Slashdot Mirror
Programming Assignment Guide For CS Students
kennelbound writes "For those students just getting started in a Computer Science degree or a career in software development, this guide has been written to help you understand what NOT to do when coding a project. Those with a little more experience should still read it to get a good chuckle (and hopefully the mistakes stated within will not seem too familiar!)"
I'm sure many people will say this, but you learn much more from making mistakes and working out the problems than by reading a book on common mistakes.
a HUGE thing is not to plagarize code. I was a TA for CS101 at my school, and plagarism is not only rampant, but really really easily detectable. besides, you don't learn anything; although, as one of my professors said, "if you can copy someone else's code and alter it so I can't tell, you may as well have actually done the assignment."
Maybe you should have used a debugger and stepped through the code. A good programmer knows when he's defeated and when he has to step through the code. Maybe that should be one of the rules. Use the debugger, it's there for a reason. Don't assume you're wiz and will fix the problem by just reading the code line by line. If you're a neophyte, chances are you're going to mess up existing working code.
Oh I just know I'm going to open up a huge bag of worms with this one, but this is why I vastly this sort of syntax:
It makes it easier to identify which statement the block is intended to begin with, and makes it easier to spot if there are un-intended characters between the condition and the block-opening (besides reducing vertical space wastage).
Yaz.
So let me get this straight, you didn't get that the article was a work of satire, yet this is the only part of it you felt needed to be challenged?
Mathematics is made of 50 percent formulas, 50 percent proofs, and 50 percent imagination.
Here's a VERY GOOD hint for those of you who are starting to program:
.h header file or UML diagram on paper.
THINK UML.
THINK OBJECTS.
THINK MULTI-TIER.
THINK BOTTOM-UP.
USE A NOTEBOOK.
If you start designing on paper the functions/object/interfaces/etc for your program, then start coding. As you begin to code, you'll start realizing that you'll need auxiliary functions (like an array searcher or something - most of the time lazy guys like you or me want to do everything in one function or method. Don't fall in the trap. If a series of steps is going to be very difficult, thing bottom-up and put it in a separate function or method. But before you start coding it, add it to a "to-do" list in your notebook.
That way you can keep coding your current function, by calling the not-yet written function that only exists as a declaration on paper. But the idea is there.
In the end, you'll end up with practically a completed
That helps a lot when programming (specially for low-termed memory guys like me). When you're finished designing the code, all you got to do is start typing and see which functions need to be coded, or which details . Why? Because you've already solved the problems in your code.
In one day i could design an OOP SQL wrapper (business tier) for my database project, and i only had to adjust minor details (i.e. bugs) when finished coding.
So, believe it or not, paper SAVES TIME. Trust me.
Do not slack at your math. You will repeat it. It often takes time. It is very important. Learn to utilize math and make it one of your more powerful tools.
Do not cheat on code assignments. Once again, it may take time but you need it. Messing up and looking through code more than writing it is what really makes you good.
Take hard CS classes. Take advantage of rare courses your school may offer in CS. Take tough classes like compilers or computational geometry. Make sure you take some diverse classes but also try and focus on something a bit that you enjoy.
Take more math. This is a skill that can really differentiate you from other programmers in the industry, If you have good math skills, you can get good paying, secure jobs in fields like computer graphics, physics, medical and other science fields that demand proficient math skills. It will also change the way you think if you really take it seriously and understand that much of the early math is indeed lame, but necessary to understand useful math that you will eventually learn.
Take other classes, like art. You can learn a lot from these things and apply them to what you are doing. Knowing about various things will come in handy at some point.
Learn more than what your school will teach you. It is up to you to read about things in the field, both theory and practical. Learn languages not needed in your school. Play around with things. Put together a cheap Linux computer at home and play around in it if you haven't already. You are interested in this anyways, so this shouldn't be something you have to do.
Maybe CS is not for you. The future is not guaranteed in this field as far as job security is concerned. You may spend a lot of time taking hard classes only to have to end up doing something else. You may not even make it through the program. Personally, I think there will always be a need for well educated, creative, smart people. The analytic skills you can learn will do more for you than anything else. Pay attention.
If you love it and are good at it and really spend the time in school to really learn this art, you could enjoy a career working in an industry you love. If you are ambitious, there will be many trails to be blazed in the future in this young, ever changing field. It's not about "computers". It's about computation, a modern subset of math that we can abstract in electronics. The possibilities are endless and you may invent the next big thing.
"If you are a dreamer, a wisher, a liar, A hope-er, a pray-er, a magic bean buyer
I've got to disagree - for beginning CS students, compiling frequently is an excellent thing to do. I'm a CS TA, and most of the panicky emails I get from students the night before a project is due revolve around small, simple problems that get blown way out of proportion. A single misplaced semicolon can make the compiler spit out dozens of errors for lines of code below it, none of which will make any sense. Errors propogated through code are terribly difficult to detect; your program's output might be incorrect due to something that you wrote two class files ago and have already forgotten about. And then there's the problem of poor program planning combined with infrequent test compilations - namely, the design sucks donkey balls and you just spent three hours laying down the foundation for a code base that is completely useless. I wrote code like this until I because a grad student, and it shows - I spend way less time designing and write far better programs due to multiple test compilations.
That's a pretty pretensions comment. Many people don't have access, either through finance or friends, etc to get interested in CS things like programming at an early age. Granted, while in it you ought to be interested and enthused about, else you should probably be doing something else that does. But, that's the whole point of college; to discover what interests you if you haven't already and experiment with things. Many of these "majority of CS students" are taking a Java class to see if they enjoy it, thus it is the first time they have touched a command line. Many of them will find it's not for them and drop. Some will stick around for some reason.
Besides, if you knew anything of CS you would know Java and command lines have as much to do with it as telescopes do to astronomy, to paraphrase Dijkstra. In fact, I had professors who could care less about Java or command lines because their interests are in theoretical computer science, algorithms, math, theory, etc.
Perhaps you are in the wrong major. Maybe you belong in a traade school for programming?
"If you are a dreamer, a wisher, a liar, A hope-er, a pray-er, a magic bean buyer
While some thought is, of course, necessary I have definitely seen the problem with new programming students of thinking too much.
Basically, they try to understand the whole problem fully before writing the first line of java or C. My feeling is that this is not possible for a new student. There is just too much. Rather you just have to write code at some point. Forcing yourself to try things in code is often the only way to really get started in your first programming language. (After the first one, you should be able to think as much as you want because you should have enough background not to get lost).
I have actually noticed this problem more in girl then guys. Guys tend to rush right in and try to hack it while girls try to understand it fully first. Sometimes the hacking approach is just the necessary one. (Of course this then flips in the second or third CS course where NOT fully thinking through the problem hurts more).
I'm a CIS student, about ready to graduate. I'd already been programming for several years before I started school (and never allowed my school to interfere with my education), and I've spent a lot of time helping my fellow students, and here is some advice based on what i've seen.
Learn to love whitespaces. I don't know how many times i've seen people try to cram their code down as small as possible by removing every possible whitespace. A few extra spaces can really help you catch mistakes when your using a lot of nested parenthesies. ( ( (th) ( (i)(s) ) ) is much easier to read than (((th)(i)(s))) if your trying to make sure you don't screw up your parenthesies.
DO NOT comment every line. Seriously. Comments are a good thing, but when you comment every single cin and cout, every single bracee and function call, then it can make it a lot harder to find what you are looking for. A good rule of thumb I tell people is to comment every line you have to think about for more than 30 seconds, comment every function and class, and comment every block of code that you have to spend more than 2 minutes pondering over.
Learn to use your editor. Whatever IDE or editor you decide to use, learn to use it well. Learn to use the debugger specifically, but also get used to the environment. I don't know how many people I've helped who's problem was not with their code, but with an improperly configured IDE.
READ Error messages. This sounds obvious, but I swear people don't read them, or don't think about what they could mean. I think a lot of this comes from programmign classes that teach people to memorize syntax, without giving them an understanding of what's going on at the machine level, or what the compiler is actually doing.
If you miracously fix something, understand why. Students seem like they can not resist randomly moving code around, and sometimes this does fix things. If this happens, take some time to understand what you moved and why it might have fixed the problem
Take Breaks. This one applies to everyone. I've seen a lot of good programmers go crazy over simple problems simply because they are too stressed out to think clearly. If you start to feel stressed, tired, or your mind starts to wander, then step away from the computer, have a cigarette or a cup of coffee, take a walk, and get your mind away from the problem for a bit. Even if you have a deadline, a 15 minute break can often save an hour of frustration at the computer.
Famous Last Words: "hmm...wikipedia says it's edible"
As of tomorrow I will have done 5 assignments in C++ this year. For 4 of those assignments I didn't have a compiler at my home (our program had to compile in Turbo C++, you could tell who used different compilers at home). Instead I had to go to school to debug my programs, so what I did was I'd write out my code at home in a day, go to school and then spend a few days debugging it. For 1 of my assignments (my last one) I had a compiler at home so I could debug as I go. Guess which assignment I actually enjoyed. That's right, the last one. Not knowing if my code will work makes coding unfulfilling. Spending days debugging is tedious and stressful. This just sounds plain lazy. Your code should (with as minimal effort as possible) always be compilable. If you've got a menu that calls 5 modules, write stubs for the modules. Utilise flags. This allows you to compile as you go along. Actually it's closer to "create 100 errors get 10000 error messages, try to sieve through all the messages to find the correct 100." Whereas it's much easier to go through 10 error messages to find the correct 1. You'll notice you tend to get the same 9 error messages each time, whereas having 1000 makes it a lot more difficult. And this is just syntax errors. We're not even talking about logical errors (and yes, there will be some, no-one's perfect). Actually you won't be implementing anything. You haven't implemented it until it's working. And focusing on the higher picture should be done in the PLANNING STAGE. NOT the CODING STAGE. Ideally you shouldn't even be on a computer when writing the algorithm (which is where you focus on the higher picture). You should have a pencil, a rubber and lots of pieces of paper. But even if you are on the computer it should be done in Pseudocode. I bow before your intelligence, for you can write code without debugging it at all and only have "a couple of errors." You must truly be a coding genius. That or you're talking out you're ass. I haven't met anyone who can do what you just described. Not any students and not any teachers. Either you write simple code or you're a liar and/or a troll. Because the chances of that being true is (IMO) very small.
Rudolf Diesel, the inventor of the engine, graduated from University and then went and trained as a skilled mechanic. Isaac Newton could use a lathe by the time he got to Cambridge. And Alan Turing could machine his own relays. Apart from the career options, acquiring both academic and practical skills makes you a more rounded person, and thus more employable generally.
Panurge has posted for the last time. Thanks for the positive moderations.
As far as I can tell, especially with people from an academic background (PhDs! argh!), this guide is way too advanced -- something like this might be more useful to start with:
-- Do not keep the entire project in
-- Use source control
-- Run the code before delivering it
-- NOT JUST ON YOUR OWN WORKSTATION, BUDDY!
Then, from that one could work up to 'try and ensure it is possible to install the software by some deterministic process', and only then would it be worth actually starting with actual code...
Well, they learn eventually I guess, in most cases.
Whence? Hence. Whither? Thither.
If you don't compile every few minutes, you're running your tests every few minutes.
If you're not running your tests, how do you know your code works?
If you're not compiling and testing, how are you figuring out what the code you're writing is meant to be doing?
I used to code for a day without compiling, then sit in amazement as my code compiled first time. Luckily computers are faster now and it doesn't take the better part of an hour to recompile - it takes a few seconds - long enough to pick up my coffee, take a sip and think about the next test I'm going to write.
Now that I compile more often, and run tests several times an hour, my code is immensely higher quality, I write it faster, I spend far less time sorting out compile errors and random bugs and I can safely walk away from the computer with just a couple of minutes notice without worrying about leaving a nasty mess that it'll take me a couple of hours to understand, finish and compile the next day.
In other words, thanks for the advice, I'll follow it if I want to return to where I was a decade ago.
I read his objection differently than I think even he intended it...
// end_if
// if (x%2)
}
should be
}
Or some such. The point being that you should have meaningful comments, and chances are with oodles of If's, sticking end-if after each closing brace will probably be less than informative.
paintball
I used to do that in the previous millenium, but its crap, you cant just // comment out the condition and run the code as is, you have to move the { later.
Besides it looks ugly the way you have it.
Liberty freedom are no1, not dicks in suits.
Your assertion is too absolute.
// writing pseudo-pascal sucks
Sometimes, especially in the real world, commenting what is just as important.
In the real world, a comment like this:
Is common - in class, it got me a D when the rest of the code was well commented and the code well-formed. It was in direct protest to the assignment I was given. In reality, my fellow coders would probably agree with me.
However, that's not my point. I've written blocks of code that were hard for me to understand after I put them away for a month, not because of poorly named variables or functions, but because they made use of side-effects. A great example of this is optimization. Optimization in many cases has the deliberate effect of making your code harder to read for a number of reasons.
Duff's device is a great example of an optimization that, without explanation of what, you certainly won't get how or why.
Well chosen names only go so far. I've worked on code that used sentences for the names of some of it's DB calls. And one could say that OCI has well-chosen names, but only the bravest database programmers tackle that mess.
The great irony, of course, is that at some time in your programming career you probably will come across a genuine compiler bug, and no-one will believe you...
If you disagree, post your argument. (-1, Overrated) isn't your personal censorship tool for views you don't like.