Mark-and-sweep tracing has its own weaknesses. Reference counting is
predictable and easier to understand, and so better fits the
REALbasic model.
Charles Yeomans
On Jan 5, 2007, at 5:09 PM, Daniel Stenning wrote:
Its a good point. Its one of the valid common problems with reference
counting approach - cyclic references. maybe it would be good if RS
were
able to offer one of the other tracing garbage collection methods
as an
switchable alternative in order to address cases where an RB app is
leading
to a large number of objects never being deleted due to cyclical
references.
Maybe RS could build in the open source Hans Boehm collector which
is used
in both C# and Objective C.
On 5/1/07 21:37, "Norman Palardy" <[EMAIL PROTECTED]
software.com> wrote:
On Jan 05, 2007, at 2:07 PM, Mars Saxman wrote:
On Jan 5, 2007, at 11:44 AM, Norman Palardy wrote:
NO garbage collection
When a reference count = 0 the object is deleted immediately
Reference counting is one of many possible garbage collection
strategies.
Just a an aside this is how I was taught the difference between
memory management (reference counting) and garbage collection (like
in Lisp, java, etc)
<http://en.wikipedia.org/wiki/Garbage_collection_%
28computer_science%29>
Note that the section on reference counting specifically mentions
cycles which is an issue we have with RB objects referring to each
other and the main application o longer having a reference to either
_______________________________________________
Unsubscribe or switch delivery mode:
<http://www.realsoftware.com/support/listmanager/>
Search the archives of this list here:
<http://support.realsoftware.com/listarchives/lists.html>
_______________________________________________
Unsubscribe or switch delivery mode:
<http://www.realsoftware.com/support/listmanager/>
Search the archives of this list here:
<http://support.realsoftware.com/listarchives/lists.html>
_______________________________________________
Unsubscribe or switch delivery mode:
<http://www.realsoftware.com/support/listmanager/>
Search the archives of this list here:
<http://support.realsoftware.com/listarchives/lists.html>
