-
swee
how do i get plasma to do screensharing here?
-
swee
and more importantly using pipewire instead of pulseaudio
-
swee
oh it works on firefox, but not on a linuxulator app like Discord
-
swee
-
radhitya
sorry i cant help you swee
-
anth
What’s the simplest way to get my install to be able to resolve .local names? Avahi looks lijenit pulls in dbus, which seems like a lot for this.
-
codegirl
Hello. Does anyone know of any C (or C++, but for the sake of god, no rust/go/js with npm/yarn stuff, etc) RDAP client that would help me when `whois` fails?
-
codegirl
Not a freebsd user (yet), but perhaps someone using freebsd knows of something
-
Gedge
codegirl:
freshports.org/dns/rdap (which is golang) or
freshports.org/net/rdapper (looks like perl)
-
Gedge
codegirl:
-
Gedge
oops
-
codegirl
Thanks a lot, gedge
-
Gedge
I just did this: ls -d /usr/ports/*/*rdap*
-
Gedge
there may be better ways :)
-
codegirl
Do you think perl would be a risk for the same kind of issue that JS&npm, rust&cargo, etc would suffer? I mean, supply chain attacks and such?
-
Gedge
That's a hard question to answer, because the answer is (always) "yes, there's a risk for [insert language here]"
-
Gedge
But, using an O/S package (whether it's perl/go/js/rust/...) probably reduces your exposure. Though, not entirely.
-
SomeVisitor
the Perl equivalent to JS/NPM and Rust/Cargo would be Perl/CPAN
-
cyric
anth: you probably want mdns in nsswitch.conf via dns/nss_mdns port
-
mns
codegirl: every language has a supply chain issue when third-party libraries are being used in the development process. I don't think there is a "language ecosystem" around assembler, but I could be wrong on that. Outside of that, most languages have an ecosystem around them and hence they have a supply chain problem. my humble opinion.
-
codegirl
mns, indeed, I think you're right, maybe I should think why I feel safer with standard shared libraries on a unix system
-
codegirl
Maybe I'm departing from broken assumptions
-
rwp
codegirl, I think standard shared libraries to me means /lib/* /usr/lib/* which are part of base and the people who can commit there are a smallish known set of historically good people. Then layered on are /usr/local/lib/* libraries coming in through ports and those have supply chain possible problems.
-
nimaje
shared libraries don't do anything to help against supply chain attacks
-
rwp
But hopefully if something has come through ports I am hoping that someone would notice a problem. Though that is not guarenteed.
-
rwp
In this case a supply chain attack would be someone puts something bad in libxz(sp?) and it gets into the ports tree and installed everywhere compromising anything that uses a standard compressor. Oh, that's already somewhat almost happened did happen in some systems.
-
codegirl
Indeed, I'm aware of the xz issue, but...
-
rwp
That was exploiting other weaknesses in other systems that we thankfully avoided by being better but it illustrates the potential for things coming through ports.
-
nimaje
well, what helps for stuff that is part of base is the slower release cycle and more reviews, but that doesn't make immune against supply chain attacks
-
codegirl
When you compare how many times shared libraries (the most actively maintained and with more eyes on it), it's a rare event that something like what happened to xz happens to, say, curl
-
codegirl
I mean, of course it's not immune, but it's quite rare that someone is able to put a malicious code inside some library that gets to do that and it goes all the way unnotice, while on stuff like rust's cargo it's quite easy
-
codegirl
Perhaps the release cycle speed is indeed something to be valued here
-
codegirl
But I think the complexity net of interdependance is also a possible weakness on the models used by rust and other languages
-
rwp
Release cycle speed is definitely a factor. Slower releases give time for reviews to happen. Fast churning releases are impractical to review because tomorrow the code is completely different anyway.
-
SponiX
This is why I'm following stabilization week for Current Again
-
ant-x
codegirl, Thanks for no supporting rustcancer.
-
» checkpoint thinks it's a matter of time, rustcancer will get to BSD world sooner or later
-
mikulvoid
why rust in FreeBSD or any BSD for that matter is a bad thing?
-
checkpoint
mikulvoid: because only few ppl can code in Rust, nobody kwows Rust en mass as C. Rust is very complex language comparing to C.
-
» checkpoint does not know Rust and do not want to lear it. This means he won't be able to fix bugs or even report them.
-
mikulvoid
what language would you like to see in BSD?
-
ant-x
How's about the principles, philosophy, and ideology of a language, e.g.: <
felipec.wordpress.com/2025/02/13/rust-not-for-linux> ?
-
ant-x
Now, Git has a dependency on Rust. :-) or :-(
-
ant-x
In BSD, I'd like to see C, Pascal, perhaps Hare, Seed7...
-
mjp
sounds like the guy just doesnt like rust
-
mjp
no good technical arguments in there
-
» checkpoint thinks that C must be used in 99% of BSD code base, maybe some Lui or Perl then heavy scripting is necessary.
-
checkpoint
C is widely spoken, everyone can join the development, report bugs, provide support. Rust is not. Rust is a barrier.
-
ant-x
The appeal to popularty is a false argument: C was no popular when it was first created. Languages are born, flourish, and die.
-
ant-x
checkpoint, I am glad you did not bring Python in. Lua, Perl, and Tcl are all better, IMHO.
-
checkpoint
ant-x: C was and is very popular since 70th. it is very easy to learn, it's easy to read and understand code in C. It's being taught in universities. Rust is much more sophisticate, it's even more complex than C++ I think, noone teaches Rust.
-
ant-x
I was shocked to learn that Universities start teaching not pure computer-science but whatever is trendy in the corporate culture :-(
-
checkpoint
ant-x: maybe because corporations fund them ? depends on univesity though. Where I studdied back in 90th they still teach Pascal.
-
ant-x
Back in the 90-ies yes. But now?
-
ant-x
They may be attracting studient by teaching the trenty sh&t.
-
checkpoint
trendy sh&t is vibe coding, it does not require teaching :)
-
ant-x
Vibe coding in Rust :-)
-
checkpoint
what's puzzle me most is that in many unis nowadays there are some tutors/teachers who require extensive use of LLMs, same time there are those who strictly oppose that. Poor students fail to carve their learning curve.
-
checkpoint
ant-x: Vibe coding in Rust - that's what Canonical is doing right now.
-
ant-x
My colleagues think those encouraging the use the Aritficial Idiot are clever and progressvie...
-
ant-x
What's Canonical?
-
ant-x
What Canon allows vibe codeing instead of celebrating human excellence?
-
checkpoint
ant-x: the Ubuntu developers, AFAIK.
-
ant-x
A modern pun: "My C is Rusty".
-
ant-x
Yikes... AI-cancer is permeating even free software...
-
ant-x
By the way, I don't know much C beyond C'89.
-
checkpoint
ant-x: c89 is pretty much sufficient to understand BSD kernel code
-
ant-x
I am not tryting to do that.
-
checkpoint
ant-x: maybe you are not just ready for that, morally. :)
-
ant-x
OT: Does anyone know whether alloca() is required to present in the standard library by all C standards?
-
ant-x
May be...
-
ant-x
* to be present in
-
checkpoint
ant-x: I don't think alloca() is in C standard, it's a compiler's feature (sub rsp, XXX)
-
ant-x
Non-standard, unlike VLAs in a later standard. Thanks. I wonder why alloca is still available in GCC with -std=c89 . Not very strict, perhaps.
-
checkpoint
ant-x: "man alloca" says: The alloca() function appeared in Version 7 AT&T UNIX/32V. Basicall it's pre '89.
-
rwp
gcc --std=c89 enables those standard features but doesn't disable non-standard features.