LWN: Comments on "2.6.39 development statistics"

Nokias role?

Jaffa — Thu, 19 May 2011 14:09:20 +0000

Harmattan is approaching (FSVO "approaching") release - as noted, Feb 11th doesn't mean an immediate layoff to everyone working on MeeGo (or on other uses of Linux)

Nokias role?

job — Mon, 16 May 2011 20:01:27 +0000

Nokia is a big company. I bet there's a lot of Linux in their telco equipment, for example.

2.6.39 development statistics

corbet — Mon, 16 May 2011 12:48:25 +0000

If you want to learn more about how these numbers are generated, remember that the gitdm source can be had at git://git.lwn.net/gitdm.git.

2.6.39 development statistics

corbet — Mon, 16 May 2011 12:46:58 +0000

An awful lot of developers do not post from company email addresses; if all we did was look for corporate domains, we'd have an awful lot more unknowns than we do. For your specific question: yes, Serge's contributions are properly attributed.

2.6.39 development statistics

patrick_g — Mon, 16 May 2011 08:42:06 +0000

Do you only use the email address to know who someone's employers are?
I ask this question because I looked at the "User namespace" patch from Serge E. Hallyn (link).
His email is serge AT hallyn DOT com but the status page for "User Namespace" is here in the Ubuntu wiki and the git development tree is in here at kernel.ubuntu.com.

I would like to know if Serge's commits are attributed to Canonical in your kernel stats.

2.6.39 development statistics

corbet — Sun, 15 May 2011 18:02:34 +0000

"None" is people known to be working on their own time - usually known because they told us so. "Unknown" is people we have no clue about.

2.6.39 development statistics

Julie — Sun, 15 May 2011 17:27:34 +0000

Yes, it would be nice to be able to elaborate on the statistics a bit more, although this is probably beyond what is practicable.

It's possible to infer or guess a few further details by knowing who someone's employers are, but what I would find really interesting would be some background on voluntary developers, which must be quite a diverse demographic - long-term veterans, consultants expanding their credentials, professional students (probably the highest figure), developers migrating from another project, hobbyists etc. It might give further clues as to why this statistic has dropped, too.

I suppose this would be a lot too complicated to compile though. How are the 'none' and 'unknown' stats arrived at?

Nokias role?

johill — Sat, 14 May 2011 15:01:15 +0000

I can't say anything about what will happen in the future, and I doubt they'll suddenly go down to absolutely nothing even though they announced a switch, but keep in mind that these numbers are "historic" already. Almost all development that went into 2.6.39 was done before the 2.6.38 release (which was in March), the stabilisation period is just a process that follows after that.

Nokias role?

kreutzm — Sat, 14 May 2011 09:57:01 +0000

I was really wondering why Nokia came up so high on the list. After their switch to the other big vendor I assumed they would no longer put (so much) ressources into the Kernel.

Anyone got some insights?

2.6.39 development statistics

Velmont — Fri, 13 May 2011 14:45:47 +0000

Cool. Maybe it'd be interesting to see where developers are living, split up by country, and maybe the US split up by state because it is big.

Also, having a time axis on this information would be interesting.

Thomas Gleixner's patches don't count for Linutronix?

corbet — Fri, 13 May 2011 13:07:31 +0000

The accounting of Thomas's patches is a bit complicated; he does some work under contract for others.

Thomas Gleixner's patches don't count for Linutronix?

ukleinek — Fri, 13 May 2011 11:03:24 +0000

How come that Linutronix only has a commit count of 276? Thomas' 442 patches should account for them, don't they?

irq changes article?

pflugstad — Fri, 13 May 2011 02:43:26 +0000

Thomas Gleixner got to the top of the per-changesets list with a massive reworking of how interrupts are managed in the kernel - a job which required significant changes in almost every architecture.

Hopefully a LWN article about this will be forthcoming? Just asking. Thanks!