[Samurai]

We have loads of non-engineers in TBHP gadget section, who know how to setup LANs without any help of BE degree. I thought these comp.sci kids would have picked up that knowledge similarly.

Originally Posted by KarthikK

So what you actually require (basic network setup) is not taught in the vtu syllabus at all for computer science engineering. Beat that!

Obviously, the BE course doesn't teach anything that is useful. I should have known that.

But there are many college professors who don't know this truth. Few years back, one MIT professor was talking to my wife at a function:

He: Now that you guys are settled so near an engineering college, your husband should consider upgrading his knowledge.
She: Huh, what does that mean?
He: He can do Master's degree.
She: My husband already has master's degree, he is an MBA.
He: No... I mean MTech in comp.sci
She: <speechless>

When she told me, I fell down from my chair. This professor who is actually my age was genuinely thinking that a professional with 20 years of industry experience can upgrade his knowledge via M.Tech in his college. What a disconnect...

[msdivy]

Originally Posted by KarthikK

I don't think we can expect all 2nd year students to know something like this, even if people are expected to try it on their own accord. Not every student might be interested in networking, not all may be able to afford a computer, and even if they could they may not have had access to multiple computers (at the same location) to try rigging them up together.

Quite true. The syllabus is so heavy on theory that it leaves little room for experimentation. . Actually if required, 1 month training from a networking institute (from Jetking or Rooman) will make anybody experts. Actually I recommend this training to any IT grad student in their semester break.

Originally Posted by Samurai

I told them I can't give access to my corporate network,

The system admin in the college wouldn't like the students to mess with the network.

Originally Posted by Samurai

All of them have laptops and have broadband connections at home. One of them has linux on her laptop. So I thought they would know that much. Anyway, I told them how setup a network.

99% of cases somebody else would have set it up. Forget fresh grads, I have fixed home broadband of 10+ years exp 'techies'. But I wouldn't expect them to know since it doesn't fall under their line of work. Some of them are experts in their area. For instance, one guy is an RF engineer and another one's hobby is photography. They invest their time in their interests, may be not in setting up home networks. This could also be true for fresh grads.

Originally Posted by KarthikK

Networking is then taught only from 6th semester and later .

One of the questions I ask in interviewing for our team is 'how to change ip address?', on Windows, Unix. Our line of work requires people to be inquisitive about these things. Not many experienced folks answer it well.

[reignofchaos]

Originally Posted by Samurai

We have loads of non-engineers in TBHP gadget section, who know how to setup LANs without any help of BE degree. I thought these comp.sci kids would have picked up that knowledge similarly.

Obviously, the BE course doesn't teach anything that is useful. I should have known that.

But there are many college professors who don't know this truth. Few years back, one MIT professor was talking to my wife at a function:

He: Now that you guys are settled so near an engineering college, your husband should consider upgrading his knowledge.
She: Huh, what does that mean?
He: He can do Master's degree.
She: My husband already has master's degree, he is an MBA.
He: No... I mean MTech in comp.sci
She: <speechless>

When she told me, I fell down from my chair. This professor who is actually my age was genuinely thinking that a professional with 20 years of industry experience can upgrade his knowledge via M.Tech in his college. What a disconnect...

I have no idea about this particular university but writing off a masters degree in comp sci as something that can be gained in 20 years of experience is incorrect. No offense to you but this is really short sighted. What a masters degree teaches you is the idea of research - it inculcates in one a spirit of deep thinking and coming out with novel ideas by applying the scientific method. This is not something one can learn in a development job no matter how many years one spends on it.

[phamilyman]

Originally Posted by Samurai

When she told me, I fell down from my chair. This professor who is actually my age was genuinely thinking that a professional with 20 years of industry experience can upgrade his knowledge via M.Tech in his college. What a disconnect...

Except you're maybe disconnected from HIS reality? Bear with me for a sec here.

You forget he gets all kind of people who have become senior through licking the rearsides of bosses, or politics or loyalty or seniority (govt) and need a *degree* for more status/claim on promotion.

I know a few such folks who took such courses and still no improvement btw

[NetfreakBombay]

Originally Posted by Samurai

I realised they didn't know how put a bunch of laptops into a network, even a wifi one. Is this not covered in the first couple years of comp.sci BE?

This is not covered and should NOT be covered.

Two reasons:
1. It is something so trivial, students should be learning on their own
2. Computer "Science" should focus on theory (Comp Arch, Set theory, Discrete Maths, Analysis of Algo....) and not on basic IT.

In third year, students have computer networks but even that course focuses on theory (In my opinion, correctly so).

[Samurai]

Originally Posted by reignofchaos

I have no idea about this particular university but writing off a masters degree in comp sci as something that can be gained in 20 years of experience is incorrect. No offense to you but this is really short sighted.

I have seen the syllabus book for M.Tech [comp.sci] for this university, so I knew exactly what I am talking about. It is geared towards providing comp.sci basics to people jumping from other branches. If you have done BE in civil, but want to switch to comp.sci, this M.Tech fits the bill. I have interviewed a few M.Tech graduates from this university and know what they are taught, and how they are taught. However, I am quite surprised about your low opinion towards industry experience. Ok, may be not so surprised, considering how most engineers end up becoming pure managers. I am still a hardcore techie, so it was stunning to me.

Originally Posted by reignofchaos

What a masters degree teaches you is the idea of research - it inculcates in one a spirit of deep thinking and coming out with novel ideas by applying the scientific method.

It is supposed to do that. But the syllabus and the quality of instruction ensures it doesn't happen. I learned all my comp.sci by self learning, still doing it, still learning new stuff every year.

Originally Posted by reignofchaos

This is not something one can learn in a development job no matter how many years one spends on it.

Depends on where you are doing it. No such hope in service companies though.

[Deep Blue]

First the disclaimers, I am not an engineer. Secondly, this isnt about employability but a query related to engineering education, so thought I will post it here.
Recently my nephew joined for MTech at VIT, Vellore. He tells me that there are about 60 students in the batch. I am not sure whether he was referring to his branch alone or across all disciplines. And that they have theory lectures in the first half of the course and the second half involves project work.
I was a little surprised because I always thought Masters programme in any field would involve much less number of students doing an indepth study of a focussed area. I was also intrigued by the concept of theory lectures in Masters. Arent graduate students capable of self study and exploring their subjects with their profs guiding them? Wouldnt these lectures hark back to our time honoured tradition of "mugging"?
I guess self study etc is reserved for Phd.

[ampere]

Originally Posted by Deep Blue

Arent graduate students capable of self study and exploring their subjects with their profs guiding them? Wouldnt these lectures hark back to our time honoured tradition of "mugging"?
I guess self study etc is reserved for Phd.

There are multiple ways of masters courses. One is the traditional extension of bachelors. In such a program you have theory subjects and labs if any. This runs for about 2 years (typically the time line/tenure of the course). They also do a project work, which can theoretical or practical depending on the scenario,

The other one is the research based masters. Here the student just goes through few selected courses, that may/can help him in his research. And the rest of the time (1.5/2/3 years) the student spends his time working on formulating and solving his research problem. If the guide thinks the problem is very good, he also can convert it to a PHD (of course that also depends on what the student wants).

[Samurai]

My first official stint as corporate trainer happened in 1992. Then I was only BE + two years of experience. I was a trainer from TCS teaching C++/OOM to newly hired staff at TISL (aka Tata-IBM), which included experienced staff as well as M.Tech trainees. That is when I realised that M.Tech trainees didn't have extra edge over experienced BE guys.

My research capabilities were ignited in 1993 when this happened:

Originally Posted by Samurai

AT&T Bell Labs then was a different world, we had lots of free time. So I used that time competing in AT&T POTM (Programmer Of The Month) brushing shoulders with giants like Doug McIlroy (he was too good), David Korn, Palith Balakrishnabati, etc. This contest was open to all AT&T staff, and we could use C, C++, ksh, awk, etc. BTW, Doug McIlroy was then the manager of certain Bjarne Stroustrup. And David Korn competed using C rather than the ksh he created. When my friend (and fellow competitor) met him and asked about it, he said he created ksh using C and there are many who could kick his butt in ksh.

These are some of the best computer scientists world has ever seen. It was a hoot to compete with them and we had to really work at it. It was very helpful few years later when I single-handedly created a custom language and a compiler, used by many of our customers even now. These were pre-XML days, so I wrote parser from scratch (didn't use lex & yacc either). And my friend (the one who spoke David Korn), went on to write a book on Compiler Design, which is now a prescribed book in many Indian universities.

[sagarpadaki]

Originally Posted by KarthikK

Are they VTU students? Networking is then taught only from 6th semester and later . The first year is anyway the common set of subjects which all branches go through, so they are into their actual CS courses only now. The third and fourth semester concentrate on basics of programming - Data structures, algorithms, graph theory, etc. Networking, DB, web programming, etc come in from 3rd year onwards. Practicals of networking was actually prescribed in the 7th semester for my batch ('08 passout).

I don't think we can expect all 2nd year students to know something like this, even if people are expected to try it on their own accord. Not every student might be interested in networking, not all may be able to afford a computer, and even if they could they may not have had access to multiple computers (at the same location) to try rigging them up together.

Originally Posted by KarthikK

I think they should bring some basic networking courses into the earlier semesters' syllabi. The way it is structured right now is a bit weird - there was a subject called Data communications in 5th sem, then computer networks in 6th sem and practicals for that in 7th sem! None of that (including lab course) taught setting up of networks. Most of the lab experiments also were dealing with packet analysis using some network simulator, no actual hardware connections. So what you actually require (basic network setup) is not taught in the vtu syllabus at all for computer science engineering. Beat that!

We learnt how to rig the computers together into a LAN on our own, but for a different reason - we wanted to play Counter strike multiplayer in the lab .

Yea, nothing much has changed even when i did my BE. I completed BE in 2011 batch(CS) . Its still the same old network packet analysis and some socket programming.

Problem is a large chunk of the syllabus formulated by the VTU does not have practical application. For example, take networks itself. Students are forced to remember many many packet transition diagrams for different protocol- Pure ALOHA, slotted ALOHA , CDMA , multiple access protocols etc. And i am sure most of the students get pissed off because of the sheer amount of diagrams they have to remember!

And the VTU paper is like a lottery. Some nincompoop question paper setter decides to snoop out a small in-congruent para from the text book and frame some Q out of it and give it for 5-8 marks! The person setting the paper does not know properly on how to test the students for the main concepts! This is the sad state evey student in VTU goes thru!

[ampere]

Originally Posted by Samurai

My first official stint as corporate trainer happened in 1992. Then I was only BE + two years of experience. I was a trainer from TCS teaching C++/OOM to newly hired staff at TISL (aka Tata-IBM), which included experienced staff as well as M.Tech trainees. That is when I realised that M.Tech trainees didn't have extra edge over experienced BE guys.

The only point is where you learn and when you learn. Its never a question of BTech or Mtech as far as learning is concerned. As you said in some MTech programs you may actually unlearn what you read in Btech and relearn the same in a better way. To that aspect entering an Mtech program may not be bad after all, esp given the state of undergraduate education in our country. (However its also sad, many seek a better post graduate program only for the brand rather than to seek a better quality of education.)

Of course it beats the genesis of a post graduate education, but still that 1-2 years of a good quality education may show its effect later to the individual. But then it truly also depends on the individual's capacity to absorb. If he/she knows what he wants, he can still get it outside the realms of a structured syllabus as well.

[Samurai]

I think people are missing the major point I am trying to make here. I am speaking very specific to comp.sci M.Tech program as it is offered today. Rest of you are talking about M.Tech in general.

Comp.Sci M.Tech is the only program which allows people from other branches to join. Is it even possible for somebody from unrelated branch to join Civil, Electrical, Electronics, or Mechanical engineering M.Tech?

Can a Civil BE grad join Electronics M.Tech? NO
Can a Electrical BE grad join Mech M.Tech? NO
Can any BE grad join comp.sci M.Tech? YES

You see the difference? Doesn't that mean comp.sci M.Tech is really watered down compared to other branches of engineering?

[Arch-Angel]

Samurai sir, i just have a basic query. For a fact, i do know hands on knowledge is valued everywhere but why isn't there appropriate emphasis given on this? In my college, even in the computer science labs for OOP/Data Structures, we were told not to question but rather memorize the stuff if we want to pass. Yes, the emphasis then was to pass and not understand. Pitiable state of affairs but then, i really did find whatever i learnt earlier to be quite useful now. This is why i am so irritated with the current system. Its just like sow the seeds, cultivate them, harvest them which ensures substandard quality.

I have a B.E from one of the good colleges in my city(i am a EEE grad) but was not able to get the CS group as i was adamant on not paying capitation fees(even though my marks were decent). Which is why i feel so lost these days...

[joe1980]

Originally Posted by Samurai

I am still a hardcore techie, so it was stunning to me.

I'm impressed with your statement. A lot of those techies in Indian MNC's end up being Managers in a short time or loose interest in doing technical stuff. Part of it is because it is demanding to be a techie. Constant change in technology and trends demands ones spare time updating his or her knowledge.

I have now 10 years of experience doing Imperative / Functional Programming and Open Source and I still feel that there is a lot for me to grasp. I love being a techie!

[mxx]

Main issues are discussed here already, like outdated syllabus, education system which is focused on marks, and no application based learning methods.

But another major problem is how people choose their career.Else where in the world people choose their career based on what interests them, what they excel in. But in India, it is about how much salary, social ego, desirability in marriage market or any combination of these. A majority has no interest in what they study. So how can we expect any sort of innovation or value to come out of them?
Sometimes I think the syllabus in our universities is made for exactly these people, all they have to do is to memorize everything a month before exam and spit it out. No need to have any interest in acquiring knowledge on the subject.