Requirements Numbering vs. ID vs. Merging

 

Is there any best practice concerning requirement numbering and Merge function ?

Consider the following scenario : 

  • Create a SysML Project "MyProject"
  • Create a "MyRequirements" package
  • Create a Requirement Table in this package
  • From the requirement table create a Requirement "MyFirstRequirement" => object numbering (and ID) is "1"
  • Save the project to TWC server => MyProject [trunk] #1
  • Create and open a branch "MyBranch" =>  MyProject [MyBranch] #2
  • From the requirement table create a Requirement "MySecondRequirement from MyBranch" => object numbering (and ID) is "2"
  • Commit change to server MyProject [MyBranch] #3
  • In MyProject [trunk] #1 from the requirement table create a Requirement "MySecondRequirement from trunk" => object numbering (and ID) is "2"
  • Commit change to server MyProject [trunk] #4
  • Use "Merge From" and select MyProject [MyBranch] => the Merge window should look like this :
  • Use the "Finish Merging" command => the window should look like this :

Variations can be derived from this scenario, but in any case it would be preferable not to ask the end user to play with numbering or with table elements management ...