How to make correct difference review to obtain full code coverage in reports?

[Anton Vorobiev]

Hello! I have the following scenario:

1. File.c created
2. File.c reviewed as file. Now the code coverage is 100%
3. New line "string 1" is added in File.c It was commited in svn
4. Another line "string 2" is added in File.c And we commited it again.
5. We want to review difference between lastreviewed and acrual state. For this we create new review, then choose option "Diffs to last reviewed version".
6. We complete this review and look in statistics. The revison with string 1 is unreviewed,this line is unreviewed and chageset is unreviewed.

This behaviour kills our statistics. It seems that we have unreviewed stuff but it is not true for lines. Is it possible to change coverage algorithms or it can be other approach to have right statistics?

By my opinion in this case we should have 100% line coverage( all lines of code in last revision is coveted), 100% patch coverage( the result of pathces is reviewed) and 100% changesets reviewed( we have looked on all changesets).

[seb · Accepted solution]

Hi Anton,

I understand your use case, and what you are expecting, however currently there is no support for such advanced "review coverage" calculations in Crucible.

Regards,

Seb