Loonix security fags on suicide watch.
> OpenBSD Will Get Unique Kernels on Each Reboot.
>This feature is named KARL — Kernel Address Randomized Link — and works by relinking internal kernel files in a random order so that it generates a unique kernel binary blob every time.
Other urls found in this thread:
phoronix.com
bleepingcomputer.com
marc.info
events.linuxfoundation.org
twitter.com
Linux has had KASLR since a long time, now it even comes turned on for retards like you
phoronix.com
>tfw even linux has zfs support
>tfw openbsd has no jail support
>using a *BSD without the only two reasons to use *BSD
You didn't even read the article but you still needed to post your shit. THIS HAS NOTHING TO DO WITH KASLR.
>"It still loads at the same location in KVA [Kernel Virtual Address Space]. This is not kernel ASLR!," said de Raadt.
Instead, KARL generates kernel binaries with random internal structures, so exploits cannot leak or attack internal kernel functions, pointers, or objects.
Also
>Kernel Address Space Layout Randomization (KASLR) turned on by default in Linux 4.12, released last week
BTFO LINUXFAGS
see or you can fucking read the article
If you think that those two are the basics of having a secure system son, you have a problem.
It's still a cool idea and if othrr oses use it then it has done well
Of course openBSD has a long way to go before it becomes a viable os for anything other than pf, but has that ever been the goal of the project?
Would like to see them try a better filesystem though, i remember there was talk of HAMMER at one point
>Of course openBSD has a long way to go before it becomes a viable os for anything other than pf,
>Citation needed
I have used it as a desktop for years. The developers all use it for their everyday needs. The only downside is the not bleeding edge packages (for stable release) but everything else is top notch quality.
No other operating systems has KARL, but the Linux kernel itself has KASLR. The difference between the two is that KARL loads a different kernel binary in the same place, while KASLR loads the same binary in random locations. Same goal, different paths.
the article doesn't say jack shit
does it randomize the layout of structs?
does it randomize the layout of the text section?
does it do both?
For prod*
It's a great desktop/laptop os. I love the userland
Are you a baby that needs to be spoonfed? Just click on the fucking links that lead to the developers discussion at openbsd-tech.
>For prod*
What exactly are your needs for prod*?
>there are people out there
>RIGHT NOW
>still using shitty Linux
HAHAHAHA
>Same goal, different paths.
The article is very nice to loonixfags because Linus HAS NEVER FOCUSED ON SECURITY. There are tons of sec joles in the linux kernel.
I like how BSD has been absolute smoking lincucks for the past few years and their reaction has been to spread FUD about it instead of recognizing that more OSS alternatives are always a good thing. They're so beholden to this tribal identity that they're willing to do the NSA's work for them. Maybe it's because so many of them unironically derive a sense of pride from maintaining an over-engineered ersatz piece of shit that the mere presence of something as simple and secure as OpenBSD makes them shite their knickers with rage
Robust filesystem
Mature virtualisation technology (ideally with containers)
Software support
To name the obvious ones
...
>Robust filesystem
FFS is aging but reliable, robust and SECURE.
>Mature virtualisation technology (ideally with containers)
You are spot on. OpenBSD doesn't have this. However this is because virt does not guarrantee sec. They recently started to support xen.
>Software support
I don't understand this. Openbsd has a huge port collection.
>hurr durr i don't even know that KASLR was developed by the people from OpenBSD.
wew lad just neck yourself straight away
>tfw zfs has no official linux support because of license issues
>tfw openbsd doesn't need jail because it's a shit version of a hypervisor running in a loosely secure chroot
>not having any fucking idea what the fuck im talking about hurr durr
you don't run containers on virtualization! containers are a replacement. literally you add an abstraction layer onto another abstraction layer onto another one. you have an OS with the hypervisor, then another OS as guest and then you have another abstraction layer which are containers. You are wasting a shit ton of resources and storage latency is a bitch, even if you'd run shit on full flash.
yeah I'd prefer a yes or no answer to my question
surely you can do that if you're familiar with this thing
>tfw no Linux asmr
Why éven live?
I'll switch to BSD when it supports my fucking 10 year old sound card.
surely you don't actually care about acquire knowledge, just the bullet points so you can pretend to know stuff about the topic. there is no need in answering your question, do it yourself you lazy fuck
Containers are a form of virtualisation
Which BSD have you tried that doesn't support it?
you're right, I don't care about openbsd enough to read up on this shit
That's another reason i admire OpenBSD. They are clean from SJWs and pretencious faggots that want others to do their work for them. Their mailing lists is a breath of fresh air full of pragmatism and no-politics.
they are, which is why i said they are a replacement and not to be put on top of it.
if you think a bit further; a lot of modern enterprise applications are java based, so you have another virtualization layer on top of that
hypervisor > OS > container > java/python/perl/ruby
not only does that sound inefficient on paper, in reality it is even worse. if your java application creates heavy I/O you are basically fucked, and enterprise grade storage ain't exactly cheap
Honestly, I've been ignoring OpenBSD before.
Is it harder to learn than other Unix OSs?
How is the desktop state of it?
I was thinking of giving it a try, have been using Windows 7 since it came out.
I never said you would put it on top.
my bad then. it sounded like you'd be running containers on top of your virtualization. that shit just triggers me
tfw linux can use OpenZFS but OpenBSD cant
why even live without zmirrors or raidz?
btfo
It's one of the easiest to use, doesnt require too much setup
Be warned that some of the stuff you are used to having on windows wont be available, and unlike linux there arent any hacky ways to get things working.
Coming from windows i'd be inclined to use linux first and see how you like it.
because if you really want to use ZFS professionally not some lousy ganoo plas leenox zfs meme without any official support. but you debianiggers and ubunturds wouldn't know about, just gobble the shit that canonigger mark shittleworth and rest serve you
they didn't add zfs because it's a bloated, poorly factored piece of shit that uses tonnes of RAM. There are some nice feature of zfs but not nice enough crap up your system with lots of unnecessary and probably insecure code. When OpenBSD gets those features, it will be done right.
it is harder in a way, I had a lot of experience hacking on linux since redhat 3 days. The difference with OpenBSD is that the man pages and docs are much better than linux and will apply to the exact OS you are using (so while it's bare bones, you do get a lot of help, that is rarely if ever wrong). I alway used to struggle with linux (back in the day) because the docs were distro specific and redhat had all these shitty scripts that did things in a non standard way...
I love linux and OpenBSD, why you fuckers have wars about this stuff? Too poor to own more than one computer...
how exactly is zfs bloat? how is it a poorly factored piece of shit?
>uses tonnes of RAM
oh now i get it, you actually don't understand how ZFS works.
>There are some nice feature of zfs but not nice enough crap up your system with lots of unnecessary and probably insecure code.
ZFS by far with ext4 and xfs are the most robust filesystems on this planet, and ZFS issince over 10 years leading on functionality, capacity allocation and redundancy.
you have just no fucking idea what the fuck you are talking about kid, go post in a desktop thread
>triggered user without ZFS
we do have official support idiot ;)
>what is OpenZFS
and OpenBSD doesnt even have unofficial support
cry some more though :D
>he doesnt understand realtime checksumming
>bloat
poor baby cant even use a archival format
>he still uses raid
Does FreeBSD have ASLR yet?
It all sounds nice, but in reality, how well BSD would work on ryzen 8 core CPU and nvidia GPU?
Oh, and of course, will I be able to use virtualization with BSD and compile my usb wifi adapter driver for it, even if driver is not updated anymore?
Checkmate, atheists.
BSD is obsolete for daily use.
OpenZFS is a fork of ZFS which includes code that is licensed under CDDL, which is why Linux and OpenBSD do not include it into their kernel.
you are just uninformed
since 10 i believe
>dedup/redup
>checksumming
>hashing
>raidz
>zmirror
>zsnap
>importing/exporting pools
>hotswap/coldswap drives
>compression
>zpools which can be "partitioned" and "formatted" without wiping the whole drive
>drive spanning
>ecryption
what do you use as archiving?
BFS?
LOL!!!!
>he doesnt know that ZFS is the longest used archival format in the enterprize
ZFS grew up in the enterprise with SUN and has decades of testing and stability
>There has been a tremendous effort in getting ASLR upstreamed over more
>than two years. We would supply a patch, a few FreeBSD developers would
>review it and make note of a few things we need to improve, we'd make
>those improvements and submit a new patch. [...]
did the freebsd devs write their own implementation?
no you have no idea you stupid fuck, do you think that OpenBSD could not import code from Solaris which is also from the BSD family? There were very good technical reasons that you seem to not understand. OpenBSD developers are far superior to you lincuck.
There is an argument to be made about the license also, which is a lot of the reason gpl people won't use it.
also, ext* filesystems are not robust at all, there are even mailing list posts from the developer showing how the linux version of fsck had to have so many more ways of fixing the filesystem in it than BSD and in some practical ways won out. (but not because the filesystem was good). How about read some code and learn something some day, and stop sucking angry finn cock.
FreeBSD > OpenBSD
You will never get ZFS support loser ;)
>uses checksumming for torrents easily downloaded off the internet.
>redundant resource usage
yes bloat.
the point is it mixes in a lot of stuff OpenBSD already has and is not written well enough to integrate in a clean way. half of the things you listed, OpenBSD already has, but in a cleaner more secure fashion.
SUN are the same faggots that produced java and look at the security mess there, enjoy your faggot zfs applet, enjoy when your false sense of security raidz and checksumming failed because they just hacked that shit into linux without thinking about it.
how do you explain FreeBSD?
face it OpenBSD is a step down
marc.info
you already look pretty stupid, just stop, because you can't damage control this anymore.
its a license issue, always have been, nothing to do with code quality. you are just as uninformed as the other faggots.
there was some fs fuzzing, maybe i can find the slides. in comparison to all existing file systems ext4 and xfs are the most securely designed
also Solaris is not BSD related, SunOS was. nigger get your facts straight, you joke. go post in an amd/intel shill thread, this is to high level for you
you can't really compare a programming language to a file system. a file system does one thing and it does it forever, save data on a disk. a programming language has a lot more complex tasks to accomplish and technology evolved around design patterns, etc.
you are comparing bananas to planets
>corporate cuck license
no thx
>slides comparing linux filesystems and ntfs using genetic fuzzer
...
what is a kernel?
you can literally just read the wikipedia page to see the heritage you wanker. I used to admin sparc stations for a living, lick my scruds pleb.
ok, just linux, i thought they included some bsd file systems as well. still, good enough measure than none at all. i'd like to see something similar for hammer, ufs, etc.
also shove your "..." up your ass.
>everything else is top notch quality
The core system is upgradable since one or two major release. The auto mounter is shit legacy crap and the scheduler is a garbage and so on...
Used it for few months and has much more compromises than on loonix.
must have not been good at it if you didn't even know the heritage
...
No one gives a shit about your "but i dont want to spoonfeed you, noob. read and learn yourself" smoke screen.
Answer his fucking question.
Why compromise for OpenBSD,
when FreeBSD has all the newest gadgets?
The people trying to down play ZFS,
are unknowledgable on the subject.
wew man you sure told me to tell you. i'm actually amazed that you managed to read past the first post
What documented attack would this have protected against? This is security theatre pure and single. Make a lot of noise about how you are making a difference without any evidence, basically trying to sway the ignorant.
The weaknes of this solution is that it requires double hdds, nobody can afford to buy fucking two NASes and waste one on raid
here in english for the uneducated
>Linus HAS NEVER FOCUSED ON SECURITY
Neither did C. C and Linux focuses more on performance.
However OpenBSD and Rust puts security first.
you miss undertand the openbsd philosophy
lack of support/features =/= increased security
:^)
youre a retard you know that?
zmirror is raid 1
theres also raidz which covers 10/5/6/0
>two nasses
you just need 1 nas with two drives or a single drive partitioned
>using a nas without redundancy or any failproofing
what a idiot
probably time based attacks during boot, kernel manipulation, maybe some memory allocation stuff too
>binary blob
>redundancy
>waste
>using a nas
>poor
the whole idea of using a nas is you have more money to spend than just using your pc
also nothing is stopping you from getting flash drives and a usb hub
systemv came related to bsd as well you gymboid. this ain't linux. SVR4 took the FS from BSD... kids these days... it would be easy enough to put zfs in openbsd, they decided not to, ... reasons...
this wouldnt prevent boot attacks that initialized before the kernel though
>security first
Peer review isn't enough for that.
>he uses a single drive as backup
>thinks he needs to use a NAS for this
why use a nas without multiple drives?
youre just complicating shit when you can just use a single flashdrive in your router
no it didn't, how about you check your cs history before you talk with grown ups, kiddo.
literally, stop posting, you making yourself look even more stupid. systemv was a separate implementation of unix next to bsd, read those fucking wikipedia articles you uneducated piece of shit
well, load order appears to be linked, therefor i assume, that you can't inject modules during boot time
Don't scar the poor lad.
lol
I want to use *BSD but openbsd doesn't support 3d acceleration my old Nvidia card, and I don't want to use proprietary software (nvidia-driver) with FreeBSD. NetBSD has basic nouveau support, which looks perfect, but is it a viable desktop OS?
> BSD is a cuck licence! if nobody gives back the project will die!
> Somehow it's superior to Linux
hhhhhmmmm...
really makes u think
>lincucks
>They're so beholden to this tribal identity
oh the ironing
Pic related.
>UFS
well....
i..
uh..
hmm..
penguins r stewpeed :^)
hue hue
so you are agreeing now that they imported features from BSD which is a seperate implementation?
ok wow it they took the FS, great you won a tiny argument by shifting from your initial statement where you implied that Solaris was *BSD based.
Isn't this a good thing? I don't understand the criticism and lulz coming from this thread.
Not the same guy, lad; I just knew what he was talking about.
im surprised you can follow this dudes conversation flow. started with zfs support on openbsd and now we are back in the 80's talking about UFS being introduced in systemv
we're talking about the FS and how easy it would be to put ZFS in OpenBSD. It's a related implementation at best, look at some code and learn unix, this is not your linux hacked together shit. (I do like linux for things though, cos hacking shit is what you need some times, just gotta realise it's inferior to doing it properly).
if you read the original point was that it would not be hard to put zfs into openbsd, but they didn't, cos they had reasons, SVR4 is close enough to BSD to do it ..
What are you on about? UFS and ZFS coexisted in Solaris for many years. You couldn't have /root ZFS until recently (using UNIX timeline scales)
/thread
It was developed by the PaX team for linux. Linux has had it available a whole decade before openbsd. But you people are the most intellectually dishonest in the world so I wouldn't expect your version of reality to accord with the truth.
it is not even hard to put it in linux but as mentioned somewhere above, it has always been an issue with the license. because CDDL is not an open source license. as far as i understand it oracle still has the last say in it, if they change CDDL to something else openzfs would die