Eclipse Community Forums: Compare » [EMF Compare] Moving items bug or by design?

Help

Home

Home » Modeling » Compare » [EMF Compare] Moving items bug or by design?

Show: Today's Messages :: Show Polls :: Message Navigator

[EMF Compare] Moving items bug or by design? [message #905695]

Thu, 30 August 2012 22:44

Stacey Hopkins

Messages: 34
Registered: March 2012

Member

I am using EMF compare (v1.3) pretty much out-of-the box with not really any customization.

All model items have a UniqueId (GUID)

I have noticed the messages for model items being moved is not how I would expect it to report

The following are three versions of models..... The first comparing v2 with v1 and the second comparing v3 with v2.

v1.

fm
|-------fg1
|	|-------f1
|	| 	
|       |-------f2
|      
|
|-------fg2

v2.

fm
|-------fg1
|	|-------f1
|	 	
|       
|      
|
|-------fg2
        |-------f2

Message comes back f2 has moved from fg1 to fg2 ....fantastique!

v3.

fm
|-------fg1
|	|-------f1
|	 	
|       
|      
|
|-------fg2
|       
|
|
|
|-------fg3
        |-------f2

Message comes back :

- fg3 has been added ... fantastique!
- f2 has been removed Sad

not so fantastique...

Think this as a message to the user is miss-leading, the user thinks the f2 has been removed when it hasn't been....

Is this by design, do I need to write own code for some cases? Am I using the tool wrongly? Or is this just a bug?

.........on a side note is this the right place to post for emf compare... the eclipse sub site takes you to the main emf forums (when clicking on the emf newsgroup) which is why I posted there originally.

Report message to a moderator

Re: [EMF Compare] Moving items bug or by design? [message #905845 is a reply to message #905695]

Fri, 31 August 2012 07:27

Laurent Goubet

Messages: 1902
Registered: July 2009

Senior Member

Hi,

This is one of the limitations of EMF Compare 1. Namely, we do not "go under" added or removed elements to check for diffs. That is why we do not even "see" the new f2 and tell you that it's been removed.

This is a limitation that will not be fixed in the 1.* stream. The 2.0 is well under way and its final release candidate should be out of the oven by the end of the day. You might want to try it out for your use case, the latest integration build can be installed from this update site. Do note that it is built against Juno.

Laurent Goubet
Obeo

Report message to a moderator