Take a look at the graph for Grails: http://www.ohloh.net/projects/grails/analyses/latest
At one point the counts drop BELOW the 0 line, then pop back up. That surely is some kind of bug.
Possible bug in code analysis
about 1 year ago
about 1 year ago
Hi Brandon,
Yes, some of our older reports were affected by a bug which has long been fixed -- however, we haven't yet recounted all of the old code.
I've scheduled a clean recount for Grails, which should clean up the problem.
Thanks, Robin
about 1 year ago
Hmm. Upon a fresh analysis, it still appears that the lines of code drops below zero.
This is going to require some deep digging. Is it possible that some of the commits have out-of-order timestamps? Has this repository always been in Subversion, or was it converted from another format?
It's also possible that there was an error in our original download of the repository, and that we need to do a clean download as well as a clean recount.
about 1 year ago
I'm afraid I can't answer your questions. I'm just a user and fan of Grails, but not a contributor. Good luck?
about 1 year ago
Hey you fixed it! Nice work. That was a quick turnaround!
about 1 year ago
Yes, it appears we had a data corruption issue. A full re-download cleaned up the mess.