Dealing With Teacher

Hi! I have been programming on and off for a while, so my knowledge is full of holes. That's why I decided to take Computer Science 201N at my high school. My teacher has said a few things that strike me as odd. First, he claimed iostream.h and conio.h were necessary for programming, which I know not to be true, I've compiled programs without them. Also, he uses void main(), not int main(). I want to question him as to these issues, but I don't want to alienate him, and I don't know if it's appropriate to go into such a class knowing so much about the subject and therefore am not looking to be seen as a know-it-all. How should I go about talking to him on these issues?

Also, he claims that some compilers won't work if you don't leave a space between things, ex. cout << "Blah"; In all the compiling that I've done with dev-cpp, I've never had this problem. I've simply written it as cout<<"Blah"; Also, he says that Visual C++ doesn't care either. Is this really an issue?

Like the Hitch Hikers Guide to the Galaxy, it seems your teachers knowledge of C++ can be classified as "contains much that is apocryphal, or at least wildly inaccurate".

I think the easiest thing to do is throw the C++ standard at them, and point out that everything they know so far is a crock.

void main is plain wrong, conio.h is a compiler specific header (have they ever programmed outside DOS before) and iostream.h is definitely old-school C++.
At best, you're going to learn an old version of C++ for a dead OS.

Windows programs have no need for conio, Linux programs (of any flavour) have no need of conio, Mac programs have no need of conio. Mobile phones, PDA's, DVD players, microwaves - nope, none of those need conio either.

I assume you're paying money for this, so make sure you're getting good value. Or take your custom elsewhere.

> Also, he claims that some compilers won't work if you don't leave a space between things, ex. cout << "Blah";
Only broken compilers would have a problem with this.

This is really like an English teacher teaching people to write exactly like Shakespeare would write. Interesting and informative it may be, but hardly appropriate for today's modern usage.

I think it's worth the money, seeing as I'll get three college credit-hours for $120 as opposed to whatever high price they charge nowadays. I guess I'll just have to stick it out then, and keep updating my knowledge through the internet.

Quote:

---

Originally Posted by Nero_Martin

Hi! I have been programming on and off for a while, so my knowledge is full of holes. That's why I decided to take Computer Science 201N at my high school. My teacher has said a few things that strike me as odd. First, he claimed iostream.h and conio.h were necessary for programming, which I know not to be true, I've compiled programs without them. Also, he uses void main(), not int main(). I want to question him as to these issues, but I don't want to alienate him, and I don't know if it's appropriate to go into such a class knowing so much about the subject and therefore am not looking to be seen as a know-it-all. How should I go about talking to him on these issues?

---

I wouldn't talk to him, I'd just drop the class. If your instructor is getting these fundamental things wrong, he's probably getting a LOT of other stuff wrong, some of which will not be so obvious to you. It can take a long time to recover from having your brain stuffed full of bogus or misleading information.

But if you do decide to bring it up, it certainly shouldn't make you look like a "know-it-all." Knowing that the return type of main() is int, not void, and knowing that any particular header file is NOT a requirement, is certainly not "all" a person could know -- far from it.

Send him here, we'll put him right!

Right, there are several types of teachers. Some take quite badly to "besser-wissers", others can deal with not being right all the time.

I'm one of those people who can't shut my mouth even at the best of times, so when I see a teacher/boss/etc do something wrong, I say so. Sometimes this gets me into trouble [more so than typing somethng on a forum, where if you get it wrong, you can edit or delete later on and hope that no one saw it].

You may have better luck staying after class and having a quiet word with the teacher, explaining that you think he's/she's wrong, and why this is. Be "gentle", and it may work better. Try on some simple issue which is non-controversial.

--
Mats

Also, if my experience with teachers is anything to go by don't use the phrase 'you're plain wrong, dolt' ... Show him he's wrong ...

Best case is you have a teacher who's willing to learn as well as teach. You can point out what you've learned elsewhere and why they are more acceptable and hopefully they aren't so full of themselves and their own infallibility that they will listen. Some teachers may treat their classroom as their own personal fiefdom and take any new ideas as a threat to their dominance/authority. If you have one of those then I'd say don't bother mentioning it to them. Odds are, the more they encourage participation of the students in the class the more they'd be willing to entertain new ideas. Teachers that simply stand at the front with their backs turned to the class and lecture while scribbling on the board are less likely to want to hear anything new. Just some personal thoughts on the matter.

The problem is, I think this teacher is probably wrong about most things.

Nero_Martin is wise to seek out other resources to verify everything which comes out in class.

> seeing as I'll get three college credit-hours for $120
How many are in the class?
Someone is making $$$ from all this http://en.wikipedia.org/wiki/Snake_oil

I took up computer science as a class 11 subject last year (now doing class 12)...they taught us everything in Borland Turbo C++ 4.5, with the end result that when I started working at home on new compilers, I didnt know head or tail of what to do...the *.h had been removed from headers, there was an obligation of putting a new line at the end of every program, using namespace std; was new to me, you cant put in void main any more...and when I joined a few C++ communities online, I found that Im the lowest ranking coder ever, despite being the best in my class.
And this is just school eh.

there are 12 in the class. I know, twelve misinformed people at the end of the year. I plan on seeing him alone at the end of the day monday.

For those who are interested, I did find the exact place where main() is defined as type int.
It's at http://www.csci.csusb.edu/dick/c++std/cd2/basic.html at section 3.6.1 [basic.start.main], so at least I have that in my arsenal when I talk to him. This is from the ANSI C++ standard.

Quote:

---

Also, he claims that some compilers won't work if you don't leave a space between things, ex. cout << "Blah"

---

There is a grain of truth to this though. Currently, std::vector<std::vector<int>> has to be written as std::vector<std::vector<int> >. Without the extra space, the last ">>" will be read as one token.

You will also find the C++ standard good for showing that <iostream> is standard whereas <iostream.h> and <conio.h> are not.

Well if you can't play, coach. If you can't coach, announce.

If you can't program, teach.

I would drop the class. Your 'teacher' has far too many fundamental concepts wrong.

Hey, what the heck, I'll throw in my two cents.

Avoid the class like the plague.

The other reason could be he is "teaching what he was taught" when he learned the language. So, he can only suggest concepts that he belives to be true, dispite them being flawed.

Quote:

---

Originally Posted by swgh

The other reason could be he is "teaching what he was taught" when he learned the language. So, he can only suggest concepts that he belives to be true, dispite them being flawed.

---

Yes, it is technically impossible to teach things you don't know, right? This does, of course, not make the teacher any more right.

--
Mats

The problem here is that you will be 'programming' your brain to use incorrect programming practices. It is very hard to remove that bad programming at will and so I would not mess with the class. Your best bet may be to wait until a different teacher teaches it or buy a good book on C++ programming and go from there.

I highly recommend The C++ Programming Language by Bjarne Stroustrup. I have about 7 different books on C++ and Bjarne's by far takes C++ to the next level far beyond what the other books show. Bjarne's book explains everything about the language and you can be sure that what is mentioned in the book is accurate and standardized. Some say the book is difficult to read and understand but I've had no issues with it.

There are a few minor issues with Stroustrup's book that I've noticed, but they're probably covered in a short errata list somewhere. One is that he talks about the export keyword without mentioning that very few compilers actually implement it, so in practice you end up putting whole template definitions in header files, not just the declarations. To be fair, this was written in 1997, and he apparently thought it would be implemented soon. He also says that the worst-case performance of sort() is O(N*N), which was true at the time, but shortly after the implementation was changed from quicksort to introsort which is O(N*log(N)) worst-case.

http://www.sgi.com/tech/stl/sort.html

Quote:

---

He also says that the worst-case performance of sort() is O(N*N), which was true at the time, but shortly after the implementation was changed from quicksort to introsort which is O(N*log(N)) worst-case.

---

From what I see the C++ standard only mandates a N log N comparisons in the average case, so perhaps the choice of introsort over the other O(n log n) sorts is implementation specific, and thus an O(n*n) worst case would be acceptable.

Quote:

---

Originally Posted by laserlight

From what I see the C++ standard only mandates a N log N comparisons in the average case, so perhaps the choice of introsort over the other O(n log n) sorts is implementation specific, and thus an O(n*n) worst case would be acceptable.

---

I hope this will be changed in C++09. According to the link, introsort is at least as fast as quicksort on average, so it shouldn't cost anything to make the stronger guarantee - and one can always take any fast average-case algorithm and turn it into a hybrid with something like mergesort to also get O(n log n) worst-case. I have code with several stages including a sort operation, none of the others being worse than O(n log n) worst-case, and if I can't count on the sorting to be the same, it throws a wrench into the works. There must be others in the same boat. Stroustrup does say that stable_sort() is supposed to be O(n log n) worst-case (with sufficient memory) - do you know if the current standard guarantees that?

Quote:

---

According to the link, introsort is at least as fast as quicksort on average, so it shouldn't cost anything to make the stronger guarantee - and one can always take any fast average-case algorithm and turn it into a hybrid with something like mergesort to also get O(n log n) worst-case.

---

If I remember correctly, introsort is precisely such a hybrid, being a variant of quick sort that switches to heap sort if a possible worst case is detected.

Quote:

---

I have code with several stages including a sort operation, none of the others being worse than O(n log n) worst-case, and if I can't count on the sorting to be the same, it throws a wrench into the works.

---

I suppose there is the option of make_heap() followed by sort_heap(), if you really want to guarantee average and worst case of O(n log n) regardless of whether there is sufficient additional memory.

Quote:

---

Stroustrup does say that stable_sort() is supposed to be O(n log n) worst-case (with sufficient memory) - do you know if the current standard guarantees that?

---

Yes, but then merge sort is the obvious choice for stable_sort() :)