IRC channel logs


back to list of logs

<azert>He made a huge rant years ago about c++
<azert>After trying introducing it in the kernel and failing
<azert>But that was an older version the lang evolved a lot
<azert>Samuel recently said something about nested function and nx bit. Can Hurd use C++ lambdas instead of nested functions ?
<gnucode>The Hurd did just get a port of Rust. Maybe we could use that
<azert>c and rust are not compatible, you’d have to rewrite the kernel
<azert>Think about c macros..
<azert>With c++ it’s straightforward
<dsmith>Rust has pretty good C interop
<azert>I think Rust could be great for translators, on the other end. But lots of work to make that a possibility
<azert>dsmith nope
<azert>It’s not even rust fault
<azert>It’s c that is a language designed to lock you in
<gnucode>rust for translators (with lots of work). I think Zig, Odin, or hare might be interesting too.
<azert>I wonder at which level one would do the bindings to allow rust for translators. Hurd libraries or ipc interfaces?
<gnucode>azert: You could do the C bindings I believe fairly easily.
<gnucode>It is usually encouraged to hook in "lower".
<youpi>gnucode: never believe what's on reddit
<youpi>mach itself is rather 80,000 lines of code
<youpi>rust-based translator should really rather plug at ipc interface
<youpi>there could be some interest in using the existing libdisk/net/trivfs, but quite a lot of plumbing, while reimplementing libdisk/net/trivfs in rust does make sense
<gnucode>I'm fairly surprized that you are ok with re-implementing libdisk/net/trivfs in rust.
<youpi>why not?
<youpi>if it's there for good, why not using it
<gnucode>fair enough. Some C developers are somewhat anti-rust
<gnucode`>hello again!
<gnucode`>ahh, bummer. I forgot to log out of my other irc client.
<gnucode`>well I am already running into problems...
<gnucode`>mke2fs in not installed grr.
<gnucode`>maybe it's not installed in my regular user path...
<gnucode`>yup that was the problem.
<gnucode`>that was silly.
<gnucode`>hmmm the crosshurd is failing due to lack of gpg keys...
<gnucode`>it even says no_PUBKEY 23902390239203C NO_PUBKEY 2320390293B36
<gnucode`>I did a gpg --recv-keys 34343940934C for both keys...
<gnucode`>then ran cross hurd again.
<gnucode`>same issue.
<gnucode`>I think I found a work around online.
<gnucode`>well I apparently need to regenerate my grub configuration.
<gnucode`>well my hurd grub file didn't seem to work. GNU Mach complained that partition 1 didn't exist.
<gnucode`>and I hard shutoff the machine, because it seemed stuck at boot.
<gnucode`>So that means I probably now have filesystem corruption. haha.
<gnucode`>nevermind. No corruption.
<solid_black>gnucode gnucode`: most of the port has to do with Lagom (SerenityOS userland libraries & services) specifics, not with the UI
<solid_black>but I did use the Qt chrome and not my GTK4 chrome, yes
<solid_black>I was going to try to build the GTK chrome next, but I haven't allocated a timeslice to hacking on that lately
<solid_black>I'd be very interested in a new Rust framework for Hurd translators! so if anyone wants to hack on that, hit me up & let's hack on it together
<solid_black>gnucode: many C developers are anti-Rust indeed, but it happens that apparently many Hurd people are also into Rust (even independently of Rust-on-Hurd), myself included
<solid_black>so it's a good match socially
<solid_black>also, your qoth made its way onto phoronix and
<solid_black>... and HN
<azert>Curious that if you look at the comment section (don’t) of phoronix it might start good, until eventually a sore loser pops out and catalyzes the attention on some bogus arguments. Everyone takes the bait
<damo22>welcome to the internet
<ThinkT510>they couldn't even stay a full minute
<youpi>opinions on the hurd are basically never moderate, it's either an all praise or an all boo, you'll have to get used to it by just ignoring it :)
<janneke>how can you ever get used to something if you ignore it :-P
<janneke>(fwiw, i'm all for ignoring)
<solid_black>I did look at the comments section of phoronix, and was pleasantly surprised by how wholesome the first few comments were, esp. compared to LOR
<Gooberpatrol66>i feel like a problem with hurd is it's just obscure enough that techie people have heard about it, but nobody actually knows much about it, so people can spread misinformation or outdated information about it with no pushback
<Gooberpatrol66>it tends to attract word-of-mouth
<Gooberpatrol66>also its past history condemns it. it doesn't matter the current state of the code, the fact that it's 30 years old means it's doomed. if it were less than 5 years old, people might see it as an exciting up-and-coming project and might look deeper into its design than "why doesn't my gaming mouse work with it?"
<emru>just change the name of the project xD
<Gooberpatrol66>or "HINH"
<Gooberpatrol66>"HINH is not Hurd"
<emru>"Hurd 2.0"
<emru>"Plan 10"
<gnucode>hello friends!
<gnucode>sneek later tell solid_black Yeah I guess I underestimated how important a qoth is for the Hurd. That's a great way to update people apparently.
<sneek>Got it.
<youpi>gnucode: actually, ideally we'd dig back in the commits to produce previous qoth :)
<gnucode>I actually started working on the 2023 q3 qoth. It would be nice to let people know that we successfully ported rust.
<gnucode>youpi: by the way, crossinstall seemed to work, but I got stuck on the "Preparing to boot" ext2fs can't seem to find the SSD. I think I can work around it by going into the BIOS and telling SATA to pretend to be IDE...
<gnucode>either way, I can actually use the Hurd on bare metal (which seems to work best for me) on a T43.
<gnucode>I'm also assuming that you fixed git send-email. If so, thanks. I am able to edit the Hurd wiki using the Hurd now.