Changes by kernel area

Posted Dec 31, 2010 11:21 UTC (Fri) by epa (subscriber, #39769)
Parent article: Who wrote 2.6.37

I see that by lines changed, Broadcom is the top contributor. But I suspect their work is only in device drivers. Can the contributions be broken down by area somehow - perhaps something as crude as top-level directory within the source tree?

Changes by kernel area

Posted Dec 31, 2010 14:09 UTC (Fri) by dwmw2 (subscriber, #2063) [Link] (2 responses)

I see that by lines changed, Broadcom is the top contributor. But I suspect their work is only in device drivers.

And device drivers that have been accused of being an ocular carcinogen, at that. A decent driver would probably have been a lot fewer lines; especially if it'd added to the existing b43 driver infrastructure instead of reinventing a bunch of stuff for itself.

These statistics should probably exclude the drivers/staging directory of the kernel. In fact, I think it would be better if Linus' tree didn't include the staging crap at all, and it was only in an external repository.

Changes by kernel area

Posted Jan 2, 2011 13:16 UTC (Sun) by wsa (guest, #52415) [Link]

+1 (again) for skipping staging in the stats

Changes by kernel area

Posted Mar 15, 2011 23:05 UTC (Tue) by josh (subscriber, #17465) [Link]

Excluding staging would exclude both the original contributors of the drivers and those who help clean them up. Excluding the latter seems uncalled for. Excluding the former denigrates the attempt to contribute at all. That a driver needs work does not make it worthless as a contribution.

As long as staging lives in the main kernel tree (which has made sense so far), the statistics should continue to count it.