Red Hat's "obfuscated" kernel source [LWN.net]

Red Hat's "obfuscated" kernel source

Posted Mar 1, 2011 2:06 UTC (Tue) by paravoid (subscriber, #32869)
Parent article: Red Hat's "obfuscated" kernel source

Maximilian Attems (and Ben Hutchings, who has commented here) are both Debian Developers with significant contributions to the kernel, in Debian and upstream, over the years.

Debian's kernel tree, with all of its modifications (which *are* significant) remains availabe in a proper format for everyone to distribute, work on it, fork it or any other purpose with the only limitations being the ones that are mandated by the GPL.

This tree is not under any corporation's control and fulfills no agenda besides the one that serves Debian's (and its derivatives) users.

Please remember that and thank these people.

Red Hat's "obfuscated" kernel source

Posted Mar 1, 2011 22:33 UTC (Tue) by dowdle (subscriber, #659) [Link] (6 responses)

I'm not trying to denigrate the contributions of Debian developers... and while they may have contributed significantly in the past and may continue to in the future... in comparison to how much Red Hat contributes... I've yet to see Debian or a Debian developer listed in the top 10 list.

It is possible I've overlooked one or two on a few releases and I haven't really been looking at it that hard... but Red Hat has been the #2 contributor to the kernel for some time.

Debian makes a fine distribution, makes it for a lot of arches, and packages up a lot of packages... but when it comes to original development work in core components, to the best of my knowledge, they do not stand out at all. On the other hand, Red Hat stands on a number of projects including the kernel, gcc, x.org, gnome, etc.

Trying to turn this into an opportunity to switch users to Debian seems a little too convenient. I hear some people tell me, because I'm an pro-Red Hat advocate, that they will never use Red Hat software... and my response is that no matter what Linux distro you use, you are using a lot of software Red Hat has developed. I don't think anyone who has actually look into it is willing to argue with me on that point.

Red Hat's "obfuscated" kernel source

Posted Mar 1, 2011 22:41 UTC (Tue) by paravoid (subscriber, #32869) [Link] (5 responses)

From Maximilian's interview:

Ben Hutchings is the Nr.1 contributor for 2.6.33. He also is top listed as author of patches on stable 2.6.32. Debian is not listed as organization as many send their linux-2.6 patches from their corporate or personal email address and thus it won’t be attributed to Debian. There is currently no means to see how many patches get forwarded for the stable tree, but I certainly forwarded more then fifty patches. I was very happy when Greg KH personally thanked me in the 2.6.32.12 release.

And this was not an attempt on switching users, but merely a give credit where credit is due attempt. Nor I disagree with you on RedHat's contributions.

Red Hat's "obfuscated" kernel source

Posted Mar 1, 2011 23:05 UTC (Tue) by dowdle (subscriber, #659) [Link] (4 responses)

Thank you for pointing that out.

When Jon Corbet first started doing his analysis of who made kernel X... and it became an ongoing feature... I emailed him to ask why Debian didn't show up... if it was because they are harder to identify... or if it was because they don't contribute as much to kernel development. It was a while ago and I'm not quoting... but I believe he said... as far as he could tell they weren't contributing to the things he was measuring.

I don't think the Debian development community cares as much about getting credit... but it wouldn't be a bad thing for them to make some effort to become more measurable in such surveys... if they are indeed contributing more than they are being credit with. I'm not sure what that would take but I'm guessing those doing the measuring could give some valuable clues.

I certainly would like to see the Debian developers get credit for the work they do.

Red Hat's "obfuscated" kernel source

Posted Mar 2, 2011 0:22 UTC (Wed) by neilbrown (subscriber, #359) [Link] (3 responses)

It seems to me that there is an apples-and-oranges comparison here.

Redhat are a software distributor and a development house and a support service (and probably other things).

Debian is a software distribution.

The kernel patches from "redhat" are almost certainly mostly from the "development house" side of redhat. There is no corresponding part of Debian.

I really don't expect 'Debian' to contribute much in the way of kernel patches (though I suspect they channel their fair share of bug reports, which are just as valuable). When individuals who happen to be Debian developers also happen to do kernel development I suspect they are not doing so as a "Debian developer" but in some other role (employee of some company, independent enthusiast or whatever).

As a software distributor you might expect Debian to appear in surveys such as "how widely used" or "user preference" or "number of packages" or "number of bug reports" or "frequency of releases", and I thing Debian does OK in most of those.

Red Hat's "obfuscated" kernel source

Posted Mar 2, 2011 5:36 UTC (Wed) by JoeBuck (subscriber, #2330) [Link] (1 responses)

Yes, and there is no organization called "Debian" paying developers so they can work full time on the kernel. There certainly are Debian developers making significant contributions all over the place, but the RHEL revenue is making it possible for many people to do nothing else but work on free software.

Red Hat's "obfuscated" kernel source

Posted Mar 4, 2011 8:07 UTC (Fri) by smurf (subscriber, #17840) [Link]

Your point being?

My point would be that they didn't invent the wheel (i..e the kernel) here. There got to where they are now because lots of others play by the rules and don't obfuscate their code, or its origins.

While Oracle doesn't play nice with Open Source, kowtowing to their level of "this is our code, whether the community likes it or not" boneheadedness and doing essentially the same thing is not the way forward: Oracle isn't going to change its business practice because of what Red Hat does or doesn't do with its kernel patches.

Summary: This step benefits nobody.

It may even hurt Red Hat directly, if a couple of their kernel people get disgusted enough about this attitude that they leave.

Red Hat's "obfuscated" kernel source

Posted Mar 5, 2011 15:32 UTC (Sat) by BenHutchings (subscriber, #37955) [Link]

When individuals who happen to be Debian developers also happen to do kernel development I suspect they are not doing so as a "Debian developer" but in some other role (employee of some company, independent enthusiast or whatever).

I work on the kernel as a Debian developer, fixing bugs where I can see how to and occasionally developing features that are important to Debian and its users. I also work on the kernel as an employee of Solarflare, but this is a pretty well separate role, as my employer does not use or claim to support Debian.