LWN: Comments on "Some 5.1 development statistics"

Some 5.1 development statistics

scientes — Thu, 09 May 2019 20:35:22 +0000

> We probably don't need 10 upstream maintainers/reviewers pointing out to their team members that they forgot a kfree() in a public forum, just because of the email volume.

But getting it published quickly so others can see it is often important.

Some 5.1 development statistics

moorray — Fri, 26 Apr 2019 16:47:08 +0000

I used to understand why patches coming to the ML with reviews already on was bad, but now, having gained more experience, I'm no longer so sure..
A major reason to review the patches internally first, is that the quality is not always great. We probably don't need 10 upstream maintainers/reviewers pointing out to their team members that they forgot a kfree() in a public forum, just because of the email volume.
As far as the reviews go - I certainly want the employees who spend their time reviewing code see their contribution acknowledged. The maintainer above me knows the people and the value of their tags, regardless of the employer.

Some 5.1 development statistics

samlh — Fri, 26 Apr 2019 15:41:37 +0000

The commit flow for soc/soc is there, it's just mostly covered by the arrowheads.

Some 5.1 development statistics

blackwood — Fri, 26 Apr 2019 07:50:20 +0000

I'm also wondering why drm-intel is talking to itself. I guess it's due to topic branches shared with other trees (we've had a pile in 5.1 of those), or maybe backmerges. But other trees do that too, but none look this self-absorbed in the graph.

Some 5.1 development statistics

blackwood — Fri, 26 Apr 2019 07:46:00 +0000

Somehow soc/soc in the commit flow graph doesn't have the number of commits flowing through that tree. Seems to be the only one.