Hi Robert, It looks like we had some trouble trying to merge the latest changes into our mirror. I've rescheduled the job, and it's running now. I'll keep an eye on it to make sure it finishes. Thanks, Robin

Hi Yaroslav, Yes, you're correct that Ohloh does not update the commits or contributors if any of the repositories fail. I have just successfully updated the CVS repository, so the Ohloh report ... [More] should update soon. If the CVS repository is removed from Ohloh, all of the old entries will be removed as well. Ohloh does not retain historical data: only data from currently active repositories can be displayed. If the CVS history was not migrated to Subversion, then the old CVS repository should stay on Ohloh. Let me know if you have more questions, Robin [Less]

Hi Michel, Ohcount is the primary piece of open source software that we support. Our source control interface code is available on my personal Github account as ohloh_scm. This bit of code is very ... [More] specific to Ohloh's needs, so there is not much public involvement with that code. If you are really interested in building Ohloh-style metrics for your private repositories, you might want to take a look at ohdb. This tool builds a sqlite database to contain commit and line count metrics for a Git repository, and works on the command line. However, I warn you that the code is old, unfinished and totally unsupported -- it was a simple weekend experiment. But it did function at one point, and if you are a Ruby hacker, you might find this an interesting take on the problem. Let me know if you have any questions, Robin [Less]

Hi Manuel, We do not prioritize work on additional languages. However, we do happily accept patches from the community, and we get a lot of new languages every month this way. The source code for ... [More] our line counter Ohcount is open, and you can find it hosted at SourceForge. Thanks, Robin [Less]

Hi trijezdci, It looks like we were unable to merge the latest changes into our local mirror. I've started a clean clone of the repository, which should clear up the problem. Thanks, Robin

Hi Aslak, It looks like we've been having trouble connecting to the Subversion server. It seems to be OK now, so I've rescheduled the job. I'll keep an eye on it. Thanks, Robin

Hi Adrien, It looks like we had some trouble on our end. I've rescheduled the update, and it's running now. I'll keep an eye on it to make sure it completes. Thanks, Robin

Yes, this is a problem on Ohloh generally. There is a lot of duplicate code in the world. It's especially true for the kernel, which is an enormous body of code that has at least a dozen duplications ... [More] on Ohloh. Reporting all of these duplicates separately is not just incorrect, it's also something of a burden on our overworked servers :-). Addressing this problem correctly is something we've been thinking about for a long time, but would require some pretty significant changes to our process. However, Git's design should make duplicate removal relatively simple because of its hash-based unique IDs. In the short term there are some hacks we've done specifically to accommodate the kernel workload, but a proper solution is probably a good ways off. Thanks, Robin [Less]

Hi Jakub, I started another clean analysis of these repository, hopefully it will clear up the problem. I'll check it again after the weekend. Thanks, Robin

I agree :-). It's a simple change; I'll see if we can't get this into the next update.