channel logs for 2004 - 2010 are archived at http://tunes.org/~nef/logs/old/ ·· can't be searched
#osdev2 = #osdev @ Libera from 23may2021 to present
#osdev @ OPN/FreeNode from 3apr2001 to 23may2021
all other channels are on OPN/FreeNode from 2004 to present
http://bespin.org/~qz/search/?view=1&c=osdev2&y=21&m=10&d=1
00:23:00 <Izem> anyone doing implementation in a concurrent language?
00:25:00 <klange> There are some Rust projects but I don't know if any of them actually make use of concurrent features. Most concurrent languages need a system backend for their concurrent functionality, which leads to a bit of a chicken/egg problem for OS dev.
00:26:00 <Izem> ohhh
00:26:00 <Izem> for some reason I though the language could provide it
00:28:00 <gog> usually the toolset relies on some library code that's necessarily tied to the OS
00:28:00 <gog> for threading and such
00:30:00 <Izem> hmm, well I guess I have to try another direction then
00:31:00 <klange> Kernels are scary places where languages doing concurrency behind the scenes is a route to disaster. Even when you're doing it yourself it can lead to awkward situations...
00:35:00 <Izem> I see, thanks
00:56:00 <Izem> is the toolchain what enables a language to be osdev friendly?
00:56:00 <zid> The language has to be turned into code, that code has certain requirements on the environment it expects
00:57:00 <zid> java bytecode expects there to be a JVM
00:57:00 <zid> python expects a python interpreter etc
00:57:00 <zid> that means your kernel, in order to use languages like this, has to also be a python interp, or a JVM, etc
00:57:00 <zid> People use languages like C because C's requirements are basically "Some amount of bytes will need to be set to zero"
00:58:00 <zid> which you can do in 4 lines of assembly, and that's the entire environment sorted
00:58:00 <klange> oops https://github.com/kuroko-lang/kuroko-efi
00:58:00 <zid> if you're thinking of something with built in threads, that environment aspect means you're now supplying a threading library as part of the runtime environment
00:58:00 <zid> which means lots of code NOT written in that language, at the very least
01:00:00 <klange> Speaking of, someone is porting their own JVM implementation to ToaruOS, they sent me screenshots of working Swing UIs under SDL.
01:00:00 <klange> But they keep trying to talk to me at 2am :)
01:01:00 <moon-child> ha, nice
01:02:00 <gog> that's fine, your cats will wake you at 3
01:03:00 <klange> My cats don't wake me at 3, they keep me up until 3.
01:03:00 <gog> ah fair enough
01:03:00 <zid> Keeping your cats up until 3am playing with them smh
01:03:00 <gog> they're crepuscular zid
01:04:00 <gog> https://www.youtube.com/watch?v=qPaXG0oEInY
01:04:00 <zid> 3am is only sunset in iceland
01:04:00 <kazinsal> I'm trying to NIH as much as I can but damn if it's not tempting to port kuroko to make rapidly developing features faster
01:04:00 <gog> nah it's more like 19:30 now
01:13:00 <klange> kazinsal: I would say "why do you think I wrote it" but then I haven't really done much with it since...
01:14:00 <klange> But then, I've been focused on kernel stuff lately with Misaka. Kuroko did get a Yutani module and test app, and I did port some of my build tools to it... and I've been using it to test some socket stuff.
01:17:00 <kazinsal> Definitely. One of my near-term goals is to get sockets "functional enough" as well, as well as some better network queuing, then get an actual proper CLI going and accesible over the network
01:21:00 <kazinsal> But this evening I have a different kind of networking task to work on: Rebuilding my actual dev network!
01:22:00 <Izem> zid: thanks, that does give me an idea for a small vm
01:23:00 <Izem> kazinsal devices drivers across a network?
01:24:00 <kazinsal> Izem: Nah, literally deploying some new VMs and re-IPing a couple others
01:24:00 <kazinsal> And LDAPifying the whole thing
01:24:00 <Izem> ah
01:24:00 <kazinsal> That's going to be the trickiest part I think
01:26:00 <moon-child> 'LDAP' whyy
01:26:00 <kazinsal> SSO across all my machines running different OSes
01:28:00 <moon-child> why would you want SSO
01:28:00 <kazinsal> Because all the machines are on my network?
06:24:00 <geist> kazinsal: allow me to sign up for your newsletter
06:24:00 <geist> i'd love to do the same thing (SSO for everything)
06:25:00 <kazinsal> the centos box is fighting me but the rest are slowly falling in line
06:25:00 <zid> pam + mysql? :P
06:25:00 <geist> any luck with windows?
06:25:00 <zid> That's how I handled it when I did it for a whopping couple machines
06:26:00 <kazinsal> the windows boxes are happy because the auth server is an AD domain controller
06:26:00 <geist> yah i'd almost like the LDAP box to be some sort of dedicated embedded like controller thing
06:26:00 <geist> like a raspberry pi, or whatnot
06:26:00 <geist> then it has no deps on anything else
06:27:00 <kazinsal> yeah, I've never tried joining windows machines to openldap
06:27:00 <geist> i wonder how windows 11 is supposed to deal with that, since they're pushing your harder and harder to using msfts loging
06:27:00 <geist> but clearly they will still support LDAP and whatnot
06:27:00 <kazinsal> yeah, it's oging to be interesting. we're moving to azure AD + intune at work
06:28:00 <kazinsal> but I'm still setting up domain controllers on prem for customers so this is also a good chance for me to play with server 2022
06:30:00 <junon> is it bad that I kind of miss seeing the long spiels of that one individual who comes in periodically?
06:31:00 <Mutabah> junon: yes
06:32:00 <geist> yes, yes it is
06:32:00 <geist> i do miss doug16k a bit though
06:32:00 <geist> we used to muse about some of the finer points of PC esoterica
06:34:00 <kazinsal> guess he never bothered adding libera to his irc client
06:46:00 <junon> :D
06:47:00 <geist> yeah, amazing how quickly i mostly forgot about all of that
06:48:00 <geist> after about 2 weeks the whole freenode thing hadn't crossed my mind
07:47:00 <junon> Yeah they did a really good job of transferring it.
07:48:00 <junon> Pretty impressed.
07:56:00 <k4m1> huh, I never even realised doug16k isn't here, used to learn a bit by just lurking & reading his stuff
08:00:00 <k4m1> doug/2
08:00:00 <k4m1> huh, soz
08:01:00 <klange> I know we had a handful of detractors, was doug one of them or just someone who never showed up again?
08:03:00 <moon-child> I'm pretty sure doug was on libera for a while
08:03:00 <moon-child> checked up on his github after he hadn't shown up for a couple of months, seems to still be alive at least
08:04:00 <zid> yea he was here for a bit
08:34:00 <vai_> well howdy
08:36:00 <junon> hi!
11:31:00 <vai_> hi
11:31:00 <vai_> hmm
13:43:00 <Oli> Hello, and good day! I want to comment you about about that I find myself thinking about doug16k, and mean to inquiry you about: Is he alright? Has he been around?
13:51:00 <gog> :| wonder where he went
13:55:00 <opios2> who made him angry? :P
13:56:00 <gog> logs indicate nothing. he just vanished
13:56:00 <gog> he's made a github commit in the last 30 days tho
13:57:00 <gog> so he's not dead it seems
15:53:00 <junon> find the commit, get the email, shoot him a message asking where he's been
15:56:00 <gog> maybe there's a reason he's taking a break from irc
15:56:00 <gog> and it could be being contacted by randos
15:56:00 <gog> i'm not exactly a rando but i'm also not going to intrude :p
15:56:00 <EtherNet> who is doug16k by the way ?
15:57:00 <gog> old regular
15:57:00 <gog> hasn't been seen for a couple months
15:57:00 <EtherNet> I hope he's okey then
17:33:00 <heat> hello
17:33:00 <clever> hello
17:33:00 <heat> sup
17:33:00 <clever> not much
17:38:00 <sortie> a lot
17:38:00 <kazinsal> just enough
17:38:00 <kazinsal> (actually sortie's right, it's always too much)
17:38:00 <kazinsal> (I should wander off into the forest and become a hermit)
17:39:00 <sortie> Just don't use any names in the forest and you'll be fine
17:39:00 <heat> _Tree
17:39:00 <heat> :O
17:39:00 <kazinsal> the place where the nameless are found
17:39:00 <sortie> Oh wait damn I violated the procedursada--- DAHSD
17:39:00 <kazinsal> congratulations, you are now part of the woods which have no name
17:40:00 <kazinsal> I should read that one again, it's really good
17:45:00 <heat> i had a bunch of uncommitted work in my project and after this long hiatus I have no idea what the status of the code is
17:45:00 <heat> welp, lets run it, see if it works
18:02:00 <heat> anyway, I've been looking at build system ideas so I can finally fix whatever broken thing I've got going on here
18:03:00 <heat> i found out chromeos actually uses portage
18:08:00 <heat> my options are: 1) try to use an existing build system and suck it up (sucks for external packages); 2) fork an existing build system and try to adapt it; 3) write from scratch
18:08:00 <heat> ideally I'd use something that spits out ninja
19:13:00 <kazinsal> ahhhh, solving problems with openbsd. my job is great sometimes
20:11:00 <geist> yeah chromeos linux is iirc based on gentoo
20:11:00 <geist> or at least was started that way, hence using portage
20:14:00 <sortie> The feeling when my OS is affected by the latest libreessl / let's encrypt / root cert expiry problem
20:14:00 <sortie> My little Sortix OS gets to join in on the cool breakage in the proper OpenBSD system
20:14:00 <j`ey_> not just OpenBSD is affected!
20:14:00 <sortie> Yeah just saying they're the source of my breakage
20:15:00 <sortie> Since I get libressl from them we share CVEs
20:15:00 <j`ey_> ah
20:15:00 <sortie> I'm still at the osdev stage where I go ‘oh cool the latest cool CVE affects my OS’
20:17:00 <j`ey_> my OSdev has been refactoring/working on a GPIO driver for the m1 macs
20:18:00 <sortie> Oh that's cool
20:18:00 <sortie> A driver for your OS or an existing one?
20:18:00 <j`ey_> for Linux
20:18:00 <sortie> Neat
20:18:00 <j`ey_> it's needed to get the keyboard working on the macbooks
20:19:00 <sortie> Sounds pretty high impact work actually
20:19:00 <j`ey_> someone else did most of the hard work, ive been rebasing and fixing it though
20:27:00 <geist> ah interesting, the keyboard isn't just some microcontroller somewhere else?
20:27:00 <geist> gpio driver sounds like you're having to scan it yourself
20:28:00 <j`ey_> ah no, the gpio driver deals with the SPI controller, which connects to the keybiard
20:28:00 <clever> that reminds me of some rossman repair videos
20:28:00 <clever> on other macbooks, the touchpad is both usb and spi
20:28:00 <geist> ah that makes sense
20:28:00 <clever> during the bios, and under linux, it operates in usb mode
20:28:00 <clever> but darwin itself, switches it into an spi mode
20:29:00 <clever> BUT, the SPI traces, are also in a worse place, in terms of water damage
20:29:00 <clever> so you can damage the board in such a way, that the mouse only works in the bios and linux
20:29:00 <j`ey_> hah
20:29:00 <clever> because darwin lacks usb drivers for its own mouse
20:29:00 <gog> lol
20:29:00 <gog> love those rossman repair videos
20:42:00 <zid> The SD card slot on my wii is like that
20:42:00 <zid> it will boot from SD card, but you can't use the SD card in any software
20:42:00 <zid> because the spi pin works and not the proper data pins
20:43:00 <clever> zid: ive considered that, for some of the rpi boot problems
20:43:00 <clever> zid: ive run into a few users where recovery.bin can load from the SD card (maskrom routines), but then reading any file from the SD card, fails the hash checks
20:43:00 <zid> yea sounds likely then
20:43:00 <zid> it took me a while to figure out the fuck was happening
20:43:00 <clever> but, the fact that it even read the correct hash from the SD card
20:44:00 <zid> like, did SD cards come in different speeds and my card was failing at its top rated speed or whatever
20:44:00 <clever> and other testing we did, like giving it a smaller file, and looking at the error hash
20:44:00 <zid> or was there a bitflip in the higher level read routines on my nand flash etc
20:44:00 <clever> in theory, you can find an old enough card, that advertises itself as 1bit only mode
20:44:00 <zid> I ended up just researching how sd cards worked and found out
20:44:00 <clever> and that should then work
20:45:00 <zid> I thankfully don't actually need the SD for anything except the bootloader
20:45:00 <zid> the bootloader supports USB
20:45:00 <clever> ah, nice
20:45:00 <clever> my bootloader still lacks a usb stack
20:45:00 <clever> so my only choice is xmodem over uart, or SD
20:45:00 <zid> well, the bootloader boots, and eventually, what I want to do, is supported over usb
20:46:00 <zid> I leave the SD card in the wii so it'd be nice if I could leave more than just the bootloader on it, a couple other vital apps or whatever
20:46:00 <zid> them having to be on the USB is totally fine, until I lose it