lies about OOP

[H. S. Lahman]

Responding to Daniel T....

Careful, the paper never claims that C++ produced more defects than C or
Pascal. It only claims that the defects found in the C++ program were
more costly to fix. That is a very big difference.

You're right. That's what I get for responding from memory of my
original reading of the paper. The mind is the second thing to go.


*************
There is nothing wrong with me that could
not be cured by a capful of Drano.

H. S. Lahman
(e-mail address removed)
Pathfinder Solutions -- Put MDA to Work
http://www.pathfindermda.com
blog (under constr): http://pathfinderpeople.blogs.com/hslahman
(888)-OOA-PATH

 

[Mark Nicholls]

Responding to Daniel T....


You're right. That's what I get for responding from memory of my
original reading of the paper. The mind is the second thing to go.

If it claims there are more defects in C++ than in C, then I am more than
willing to believe it, I would struggle to get a 3 line c++ program to
compile let alone run.

The leap that equates C++ to OOP, and C to SP would seem to be tenuous
though.

 

[Steve Holden]

Maybe I'm reading this wrong, but it seems like you're implying that
procedural code is inherently unmodular. Why would just one class/instance
suffice to represent a procedural program? I think of namespacing and code
organization as broad concepts that cut across all styles of programming.

Well I can't speak for Terry, but I've observed in the past that
object-oriented system complexity can grow beyond the programmer's
ability to manage it when objects sprout too many methods.

Effectively each instance's namespace becomes a set of global variables
that the methods use to intercommunicate. This style of OO programming
sometimes drives me nuts (IIRC it was the library modules based on
SocketServer that drove me to this conclusion).

Interestingly, Python's main unit of modularity is the "module", not the
"class". A direct translation of a typical Python module to Java would look
something like a Java class containing only static methods and inner classes.
Even this would be considered rebellious within the context of ordinary Java
programming.

Well Java got some things right, but one of the things it got wrongest
was the absence of functions and procedures and the consequent need to
provide class methods. Though for the life of me I don't *really* know
why I find it acceptable in Python to have to write module.func(args)
when I dislike having to write class.meth(args) in Java.

perhaps-i'm-just-old-and-crotchety-ly y'rs - steve

 

[Steve Holden]

Obviously -sen is the only real suffix, as it's -sen in Dutch as well,
as in Faas-sen.



That's not the Swedes, it's the Finnish that did that. They typically
don't like being mistaken for Swedes.

That's because their language is derived from Serbo-Croat. But both the
Finns and the Swedes will tell you it's the Norwegians who are
alcoholics. They're all dreadful liars, these Scandinavians ;-)

regards
Steve

 

[Steve Holden]

You wish! If only ending a thread were that easy - we wouldn't be hearing
about "Why is Python slower than Java?" anymore!

But I agree with Martijn (note spelling) - I didn't notice any flames in the
thread, it seemed civil enough to me.

-- Paul

I suspect Jive just got bored with the thread. Easy to do when you are
seeing the same arguments from the same people for the umpteenth time,
but when that happens I try to remember there are people joining the
group all the time, so even my rantings are new to someone.

All the same, I'd like to thank Jive for his indirect responsibility for
triggering Martijn (who, by the way, SUX) to produce his last outburst.
Very few postings do actually make me laugh out loud, but that did.

regards
Steve

 

[Peter Hansen]

They're all dreadful liars, these Scandinavians ;-)

So, you're saying they are not very good at lying?

(And be careful what you say about Scandinavians. I may
have to bring my axe to the next PyCon... "You! Where
is your payment voucher! Thwack! Everybody line up
here and deposit your wallets, watches, and villages
in this basket and you will be allowed to live!")

-Peter

P.S.: I'm only half Danish, but the other half is from
a particularly bloodthirsty line of Canadians.

 

[Peter Hansen]

... I try to remember there are people joining the
group all the time, so even my rantings are new to someone.

You wish, Steve. ;-)

All the same, I'd like to thank Jive for his indirect responsibility for
triggering Martijn (who, by the way, SUX) to produce his last outburst.
Very few postings do actually make me laugh out loud, but that did.

Agreed. A good belly laugh is a good thing, provided the milk
dripping out of your nose can avoid the keyboard.

 

[Jon Perez]

A co-worker considers himself "old school" in that he hasn't seen the
light of OOP.(It might be because he's in love with Perl...but that's
another story.) He thinks that OOP has more overhead and is slower than
programs written the procedural way.

The problem with OOP is not overhead in the sense of slower programs.
It's overhead in terms of complexity. OOP frameworks also make doing
simple things require way too much effort and study.

Inheritance is a very inflexible way of organizing code - it makes it
impossible/extremely difficult/expensive to change the design of base
classes. Hence, the recent popularity of interfaces as an alternative.

The thing I love about Python is how it manages to hide its OOP-ness
most of the time compared to languages like Java and C++. Python
concepts like 'protocols' greatly lessen the need to deal with
awkward OOP ideas.

 

[Roel Schroeven]

This, of course, means that DJIJSTRA (there, one of the J's is right)
SUX!!!!!!!

And so do DONALD NUTH, BRIAN ERNIGHAN, ROB PIE, EN THOMPSON, JOHN
CARMAC, FRED BROOS, NIKAUS WIRTH, ONRAD ZUSE AND STEVE WOZNIA!!

 

[Terry Reedy]

provide class methods. Though for the life of me I don't *really* know
why I find it acceptable in Python to have to write module.func(args)
when I dislike having to write class.meth(args) in Java.

I have the same reaction and I strongly suspect it is because the concept
of OOP class (of instances) has more mental baggage than module as
convenient name-grouping.

Terry J. Reedy

 

[Jeff Shannon]

P.S.: I'm only half Danish, but the other half is from
a particularly bloodthirsty line of Canadians.

I thought it was physically impossible for Canadians to be bloodthirsty
outside of hockey games...

Jeff Shannon
Technician/Programmer
Credit International

 

[Adam DePrince]

I thought it was physically impossible for Canadians to be bloodthirsty
outside of hockey games...

When I read that I parsed it as reading "I'm Danish; my spouse plays a
mean game of hockey."

Adam DePrince

 

[Andy Salnikov]

I think that costs(=time) to develop and maintain software depends not
on wheter it is based on OOP or not but on two factors:

* Number of NEW Code lines to solve the given problem
* Complexity of this new code

Complexity of the code in turn may depend on too many factors.
For me OOP is good because it gives me the means to manage the
complexities. Remember that OO is not only about objects and
things like encapsulation, important part of OO analysis and design
is making dependencies "right" which in turn reduces complexity.

The anwser to the question if OOP is better is: it depends

If the given problem is solved with less code and complexity in OOP then
it is the better approach if not the reverse is true.

For any given problem there could an infinite set of solutions
expressed in any methodology. If you take best non-OO and compare
it with he worst OO, the winner is pre-determined And then
chosing a winner is always subjective, we are all people When
programs will be written by robots they will chose objective
criteria which will tell if OO is better for particular task.
But then we won't need to disscuss itin c.l.py

Thats why I like python because it does not force to use OOP or
procedural programming.

But isnt the main argument to code in Python (or other high level
languages) easy of use and compact code?

Therefore should Python code be less buggy and cheaper to develop and
maintain. Are there any papers on that?

Compact is not a synonim for less complex. It just means that you
can express a solution in fewer lines of code. COmplexity of the code
will remain the same as in any other language. And there is no bug-free
code, there is well-tested code.

Andy.

 

[Alex Stapleton]

To canadians there is no "outside" of hockey games.

 

[Peter Hansen]

To canadians there is no "outside" of hockey games.

Some Canadians aren't so fanatical about hockey, or any sport.

For example, I've still never figured out how "conversions" work...
or switch-hitters.

 

[Aahz]

P.S.: I'm only half Danish, but the other half is from a particularly
bloodthirsty line of Canadians.

Oh, you're part Quebecois?

 

[not [quite] more i squared]

Imagine, for instance, what if he wants to egosurf, google for his own
name and finds nothing because everybody was saying Djikstra all the
time? That'd be terrible!

Fortunately, not in our time stream :

Dijkstra - 892 000 hits
Djikstra - 5 500 hits

"Edsger Dijkstra" - 25 500
"E. Dijkstra" - 8 080
"E. Djikstra" - 248
"Edsger Djikstra" - 36
"Edgar Djikstra" - 10

 

[Paul Foley]

That's because their language is derived from Serbo-Croat.

No it isn't.

[It's distantly related to Hungarian, if that's what you were
thinking of. Hardly "derived from", though]

 

[Fredrik Lundh]

No it isn't.

time to tune your "absurd humour" sensor somewhat slightly? I thought
the next sentence was a pretty obvious giveaway:

"But both the Finns and the Swedes will tell you it's the
Norwegians who are alcoholics."

</F>

 

[Steve Holden]

time to tune your "absurd humour" sensor somewhat slightly? I thought
the next sentence was a pretty obvious giveaway:

"But both the Finns and the Swedes will tell you it's the
Norwegians who are alcoholics."

;-)