What is up with all of the group projects in graduate CS programs today?

What is up with all of the group projects in graduate CS programs today?

Post by no.on » Thu, 13 Feb 2003 06:37:15



I returned to school, part-time, to earn my MSCS this year after being out
of the university loop for the better part of ten years, and cannot get over
the number of group projects that are assigned in each class.   It is not
like these projects are so hard that they cannot be completed individually
by a properly prepared graduate student (i.e., one that has completed either
an undergraduate program in computer science or the non-CS graduate
prerequisites).    I know a few people that completed their MSCS from the
same school a few years back, and the only group projects they had were in
courses that emphasized "software development in the large" (i.e., software
engineering courses).

The one thing that I find truly amazing about the program is that something
like seventy-percent of the students are foreign born.   Most of these
foreign-born students do not hold undergraduate degrees in computer science,
and few speak English clearly (the bulk are from China with the remaining
fraction being mostly from India).  Not only do I have to decipher Chinese
and Indian English, I end up having to teach these people fundamentals that
they should have mastered before enrolling in the class.  I get a headache
every time I get assigned to a group in school!

Where have all the native-born computer scientists gone?

 
 
 

What is up with all of the group projects in graduate CS programs today?

Post by Scott Hightowe » Fri, 14 Feb 2003 02:25:41



> I returned to school, part-time, to earn my MSCS this year after being out
> of the university loop for the better part of ten years, and cannot get over
> the number of group projects that are assigned in each class.   It is not
> like these projects are so hard that they cannot be completed individually
> by a properly prepared graduate student (i.e., one that has completed either
> an undergraduate program in computer science or the non-CS graduate
> prerequisites).    I know a few people that completed their MSCS from the
> same school a few years back, and the only group projects they had were in
> courses that emphasized "software development in the large" (i.e., software
> engineering courses).

> The one thing that I find truly amazing about the program is that something
> like seventy-percent of the students are foreign born.   Most of these
> foreign-born students do not hold undergraduate degrees in computer science,
> and few speak English clearly (the bulk are from China with the remaining
> fraction being mostly from India).  Not only do I have to decipher Chinese
> and Indian English, I end up having to teach these people fundamentals that
> they should have mastered before enrolling in the class.  I get a headache
> every time I get assigned to a group in school!

> Where have all the native-born computer scientists gone?

<bitterness>We're unemployed and can't afford to go back to school.</bitterness>

Maybe the thinking is the group projects will be more like the real world.  You
know, a bunch of recent grads (i.e. no experience) sitting around trading
ignorances about how to implement the project.  Meanwhile, any senior people
that are still around have been promoted to managers, and they're too busy
filling out paperwork and attending meetings to share their hard earned
knowledge.

I'm just waiting until it all blows up and "they" realize they need the senior
people back to clean up the messes.  (Can you spell "consultant fees?")

 
 
 

What is up with all of the group projects in graduate CS programs today?

Post by jstp » Fri, 14 Feb 2003 04:27:52


They've all gone to Business Schools? (I would actually be curious to know)


Quote:> I returned to school, part-time, to earn my MSCS this year after being out
> of the university loop for the better part of ten years, and cannot get
over
> the number of group projects that are assigned in each class.   It is not
> like these projects are so hard that they cannot be completed individually
> by a properly prepared graduate student (i.e., one that has completed
either
> an undergraduate program in computer science or the non-CS graduate
> prerequisites).    I know a few people that completed their MSCS from the
> same school a few years back, and the only group projects they had were in
> courses that emphasized "software development in the large" (i.e.,
software
> engineering courses).

> The one thing that I find truly amazing about the program is that
something
> like seventy-percent of the students are foreign born.   Most of these
> foreign-born students do not hold undergraduate degrees in computer
science,
> and few speak English clearly (the bulk are from China with the remaining
> fraction being mostly from India).  Not only do I have to decipher Chinese
> and Indian English, I end up having to teach these people fundamentals
that
> they should have mastered before enrolling in the class.  I get a headache
> every time I get assigned to a group in school!

> Where have all the native-born computer scientists gone?

 
 
 

What is up with all of the group projects in graduate CS programs today?

Post by Randall S. Becke » Fri, 14 Feb 2003 07:24:49



|
|
|
| > I returned to school, part-time, to earn my MSCS this year after being
out
| > of the university loop for the better part of ten years, and cannot get
over
| > the number of group projects that are assigned in each class.   It is
not
| > like these projects are so hard that they cannot be completed
individually
| > by a properly prepared graduate student (i.e., one that has completed
either
| > an undergraduate program in computer science or the non-CS graduate
| > prerequisites).    I know a few people that completed their MSCS from
the
| > same school a few years back, and the only group projects they had were
in
| > courses that emphasized "software development in the large" (i.e.,
software
| > engineering courses).
| >
| > The one thing that I find truly amazing about the program is that
something
| > like seventy-percent of the students are foreign born.   Most of these
| > foreign-born students do not hold undergraduate degrees in computer
science,
| > and few speak English clearly (the bulk are from China with the
remaining
| > fraction being mostly from India).  Not only do I have to decipher
Chinese
| > and Indian English, I end up having to teach these people fundamentals
that
| > they should have mastered before enrolling in the class.  I get a
headache
| > every time I get assigned to a group in school!
| >
| > Where have all the native-born computer scientists gone?
|
| <bitterness>We're unemployed and can't afford to go back to
school.</bitterness>
|
| Maybe the thinking is the group projects will be more like the real world.
You
| know, a bunch of recent grads (i.e. no experience) sitting around trading
| ignorances about how to implement the project.  Meanwhile, any senior
people
| that are still around have been promoted to managers, and they're too busy
| filling out paperwork and attending meetings to share their hard earned
| knowledge.
|
| I'm just waiting until it all blows up and "they" realize they need the
senior
| people back to clean up the messes.  (Can you spell "consultant fees?")

Or <wicked grin>, just do the work for them, so they'll learn nothing and
take only that back with them ;-)

20 <cough> years ago, the only group project I had to content with was in
the Compiler Design course - and let me tell you, it wasn't something to do
alone. However, I partnered with two foreigners, both of whom were extremely
capable, and the mathematics of formal languages was what was important.
*WE* completed where all others failed, mind you. So I can't say the
experiment was entirely successful.

But a group thesis? Wait a minute.

 
 
 

What is up with all of the group projects in graduate CS programs today?

Post by Bill Burri » Fri, 14 Feb 2003 08:58:36


What are they teaching these days?

From the books I have been reading, it looks like the industry is about to
move on to the next stage beyond OO.  The emphasis will shift towards Domain
Analysis, Frameworks, Program Generators, & Components.  The industry is
getting serious about reuse.

Bill


Quote:> I returned to school, part-time, to earn my MSCS this year after being out
> of the university loop for the better part of ten years, and cannot get
over
> the number of group projects that are assigned in each class.

 
 
 

What is up with all of the group projects in graduate CS programs today?

Post by Liz » Fri, 14 Feb 2003 09:31:21



Quote:> What are they teaching these days?

> From the books I have been reading, it looks like the industry is about to
> move on to the next stage beyond OO.  The emphasis will shift towards
Domain
> Analysis, Frameworks, Program Generators, & Components.  The industry is
> getting serious about reuse.

think we're already in that stage, no ?  what are the copyright dates on
your books ?
 
 
 

What is up with all of the group projects in graduate CS programs today?

Post by Randall S. Becke » Fri, 14 Feb 2003 10:18:15



|


| > What are they teaching these days?
| >
| > From the books I have been reading, it looks like the industry is about
to
| > move on to the next stage beyond OO.  The emphasis will shift towards
| Domain
| > Analysis, Frameworks, Program Generators, & Components.  The industry is
| > getting serious about reuse.
|
| think we're already in that stage, no ?  what are the copyright dates on
| your books ?

We were at that stage before OO. This is early 1980's stuff. OO came later.
The key now is to start applying bigger architectural concepts to OO.

 
 
 

What is up with all of the group projects in graduate CS programs today?

Post by Bill Burri » Fri, 14 Feb 2003 15:18:33


I studied Electronics Engineering Technology back in the 70s.  I became a
software developer by reading books and banging away at the keyboard.  The
few SE books I read back in the 80s didn't make any sense to me as a 1
person development team.  I started reading about OO back in the 80s.  The
idea behind OOP made sense to me and I have been using C++ since 1990, have
tried Java, & now use C#.  I never did catch on to doing OOA & OOD, the
books and techniques are not designed to be of much use to the individual
programmer.  My approach has always been to sit down at the keyboard and
start witting code.  At the start of a project there was always too many
unknowns to make any headway doing analysis & design.  The only was to make
progress was to write experimental code.

Recently I read "Program Generators With XML & Java" by J. Craig Cleaveland.
Now I am reading "Generative Programming" by Czarnecki & Eisenecker.
Another book I have been looking at on books24x7, is "Practical Software
Reuse" by Donald J. Reifer.

The impression I got was that bits and pieces of Software Reuse techniques
have been around for a long time, but a complete and standardized
methodology is yet to appear.  The Generative Programming book talks about
attempts to extend UML to integrate Reuse & OO techniques.

What I am doing now is thinking about how to use Domain Analysis,
Frameworks, Program Generators, & Patterns.  My plan is to pick a simple
component (maybe histograms), and do a feature analysis.  I will then create
the Active Library (feature DB, Program Generator, & Test code).  I will do
as many iterations as it takes to come up with a working process.  The next
stage will be to pick a slightly larger domain and start over.

Does anyone have suggestions for what books I should be reading next.

I have discovered a number of books about Architecture, Domain Analysis,
building Frameworks, & Patterns, but haven't decided which ones to purchase.
The data acquisition systems I develop, typically involve concurrency &
real-time.

There is the question of modeling tools.  I have Visio, and should be able
to learn how to create add-ons in a reasonable amount of time.  Then there
are a few open source tools.  A couple I have identified but haven't tried
yet are:
Toolkit for Conceptual Modeling   http://wwwhome.cs.utwente.nl/~tcm/
&
ArgoUML   http://argouml.tigris.org/

Two books that I have just purchased but haven't read yet are:
Modern C++ Design by Andrei Alexandrescu
&
C++ Templates by Vandevoorde & Josuttis

I ordered these 2 books with the idea that they would help modernize my C++
skills.  C++ is a different animal then it was back in 1990 when I started
using it.

Bill
--
http://www.componentsnotebook.com/





> |


> | > What are they teaching these days?
> | >
> | > From the books I have been reading, it looks like the industry is
about
> to
> | > move on to the next stage beyond OO.  The emphasis will shift towards
> | Domain
> | > Analysis, Frameworks, Program Generators, & Components.  The industry
is
> | > getting serious about reuse.
> |
> | think we're already in that stage, no ?  what are the copyright dates on
> | your books ?

> We were at that stage before OO. This is early 1980's stuff. OO came
later.
> The key now is to start applying bigger architectural concepts to OO.

 
 
 

What is up with all of the group projects in graduate CS programs today?

Post by Phli » Fri, 14 Feb 2003 15:28:54



> Does anyone have suggestions for what books I should be reading next.

Refactoring
Test Driven Development
Agile Software Development: Principles, Patterns & Practices

They are the right way to "sit down and start writing code" AND "analyze
until every possible axis of re-use appears."

Quote:> Modern C++ Design by Andrei Alexandrescu
> &
> C++ Templates by Vandevoorde & Josuttis

Forgive me if you already have such elemental works as:

Effective C++
Large Scale C++ Software Design
Design Patterns

--
  Phlip
       http://www.greencheese.org/LucidScheming
  --  All sensors report Patti having a very good time  --

 
 
 

What is up with all of the group projects in graduate CS programs today?

Post by Jens Christian Larse » Fri, 14 Feb 2003 18:39:23


I believe it is less expensive to assign students to groups rather than
having them do individual projects. This might be part of the reason.
Furthermore, working in project groups gives valuable experience that can
be applied to your professional life afterwards.

It is unfortunate that you are stuck with students that are less than
good. During my education as MSSE I have had encounters with students from
India and China and I was amazed by their lack of knowledge. For some
reason, the educational system held their hands over them, and I didn't
even see one of them failing to get their degree. Thats sad.
After I got my degree I started working in a company what had employed a
chinese graduate from my university. He was fired again within 6
month.

Jens-Christian

 
 
 

What is up with all of the group projects in graduate CS programs today?

Post by no.on » Sat, 15 Feb 2003 01:20:05


Actually,  Bill is right on the money about the current trend in research!
The really negative side effect of OO has been code bloat.  OO systems are
overly complex and bloated.   The level of abstraction is far too low, and
the target languages are too general in nature.   OO has been the largest
contributor to skyrocking software development costs (not to mention the
fact that properly-prepared OO developers are really expensive here in the
U.S.;  thus, the push to using offshore labor sources).  We are going to see
a move back to domain-specific languages.   However, this time around, they
will be functional, not procedural languages.   These languages will make it
much easier for designers to formally-prove their code; thus, hopefully,
making extensive regression testing a thing of the past.


Quote:

> Recently I read "Program Generators With XML & Java" by J. Craig
Cleaveland.
> Now I am reading "Generative Programming" by Czarnecki & Eisenecker.
> Another book I have been looking at on books24x7, is "Practical Software
> Reuse" by Donald J. Reifer.

> The impression I got was that bits and pieces of Software Reuse techniques
> have been around for a long time, but a complete and standardized
> methodology is yet to appear.  The Generative Programming book talks about
> attempts to extend UML to integrate Reuse & OO techniques.

 
 
 

What is up with all of the group projects in graduate CS programs today?

Post by Liz » Sat, 15 Feb 2003 02:46:54



Quote:> Actually,  Bill is right on the money about the current trend in research!
> The really negative side effect of OO has been code bloat.  OO systems are
> overly complex and bloated.

gee, a critic of OO ... so, maybe I won't bear wearing object oriented shoes
in a few years after all ..

but why do you see OO per se as causing code bloat ?

 
 
 

1. CS Graduate Programs

Looking for recommendations for Computer Science graduate programs
with strong Artificial Intelligence program.  Would also like
the following attributes:

        - a more "flexible" program, one that allows the students
          to make a lot of choices, rather than taking MANY
          prescribed courses

        - in the "Midwest" (Missouri, Kansas, Illinois, Iowa, Nebraska,
          or maybe even Colorado)

        - the person I'm posting for is also studying psychology - so
          any program that actually allows one to work with another
          department would also be good

Reply to druhi!rmm, (303)538-4899 or

        Rose M. Marra
        12110 N Pecos
        Rm 8Z106
        Denver CO 80234

Would appreciate any and all information . . . thanks in advance!

2. Backing up to a cd

3. SoftEng or CS Graduate Program Search

4. Memory Upgrade Problem - Fatal Exception

5. Good CS graduate schools?

6. Wanted: Source for Telephonics Info

7. Please share your opinions about graduate studies in CS/CE

8. Disassemblers

9. Good CS graduate schools?

10. Free check-ins or stable check-ins?

11. H-1B PROGRAM REPLACES PERMANENT IMMIGRATION: INS/IEEE-USA DATA

12. UPS - UPS 6/99

13. Project Write-ups