Skip to main content


Eclipse Community Forums
Forum Search:

Search      Help    Register    Login    Home
Home » Modeling » EMF » [EMF Compare] Which version for 4.3M3?
[EMF Compare] Which version for 4.3M3? [message #986311] Mon, 19 November 2012 21:14 Go to next message
Ed Willink is currently offline Ed WillinkFriend
Messages: 7655
Registered: July 2009
Senior Member
Hi

Is EMF Compare 2.x yet ready for interactive use on 4.3M3? It is in the
staging repo so I gave it a try, but it reports differences that are
only vaguely correlated with actual differences. Version 1.3.2 for Juno
seems much better.

Regards

Ed Willink
Re: [EMF Compare] Which version for 4.3M3? [message #986350 is a reply to message #986311] Tue, 20 November 2012 08:15 Go to previous messageGo to next message
Laurent Goubet is currently offline Laurent GoubetFriend
Messages: 1902
Registered: July 2009
Senior Member
Ed,

What are differences "vaguely correlated with actual differences"? What kind of models are you comparing? What result do you obtain that is not what you expected?

M3 had a number of known issues when comparing models that did not use IDs of any sort, though it should already yield better results than any of the 1.x builds for most models, even when not using IDs. Most of the 'known' issues have now been resolved, though these improvements aren't yet accessible in an integration build. (You can make an idea for yourself with the test results.)

Laurent Goubet
Obeo
Re: [EMF Compare] Which version for 4.3M3? [message #986363 is a reply to message #986311] Tue, 20 November 2012 09:01 Go to previous messageGo to next message
Cedric Brun is currently offline Cedric BrunFriend
Messages: 431
Registered: July 2009
Senior Member
Ed, content match is subject to "threshold effect" and we need to tune the algorithm and the ratios so that we provide the comparison the end user expect.

EMF Compare 1.x was quite good in this regard as we tuned it for quite a long time with many examples. 2.x clearly need some more work, but to be as efficient as possible we would need your test data and what you would have expected.


http://cedric.brun.io news and articles on eclipse and eclipse modeling.
Re: [EMF Compare] Which version for 4.3M3? [message #986472 is a reply to message #986363] Tue, 20 November 2012 15:11 Go to previous message
Ed Willink is currently offline Ed WillinkFriend
Messages: 7655
Registered: July 2009
Senior Member
Hi

Example files that compare very badly on EMF Compare 2.1 attached to

https://bugs.eclipse.org/bugs/show_bug.cgi?id=394668

Regards

Ed Willink


On 20/11/2012 09:01, Cedric Brun wrote:
> Ed, content match is subject to "threshold effect" and we need to tune
> the algorithm and the ratios so that we provide the comparison the end
> user expect.
>
> EMF Compare 1.x was quite good in this regard as we tuned it for quite
> a long time with many examples. 2.x clearly need some more work, but
> to be as efficient as possible we would need your test data and what
> you would have expected.
Previous Topic:emf.sdk feature on Kepler milestone ?
Next Topic:Export model to XMI with direct mapping, not by id
Goto Forum:
  


Current Time: Thu Mar 28 08:39:21 GMT 2024

Powered by FUDForum. Page generated in 0.02707 seconds
.:: Contact :: Home ::.

Powered by: FUDforum 3.0.2.
Copyright ©2001-2010 FUDforum Bulletin Board Software

Back to the top