[M3devel] language question -- locals uninitialized or zero?

[Daniel Alejandro Benavides D.]

Hi all:
if you run such a program and violate safety abstraction of uninitialized programs, you could run the program safely, but being honest means, you run the program and bypass its initialization and get the same results (which is still under first amendment argument, that is violate safety abstractions or RT).
Let's suppose you run the program on different runtime environment with different initialization values the fact is you can get different results, which makes your program implementation dependent. This is UNSAFE.
Thus if you don't have RT environment to choose from, you don't have any tool to verify that your program is safe and ignoring a safety abstraction is a violation always, because *******RT assumptions must always be truth**********.
DEC ALPHA allowed to detect use of unitialized variables  on RT, which is still better than a debugger for my own semantic model. in VAX you could get RT exception handling which is more safe, if you will to allow such program behavior in RT, but not unsafely ignoring it like you are trying to do it.

Thanks in advance
 

--- El dom, 24/2/13, Jay K <jay.krell at cornell.edu> escribió:

De: Jay K <jay.krell at cornell.edu>
Asunto: [M3devel] language question -- locals uninitialized or zero?
Para: "m3devel" <m3devel at elegosoft.com>
Fecha: domingo, 24 de febrero, 2013 03:10



PROCEDURE F1()
VAR a, b := 1;
    c: INTEGER;
    t: TEXT;
    r: TRACED REF...
    u: UNTRACED REF...
BEGIN
END F1;



Is a defined? To 1 or 0?
Is c defined? 0?
t? NIL?
r? NIL?
u? NIL?



 - Jay

 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://m3lists.elegosoft.com/pipermail/m3devel/attachments/20130224/de48aeca/attachment-0002.html>