We are gonna weekly track the progress of GNOME by committing with charts just ..for the curious in you!
And we are starting right now, watching the road map from 3.4 to 3.6. We take as a starting point issue 181, March 25, 2012, where the 3.4 was about to be released, and the road to 3.6 was starting. Since then we track 10 issues (181-190) and 10 weeks (March 25 – May 27).
Commits per week, commits total. A crazy 20021 commits during a 10 weeks period! 2002 commits per week, or 286 per day, or 12 every hour. Not bad! Most active week had 2870 commits while less active had 1323.
Commits per week, per hacker, per project. This is a quite interesting chart. Seems like one hacker corresponds to one project. Unfortunately I couldn’t count different hackers and projects, so I will make a “dirty” average. So 232 hackers a week, contribute in 201 projects by 2002 commits.
Alright, these numbers seem incredibly high, but we need to compare them to something. I’ll do the following only once, just because I was curious about whether other projects could score as high or not. So I picked the “rival” KDE. Could KDE do better?
Commits per week, commits total, per project. KDE scores a higher 22114 commits over 20021 of GNOME (!), and that is a weakly average of 2211 over 2002 of GNOME or 315 a day over 286 of GNOME. That leads to an impressive ~10% clean ahead for KDE!
Commits per week, per hacker, per project. What we see here is that GNOME with more hackers score less commits than KDE. GNOME has 232 hackers that contribute 2002 commits during a week, while KDE with only 168 hackers scores 2211 commits. A GNOME hacker does average 8.6 commits when a KDE hacker does average 13.1 commits! That definitely shows that KDE hackers are more committed to their project.
While this is only a “just for fun” comparison, and isn’t meant to be useful for serious analysis, as the only parameter I use is the commits and the hackers, it reveals something that is well known. GNOME is more popular, while KDE is more advanced.
* KDE’s issue 233 was broken and I used issue 226 instead.