Details
-
Type:
Improvement
-
Status:
Closed
-
Priority:
Minor
-
Resolution: Won't Fix
-
Affects Version/s: None
-
Fix Version/s: None
-
Component/s: Infrastructure: CatTrack, Infrastructure: Test
-
Labels:None
-
Number of attachments :
Description
Not sure if this is easy or hard....it's easy to describe 
it would be nice to have the svnversions that show up in the regression email shown in two lists. Those that are in rvmroot/trunk (ie, those that are very likely to be the things being tested) and those in some other part of the repository (that although they could in theory cause a regression, the odds are very low).
In one sense this is relatively easy as we could just grep output of svn commands. However this could mean that the generation of web pages and mails took a VERY long time (especially given the speed of sourceforge servers).
One thing I had considered as a long term option would be to import svn log information into the database so we could correlate information between users and results. (i.e. who broke the build and how often, speed up and who caused them etc). A more generalized but less fancy approach could be taken and I could just start importing minimal amount of data to enable filtering of potential change causing revisions +author.