Subversion and Trunc/Branch merging

Question

Hi All,
We have been trialing data modeler for a single project, but we are so impressed, we want to use it across the IT estate.
Therefore we need to propose a process that is robust and allows us to co-ordintate teams and versions of the databases.
We have come up with the following - I am hoping for advice on whether there are easier methods to achieve the below.
When moving from branch to trunc, I need the following process (which I suspent most people need in some guise or another).
1. Compare branch changes to trunc.
2. Identify all changes between trunc and branch (and document them into a release document)
3. Resolve any off-project but neccessary trunc changes into branch and re-integrate into development cycle (other project changes, live fixes etc....).
4. Generate DDL to move from trunc to branch (ie database upgrade script to branch level).
5. Merge back into trunc and report merge details (so we can match them up to 2.).
6. Generate full DDL for latest trunc.
Data modeler seems to allow me to do most of this, but I am wondering if there is a better way for some of it:
1. Using Versioning "Merge Design" I can compare and resolve conflict.
2. Is there an wasy way to report into a document, the differences between 2 models?
3. I have to open up branch and import changes from trunc, before re-opening trunc and importing from branch - even though they are the same changes.
4. Checkout both models and do "model compare". Is there a way without having to checkout both models? - ie within subversion repo
5. Ok, easy enough - open Trunc and merge back in.
6. Generate fresh baseline DDL from trunc.
Outputs:
Branch=Trunc
DDL to move from Old Trunc to New Trunc
DDL to build new Trunc from scratch
Report on difference between Trunc versions.

Answer 1

Hi,
I assume you are using DM 3.3.
1. Compare branch changes to trunc.
2. Identify all changes between trunc and branch (and document them into a release document)You can use SVN repository tools to get reports about changes. Otherwise you can get changes comparing two relational models (part of two designs) - you need to check out designs at point(revisions) you
want to compare. Compare dialog has "Tree view" and "Tabular view" - you can generate report about all changes in "Tabular view"
3. Resolve any off-project but neccessary trunc changes into branch and re-integrate into development cycle (other project changes, live fixes etc....).If you use DM 3.3 as part of SQL Developer you can use "View>Files" to get window with directories structure and versioning functionality is available for versioned files. This functionality is limitted to less functions in Standalone version of DM.
Or you can use third party tool for that - DM 3.3 comes with 1.6 SVN client so you should take care to use tool for 1.6 SVN.
4. Generate DDL to move from trunc to branch (ie database upgrade script to branch level).Using compare of two models at related repository revisions.
5. Merge back into trunc and report merge details (so we can match them up to 2.).
3. I have to open up branch and import changes from trunc, before re-opening trunc and importing from branch - even though they are the same changes.
5. Ok, easy enough - open Trunc and merge back in.1) You need to merge trunk into branch first in order to see that design in branch has no problems and then
2) use "Merge reintegrate" to get changes (in branch only) back to trunk - after successful merge reintegrate you should not use branch any more, another branch can be created if you want to split work again
Of course you can choose to not follow that path and do merging of different objects or revisions.
4. Checkout both models and do "model compare". Is there a way without having to checkout both models? - ie within subversion repoNo
Philip