ori_b

29 ms·

661.
by ori_b 4y ago
The only way a program can interact with the system is through system calls. Can you explain why irreversibly taking away a program's permissions to do anything interesting, especially as root, is a bad approach?
662.
by ori_b 4y ago
I'm not being smug or insulting when I ask this, but are you familiar with pledge and unveil at all, or have you written any code? Pledge and unveil maintains it's own set of system calls and allowed paths that a process may acces
663.
by ori_b 4y ago
Pledge means YOU HAVE NO SYSCALLS. There is no amount of root magic that can allow the code you've exploited to do IO.
664.
by ori_b 4y ago
If you believe that there's a way to become root, why don't you believe there's a way to become some other non-root user? Possibly by using the same exploit you'd use to get root, or possibly by using the permissions roo
665.
by ori_b 4y ago
[flagged]
666.
by ori_b 4y ago
By your logic elsewhere -- if you're root, you can just remove the selinux policies.
667.
by ori_b 4y ago
[flagged]
668.
by ori_b 4y ago
How? you don't have access to the devices. They're not in your namespace. Again, proof of concept please.
669.
by ori_b 4y ago
> If someone has remote root you are telling me they can't unhide a filesystem? Yes. There is no mechanism at all to unmask the vnode. If you believe otherwise, I'd like to see a proof of concept.
670.
by ori_b 4y ago
> Hiding filesystems is security via obscurity more than anything else, and probably won't hold up if someone has root. Can you show me a proof of concept here?
671.
by ori_b 4y ago
"When you've already been compromised, OpenBSD doesn't prevent you from being compromised"
672.
by ori_b 4y ago
It ships with many of the services you may want by default (httpd, relayd, bgpd, etc). Major external packages are patched to improve security. For example, a compromised chrome or firefox can't observe any directory other than ~/
673.
by ori_b 4y ago
It's comfortable. And it works with the hardware I own just fine.
674.
by ori_b 4y ago
Posted from 9front (using netsurf). I also run https://shithub.us , which is hosted from plan 9, and more or less needs plan 9 for push access. It uses the hjgit:// protocol for push access, which requires plan9 auth to
675.
by ori_b 4y ago
Can you explain how a compiler topologically sorting the entries in .init_array makes a library unusable? What breaks?
676.
by ori_b 4y ago
Libraries exist with a well defined compiler. The solution is at the compiler level, using information coming from the import graph that exists in languages with a module system. And I'm not sure how you can forget to import a library
677.
by ori_b 4y ago
??? If you generate a single initialization symbol that guards against multiple runs in the init section, this works just fine with dlopen. This isn't even hard, but you need to be able to form a full dag of imports. The only problem i
678.
by ori_b 4y ago
> If it was "relatively easy" then there wouldn't be a million dependency injection solutions for Java of ungodly runtime cost to try and solve this problem. You realize Java has the feature you're discussing, without
679.
by ori_b 4y ago
def loadlib() { lib = dlopen("libfoo.so") sym = dlsym(lib, "pkgname.__init__") sym() return lib } Hell, you can even do this with a single constructor attribute that runs
680.
by ori_b 4y ago
No, you're not. The fact that a misfeature exists in the toolchain doesn't force you to use it. You can order the init sections on your own when you generate the binaries and add your own constraints, and if your toolchain doesn&#
681.
by ori_b 4y ago
It's relatively easy to solve the problem in a language with acyclic imports by doing a topological sort of the imports, and initializing in that order. In C++, with textual includes and no standardized control over the linker, you don
682.
by ori_b 4y ago
> But there's also less of a reason for anyone with permission to read DMs @ Twitter to do so, possibly with logging for any audits into unauthorized access. https://www.telegraph.co.uk/news/2021/07/
683.
by ori_b 4y ago
Probably. But not easily.
684.
by ori_b 4y ago
Quite often, yes. ML is unlikely to be one of those places, but appealing to the efficiency of large, bureaucratic companies is a poor argument.
685.
by ori_b 4y ago
Keyboard navigation is one of the features of this release. And it's fully scriptable via /dev/kbtap: you can write your own key bindings and compose them by piping the keyboard events through your scripts. The usual program
686.
by ori_b 4y ago
Depends on what you find notable. An IME for CJK languages, as well as some others, didn't make the list. ( http://man.9front.org/1/ktrans )
687.
by ori_b 4y ago
Sort only uses '<'.
688.
by ori_b 4y ago
> Even FreeBSD lacks anything but a half-baked 802.11n implementation :/ OpenBSD does better here. AC is supported on Intel cards.
689.
by ori_b 4y ago
Ah. I see How much housing should be built before you decide to go homeless instead of paying what colluding landlords decide they can extract from you? The issue here is the collusion in place of competition.
690.
by ori_b 4y ago
Yep. In practice, make gets bigger. cmake also ends up looking like this: https://github.com/opencv/opencv/tree/4.x/cmake Trivial examples look nice enough in either system, big ones look bad in make and

More ›