Baseline shows difference against model without any changes on the model

[crumble]

Hello all,

I'm trying to use the baseline diff to find changes on the model because of removing a whole package.
I got unexpected number of changes on my first try. Obviously not related to the removed package.
So, I created a baseline and directly compared it with the model. Surprisingly the diff shows a lot of differences (expected: 0).
Does anybody know how to get rid of this automatic changes on baseline creation. The model seems to stay unchanged, but the baseline is different.
If I get a second baseline (again without any model changes) and compare the model again, the same differences are shown as in the first run.
Example of such a diff: The scope of a 'dependency' link is set to 'Public' in the baseline, but it is empty in the model.

Any idea, why this happens and how to avoid it?

[Paolo F Cantoni]

Hello all,

I'm trying to use the baseline diff to find changes on the model because of removing a whole package.
I got an unexpected number of changes on my first try. Obviously not related to the removed package.
So, I created a baseline and directly compared it with the model. Surprisingly the diff shows a lot of differences (expected: 0).
Does anybody know how to get rid of these automatic changes on baseline creation? The model seems to stay unchanged, but the baseline is different.
If I get a second baseline (again without any model changes) and compare the model again, the same differences are shown as in the first run.
Example of such a diff: The scope of a 'dependency' link is set to 'Public' in the baseline, but it is empty in the model.

Any idea, why this happens and how to avoid it?

I don't have a solution but an observation that "plus ca change, plus la meme chose".  MANY years ago (possibly over a decade) I spent months proving to Sparx that their XMI round-tripping was broken.  The same kind of thing as you - I export a package and re-import and get a different result!  Apparently, it was beyond the ability of Sparx to create a testing regime both conceptually and physically to create such a simple "smoke test".  Since this ability was fundamental to our business at the time, WE quickly developed the appropriate tests and mechanisms to confirm for us that we could successfully transmit a package from one instance to another and retain all the relevant information.  It took months and many EA releases to finally get import/export under control.

Paolo