Subject: Re: final question again
From: Paul Freeman (pfreeman@ccs.neu.edu)
Date: Tue Dec 10 2002 - 21:03:57 EST
Wei -
You are correct. This is a bug in the class form checker. I noticed it
this morning. You should point this out in your write-up. There is not
a pointcut that will get rid of this unfortunately. There are a few
interfaces that are used which you can define as stable however.
Defining various calls to the interfaces as stable will limit the
number of violations produced and you will eventually be able to verify
whether or not there are indeed any LoD violations.
Paul
Wei Hua Chen wrote:
>Hi, Paul. I ran the class form of you check with the aplib example. I
>think I only got one kind of violation that looks suspicous: the preferred
>supplier contains the interface and the target object is the class
>implementing the interface, a violation is raised as it can't find the
>concrete class in the preferred supplier hashset. Is such case considered
>a violation (maybe in strong form)? Is the question asking us to define
>either custom stable or global preferred point cut that will get rid of
>such violations? I can't figure out how. Maybe I spent too much time with
>eclipse and forgetting all about aspectj :-)
>
>Thanks for your help.
>
>:Wei.
>
>
>
This archive was generated by hypermail 2b28 : Tue Dec 10 2002 - 21:08:38 EST