TIP: Click on subject to list as thread! ANSI
echo: c_plusplus
to: THOMAS MAEDER
from: CAMERON CLARK
date: 1997-05-05 17:35:00
subject: Re: Rating of C++

TM> Paradigms as defined by Thomas S. Kuhn are  commonly  applied  problem
TM> solving  approaches.  I  referred  to  structured  or modular software
TM> development.
    I never learned true structured programmer; however, I learned how
    to use data flow diagrams and McCabes number to attempt to reduce
    complexity. I learned a bit about fan-in/fan-out but more along the
    lines of metrics and reducing coupling.
    I feel that a program (using OOP & keeping an eye on modularity)
    should be topdown designed and bottom up coded with integration
    testing between the first finished related classes. 
Basically*[A*[A*[A*[A*[A*A*[A*[A*[A*[A*[A*[A*[A*[A*[A*[A*[A*[B*[B*[B*[B*[B*[B*
[B*[B*[B*[B*[B*[B*[B*[B*[B*[*[B*[B*[B*[B*[A*[A*[A*[A*[A*[A*[A
*[B*[B*[B*[B*[B*[B*[B*[B*[B*[B*[B*[B*[K*[B*[B*[B*[B*[B*[B*[B/?
q
*[C**[B**[A*[A*[A*[A*[A*[A*[A*[A*[A*[A*[A*[A*[A*[A*[A*[A*[A    , write a main 
proy and then
    write each class and a test driver for each. Then integrate each
    into the main procedure piece by piece. This add for a little bit
    of integral programming - which I think is one of the best ways
    to imediately expose bugs.
TM> I wouldn't be so strict. Even now, software is  developed  succesfully
TM> without  application  of  object-orientation.  But as you, I think the
TM> object-oriented methods are the best I know.
    I feel that interface errors are too common without OOP. For me,
    each variable should have a set of rule included as source code.
    Each variable should have a list of methods allowed to alter the 
    variable. EG:
    int count - range 1-35;
        access .IncCount();
        access .DecCount();
        access .SetCount();
    If any methods assume that a atomic variable will never be outside
    a specific range, then, a method should be responsible to make sure
    that the variable adheres to the rules.
    I think that variables should also be "linked" to others.
    int window_width - linked .window_title;
    The compiler could spit out a warning if window_width is altered and
    *[A*[A*[A*[K*[D*[C*[C*[B*[B/*[B*[D*[D*[D*[D*[D*[D*[D*[D*[D*[D/
*[*[A*[A*[C*[C*[C*[C*[B*[Ch*[A*[A*[B*[B*[B*[B*[B*[B*[B*[B*[B*[B*[B*[B*[B*[B*[B
*[B[A*[A*[D*[D*[D*[D*[D*[A*[A*[A*[A*[A*[A*[A   int win_width -linked- 
.title_len;*[B[B*[B*[B*[B*[B*[A    the the windows title length should be 
altered to reflect th    window width.
    I could make a pre-processor that check the semantic 
rules*[D*[D*[D*[D*[D*[D*Ds the semantic rules
    and then strips the the non c++ code out and compiles it from there.
--- GEcho 1.00
---------------
* Origin: Digital OnLine Magazine! - (409)838-8237 (1:3811/350)

SOURCE: echomail via exec-pc

Email questions or comments to sysop@ipingthereforeiam.com
All parts of this website painstakingly hand-crafted in the U.S.A.!
IPTIA BBS/MUD/Terminal/Game Server List, © 2025 IPTIA Consulting™.