Yes, I'd think that in this particular case it is rather DLP1138 related (which is generic fix). Brendan with his BJB266 fixed more specific case with 3rd class involved.
BTW, these both changes are in 5.0 (BJB266 since 5.0.3, and DLP1138 since 5.0.7)


Regards,
Timur Safin

On 12.05.2004 14:41, Wolf Koelling said the following:

Or possibly DLP1138 in 5.0.7?
(http://www.intersystems.com/cache/technology/product-tables/releasenotes/50
7/index.html#DLP1138)

Wolf Koelling
Slaughter and May

"John Murray" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]

Sharon,

What Cache version?

Sounds like the issue fixed by BJB266 in 4.1.16. See
http://www.intersystems.com/cache/technology/product-tables/releasenotes
/4116/index.html

John Murray
George James Software


-----Original Message-----
From: Sharon Hershon [mailto:[EMAIL PROTECTED]
Posted At: 11 May 2004 08:25
Posted To: intersystems.public.cache
Conversation: %IsModified and relationships
Subject: %IsModified and relationships


I'm curious as to folks take on this. I have class named Site that has a one to many relationship to a class named Records (so one site has many records). The following code causes the %IsModified flag to go to 1:

w  oRec.%IsModified()
0
w oRec.recToSite.internalNo // Relationship recToSite As Core.Site [

Cardinality = parent, Inverse = siteToRec ];

0
w oRec.%IsModified()
1

To get the value of  oRec.recToSite.internalNo site needs to
be swizzled, but that causes oRec's modified flag to go to 1.
What I really want is to rely on IsModified for when
properties actually change value on oRec.  It could be argued
that a relationship property has changed value, thus
%IsModified is true, but I sure wish I could take
relationships being swizzled out of the picture.

Has anyone been down this road?  Do I have to roll my own here?

MTIA,

/Sharon






Reply via email to