Linux 4.3-rc4 Kernel Released: Adds A New & Better String Copy Function

[phoronix]

Phoronix: Linux 4.3-rc4 Kernel Released: Adds A New & Better String Copy Function

Linus Torvalds has tagged the Linux 4.3-rc4 kernel a few minutes ago in Git. Besides bug-fixing, this new kernel update adds a post-merge-window feature of a new and more secure string copy function for developers to utilize in the future...

http://www.phoronix.com/scan.php?page=news_item&px=Linux-4.3-RC4-Released

[mark45]

"if it isn't broke, don't fix it" - like the Koran, this phrase is used by demagogues so often that I developed a knee-jerk reaction to object to it by default.
Be it about systemd, or wayland or lots of other good stuff - all the idiots who were against them used this very idiom extensively.
And only rarely did I see anyone using it and being right.

[Dorsai!]

"if it isn't broke, don't fix it" - like the Koran, this phrase is used by demagogues so often that I developed a knee-jerk reaction to object to it by default.
Be it about systemd, or wayland or lots of other good stuff - all the idiots who were against them used this very idiom extensively.
And only rarely did I see anyone using it and being right.

Concerning systemd the phrase was probably more like "If it has been sufficiently fixed before, don't fix it again in a much more monolithic and inflexible, overengineered and bug-prone manner". OpenRC for example, with it's simple and elegant design fixes all the major issues of legacy inits (like sysV init), without all the new problems systemd is introducing.

But I guess we'll talk again once the bloated tetsuo-blob of systemd's codebase implodes on its own weight and blows up in the face of the Linux/Unix ecosystem.

[Luke_Wolf]

"if it isn't broke, don't fix it" - like the Koran, this phrase is used by demagogues so often that I developed a knee-jerk reaction to object to it by default.
Be it about systemd, or wayland or lots of other good stuff - all the idiots who were against them used this very idiom extensively.
And only rarely did I see anyone using it and being right.

Pretty much agreed. Just like the non-developers who clamour on about the "UNIX philosophy" without having the slightest clue what it actually really means. What makes it all the worse is that this phrase is used to defend the most broken, ducktaped together software out there like x11 and SysV style init, where the developers of which will tell you that there's only a handful of people in the world that really understand how it works, and thus it really is quite broken, just the peanut gallery just doesn't want to admit to it.

[dungeon]

Pretty much agreed. Just like the non-developers who clamour on about the "UNIX philosophy" without having the slightest clue what it actually really means.

Unix philosophy means to me that couple senior Debian developers and members of commitee resigned during systemd debate.

So i am quite sure that Unix philosophy is nothing else but - Do One Thing, Make Sure It Broke Community and Do That Well.

[michaeljt]

I very much like this comment on the subject of strscpy by Linus[1]:

"I can't think of a single strncpy (or strlcpy) in kernel space that is actually worth even optimizing for. They just don't tend to be in the critical path. So correctness is likely *much* more important than worrying about performance."

[1] https://lkml.org/lkml/2015/4/28/766

[rtfazeberdee]

Oh dear, Dorsai!. if you are going to troll systemd, use something original that looks possibly like a fact instead of a deluded and moronic opinion like "more monolithic and inflexible, overengineered and bug-prone manner" otherwise it makes you look stupid.

[s_j_newbury]

I very much like this comment on the subject of strscpy by Linus[1]:

"I can't think of a single strncpy (or strlcpy) in kernel space that is actually worth even optimizing for. They just don't tend to be in the critical path. So correctness is likely *much* more important than worrying about performance."

[1] https://lkml.org/lkml/2015/4/28/766

I've often wondered how much performance is indirectly affected by non-critical path code using up CPU cycles that could otherwise be used by performance critical code paths, or else at least getting the CPU to idle more quickly and giving better energy savings. Probably something micro-benchmarking wouldn't pick up, but over an entire system it must accumulate to being quite significant..?

[unixfan2001]

Concerning systemd the phrase was probably more like "If it has been sufficiently fixed before, don't fix it again in a much more monolithic and inflexible, overengineered and bug-prone manner". OpenRC for example, with it's simple and elegant design fixes all the major issues of legacy inits (like sysV init), without all the new problems systemd is introducing.

But I guess we'll talk again once the bloated tetsuo-blob of systemd's codebase implodes on its own weight and blows up in the face of the Linux/Unix ecosystem.

I'm sorry. I didn't realise this forum was reserved for up and coming comedians.