KDBUS To Be Included In The Linux 4.1 Kernel

[uid313]

Yep. It is his call.

I hope he goes wild and rants and calls everyone dumb, incompetent sons of bitches, then hacks away all weekend and announces his own god-sent IPC mechanism the following Monday. The world goes wild, just like it did when he made Git, and then it gets adopted by FreeBSD, OpenBSD, NetBSD, DragonFlyBSD and OpenIndiana/Illumos.

[RahulSundaram]

I found the reply by GKH very interesting (read here), so... are we waiting for the Linus opinion or what?

Yep. It is his call.

[valeriodean]

I found the reply by GKH very interesting (read here), so... are we waiting for the Linus opinion or what?

[Anarchy]

i seriously doubt this reading this answer http://lkml.iu.edu/hypermail/linux/k...4.1/03957.html , especially how good track record of GregKH is. it is more like other guy is complaining... just complains and never states reasoning. based on this answer of GregKH this is not first time same person complained and each time no specifics.

It appears that the other guy doesn't just complain about kdbus, he doesn't like d-bus in general. He thinks it's badly designed and seriously flawed, and putting that into the kernel is a bad idea. But, he doesn't offer any constructive criticism on how to improve d-bus and kdbus in particular. I haven't read the entire correspondence between them, but if he knew how shitty dbus was from the start then why didn't he actively try to influence the work done by GKH et al...

[pininety]

http://en.wikipedia.org/wiki/D-Bus#Implementations

So were are at least four implementations and I remember intel also having their own d-bus implementation being mentioned in a talk.

[yoshi314]

from the email :

this is something that finally fixes a number of almost unfixable races in the current dbus implementations.

wait, there are multiple implementations?

[justmy2cents]

I'm sure he knows, but I'm questioning whether he's applying his knowledge of "what kind of code belongs in the kernel" at this point. If he gets the same (and valid) criticisms every damn time he submits the PR, and half of them are security issues, why doesn't he fix them? Because some people (systemd, gnome, who??) don't want to rewrite their code to support a BETTER D-Bus API? Fuck the lazy people, Greg, I want a secure Kernel... and that's all it really comes down to.

i seriously doubt this reading this answer http://lkml.iu.edu/hypermail/linux/k...4.1/03957.html , especially how good track record of GregKH is. it is more like other guy is complaining... just complains and never states reasoning. based on this answer of GregKH this is not first time same person complained and each time no specifics.

[RahulSundaram]

I'm sure he knows, but I'm questioning whether he's applying his knowledge of "what kind of code belongs in the kernel" at this point. If he gets the same (and valid) criticisms every damn time he submits the PR, and half of them are security issues, why doesn't he fix them? Because some people (systemd, gnome, who??) don't want to rewrite their code to support a BETTER D-Bus API? Fuck the lazy people, Greg, I want a secure Kernel... and that's all it really comes down to.

Can you point me to any specific security issues? I read the thread and I see a bit of ranting from the same couple of folks who seem opposed from the start but they are very generic.

[Daktyl198]

The lead dev is GregKH-- he's the maintainer of the linux-stable tree.. trust me, he knows.

I'm sure he knows, but I'm questioning whether he's applying his knowledge of "what kind of code belongs in the kernel" at this point. If he gets the same (and valid) criticisms every damn time he submits the PR, and half of them are security issues, why doesn't he fix them? Because some people (systemd, gnome, who??) don't want to rewrite their code to support a BETTER D-Bus API? Fuck the lazy people, Greg, I want a secure Kernel... and that's all it really comes down to.

[Ericg]

There are well thought out valid criticisms, and I'd hope they get addressed. I'm not sure the dbus guys understand that once this is merged all its flaws are locked in place.

The lead dev is GregKH-- he's the maintainer of the linux-stable tree.. trust me, he knows.