Re: Cobol data protection? Get a dog...

A hard working student,.

[EMAIL PROTECTED]  (KELLIE FITTON)
Date: 10/1/03 12:08 PM EST
Message-id: <[EMAIL PROTECTED]>


Writes:

<<
Hello everyone,

according to the cobol book that I am reading, some Submodules that are called
in the computer's open environments memory that have "active components" or
"Java applets" can overwrite the working areas of my programs or even clobber
the global variables. Is there a way to protect my cobol's working area? should
I use win32 APIs to LOCK the memory heaps? should I get a german shepherd?

thanks a lot for your kind help, Kellie.

>>



As you know by default, COBOL 
passes values BY REFERENCE. USING clauses can
be coded to achieve passing BY VALUE, as well 
(that is fairly recent in COBOL). 


You may wish to take note of 



Standford Artificial Intelligence Laboratory,
October 1973, Memo AIM 224
Computer Science Department Report No CS-403

"Hints on Programming Language Design"
By C. A. R. Hoare


"Section. 8 Variables"

(snips)

An assignment to x
"... if x is a reference variable may change any
other variable (of appropriate type) in the whole machine. ""

(snips, then at the end of the section)


"References are like jumps, leading wildly from one part of a data
structure to another. Their introduction into high level languages has
been a step backward from which we may never recover."


The early languages of interest to Hoare had parameter passing
by value, by reference and a thing they called BY NAME. Phrases similar
to BY NAME can mean different things.

I am a student of ALGOL and not an experienced practitioner. My summary
of an ALGOL parameter passed BY NAME is that it is a two level reference.

The problem this solved was that things could move, so if you had 
a mere reference, the thing you want might not be there anymore. 

Imagine shopping carts at an online bookstore, either your cart has 
moved to a new 'service station', or the cart at the station of interest
is the next cart instantiation. 

They did try to achieve what they called pseudo parallel systems with
the work leading from ALGOL to SIMULA.

So the ALGOLites used a pass BY NAME feature to pass you a reliable 
indication of where yet another pointer was. And you would have to double
dereference you pointer to pick up the real intended data.

Hoare thought even less of BY NAME than of BY REFERENCE.  There is 
another factor of early concepts that shows up in the cited paper. 
Early work in some cases tried to make subroutines as pure functions. 
This seems esoteric at first, a silly, boring distinction that is 
now actually rare and seems not related to COBOL or business.

Yet it is really quite important if you want to understand the departure
that was occuring that caused Hoare so much concern. (That departure
now has arrived at a net that connects very untrustworthy individuals
to business data storage).

Some uses of the term 'function' are very restricted. The word can mean 
something like a sub program that we call. But the code is not allowed
to have any 'side effects.'  For the sake of introducing the idea and
relating it to german shepherds. A side effect is simply any permanent 
record of the execution of the code in the sub module, like a move of
a value to memory. So you will occassionaly encounter this use of the term
function.
The only thing that happens in a 'function' is that it returns its 
value.

This is still a very power theme in some development environments. The gist
of this thing is that they wanted to extend mathematical formulas beyond the
simple arithmetic type operators. Functions were engaged to do calcs, these 
were not so much means to structure a system by modularizing it.

So Hoare's view of the 'interface' into a submodule is somewhat influenced
by a residual orientation to that kind of purpose.

>From that earlier perspective, miscreant active attachments that interrogate
your Windows mailing lists and audaciously emit flurries of wormy cyberbytes 
out onto the net seem gastly indeed. Worms are geometrically expanding
side effects. Gastly, but to Hoare, maybe not such a surprise.


It is noteworthy that the title page to Hoare's published notes
indicate

Supported By
Advanced Research Projects Agency
ARPT Order No 24y4
Project CODE 3d30

Maybe the title of the quoted section  should have been
"References Considered Harmful."

Best Wishes
Bob Rayhawk
[EMAIL PROTECTED]