LWN: Comments on "Some 3.18 development statistics"

Novell vs SUSE

corbet — Tue, 09 Dec 2014 16:28:40 +0000

Oops...that crept in when I fixed up the employer table. Fixed.

Novell vs SUSE

fcrozat — Tue, 09 Dec 2014 16:26:56 +0000

It looks like the first column is using "SUSE" as employer but the second one is still using "Novell" instead of SUSE (this is not the case on https://lwn.net/Articles/624330/ btw)

McAfee vs. Intel

robbe — Sat, 06 Dec 2014 21:22:57 +0000

According to mcafee.com the company line is (still) to identify as a seperate entity. (But the "Intel Security" name has been looming on the sidelines for some time.) I have a hard time imagining individual developers that would identify *more* with Intel than their corporate overlords do.

Some 3.18 development statistics

josh — Thu, 04 Dec 2014 01:46:13 +0000

I think we should draw the line wherever the entity in question wants it drawn. Subsidiaries that want to be counted as a completely separate entity (and that make it easy to do so, such as by using a separate @othercompany.example.com) can get counted separately; subsidiaries that consider their contributions part of the parent company should get counted as such.

So, in this case: do the McAfee folks *want* to be counted as part of Intel or as a separate entity?

Some 3.18 development statistics

aryonoco — Wed, 03 Dec 2014 05:40:31 +0000

It is.

Do we count subsidiaries as a different company or as part of their parent? I think the line is very arbitrary.

Some 3.18 development statistics

robbe — Sun, 30 Nov 2014 20:18:40 +0000

I am similarily intrigued. Maybe LWN would be interested to run a feature about Mr. Sweeten and the Comedi drivers?

Some 3.18 development statistics

dberkholz — Sat, 29 Nov 2014 19:42:11 +0000

Isn't McAfee part of Intel?

Some 3.18 development statistics

nix — Thu, 27 Nov 2014 17:36:28 +0000

My thoughts exactly -- at least in its original state.

Perhaps it's taking so long because it supports about sixty million different pieces of hardware: <http://www.comedi.org/hardware.html>. i.e. this is not "a driver", it's *lots* of drivers, and they were all horrible.

Some 3.18 development statistics

neilbrown — Wed, 26 Nov 2014 20:52:07 +0000

A Comedi of errors, perhaps?

Some 3.18 development statistics

nix — Wed, 26 Nov 2014 19:52:34 +0000

It seems like it would have been easier to write a new driver. Heck, it seems like it would have been easier to write a new *kernel*.

Some 3.18 development statistics

corbet — Wed, 26 Nov 2014 15:29:03 +0000

Send a note to me or to Greg.

Some 3.18 development statistics

Cyberax — Wed, 26 Nov 2014 08:45:05 +0000

Looks like it's been close to 4 years by now. I wonder, where can one get a Comedi device for cheap?..

Some 3.18 development statistics

jezuch — Wed, 26 Nov 2014 07:28:07 +0000

> As is usually the case, H. Hartley Sweeten tops the by-changesets list with the epic task of getting the Comedi drivers into shape in the staging tree.

That must be some ungodly mess. How long has he been at it?

Some 3.18 development statistics

mathstuf — Wed, 26 Nov 2014 04:47:17 +0000

My kernel patches prompted an email from Greg K-H about my employment once a release was getting close. You could try emailing him about it.

Some 3.18 development statistics

Magni — Wed, 26 Nov 2014 03:45:19 +0000

So totally off topic but how do we change $company associated with $email_address?

Some 3.18 development statistics

Trelane — Tue, 25 Nov 2014 23:35:55 +0000

> So it might be safe to say that, for now, the decline in volunteer contributions appears to have leveled out.

Was the fraction decreasing it the absolute number? The previous sentence discusses fraction, but this, that allegedly builds off of it, is absolute.

Some 3.18 development statistics

josh — Tue, 25 Nov 2014 22:32:52 +0000

If you do end up producing statistics for Mesa, please consider posting those for inclusion in LWN.

Some 3.18 development statistics

Kayden — Tue, 25 Nov 2014 21:24:51 +0000

Awesome, thanks!

Some 3.18 development statistics

corbet — Tue, 25 Nov 2014 21:21:47 +0000

It's done with gitdm, same thing Greg uses. git://git.lwn.net/gitdm.git

Some 3.18 development statistics

Kayden — Tue, 25 Nov 2014 21:20:33 +0000

Out of curiosity, what scripts do you use to produce these statistics? Is it https://github.com/gregkh/kernel-history/ ?

I've hacked up a few of my own to produce commit and review statistics for Mesa and Piglit, but I don't yet have any infrastructure for dealing with changes in employment, and the code needs a total overhaul if I'm going to add any more features.

Some 3.18 development statistics

wsa — Tue, 25 Nov 2014 17:08:29 +0000

"For now, there doesn't appear to be any reason to believe that we will not continue to crank out kernel releases at a high rate indefinitely."

Sure, there is. We do not only need code contributors, but also reviewers. The gap is already huge, and as things look it will get even bigger in the future. I know that the problem is discussed here and there and solutions are not known. Still, for starters, I think we shouldn't say that all is perfectly fine at least...