[Samurai]

Originally Posted by APJace

1. Examiners are so pathetic, they can't comprehend anything other than the rote answers.
2. Students have a disincentive against creative thought, if they want to graduate/not fail.

Originally Posted by Arch-Angel

APJace, you aren't alone there. The very same thing happened to me while i used to code programs in c++,c and try out alternative approaches. Once i had scored a zero (yes the one that reads '0') in the programming section because i had written the code on an alternative approach i found out. The justification given by the examiner was that the code should be exactly the same as per the manual or you won't get the output.

Exact same situation happened 25 years back with me. So nothing has changed.

http://www.team-bhp.com/forum/shifti...ml#post1621617

[Arch-Angel]

Originally Posted by Samurai

Exact same situation happened 25 years back with me. So nothing has changed.

http://www.team-bhp.com/forum/shifti...ml#post1621617

Sir, i really think the system will continue to languish and produce engineers who cannot think of anything other than marks or rather, let me put it this way: They cannot think out of the box. How can i make a difference? I felt the best place to start changing is yourself, adapt in such a way that you can survive, but do not give up what makes you, you. I do feel the system has taken its sweet time to get to nowhere but i do feel that if i do end up teaching people what i know, this knowledge will flow and therefore, it will help someone, atleast i hope it does. But, i am not optimistic about how things are at the moment, the future generation will have less of innovation and more of modification.

[girimajiananth]

Originally Posted by Samurai

Exact same situation happened 25 years back with me. So nothing has changed.

http://www.team-bhp.com/forum/shifti...ml#post1621617

YES , you are right , nothing has changed . You need to write answers from "prescribed" text book only otherwise you will not be awarded marks . Not sure about other universities , but this is true in VTU . The university considers local author or non prescribed text books as a taboo and writing answers from it is a big sin.

[silversteed]

During my engineering days, we were taught a common syllabus in the first year. One of the papers was on programming with C. Since I had made the extra effort to learn programming and not mug up only the topics specified in the syllabus, I always found different ways to solve the same problem. And the lecturer (probably a fresh grad from another engineering college) always found different reasons to chide me. The person in question, after seeing that I'd used pointers in a seemingly simple program involving some String operations, didn't hesitate to proclaim that my program will not work without even checking the output. I proved the claim wrong is a different story, and also managed to score the highest in that subject, despite having to get quite a few reprimands for being "oversmart" with the workstation.

***

I was asked to wait (indefinitely) for my date of joining by one of the Indian IT majors after my course (yes, I attended campus-recruitment drives for the desperate want of a job), while most of my friends joined various organizations and started discussing which "vertical" has more scope and which "horizontal" they were put into. Though I had to live with the pressure from the bank that loaned me money for my engineering course, the wait was worth it, as I managed to find a job where I could do what I studied, and what I wanted to do. The only downside I see here is that the organization I work for, gets counted among the many not-so-famous "IT" or "Software" companies in the metro cities.

Some of my friends, instead of working, decided to pursue M.Tech courses in Tier-2 and 3 colleges. I'm glad that I didn't choose that path, because some of my colleagues who have pursued M.Tech from similar institutions do the same kind of work that I do, and earn almost the same, despite having an additional qualification. Neither do they bring any value addition to the work they do, nor do they get treated any better than a regular Grad.

I've also had the misfortune (?) of interviewing candidates of the same age as mine, for a lower or identical post. The vast majority give me canned responses to technical design questions - without an iota of creative thinking going into it. Some freshmen, even lie that their academic projects are some piece of art which they did themselves, while all they would've done was to pay $$$ and 'buy' a project. It's quite easy to catch them, if you ask me!

[Samurai]

Originally Posted by Arch-Angel

Sir, i really think the system will continue to languish and produce engineers who cannot think of anything other than marks or rather, let me put it this way: They cannot think out of the box. How can i make a difference? I felt the best place to start changing is yourself, adapt in such a way that you can survive, but do not give up what makes you, you.

Well, you are right, don't change and become one of them. I didn't, this is what I did: http://www.team-bhp.com/forum/shifti...ml#post1618768

Originally Posted by silversteed

The person in question, after seeing that I'd used pointers in a seemingly simple program involving some String operations, didn't hesitate to proclaim that my program will not work without even checking the output.

For some reason most people seem to be born without the part of the brain that understands pointers. I didn't say this, but a co-founder of stackoverflow said it.

Originally Posted by silversteed

Some of my friends, instead of working, decided to pursue M.Tech courses in Tier-2 and 3 colleges. I'm glad that I didn't choose that path, because some of my colleagues who have pursued M.Tech from similar institutions do the same kind of work that I do, and earn almost the same, despite having an additional qualification. Neither do they bring any value addition to the work they do, nor do they get treated any better than a regular Grad.

I had a classmate who underwent this humiliation. He was a distinction student throughout, while I was ordinary in academics with ample extracurricular stuff. In the 80s, it meant he had a bright future unlike me. After BE, he wrote GATE exam and joined M.Tech next year. In 1993 he joined TCS as ASA trainee, only to find me there as a confirmed ASA with 3 years experience, about to leave for US trip. He was very hurt since he had always thought me to be a nobody.

Originally Posted by silversteed

The vast majority give me canned responses to technical design questions - without an iota of creative thinking going into it. Some freshmen, even lie that their academic projects are some piece of art which they did themselves, while all they would've done was to pay $$$ and 'buy' a project. It's quite easy to catch them, if you ask me!

Looks like you are new to this thread. Start from the beginning, you will find all kind of examples.

[silversteed]

Originally Posted by Samurai

For some reason most people seem to be born without the part of the brain that understands pointers.

You bet! Always wondered what that meant, until I came across so many people who found it quite difficult to comprehend and use pointers. Perhaps, the instructor who taught me C was so clear about it, that I never felt it's a rocket science

Originally Posted by Samurai

Looks like you are new to this thread. Start from the beginning, you will find all kind of examples.

I am not new to this thread, Samurai san. I merely chose to narrate my experience when I saw a few of the latest posts

[acurafan]

Originally Posted by silversteed

You bet! Always wondered what that meant, until I came across so many people who found it quite difficult to comprehend and use pointers. Perhaps, the instructor who taught me C was so clear about it, that I never felt it's a rocket science

Count me as one of these people, who was never comfortable with pointers. One of the reason that I never applied for a job that needed C coding skills Fortunately there are few tech jobs for such C-illeterate people !!

[vivekiny2k]

It should be the same for OOPS concepts. Even when I code in OO languages, I always write my code procedural. CPU always processes one instruction at a time. I can't live in an illusion that software can act like objects

[NetfreakBombay]

<nitpick>

Originally Posted by vivekiny2k

CPU always processes one instruction at a time.

Almost all modern CPUs are "superscaler" and support "SIMD".

With this, CPU can run many instructions in parallel (E.g. 3rd Gen Xeon runs 12 instructions in parallel in each core).

https://en.wikipedia.org/wiki/SIMD

This link has details. http://archive.arstechnica.com/cpu/4...jc/majc-2.html
It was hot new thing in 90s.

Originally Posted by vivekiny2k

I can't live in an illusion that software can act like objects

Even some CPUs understand OOPS. For example, Intel CPUs used in defense. HAL Tejas also uses this CPU (i960 Extended).
</nitpick>

[amitoj]

Originally Posted by vivekiny2k

It should be the same for OOPS concepts. Even when I code in OO languages, I always write my code procedural. CPU always processes one instruction at a time. I can't live in an illusion that software can act like objects

OO is more of a design approach than a coding method.

[vivekiny2k]

Originally Posted by NetfreakBombay

<nitpick>


</nitpick>

thanks for the nitpick, I know about the multi processors. that has nothing to do with simulating real world objects.

Originally Posted by amitoj

OO is more of a design approach than a coding method.

I know. Not that I have never used it, just that I prefer writing code as a program will process it.

[Samurai]

Originally Posted by vivekiny2k

It should be the same for OOPS concepts. Even when I code in OO languages, I always write my code procedural. CPU always processes one instruction at a time. I can't live in an illusion that software can act like objects

You are indeed confused about OOPS.

As Amitoj already mentioned, it is a design technique.

There are lots of people who think OOPS or OOM is fake or fad. My boss (TCS VP) in 1997 told me it was a fad, even though he was a Mainframe guy and I was a trainer for C++ and OOM. So he wanted me to get into Y2K projects.

I have to confess that I didn't understand the whole point about OOM when I started using C++ in 1991. I liked the improved syntax though. First couple years I wrote lots of bad C++ code. Mainly because C++ books then were only teaching syntax. In fact every C++ book covered C for 75% of the book and then taught the C++ syntax in the rest 25%. I very rarely used to run into other C++ programmers. Once I ran into a gentleman who wanted to know why Turbo C++ can't handle inheritance beyond 17 levels. I was deeply impressed because I hadn't gone beyond 4 levels of inheritance. That was all nonsense of course. I became a faculty for C++/OOM in late 1992 in TCS. That is when I was really forced to learn, considering I had to teach. Meanwhile, in my project I also used one OO language called Actor, which was very similar to smalltalk. That one really help clarify my ideas. I guess it took me nearly two years before I could write a decent C++ program that was actually object oriented.

Is OOM a fad? Well, consider this. Most non-software products are designed using OOM, whether it is cars, machinery or office furniture. Software took a very long time to adopt the same design principle. But it is not easy to learn or understand. I still teach it, but very few manage to understand it. But for those few who understand, it feels like poetry.

I know. Not that I have never used it, just that I prefer writing code as a program will process it.

I know what you mean. But this limits your imagination during design. I don't know whether you coded in assembly. When you move from assembly to C, it releases you from lots of bookkeeping, like not having to worry about registers, keeping track of stack, managing your own heap, etc. So you can focus on the application, and also write more flexible code, using much complex designs. One could say C enables you to write 10 times more powerful design than assembly language. Similarly, one can say C++ enables you to write 10 times more powerful design than C, provided you understand OOM.

[vivekiny2k]

Originally Posted by Samurai

You are indeed confused about OOPS.

Yup, this is how the conversation started. You are agreeing with me.

Originally Posted by Samurai

For some reason most people seem to be born without the part of the brain that understands pointers.

Originally Posted by vivekiny2k

It should be the same for OOPS concepts.

[k_nitin_r]

Originally Posted by Samurai

As Amitoj already mentioned, it is a design technique.

I'm glad someone mentioned on the thread that object-oriented programming is a design technique. There are programs written in C that use the concept of classes and inheritance, and there are programs written in C++ that do not embrace object-oriented programming.

Originally Posted by Samurai

There are lots of people who think OOPS or OOM is fake or fad. My boss (TCS VP) in 1997 told me it was a fad

Core Java programmers will tell you that object oriented development is far from being a fad, but I can imagine why someone would have said that in 1997. Today, we are seeing the adoption of aspects, a concept that was seen as a mere idea about a decade ago.

BTW, why do a lot of Indian programmers call it "OOPS" instead of "OOP"? I asked a college lecturer at IIIT-H who replied, "It is what I was taught so I continue to call it OOPS, but I have no idea what the 'S' stands for. Maybe it stands for systems."

Originally Posted by Samurai

In fact every C++ book covered C for 75% of the book and then taught the C++ syntax in the rest 25%.

This is perhaps one of the reasons why so few programmers actually use the Standard Template Library. I've always seen the last chapter of most C++ text books briefly mention the STL.

Originally Posted by Samurai

But this limits your imagination during design. I don't know whether you coded in assembly. Similarly, one can say C++ enables you to write 10 times more powerful design than C, provided you understand OOM.

To a new developer who has been taught object oriented development, everything looks like a problem that can be solved with object oriented development. For instance, I sometimes see a small low-budget one-off program being written with classes simply because that is what the developers were taught - something for which one would normally use a script instead. However, I agree that when used appropriately, one can build flexible and extensible systems much more quickly with C++ when compared to C, despite being able to accomplish the same goal with both programming languages.

[Samurai]

Originally Posted by k_nitin_r

I'm glad someone mentioned on the thread that object-oriented programming is a design technique. There are programs written in C that use the concept of classes and inheritance, and there are programs written in C++ that do not embrace object-oriented programming.

That's true. But C++ is designed for OO design, unlike C. I have seen OO code using C, the readability really sucks.

Originally Posted by k_nitin_r

Core Java programmers will tell you that object oriented development is far from being a fad, but I can imagine why someone would have said that in 1997. Today, we are seeing the adoption of aspects, a concept that was seen as a mere idea about a decade ago.

No, it was more than mere idea much before that. I was writing good OO code before Java was invented. Heard of smalltalk? I was using a Windows variant of that in 1992, it was called Actor. This really clarified my OO design concepts.

http://en.wikipedia.org/wiki/Actor_(...mming_language)

Originally Posted by k_nitin_r

BTW, why do a lot of Indian programmers call it "OOPS" instead of "OOP"? I asked a college lecturer at IIIT-H who replied, "It is what I was taught so I continue to call it OOPS, but I have no idea what the 'S' stands for. Maybe it stands for systems."

I can answer that since I was teaching that course in TCS during 92-93, it stands for Object Oriented Programming Strategies.

Originally Posted by k_nitin_r

This is perhaps one of the reasons why so few programmers actually use the Standard Template Library. I've always seen the last chapter of most C++ text books briefly mention the STL.

I am talking about books that pre-dated STL. In fact, these books didn't even address templates, let alone STL. There is a reason for that, then Borland C++ was the only compiler that supported templates. Guess in which compiler STL was developed? Yup, Borland C++. So I had started using Borland's template library even before STL was created.

Check this: http://bitsavers.trailing-edge.com/p...uide_Oct93.pdf

Then I had the rare privilege of attending C++ day (11th birthday) at AT&T Bell Labs (Holmdel building) during late 1994. There I saw Bjarne Stroustrup officially present STL to a puzzled audience, probably for the first time. The next speaker walked in and said "I wonder whether even 10% of the audience understood what Bjarne said!". I was stunned at that comment, because I was among that 10%. Only because I had used templates before that. The next speaker spoke about Design Patterns, which just flew over my head (then).

By 1995 I was using STL, and even started teaching it at TCS and HP. TCS didn't have a syllabus for it, so I made up my own. And TCS didn't have a compiler that supported STL, so I had to carry my own BC++ CD set. But every CD drive in TCS was disabled, so installing it for the class was even bigger head ache.

After doing STL for 15 years, then I delved into boost c++. Wow, what a brave new world that is... coding has become a performance art with boost c++.

Originally Posted by k_nitin_r

To a new developer who has been taught object oriented development, everything looks like a problem that can be solved with object oriented development.

That's like a guy with a hammer, everything looks like a nail. I feel programmers should be familiar with linear and structured programming, before learning OOP.

PS: Sorry about the long post on C++. Memories just came flooding...