Posted Jul 9, 2009 7:40 UTC (Thu) by Frej (subscriber, #4165)
Parent article: Some ado about zero
The interesting datapoint is that it takes 18 months for a stable kernel to get enough widespread use before this kinda of problem emerges.
Ofcourse it's not a statistically valid sample (a single datapoint....) but still interesting :).
But what's the actual delay between release and actual distribution? Nobody outside the distro's know how long it takes before fixes (x.y.z+1 releases) reach the user. In many cases that includes the maintainer being left in the dark. But what's the average? I guess availability of said release in stable software is good enough, not actual penetration among users.
Maybe it's worth an article? (Although i guess it's hard to get data from redhat/novell, but they would know when actual customers upgrade).