easy conceptually and to implement and to execute what's really nicest is when they're also optimal. But as you'll see in a moment for the hacker version of this problem set we throw a wrench into the works and you don't necessarily have all of the bills and coins that you need. So in this program here you are asked, you implement its program, that has to figure out how many quarters, how many dimes, how many nickels and how many pennies to hand someone given a finite amount of change and its going to reduce some of those basic building blocks from scratch. So some if conditions, some subtraction, maybe some variables but this too will get you familiar with this new environment and this new language and then we thought we'd try to get trendy and so we pulled up, isawyouharvard.com, the final project from CS50 last year, and we thought, wow, this is perfect. This is a graph, this happens to represent who is spotting whom on campus based on the data she's collected, and its really nice and really simple bar graph that we can plot so we're not going to actually use the same API that Tej is using, which happens to come from Google, it too is free, just like the maps we looked at a moment ago but we can simulate this using ASCII art. Using characters on the keyboard to give a decent approximation of what a bar graph is and so here will you be asked to ask the user for the number of people spotting whom, based on gender and then you'll plot this graph horizontally and this is clearly going to involve some kind of looping structure, some conditions, maybe a bit of math because we specified too how big the users screen is going to be so you're going to have to figure out the proportions accordingly. Now for those of you who decide to tackle the hacker addition of this P-Set and realized this is generally the minority of the students in the class. Anyone is comfortable, less comfortable more comfortable like is welcome to do this. It doesn't matter what you declared yourself as, you can decide week to week. We have the same warm up exercise involving some candy for you but then we hand you credit card numbers. So in the second portion of the hacker addition are you handed a definition of what it means to be a credit card number. It turns out these things have patterns and there are algorithms, procedures with which you can verify if credit card number is legit or not. And so in this part of the problem set you implement a validity checker for credit cards to actually ascertain yes or no, this is a credit card or this is just some random sequence of numbers that someone is trying to use to pay for something on the phone. Toward the tail end of the hacker addition do you implement the same idea of graphing with some ASCII art this chart but this chart we compel you to rotate it 90 degrees. You actually have to make a vertical bar chart which, if you think about how you would implement this in screen, thus far we've only seen PrinTif. Of printing to the screen and PrinTif's gotcha is that it pretty much prints line by line by line which lends itself to horizontal bars. Print a row, print a row. But if now these bars are stacked up high you're going to have to be a little more clever when it comes to figuring out, when do you draw this bar versus this bar and that will be the hacker addition this week. So those pdfs are on line and just FYI, if you were one of those like me who downloaded the pdf for hacker at 7:01 PM do redown load it because I screwed up in the middle of the night and I made the bars the wrong shape in the sample output so grab the latest pdf. So what exists in the way of support? So we've echoed this multiple times. Helpatcs50.net is manned by the CS 50 staff and our response time tend to be remarkably fast. But in the interest of economies of scale, what we also released this weekend is the courses discussion forms or bulletin boards. There is certainly redundancy in the types of questions we get because there are absolutely some frequently asked questions so on the course's home page now on the left hand side will you find the link to discussion aka help.cs50.net any time thenceforth you have a question that you think probably someone else might have and its not something that's particularly personal or particularly unique to your own code, absolutely turn to the bulletin board first for two reasons. One, to check if someone else has asked the same question, its nicely browzeable and searchable, you can save yourself sometime, get an immediate response or, two, post the question if you think it might benefit others and we spell out in the main page of the discussion forms exactly where that line is between oversharing and asking the right kind of questions. Posting a few lines of your code saying here I'm having trouble figuring this out, that's probably fine, saying here's my program, doesn't work, help, too much. The idea is to actually engage us more privately with such things hence the email address. So let me for just one moment put on my scary voice now that we're at the start of the semester. So the course takes very seriously issues of academic honesty. It is very clearly spelled out in the courses syllabus and for your convenience we include it on the second page of every single problems set throughout the semester. It explains as best we can where that line is between collaborating and working and learning and working with and learning from your classmates which is absolutely encouraged and where that line is between actually taking their ideas, taking their code and then presenting it as your own. as you'll see in the syllabus and do read this paragraph if you haven't already, the line essentially boils down to its absolutely okay to talk with your friends, talk with people in your section, stand up at a white board together, sort of brainstorming ideas, writing things down in pseudo code and tackling the problems conceptually together but where we draw the line in the syllabus is, don't go to the point of talking in exact C code. Don't look at the person's screen and certainly don't say hey do you mind sending me your code? To be honest, almost every time there are instances of academic dishonesty in this course its not because it was this fine line between oversharing, talking a little too much, it was 2 a.m. and someone decided to copy the code from their friends computer, their friend made the mistake of actually sending them their code in poor judgment and its those kinds of issues, so realize its absolutely the type of course where it's a community, you're certainly encouraged to learn from and work with each other but do mind that line and in conclusion in scary voice, I made a little animation. 

[ background noise ] So that is the number of people this course has add boarded over the past several years. recent years so this is perhaps more effective I can say verbally so realize we are computer scientists. Realize we retain the archives of all past problem sets that are submitted. Realize we have the capabilities to cross check every problem set submitted against every problem set submitted and we're very good at what we do and it gives us no pleasure to escalate things to the administrative board but we do it in fairness to your classmates who are really giving it their all, so late days exist for reason. When you're at that point at 2 a.m. when your feeling yourself stressed and you've got so much going on in your life, honestly that's when you cash in the late day, go to bed or worst case reach out to me or Jan Sorukki and we'll work something out if you're feeling that temptation that frequently happens with stress and we have these mechanisms in place so you can avoid it. so hello world was the first program that we began with, I have to make this awkward transition now to happy time again, so we began looking in quite detail at first scratch and then see and it turns out you can do a bunch of things in all sorts of languages and just so that you've realized that the ideas we're talking about now are absolutely applicable to other languages and event the tail end of the same course, what I thought I would do is just grab a couple of snippets of code, and actually let me not take for granted this work flow. So I'm on a Mac. I decide its time to connect to the cloud.cs50.net so I open on my Mac a program called terminal, I can zoom in on my font size here, I'm going to type the command as the problem set 1 walks you through cloud.cs50.net, I'm going to hit enter, type in my password and bam. Now I'm connected from my Mac to the remote server. For those of you with pc's realize the problems that actually walks you through the same process, you will see as the P-Set directs there are some how-tos which is sort of geek speak for tutorials, user manuals or what not for both terminal which I just used and If your finding this a little too fast realize we took nice screenshots and walked you through the exact same process and sequence of steps here and just so you've seen it before, the half of some of you with PCs realize that putty is a free program that you can download off the internet, it is an SSH client, it looks a little different, it doe have a more interesting guy graphical user interface, but at the end of the day once you follow this sequence of steps you end up seeing the same thing. Font might look a little different but the result is actually the same. and those of you running Linux probably know how to do this with your own computers anyway but let us know if not. Yeah? 

>> [inaudible] 

>> Why did I tell them that? Oh so why did I tell the Mac people to change your font colors and background? So when you read through this tutorial, we strongly suggest that you configure terminal or configure PuTTy in a certain way so they graphically things look the same in terms of color coding, syntax highlighting, so things look a little prettier. You do that on your Mac but those setting then sort of put a lens in front of the server through which you look at your code. So we just change the colors locally but the effect is when you look at remote things, they too are different colors. That's all. Is it necessary? No, if you prefer just looking at the more generic white screen that certainly fine and even I, in lecture today am switching to white for readability sake and its really to each his or her own, its up to you, but the structure is there. so I am on the cloud, I'm going to go ahead into source, my source directory which you have print outs of today and what I thought I'd do is open up, for instance this, which might make some of you feel more at home. So this is hi.java. so stupid name for a program but this is a canonical program in Java, some of you already know what this looks like, if you've ever wondered what Java's source code looks like, it looks like this. There's a few different key, oh sorry, I didn't print this one out because its so useless for moving forward. No, not as a language, just as an example. You'll see that the syntax kind of similar in spirit, there's a mention of main, there's parenthesis, there's semi colons, there's closed quotes but there's some other jargon there. there's the word class, there's some additional bracket syntax here so just realize at the end of the day this program too get compiled in a similar way. Instead of running gcc, someone writing this program would run Javachi.java and then what you would do is run Java on this program hi and there is in fact the same program. I think I have another one here in, let's say, hi.php, nope, that's not it, okay so here's perhaps the simplest php program you can write and we'll be using php toward the town of the semester to actually implement things like isawyouharvard.com or any web site based application you'd like to make you can kind of infer how this thing works. Echo, sounds like its pretty similar to print, the  are familiar, the, are the same. There's some new syntax there with the open bracket and then the ? but again the point of showing you these little teasers of other languages is that once you have a solid foundation in one language, maybe two languages or different types of languages as might be covered in this course and follow on courses, you can very easily bootstrap yourself to other environment. To be honest, I learned in 50, the language C, I took CS51 the next semester and learned the language called lisp and C++ at the time and that was it for my formal introduction to programming languages. Every other language I've learned since has been completely on my own and from friends and from projects and the likes so realize its very empowering just getting your foot in that door. Well what about this issue of imprecision? So let me go ahead and whip up a program like, let me take a look at this for one sec, let's do, yeah. So I'm going to go ahead and create with nano, imprecision.c this is not among your print outs because we're doing it in real time, first thing I'm going to start off with is include standardio.h and what's the point of that? Quick sanity check? It will give me PrinTif so I can actually interact with standard input the keyboard and output the screen such as through printf, I'm going to say int and we'll explain eventually why you say int, main, I'm going to say void because main at the moment does not take any arguments but we'll soon peel back that layer and I'm going to get myself ready to start coding by putting the curly braces there in advance. So now I'm going to do a couple of things. I'm going to say float, to get a floating point variable, x gets let's say .088 which just so happens to be a floating point value one of your classmates mentioned before class when we were chatting so it's a perfect example, we'll go with that one. Let's say float y= and I have a .01 and now float z is going to equal X plus y. so now to the astute, the mathematicians in the room, this value of z is now .89, at least it should be. Now let's make this more relatable to the actual piece at hand, this is like change so one of the problems in that problem set is again change making where you have quarters which might be .25, dimes which might be .10, pennies which might be .01 so surely there's an application already for floating point values so suppose I'm trying to add 88 cents and 1 cent here to get 89 cents and then display this value. Well let's do that. Let's do PrinTif % f for the floating point format string if you don't remember these as an aside one of the best resources early on if we don't mention things in the problems set at hand if you go under C in the resources page of the courses web site one of the best references here is this thing called c reference and all you have on the left hand side are links to the categories of functions that come built into C so for instance if I'm dealing with PrinTif you should probably know intuitively now it has something to do with IO right? because it standard IO so if I follow this link and then scroll down I'll see a whole bunch of functions we've not looked at yet and we're not going to bother looking at all of these, you pick them up as you go but there in fact is printf. So here is some nice documentation with all the format codes for printf, there's some example code on this website which is why we like it so this should help get you going as well if you feel that you didn't quite see the right approach in say, some class. So here we go percent F, let me do a new line F so that things don't blend onto the multiple, don't stay on the same line, ; let me go ahead and quit with control X, Y for yes, enter for save and now to compile this I've got a few options, I'm going to start getting more efficient and just say let make do it for me. figure out all the flags, figure out the -o and all that mess and hit enter. And nothing bad seems to have happened, which presumably means something good happened, so now let's go ahead and run imprecision with ./imprecision. Okay. so that actually looks pretty good. Answer looks right, I seem to have gotten some free zeros at the end of it but remember from last time that we can actually exercise a bit more discretion when it comes to seeing the places after that decimal point and I can do that by saying .and let me see 10 decimal places. That was like 6 or 7 there, let me go a little farther to the right and see what's really inside this variable Z, I'm going to go ahead and hit control X Y and enter. I'm back at the prompt and I'm going to rerun it and yet I see the same thing. Sanity check? 

>> [inaudible] 

>> You have to recompile, right. Again, it's sort of a neophyte mistake, you get used to it very soon. Let's remake this with imprecision, enter. I've recompiled my code, now if I rerun it and we move the cursor to the top, enter. Interesting. So that was kind of a lie, so when we've said before that computers only have a finite amount of memory on this a finite amount of precision which is another way of saying if you've only got a fixed number of bits, surely you cannot represent with a fixed number of bits, 32 or 64 the infinite number of real or rational numbers that exist in the world so there's got to be some approximation going on and there you have it. at 88 cents plus one cent you do not in fact get 89 cents you get just shy of 89 cents so again those of you who are fans of Office Space or Superman 3 this is the kind of stuff that they were literally referring to in that movie. These rounding errors, you start to add all that up after billions of transactions and this is how these guys start to make some money. So why is this relevant? Well suppose I add to this program a little condition. Again, just to hammer this home, if Z is equal to 0.89 because that's a reasonable check to make, let me say printf and just say equal. I'm just doing a little bit of diagnostic work here. now I'm going to say, let's say else if Z is greater than 0.89, let me go ahead and printf greater than .089 print greater than backslash N else let's go ahead and printf it should be obviously less. I could say else if Z is less than but its redundant, I can just say less because there's only three possible scenarios here but I could, to be clear, say Z less than 0.89 and we'll get the same result. So let's see what's actually going to print. Let's save, quit, remake it, rerun it and in fact it's less. so you'll see how to deal with this. Well in the problems that we walk you through the process of suggesting, well you know, if you've got floats, and the problem is floats well why not work with cents, why work with ints so we propose very simply and this does actually work out to solve this particular problem in this case, multiply those dollars, cents by 100 and what do you get back? Well $1.99 becomes 199 pennies and if you round appropriately as you'll see in the spec than it becomes 89 which is an actual int. but therein lies the reason and this isn't quite the same bug, certainly that resulted in the explosion of that rocket but its this kind of idea, not understanding what's really going on under the hood and not understanding what the machine is doing for you and writing one stupid line of code and $300 million later, you've got a problem. Now just to push back, in case you're thinking, well this is ridiculous, I should come back in 20 years once mankind's gotten better at providing me with tools to avoid these problems and we have. There are other languages, we get to them toward the tail end of the semester, but again, too many people don't understand these underlying implementation issues and that's why we continue to have bugs and security holes and more on that in just a bit. So where did the first bug come from? It was, as legend have it, a bug inside of a very old machine, one of which you can actually see in the science center, not the bug but the machine, that big Mach 1 computer, take a look if you've never, at the little touristy plaques that are on it. its quite a part of this schools and computers history but this is supposedly the first bug that was taped, with some cellophane tape by a computer scientist named Grace Hopper to a piece of paper in her lab notes documenting why the computer had malfunctioned that day. These days bugs clearly has a different meaning but derived from that spirit so let's see if we can take a look at a bug ourselves. I'm going to open up buggy1.c. This is the first of the programs you do have a print out of, so if you'd like to follow along at home there, go ahead and take a look at this, take say, 30 seconds, 60 seconds and decide for yourself and confirm with a partner what in fact the bug is with this thing here. Don't worry we won't do this every time you have a question. 

[ Music ] All right, so short program, if you didn't catch this it's okay. You'll very quickly get better at spotting these things and one of our roles at office hours is to glance at code like this is, bam, unstick you and move you along your way, what's the bug? 

>> [inaudible] 

>> Where is it? Hopefully alphabetically in the handouts. Is it not there? It should be. Buggy1.c. Okay what's the bug? Its supposed to print 10 stars but it does not because it actually prints all right, it prints 11. So this is again one of these newbie mistakes. You get used to it too and I point this out again, never to poke fun but because these are the kinds of stupid little bugs you trip over at first, you sink too much time into. So once you get past some of these frequently encountered bugs can you really start focusing on the juicier problems. So the problem here is that I started counting at zero like a good computer scientist or a computer but I then went up to and equal to 10. So if you actually think through the iterations there, because recall that in a 4 loop the initializations happen first on the left, the condition gets checked, then something happens and then the incrementation or the update happens that will, in this case actually print 11 stars. And so the fix could be a couple of things. The quickest fix and perhaps the stylistically common fix is to just do that. When you're iterating from something to something you start at zero and you go up to but less than that other value. If you really want to be more consistent with sort of normal human beings, we could absolutely count from 1 to 10 literally, but just realize that probably shouldn't start this habit because there are some lower, there's some other features of programming languages, arrays being one of them which we'll talk about this week but your just going to get in trouble if you insist on counting like normal human beings. Start counting in this way. So let's take a look at a second little example here. Let's go ahead and look at buggy2 which should follow that one. This one too is supposed to print 10 stars but this time one per line because I whipped out a backslash N but it does not. So a little sanity check? What's the bug here? 

>> [inaudible] 

>> Right. so another frequently encountered issue. Even though you typed this code out so prettily, you hit tab a couple times or the space bar to nicely indent things. It's very readable, seems consistent with the CS50 style guide which is discussed in problem set 1. Unfortunately the compiler only knows to execute one line of code after a 4 loop unless you have encapsulated it with curly braces like this and then one down here. So this would be correct if I want to print a star and then print a new line, print star new line, star new line. But let's take a look at what actually happens if I go with the original version, which is just this one. So let me go ahead and make buggy2 and then let me go ahead and run buggy 2 and in fact they're all on the same line but if I go back into nano, scroll down and obviously I could just fix it like this, but demonstrating a little point here. So I could just go in here, put that curly brace, put this curly brace, save, quit, recompile, rerun and there we go. Stupid program one star though is now per line and let me point out one thing too, because when we get questions a lot of the time on help at cs50.netinthelight very commonly we see code that clearly won't compile because you've got a ; over here, a parenthesis over here, and that's normal because it's a lot of new syntax for a lot of folks but I can't emphasize enough, before diving in or before wondering wherein lies this problem, look back to some of the examples from last week or even this week. They're very short very deliberately so that you can actually look at them, get a sense of what correctly structured code is like, look at this one and say, okay it starts with int, main then open curly brace and just make sure you're mimicking that kind of syntax and then you'll realize that there is some logical sense to this kind of structure. All right, so moving on to something more interesting involving type casting as that teaching fellow alluded to in his email to me. so typecasting is just this process of taking one data type like an int and casting it, converting it to a char or vice versa, you can do this with a few different data types, we've talked about char int like a letter A going to 65 or 65 going to the letter A. this will be useful in applications of cryptography and other things but you can even cast from a float to an int. you can take 1.99 put parenthesis int in front of that and that casts it to an actual int but what's the answer? So it's just 1. you throw away everything after the decimal point which may be what you want, in fact as you'll see in the standard edition of the problem set for this week we tell you that if you ever do have to round these bars when figuring out, sorry, their width, go ahead and round down so its not always 100 percent of the length, we tell you to round down and if you need to round down casting can do the trick but we're going to see other data types before too long. Certainly toward the tail end of the semester when we're implementing more sophisticated programs we might want to implement the notion of a course, right? The Harvard courses web site of the CS50 app, we clearly have inside our server's memory the notion of a course. The notion of an instructor. The notion of a Q guide rating. There's all these conceptual entities and yes, we could represent all of those things with just numbers and strings and the like but as we'll very soon see, you can make your own data structures. You can make your own data types so if you want to actually represent the notion of a course and a course has a name, a catalogue number and ... some other stuff. Well you can essentially join a whole bunch of variables together, call them something new. The course data type and then start manipulating that but if you choose to do that, typecasting doesn't mean you can take a course and cast it to faculty member or vice versa, it has to actually make sense so in this context of primitives where at the end of the day its just bits, doesn't make sense some time to cast one thing to another. So let's take a look at ASCII1.c This too is among your printouts toward the beginning, let me go ahead and scroll down here and I've left the comments into this one, just because it's a few more lines of code, feel free to mark it up if you like, but what does this actually do? Well this is just to hammer home the point from week zero and if you want to represent letters and sentences and words on the screen, well it turns out you can do this using just the underlying integer representations but if you cast things to a char you can very explicitly see that 65 is in fact capital A. 66 is in fact capital B and so forth and so what is this really going to do? Well here we have a 4 loop, certainly by the time you've dived into problems that one, this kind of syntax will look more familiar, even though we've started counting at zero before, it's totally legit if I want to start at a fixed value like this and iterate from 65 up to 65 +26. Now arguably this is a bit of a cheat, I just didn't want to do the math that late at night so I just did the addition here because the computer can do it but certainly we could add those together in advance an probably save ourselves some cpu cycles. What's going to happen here? Well this line of printf prints a character followed by A semicolon and a space and then one or more digits. What am I putting in there? Well I'm putting I both times. I'm plugging I into both of those placeholders but I'm first telling the compiler, you know what, treat I as a char. Treat it as a character so that what's printed is not 65;65 but hopefully A;65 and as an aside, it turns out its not strictly necessary to explicitly cast or explicitly say, open (char) here. I'm doing it really just to be deliberate and clear but the compiler or printf is really smart enough to know that its this thing over here that dictates how those bits are rendered and not necessarily this so realize that. As for the bottom half here 97 happens to represent little a, 98 little b and so forth so there are charts that show you what these mappings are but now knowing 65 is big A 97 is little a we've got 52 letters, representations that we can figure out and every other character on the keyboard has its own. so let's run this. MakeASCII1 seems to have compiled okay, let's go ahead and run it and what do I get? Well just get a big long chart showing me the so called ASCII table so we've pretty much scratched the surface of this verbally only but A is 65, B is 66, T is 84 and then sure enough, little a down here is 97. all together this is a pretty unwieldy program. If you spit out output like this its already annoying to me that I have to scroll up just other see what the program told me, so let's see if we can do a little better. Let me go ahead and open the second version of this, nanoASCII2 and scroll down and notice I've done something a little differently here. again just a little trick especially for something like problem set 1 or future P-Sets where you want to start to control the aesthetics of your program a little more, not just dump stuff to the screen like printf makes so easy, well here, same kind of iteration. But what am I doing? Well I'm printing out character, some white space digits, some more white space and then the 3 D. What does this denote here? So 3 spaces so I figured this out, frankly by trial and error or if I were more astute I'd sort of anticipated this but once I get to the lower case letters, it starts at 97 and then 98, 99 it eventually becomes three digits so just to be anal and aesthetically pretty I wanted to make sure that all those numbers take up three columns of text so that everything aligns nicely so I eventually went back in and said, all right, make this 3 D and not just D for reasons we'll see literally in a moment and then we have another character, so what's over here? well let me scroll over to the right, since the lines a little long actually, let me just scroll back, shrink the font a little bit, so what do we have here? the first place holder is going to get the character equivalent of I, that's copy paste from last time then literally I, copy paste. Well here I just did some trickery. I realized, you know what, big A and little a they're always the same distance apart, they're always 32 characters apart, 65 to 97 so let me just do this all at once, the current value of I plus 32 will take me from the upper case to the lower case range and lastly over here I'm casting to a char the result of doing that same calculation, I plus 32 so in the end, what do we have here? ASCII 2, let's go ahead and run ASCII 2 so its not perfect but my screen, my fonts. My screen's a little small, my font's a little big but now I feel like, okay at least I'm now controlling this machine a bit more. so maybe a little underwhelming in the context of today's guise and what not but again the take away for now is control and specifying to the machine what you want it to do for you. any questions? On this little ASCII exercise? Yeah? 

>> [inaudible] 

>> Oh, good question. And when in doubt we'll always try so let's scroll down here and let me get rid of what I thought was a fairly clever trick, let me save, quit, recompile, and now rerun and so you know, its not that bad, I'm being a little uptight here but that's the problem I was trying to solve. I wanted to make sure each of those numbers took up three places and not just two potentially. All right so one last variant of ASCII. Let's scroll on down and realize if you really have free time you can transform these very basic primitives into the entirety of movies. Well now notice two here, this isn't necessarily the way you should do things, but just to hammer home this point that at the end of the day all we're doing is manipulating bits. Zeros and ones and we happen to group those bits as bytes like a char is one byte, recall, an int is 4 bytes recall generally. Turns out we can even iterate over characters just to make all the more clear what, that there is this equivalent. I can iterate from the character c starting at capital A, I can do this as long as that's less than or equal to capital C and then I can do this instrumentation, again here I'm being very deliberate with all of my parenthesisation the compiler does not need all of these hints necessarily but notice let's start on the inside of the parenthesis just like you would in grade school math, focus on the deepest ones first. So here's c, I cast it to a number and then I add 1, because again the goal is to just start incrementing on each iteration, then I don't want to print a number I want to print a char so I cast c back to a char and then this line of code is the same and at the end of the day, sure you want to iterate and call it a character, that's fine, its still just a byte, its still just a number underneath the hood. Well you might recall from childhood a game called Battleship. Now I can't do justice to this game with my little command line programs yet but I can print out the board thereof. So when you start using interfaces where the user has to now see data and isn't really tolerant of just a blinking cursor and line after line of text. We can absolutely begin to implement what we would call graphical interfaces, albeit in a command line environment so if you don't, if you've never played this, this is essentially a plastic board you get as a kid, you've got numbers on the top, letters on the side and you put these little ships inside these holes that I chose to represent with the lower case zero there. what's interesting about this particular exercise is just how nicely formatted everything is and also there's some kind of looping going on here and that certainly might be a trick that proves handy in problem set 1, let's take a look for a moment at battleship.c. it turns out to print all that out there's not all that much going on but again, there's some decisions I need to make at each point so I claim with this comment up here that I'm first going to print the top row of numbers. Well first I print some white space so new line, just because I want to move it down a line just because it looked prettier to me, then indent three spaces. Again, just to keep things pretty, but then what am I doing? It looks like I'm just printing 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 on top of the little battleship board like we saw. Well then what do I do after that? Just a new line, so this white, the problem I've taken out first, it just prints the row of numbers and so this is again to echo the point from last week, especially when you're diving into P-Set 1, do not try to implement the entirely of the Skittles problem, the entirety of the change making problem, the entirety of the I saw you program because inevitably you'll have one or two or three bugs and you'll have to chase them down altogether it just adds complexity and wastes time, start by just implementing the numbers, save, compile, run, looks good and now move on to the next part. That next part is this. Print the rows of holes represented by lower case o and the letters in the left most column. So I had to do some thinking here. now I don't want to print one row, I've got to print one row than another and then another and another but in each row I kind of have to iterate this way so I have to go this way and then this way so that's just looping in two different directions so you can absolutely as you could in scratch, loop using nested loops. Using wild, do wild with 4 loops, I decided to go with 4 here and as follows. So 4 int I gets 0 up to 10 so this loop here is this controlling the cursors top down movement or left right movement would you say? so it is in fact top down. Now why is that? Well on each iteration of the outer loop just conceptually you're probably going to print as we've seen printf behave before, one line at a time so to print this and this and this that's got to happen on the inside because the only one I've printed this and this and this am I allowed to wrap around to the new line? and that's the roll being played by this outer loop so indeed this outer loop, I, represents the rows not the columns per se. now what am I doing here? I'm printing out a character place holder followed by literally a plus I so here I started to understand that I know that this cast can happen and I don't have to be so explicit any more and so this is going to print A when I is zero, B when I is 1, C when I is 2 because you're just adding 65 plus that value. Then what do I print next? Well I just need to print 10 of these holes in the board so I print a lower case o space, space to line things up and then a new line eventually, notice that its not inside the 4 loop its at the end nor is this thing inside the 4 loop that's at the end and the end result is to print this, this, this and this but before you advance to the next line you're going this, DDDD this DDDDDD and so you get this looping structure and this is very similar in spirit to a little something that you might do for instance in say, a picture like this. Why don't we go ahead and take a five minute break. All right. so we are back. Let me verbally answer two frequently just asked questions. So some of you noticed that all of a sudden I started making main look a lot more complicated than it used to be by including int, part, C char, star, art B, bracket. So the reason for this is that this week when we start action teasing apart what a string actually is. So far we've talked about a sting as being a word, hello, a sentence or what not but it ultimately it's a sequence of chars so it turns out that program like, that you write that have main functions can take strings, can take words as arguments themselves so this is a foreshadowing of exactly the syntax that we're going to start using this week and beyond, it's a little more sophisticated then just saying void but it useful when you actually want to take input from the user that moment they run the command. Case in point, you've been running GCC or soon will, you've been using nano or soon will, those programs take what we've been calling command line arguments. Nanohello.c make hello. So that word has to somehow be handed into your program and the means by which you're going to empower you to hand inputs, words, strings and the like. Add the command line to your programs is by way of these inputs to main. The thing inside the parenthesis but we'll come back to that before long. As another FAQ I suddenly seem to have stopped using double quotes and instead using single quotes, the reason for this is, again, it reduces to an issue of types. When I'm using double quotes as an argument as an input to PrinTif so that is a string, what we've been calling it. a string, recall from last week is just a sequence of chars, a sequence of characters. When you have just one character and you want to be sure that that one character is in fact being represented with just 8 bits, a single byte, must you use single quotes around it. if you don't as we'll see next week and perhaps this Wednesday, if you don't that's A, if you use double quotes it will actually be treated as a string or really a one character word, a one character sentence and that has implications for how much memory is being used and also for potentially the security of the information being stored there. 

>> [inaudible] 

>> Ah, good question. When you are printing backslash N can you use single quotes? No this needs to go inside of double quotes so I have been very particular in using my double quotes and single quotes and pretty much you should, just a safe generalization? Almost always use double quotes unless you know what you want to print is and individual char and you've hard coded it into that program as such. All right. so a little bit of a familiar if annoying song. So I needed something that had a lot of simplicity in it, just repeats ad nauseum and something ideally that starts counting downwards so you have this song the kids used to sing on the school bus for instance and what's interesting that there's a couple of changes that happen over time with each annoying verse of this song, the 99 becomes 98, the 98 becomes 97 and you count down to one or zero or whatever. But notice too inside of each verse there is a change in let's call it I itself and it doesn't happen on every loop, it actually happens inside the verse itself and then if I really want to be anal, this last line when I say, something bottles of beer on the wall, when I get down to one, I probably want to fix the English so that it just says one bottle of beer on the wall so it seems like there's an opportunity for a condition and if else there even though its only useful at the very tail end of that program. So let's go ahead and open up a little program called beer1.c first incarnation of this. I'm again using cs50's library and what's going on here in this program? So the first thing I've decided to do is ask the user for a number. All right, how do I do that? Well I literally use printf an ask them for a number and then I use getin from cs50's library and recall that get int is a function and a function is kind of this black box that might take input, it doesn't have to but might take input and might return output and if it does in fact return output that you care about you better do something with it and in this case you better store that return value, that output in a variable end so we can use it later. Now here's what we'll start calling error checking or rigorous error checking and they won't necessarily see it specified pedantically in every problem set, check for this, check for that, make sure the user can't do this. This should become natural before long. If you're writing a program that takes and form of user input it is ultimately and should be up to you to make sure that that users input cannot in any way break you're program. Its not an excuse to say, oh, well they shouldn't have inputted in a negative number and a case in point, not to over emphasize that explosion of that rocket ship you make assumptions and bad things eventually happen, especially when lots of people and lots of users become involved. So I'm doing a sanity check here. if the user has typed in an int, which is allowed by get int, but that int is 0 or negative, I don't want to deal with this user so I'm simply going to say, sorry that makes no sense and then I'm going to return 1. So this is a piece of new syntax. It turns out that when you are writing a function which we are, this thing called main, function tab can have return values, this is that point I was just making, the return value of get int is what data type presumably? So its an int and you know, I kind of have been waving my hands at it, what's the return type of main? Its been always int, even though I've not used this explicitly so it turns out that even the function called main which is the one you have to start your programs in c, even it has a return value and thus far every program I've written has had a return value, every implementation of main that I've written actually has a return to value but it returned implicitly a value of zero. So if you don't specify a return value at least in main, the value of zero is just assumed so you don't have to waste time and effort at the very bottom of your function. Now why is this useful? Well there's a lot of things that can go wrong when you write a program, we've seen visually an example of that, we've certainly seen a couple of buggy examples today of that, so the idea that a program or function can return a value is useful because its only mechanism for informing the programmer or the human, the user, that or whether or that something went wrong. So zero is good. Zero means nothing went wrong, I'm not going to tell you that nothing went wrong but unfortunately with programming and with computers can we make a heck of a lot of mistakes. There's as many as four billion mistakes exist when it comes to programming, even that is technically an over simplification. But this is good because zero means good and I can assume that zero always means success. You return zero, good stuff happens, but in case you made any number of the billion possible mistakes that someone can make, you got a billion other integers you can return so that you can signify those errors. Now I'm exaggerating, hopefully your not hand error checking for a billion different things in a program. Usually you check for one or two that can actually happen but the point here is this. I've decided arbitrarily by human convention if this thing goes wrong I'm going to return this so called sentinel value, just a special arbitrary number, the number 1 why, because its different from zero and that's really all it is. So if you've ever been using your Mac or your pc and you all of a sudden get some popup that says an error has occurred or it freezes, very often in these little dialog windows do you see a number. Maybe it positive, maybe it's negative, it generally means nothing to you but guess who it does mean something to? Right the programmer. The person at Microsoft, Apple or whoever who wrote it, so not all that useful for you unless you call tech support and tell them what you're seeing on the screen but these error codes. These return codes are generally used by the developer, the programmer to actually understand what's going wrong. So here by convention I'm returning 1. Really this program is so simple it doesn't matter what I return but again, convention, return a non-zero value. All right so let's assume that the user is not being difficult and I'm proceeding to sing this annoying song so I'm going to print a new line first, just for aesthetic reasons and it turns out this program is not that hard to implement. Let me go ahead and iterate from I equals, oh this is interesting, so even though thus far, just by habit I've been incrementing from zero on up or one on up, there is nothing stopping you from doing it in the reverse direction, in fact here, because conceptually I'm starting at 99 and going on down, I don't want to confuse myself by starting the loop at zero by printing the opposite, 99, let's just start at 99. so int I gets 99 or whatever the value of n is, I'm going to sing this song so long as I is greater than zero, maybe the song is supposed to end at zero or one, I wasn't sure so I said keep going until its zero, we definitely don't want to go negative and then on each iteration do I minus minus and this recalls a shorthand quotation. We saw plus plus even as far back as our sock changing example, but really this is shorthand notation for I gets the assignment operator, I minus 1, right? or if you do the plus plus that is equivalent to I plus one or there's yet other shorthand notation you might see in on line resources or the TF's code or in books, you can do I minus equals 1 to decrement or I plus equals one to increment, frankly this is why most people just do I plus plus or minus minus because its much more straight forward. As an aside there is also a minus minus operator and a plus plus operator. In many context they do the same thing, in some context they don't, we'll see over time when and if you should actually care but generally put it after and you'll be pretty safe. So let's take a look at this song, its got four lines, prints some number of bottles and notice here, I'm already kind of taking a shortcut, I didn't want to have to write too much code. What's my cheat here? 

>> [inaudible] 

>> Yeah bottles, like cut this corner, right? it's not wrong grammatically its just the parenthesis are only useful eventually so we'll get there right? baby steps. so the first line is print some number of bottles of beer on the wall and I'm inserting to that format string, that placeholder I, some number of bottles of beer, same thing, , I, take one down and pass it around, that's just text so notice printf again, it doesn't have to take any commas, or anything after it, now I have to do a little bit of math but not much. Some number of bottles of beer on the wall but now I want to do the decrematation. So now is there a risk that I'm actually going to say negative one at any point? 

>> [inaudible] 

>> Disagreement so push back. What's, how is this happening? Well let's see. There's a loop here, we're starting at n so we're definitely starting big but I'm checking this condition on every iteration is I greater than zero so its only going to be the case, even at the very last iteration that I has to be greater than zero, so on my very last pass through this loop I is going to be minimally. So 1, so if I is one I think I'm safe, so 1 gets printed here, because its I, 1 gets printed here because its I, 1 gets printed here, or rather, no, zero gets printed here because its I minus one but recall what happens with a 4 loop after each iteration, you don't go back to the left of this thing, its the update that happens, I becomes 0, the condition is checked and so in fact we are safe. So its close but we are in fact safe because I'm stopping so long as I is greater than or equal to 0, the error would be if I did something like this or any other number of logical quirks. So now what's one difference toward the end of this program? Well at the end I decide to exit and no because I've used a non-zero return value technically I don't need this return 0 because 0 will be assumed but as a matter of good style, if I'm going to start taking the trouble to return values explicitly let's be consistent and actually return 0 at the end. For all intents and purposes, certainly in this course, its not generally going to matter but when your programs do get larger and more sophisticated is this a very useful habit to already be in. but let's see if we can, well let's actually sing the song, so let's, well that would be awkward, but let's just do the program version. Make beer, let's run this first. So make beer one, let's go ahead and run beer one. How many bottles would there be? Well its not going to fit on the screen but let's try it. 99 and sure enough, if I scroll back up in time it's a whole lot of iteration going on, seems to check out at the top, 99 goes down to 98, 97 and at the very bottom 0 bottles of beer on the wall so not a great program, because I'm totally cutting corners with the parenthesization of the s, but its not bad. Yeah? Oh , good question. So when in doubt let's just answer it definitively. So let's go ahead and do what's pretty reasonable, maybe I'm not even quite sure early on of the logic so let's resave that, go down, sorry, that's not correct, greater than or equal to, so the syntax follows the English, greater than or equal to 0 so I argue this is going to take us one step too far. It will compile, so this is not a compilation error, this is s logical error. This is a programming conceptual mistake, not a syntax error, so let me go ahead and no rerun beer one, doesn't matter what it is, I could do 199, its still going to run fast and in fact now I see this aesthetic quirk. So I see in fact a negative 1 because percent d just means one or more digits whether positive or negative. Yeah? 

>> [inaudible] 

>> Oh, I'm sorry, okay I'm sorry, I misunderstood, so let's do that too, so let me go back and let's say, let's just rerun it with this so negative 99 beers, I should catch that. Because of my if condition, my error checking at the top of the function. 

>> [inaudible] 

>> Oh, okay. so this is where I say for all intents and purposes it doesn't matter in this course because you don't see the return values but in a week or two's time you're going to find that PrinTif as a debugging tool, a tool with which to find mistakes in programs is only so useful. When your programs get, start taking input from users and you might have a lot of variables in mind, you can't just rely on PrinTif as a useful tool so we're going to start using what's called a debugger. Something called GGDB. You may recall from last week I just mentioned terribly briefly that there's this other flag GGDB that includes some special stuff in your program when you compile it. well, long story short, we're going to get to the point very soon of not only being able to run programs you write but be able to run them in a simulation mode whereby you can step through your own code line by line by line. Interact with it, change it and in that context will be more useful to know what the return value is because you'll be able to see it in that simulator. 

>> [inaudible] 

>> So it's a good question. In the first one, it's not really necessary. Your welcome too so long as your consistent with these lessons, return 0 for instance by default but at some point soon will you need to get to that point, and if you are, in fact already in problem set 1, doing any error checking which you might very well be, check the number of Skittles that are in the machine and so forth, then absolutely. Start returning 1 then 2 then 3 to represent arbitrarily your various return values. But again, and we'll talk more about some of these details in section but let's introduce a pretty bad bug here. so let me go back into beer 1.c and let me scroll on down, and you know what? Let's actually omit this error check at the very top so let me save that, and actually let me fix the other mistake, so we're not changing too many things at once, I'm going to go back to the original but notice, I've not done my original check, save, recompile beer 1, rerun beer 1, whoops, rerun beer 1, 99, seems to work, let's try negative 99. Hm. So okay, so its not quite right, right? so in fact let's make one other mistake just to see how well we can really break the cloud here. so let's actually suppose that I goof and I do, while I is less than 0. so let's recompile, and again this bang trick I'm doing, again, exclamation point XYZ will actually get the last command that started with XYZ so let me recompile this, oh, and this is another message, if you get this and you, I just want to recompile my program, well you already did is what that message is telling you. all right. that ones come up before too. So let's rerun beer1, 99 bottles. Okay I've really broken this now let's be difficult, negative 99. so what's going to happen. Can we reason through what's going to happen? Say that's a lot of ideas. So a couple things could happen. One, the assumption you have to make is this is an int, even though I'm messing around with the input its still and int and we know and int, like any data type in a computer is going to be a finites size, probably 32 bits, maybe 64 bits but in this case, 32 bits, the biggest negative number we can represent with an int is so negative 2 billion so frankly we might be here a while if we keep waiting for this thing to go but eventually you're going to go to negative 2 billion and then you're going to keep going a little too much and what's going to happen? You're not going to overflow to another int in memory, what's going to happen if you only have 32 bits and now you're at the point of representing those, the biggest negative one possible. Its probably, something's going to change, right? if you're just flipping 0's and 1's as zero becomes a 1 and 1 becomes a 0, just like the board exercise we did and the hands being raised the other day. If you only have so many premutatons of hands in the air or 0's and 1's on the board, well eventually if you keep adding 1 your probably going to wrap around. Its wrong, but you're going to wrap around from what appears to be a negative number to what's an arbitrary positive number probably positive 2 billion and you'll get in this loop. Now this really will take a while so let's see if we can't speed up this bug, why don't we, instead of doing this, we could do minus equals how about 1000 at a time. this program is 1000 times faster now, negative 99 let's see if that's fast enough. No that's still going to take a while. Let's do, that's pretty fast, let's recompile, rerun, be difficult again, okay. I fixed it right? it stopped. So it didn't go positive it seems, I lied right? so it did become positive, right? the only condition in which this song ends and says, wow, that's annoying, is when that condition is no longer true but the condition I just changed was to be sure that I is always less than 0 so logically if this program stopped and I didn't kill it like with control c, it just stopped on its own, it said wow that's annoying, it must be that I became 0 and in fact it did. It simply wrapped around back to a positive value, but condition exited so it kind of fixed the programs execution but it didn't fix the logical error at hand. So you have the correct version, I have made a complete mess of this particular version, but the take away here is again, not only do computers have a finite amount of precision, they also have a finite amount of capabilities when it comes to storing a large or potentially large number. Let me point out one little teaser, if I may. I pulled up this web site in advance, some of you might have a ps3 or Wii or Xbox. If you have a Wii you can crack these things, there's this so called homebrew club that allows you to do fun things with your Wii and I thought I would read this and it related to this idea of computers only having so much finite memory space and we'll see more on this when we discus a raise. How the Zelda TP Exploit works. Stack, buffer over flow. During a stack buffer over flow a program writes outside of the intended program data structure in the memory address, so not the same problem we just induced, but definitely related to this idea of finite storage and maybe not checking if we're using memory that doesn't belong to us. This is used in the twilight house so this is the twilight Zelda game so for the familiar, when you're in a game, a lengthy name of Opona the horse is used to generate a stack overflow, this occurs during a conversation with a man when you load your save, when he loads the name and in order, when he loads the name to use in the dialog box or when you're entering the next zone where you have to ride a horse. This is to say, we will start Wednesday's lecture not with an explosion but with this particular gaming error. See you then. I ?? ?? ?? ?? 

==== Transcribed by Automatic Sync Technologies ====