gdgghhhhh

So, they "decided" it's not a derivative and thus can be listened under MIT instead of GPL....

https://nvd.nist.gov/vuln/detail/CVE-2026-21250

Oh. I was not aware of that

/proc is a Linux thing.

It's amazing how many of the "experts" here don't get that bcachefs != btrfs

Also consider putting Firefox itself into a jail. E.g. using bubblewrap on Linux: https://gist.github.com/richardweinberger/cae9edeafeec4cdf65...

FWIW, the missing source has been released.

In this context, this is worth a read: https://hachyderm.io/@josefbacik/114755106269205960

politely, yes.

I wonder how well it works with heavily nested files. AFAIU, this is a recursive parser.

The title really confused me until I realized this has nothing to do with https://millcomputing.com/ :-)

Speaking of helix, does is support vim key bindings? Last time I tried, it was kinda disappointing to be forced to learn a completely key binding set.

I wonder how the code looks, it must be a maze of "on error goto" lines to not crash at the slightest error.

The program doesn't look like a typical VB6 application, I guess the author draws all UI elements himself?

No Kent, really.

I suggest reading the whole mail thread before judging. The CoC decision was the last resort.

Don't get me wrong, OSS does not mean it will build in any random configuration on any distro. The source is there. Maybe 7zip could do better, but still...

Perf trace is also worth noting. It was much less performance overhead than strace. https://man7.org/linux/man-pages/man1/perf-trace.1.html

Since you are writing to the location where rip is pointing to, you need to rewind to the beginning of the page first. Otherwise the program will crash if you're unlucky and rip is currently exactly at the end of an executable mapping. Low chance, but still...

One use case I have faced more than once are stuck processes in legacy build systems. For example, generating docs takes ages. Just killing the generator process will fail the build because the parent will notice. And since the build system is buggy/old, restarting the build will start from zero.

To support a block based filesystem on top of MTD an advanced FTL (Flash Translation Layer) is needed. Sadly most algorithms for decent FTLs are patented. SSD vendors try hard keeping their magic sauce there own. In Linux we habe some trivial FTLs implemented, but even these are toxic.

I don't get it. Isn't this just a live CD that setups nvmt like described here? https://blogs.oracle.com/linux/post/nvme-over-tcp

My best guess it's the animated SVG. Maybe it should get lazy loaded.

There is also smatch, it is used by the Linux Kernel community https://github.com/error27/smatch

Funny to see them used as example by a tutorial from Red hat:

https://www.redhat.com/sysadmin/linux-access-control-lists

Just share an open FD to a socks proxy with the isolated service.

So what? Don't get ne wrong, tuxera is cool but this still does not make it free software.

I fear it's closed source?

I wonder how it compares to ELBE. https://elbe-rfs.org/

That's why Linux has the ORC unwinder. It is fast and reliable, in contrast to frame pointer unwinding.