klange changed the topic of #osdev to: Operating System Development || Don't ask to ask---just ask! || For 3+ LoC, use a pastebin (for example https://gist.github.com/) || Stats + Old logs: http://osdev-logs.qzx.com New Logs: https://libera.irclog.whitequark.org/osdev || Visit https://wiki.osdev.org and https://forum.osdev.org || Books: https://wiki.osdev.org/Books
Lucretia has quit [Remote host closed the connection]
<cloudowind> most the packages ive compiled for my own little linux distro belongs to gnu
<cloudowind> i do see gnu as other part of linux , it completes linux am i wrong?
<heat> two categories of GNU packages: 1) maintained and worked on by unpaid voluntary contributors
<heat> 2) maintained and worked on by big corporate backing (gcc, glibc, etc)
agent314 has joined #osdev
lukflug has quit [Quit: Leaving]
saulosilva has joined #osdev
saulosilva has quit [Quit: Client closed]
Fingel has quit [Quit: Fingel]
heat has quit [Ping timeout: 246 seconds]
dysthesis has quit [Ping timeout: 264 seconds]
frkazoid333 has joined #osdev
m5zs7k has quit [Ping timeout: 265 seconds]
m5zs7k has joined #osdev
agent314 has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
<cloudowind> those volunteers i do admire
<cloudowind> uhm hes gone
<kof673> heat reads logs or similar, so you can badmouth him while he's "gone" and he will be fine with it, you are not talking behind his back
<kof673> fsf stuff i never had a problem with, but it is all assuming their are courts and "law" in place, that they are followed, that the normal person has access to them, etc.
<kof673> *there
<kof673> and AFAIK because they let an organization/company/corporation/group be an "individual" ....they are presumably just following "the law"....one can argue for or against that, but i am unaware of any license that does not allow that
<kof673> so an "individual" can "share" amongst itself and this is not "distribution" -- is this the intended behavior, or just a quirk of "the law" allowing various group entities to be "individuals" ? "you are individuals!" -- <chanting> we are all individuals! </chanting> </monty python>
pross has joined #osdev
<cloudowind> hmm
frytaped has joined #osdev
<cloudowind> yes , even our names are pointers they point to us:)
edr has quit [Quit: Leaving]
<cloudowind> geist probably pssed of with our offtopic chat:) i need to finish this installer script anyway , wishing you the best one today
zid has quit [Read error: Connection reset by peer]
zid has joined #osdev
Fingel has joined #osdev
dysthesis has joined #osdev
housemate has joined #osdev
duderonomy has joined #osdev
<geist> wat?
<geist> nah i'm not an angry person
dysthesis has quit [Remote host closed the connection]
<Ermine> volunteers are more oppressed by gnu/fsf than supported; and big corporate employees obviously don't need support from fsf
housemate has quit [Read error: Connection reset by peer]
craigo has joined #osdev
housemate has joined #osdev
cloudowind has quit [Ping timeout: 244 seconds]
cloudowind has joined #osdev
* nur wants to travel soon after 5 years of not leaving the country, or even my house. I wonder if there any osdev meetups
<childlikempress> some stuff at usenix
goliath has joined #osdev
<bslsk05> ​twitter: <AviSchiffmann> Microsoft has never made a good product. Windows is absolutely dogshit - its only use case is to play video games but even then just get a playstation. ␤ ␤ Github is kinda cool but they didn't make that. HoloLens sucked. At one point, Microsoft did sponsor me and sent me their best [https://twitter.com/luishxyz/status/1876590985385554400 <luishxyz> @AviSchiffmann
<bslsk05> ​Only B2B company is Microsoft. Nvidia has too little focus on sales and too much product edge. Not typical for B2B ]
<fedaykin> and don't get me started on bill gates and what he has become with his vax mandates, get a jab or get fired =) those who got the jab later got fired anyway, because "economy was weak" and interest rates, and now, nothing really changed but "economy is strong" all of a suden, and they are hiring, clown world
agent314 has joined #osdev
<ring0_starr> did somebody say Vax???
<fedaykin> =)
kof673 has quit [Ping timeout: 240 seconds]
duderonomy has quit [Ping timeout: 272 seconds]
kof673 has joined #osdev
* geist is summoned
<geist> but also lets not gripe about companies and whatnot
agent314 has quit [Ping timeout: 244 seconds]
Fingel has quit [Ping timeout: 252 seconds]
<ring0_starr> Being an anti-VAXxer is RISCy business.
agent314 has joined #osdev
GeDaMo has joined #osdev
<fedaykin> touche
* Ermine preemptively ticks yotld off today's #osdev bingo card
<Ermine> Thought of installing linux on my granddad's laptop and found out there's no suitable DE or WM for that
<GeDaMo> How are you defining "suitable"?
<Ermine> for starters, it should run smoothly (conditions are rather spartan there) and properly support hi-dpi
<GeDaMo> Ah
netbsduser has joined #osdev
<Ermine> and be at least somewhat accessible to non-nerds
Lucretia has joined #osdev
hwpplayer1 has joined #osdev
housemate has quit [Ping timeout: 276 seconds]
<nikolar> geist: it's well known that you are a dictator with an iron fist around these parts :P
Left_Turn has joined #osdev
<nikolar> Ermine: why wouldn't DEs run smoothly on your granddad's laptop
alifib has joined #osdev
<nikolar> can't say i've seen them run particularly non-smoothly even on shittiest hardware
<nortti> Ermine: does gnome not fit the bill? I was under the impreasion that, any personal gripes aside, they do prioritize non-nerd use case and should support hidpi
Turn_Left has joined #osdev
hwpplayer1 has quit [Remote host closed the connection]
Left_Turn has quit [Ping timeout: 276 seconds]
frytaped has quit [Ping timeout: 265 seconds]
<Ermine> nortti: the last time i've tried it, their hidpi support is poor (the fractional one)
<Ermine> and, even if we cast aside the fact that gnome's ui is an insult to ergonomics, i also want ui to be windows-y to simplify moving from windows (we're talking about old man after all)
<Ermine> nikolar: plasma somewhat is lagging on my gemini lake laptop, and it's more powerful than my granddad's laptop
<nikolar> To be fair, it's been a while since I've ran plasma
<Ermine> (ftgj it's absolute cheapass potato laptop and I wouldn't approve bying that)
<Ermine> I'm inclined towards IceWM (which I can configure to suit granddad's needs and adjust as needed), but hi-dpi in x11 world works fine only on plasma
<Ermine> and maybe gnome, but that's not an option anyway
<nortti> there was an article about getting hidpi going on openbox + x11 recently on lobsters, lemme see if I can find it
frytaped has joined #osdev
<bslsk05> ​horstmann.com: Cay Horstmann's Unblog
<Ermine> I think it involves setting Xft.dpi, then going to gtk config, then something else
<Ermine> Oh gosh, that's a lot
<Ermine> thank you nortti !
<Ermine> The thing I like about wayland is that all of this is now a single line in a compositor configuration
<Ermine> It took several years and one Kenny Levinsen to make a protocol for that, but now it's done
duderonomy has joined #osdev
goliath has quit [Quit: SIGSEGV]
teroshan8 has joined #osdev
teroshan has quit [Ping timeout: 252 seconds]
pie_ has quit [Ping timeout: 252 seconds]
vancz has quit [Ping timeout: 252 seconds]
teroshan8 is now known as teroshan
pie_ has joined #osdev
vancz has joined #osdev
vdamewood has quit [Quit: Life beckons]
lukflug has joined #osdev
hwpplayer1 has joined #osdev
bauen1 has quit [Ping timeout: 260 seconds]
duderonomy has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
netbsduser has quit [Ping timeout: 252 seconds]
anonpreet has joined #osdev
muffin has joined #osdev
anonpreet has quit [Quit: Leaving]
goliath has joined #osdev
netbsduser has joined #osdev
m3a has quit [Ping timeout: 265 seconds]
bauen1 has joined #osdev
edr has joined #osdev
muffin has quit [Quit: leaving]
frytaped has quit [Ping timeout: 260 seconds]
heat has joined #osdev
frytaped has joined #osdev
frytaped has quit [Quit: WeeChat 4.4.2]
mpetch has quit [Quit: Client closed]
<heat> cloudowind, sure, but the point is that even shit like coreutils is ran by *volunteers*
<heat> afaik the fsf money doesn't hit the people actually running the projects, just the fsf infra
<heat> which, imo, could _easily_ be replaced by any other git service out there
saulosilva has joined #osdev
housemate has joined #osdev
<heat> for big stuff like e.g glibc or gcc the GNU people (@gnu.org) only pop up when it comes to removing abortion jokes or being purposefully incompatible with LLVM and LLD
<pog> abortion jokes????
<pog> it's not a joke i'm pro-life ok
<Ermine> savannah at least has bugtracker *giggles*
saulosilva has quit [Quit: Client closed]
housemate has quit [Ping timeout: 252 seconds]
<Ermine> Though, with such minimalist infra, why do they need $400k on it???
<pog> to pay me
<Ermine> we can pay you directly, no need to involve fsf
housemate has joined #osdev
<Ermine> and there's fsfe also
housemate_ has joined #osdev
housemate has quit [Ping timeout: 252 seconds]
housemate_ has quit [Remote host closed the connection]
housemate has joined #osdev
housemate has quit [Remote host closed the connection]
housemate has joined #osdev
<heat> Ermine, infra is costly, plus im sure they pay _some_ people up in the fsf hierarchy
<pog> tjhen give me money
<pog> subscribe to my onlypogs
<heat> subscribe to my onlykernal sexy kernal pix
goliath has quit [Quit: SIGSEGV]
* kof673 gives gog magical golden dorado swordfish/dolphin > "Over the Mountains Of the Moon, Down the Valley of the Shadow, Ride, boldly ride," The shade replied— "If you seek for Eldorado!"
muffin has joined #osdev
<sortie> help, my kitchen scale says -1 but didn't set errno
<Ermine> seems like kernel bug
<Ermine> you should port sortix to it, maybe it will help
hwpplayer1 has quit [Ping timeout: 252 seconds]
<sortie> Sortix. An operating system for kitchen scales.
<heat> sortix would never run on a kitchen scale
<heat> prove me wrong.
<sortie> onyx is web scale, sortix is kitchen scale
<pog> does onyx have .net
<pog> i will run it in my kubernetes containers for my azure web api
<sortie> My customers walk into the room and immediately go “oh hey what's that it smells delicious is that sortix”
<heat> not yet
<heat> .NET would be crazy
<heat> i can find more uses for java which i still do not have
<sortie> Mono C# was an easy port back in 2013
muffin has quit [Quit: leaving]
<heat> mono's ded
<sortie> Oh?
<Ermine> mono is bootstrappable somehow
<heat> .net is FOSS now
* sortie wakes up in a weird future
<Ermine> dotnet builds download some binaries
<heat> microsoft actually bought mono and are now spitting it back out because it's useless
<sortie> mono seems to be developed as part of wine now
<sortie> It does look maintained
<Ermine> that's wine-mono
<sortie> "August 2024, Microsoft transferred ownership of Mono to WineHQ."
<sortie> So, well, it's mono?
<Ermine> Also, mono implements .NET framework, while dotnet doesn't
<sortie> Is there any other one?
<Ermine> (and framework has stuff .NET doesn't, and vice versa)
<sortie> https://www.mono-project.com/ ← Oh yeah I see that it's shut down by microsoft
<bslsk05> ​www.mono-project.com: Home | Mono
<sortie> Is Wine's mono significantly different or just a continuation?
<heat> ok actually dotnet seems to possibly Just Build for a normal UNIX system
<bslsk05> ​github.com: Haiku: Initial CoreCLR support (#109580) · dotnet/runtime@bf44aaa · GitHub
<sortie> Because at least a decade ago, mono was an easy port (but I wasn't ready for it and it was too big and had no need and it was unstable) and bootstrapped cleanly
hwpplayer1 has joined #osdev
<sortie> If Haiku did it, that means it's probably quite doable
muffin has joined #osdev
housemate_ has joined #osdev
housemate has quit [Ping timeout: 265 seconds]
bauen1 has quit [Ping timeout: 244 seconds]
housemate_ has quit [Ping timeout: 248 seconds]
Fingel has joined #osdev
Lucretia-backup has joined #osdev
xenos1984 has quit [Ping timeout: 252 seconds]
Lucretia has quit [Ping timeout: 248 seconds]
xenos1984 has joined #osdev
muffin has quit [Quit: leaving]
Lucretia has joined #osdev
Lucretia-backup has quit [Ping timeout: 252 seconds]
m3a has joined #osdev
housemate has joined #osdev
duderonomy has joined #osdev
xenos1984 has quit [Ping timeout: 272 seconds]
alifib has quit [Quit: .]
xenos1984 has joined #osdev
craigo has quit [Quit: Leaving]
saulosilva has joined #osdev
housemate has quit [Quit: Nothing to see here. I wasn't there. I take IRC seriously.]
housemate has joined #osdev
housemate_ has joined #osdev
housemate_ has quit [Remote host closed the connection]
housemate has quit [Ping timeout: 265 seconds]
<geist> nice burn there
<heat> geist!
* geist feels the heat
<sortie> geist: It's actually not lol. Sometimes just knowing that some of the weirder systems pulled it off means the upstream project isn't doing a ton of unportable stuff
<geist> okay fair nuff
<sortie> Like if I see Haiku in a port pop up, the first thing is to grep for any mention of Haiku
<heat> isn't haiku kind-of unix?
<sortie> It's a rare word and likely every mention would be something I should consider too
<sortie> Yeah. It does have some special stuff.
<sortie> So some of their patches are often not needed
<geist> it's unix in the way beos was unix, but more so than that
<geist> ie, has mmap (beos didn't) and i think it has fork (beos didn't)
<geist> OTOH beos was pre POSIX 2003
saulosilva has quit [Quit: Client closed]
<geist> er 2001
<geist> ah though looking at wikipedia it was well established before then
<heat> yeah POSIX is ooooooooooold
<geist> haikus on that
<pog> hi geist
<bslsk05> ​es.wikipedia.org: Sortix - Wikipedia, la enciclopedia libre
<nortti> geist: how did process creation work on beos?
<geist> hola!
<geist> it had a straight spawn style call
<sortie> I'm pretty sure Sortix was deleted from the russian wikipedia
<geist> wasn't posix_spawn (dont know if that existed at the time) but something pretty similar
<heat> nah posix_spawn is newish
<heat> maybe 2011
<heat> or was it 2008?
<sortie> I suppose whatever spawn mechanism it is, it must be in Haiku
<geist> so i remember there was some local patches for a lot of the gnu tooks (gcc, etc) to use the straight spawn
<sortie> No POSIX 2011
<sortie> I think there was 2001 and 2008. Maybe a 2004 variant, I forgot
<heat> STANDARDS top
<heat> POSIX.1-2008.
<heat> HISTORY top
<heat> glibc 2.2. POSIX.1-2001.
<heat> confused.
<heat> might be 2001
<heat> i think there was a POSIX 2011? and then a POSIX 2017
<sortie> 199506L, 199506L, 200112L, 200809L, 202405L
<heat> as interim versions
<geist> that wikipedia page lists em
<sortie> heat, those were just editorial fixes
<sortie> No content changes besides language lawyering
<sortie> And even most of that got deferred
<heat> i know
<sortie> Which is why releasing POSIX 2024 was such a big deal after 16 years
<heat> <citation needed>
<sortie> It's an overall small release but it matters a lot to just apply so many tons of small fixes
<sortie> So we didn't have an old 2008 standard full of known issues
<geist> but in general remember that beos was mostly a UI OS. it had an underlying posixy like layer, but kinda like windows you were generally supposed to be writing ui programs
<geist> and that was all under a layer where you created a new BApplication object, etc etc
<heat> https://en.wikipedia.org/wiki/POSIX#POSIX-certified look at these amazingly important operating systems
housemate has joined #osdev
<geist> ah load_image() out of the kernel kit, it seems
<geist> https://www.haiku-os.org/legacy-docs/bebook/TheKernelKit_Images.html#:~:text=valid%20image%20ID.-,load_image(),-thread_id%C2%A0load_image
<bslsk05> ​www.haiku-os.org: The Be Book - Classes And Methods - The Kernel Kit
<bslsk05> ​www.haiku-os.org: The Be Book - Classes And Methods - The Kernel Kit
<geist> oh what a simpler time. remember this was like 1998
<geist> so dont rip on it too hard
<geist> even having threads or being SMP capable was at the time sort of a Big Iron thing
<geist> also never forget https://www.haiku-os.org/legacy-docs/bebook/TheKernelKit_SystemInfo.html#:~:text=returns%20B_OK.-,is_computer_on(),-int32%C2%A0is_computer_on
<bslsk05> ​www.haiku-os.org: The Be Book - Classes And Methods - The Kernel Kit
<heat> dont worry mjg isn't here
<heat> no CRAPPER
<nikolar> oh is mjg even alive
<nikolar> he hasn't been around for months now lol
<nortti> was the terminology not established that much yet, or is there some other reason why beos calls them teams instead of processes?
<heat> yeah he's alive
<heat> i see him in #mm
<geist> nortti: i dunno. even at the time i thought it was a little weird, but i have a suspicion if i know more about macs or amigas it was started there
<geist> a lot of the beos team was ex-mac or ex-amiga demo coders and wahtnot
<geist> so there was a hodgepodge of ideas from those worlds that made it into the apis
<geist> (like file reference handles, etc)
<sortie> Granted a lot of people were really, really weird about threads back in the day when threads were new
TkTech has joined #osdev
<geist> and i think there's some oh that's the multithreaded early kernel..... some of that influence
<geist> like threads having an implicit message queue
<geist> mach has that but also... there was some sort of example kernel at the time that i am forgetting
<heat> linux? :v
<geist> heh, no. heck did linux have threads in the mid 90s?
<geist> i know the BSDs only had user space threads then
<nortti> I wanna say linux threads were a 00s thing
<geist> by 98 beos was like version 3 or 4. the main kernel design was probably 1994 i think
<nortti> oh apparently https://en.wikipedia.org/wiki/LinuxThreads was there by 1996 already
<heat> linxuthreads was totes fucking broken though
<geist> yeah linux was far more of a toy then
<nortti> "As of 2006, LinuxThreads may still be seen on production systems, particularly those using version 2.4 or lower of the Linux kernel, as NPTL requires facilities which were specifically added into the 2.6 version of the kernel for its use."
<heat> NPTL was a huge step forwards
<geist> if nothing else because threasd and posix i think were in pretty bad shape then. only really solaris was pushing in that direction
<nortti> I think this sentence might benefit from some updating
<Ermine> is nptl a part of kernel
<Ermine> ?
<heat> no
<heat> nptl is glibc's pthreads implementation
<geist> oh yeah i hadn't heard that acronym in a long time
<heat> it directly depends on clone() and futex() and all that shit
<geist> i remember that being a thing also around the libc5/glibc stuff
<heat> some parts were only really laid out in early 2.6
<geist> which distro used what, which distro had nptl, etc
<sortie> It's wild in Sortix to just directly implement real threads without all of the old style threading bullshit
<heat> i've been strongly considering linux's threading model
<Ermine> libc5 is iirc a debian's glibc fork?
<heat> it is kinda cool
<heat> Ermine, nope
<geist> yeah see i learned on beos, which also just straight up implemented threads and the kernel was also just straight fully preemptable
<geist> no preemptin points etc
<heat> that would be eglibc i think?
<geist> so to me that's the natural way to do it
<geist> it's primitive by todays standards, but its also 'modern' in the sense that there was no backwards legacy stuff in that regard
<sortie> It always struck me as weird to read books like Tanenbaum's that go through all these wild crazy threading models like they're reasonable choices
<sortie> It really was a weird world back then
<heat> linux threads (as well as plan9 threads, and earlier freebsd threads i think) are kinda cool
<heat> you can share whatever you'd like
<sortie> Of course preemptive kinda-single-threaded stuff has come back massively since then, sooo
<sortie> *cooperatively scheduled
<heat> some variations are nonsensical but oh well
<sortie> async await like it's Windows 3
<nortti> Ermine: my understanding is that glibc proper didn't use to support linux, so it got forked into linux libc which had its own versioning, and finally got merged back into glibc proper after version 5
<Ermine> it worths it to look at plan9 threads
<sortie> libc6 to rule them all
<Ermine> nortti: well idk, i did not exist back then
<geist> hmm my brain is wracking to remember this. there was some academic os back then... i think it had a 3 letter acronym, kinda like xnu
<geist> but *not* xnu
<geist> and not mach and not hurd
<nortti> xinu?
<geist> yeeeeeees i thin kthat's it
<heat> BSD!
<Ermine> xinu is not academic though
<geist> https://xinu.cs.purdue.edu/ well it came from purdue apparently
<bslsk05> ​xinu.cs.purdue.edu: The Xinu Page
<Ermine> well ok
<geist> but yeah i think beos had a bunch of xinu influence, but i've never looked at it
<Ermine> i had a feeling it was backed up by a company
<geist> there was a textbook about it and everything
<heat> Ermine, plan9 threads AFAIK are literally just rfork()
<heat> rfork() is a worse clone()
<geist> but yeah according to their page it got used in commercial stuff in the 90s
<sortie> tfork ftw
GeDaMo has quit [Quit: 0wt 0f v0w3ls.]
<heat> clone() can be used in full effect in posix_spawn
<heat> CLONE_VM | CLONE_VFORK for vfork behavior, but it changes stack which allows it to actually do things
hwpplayer1 has quit [Remote host closed the connection]
goliath has joined #osdev
<nortti> "(If the child shares the parent's memory because of the use of the CLONE_VM flag, then no copy-on-write duplication occurs and chaos is likely to result.)"
<clever> but thats also how threads work in linux
<clever> threads are just a special flag to also clone the "pid", and share most everything else
<nortti> this is specifically describing the use case where you don't provide a new stack
agent314 has quit [Ping timeout: 244 seconds]
<clever> ah, yeah, thats where vfork comes into play
<clever> leave the parent on pause until the child has stopped messing with things
steelswords9436 has quit [Quit: Ping timeout (120 seconds)]
steelswords9436 has joined #osdev
<heat> >chaos is likely to result
<heat> if you're an idiot, and if you're writing the libc you're hopefully not an idiot
<geist> or you are an idiot for writing another libc
<heat> CLONE_VFORK is cool because it just stops the parent process until execve or exit
<heat> which is *exactly what you want*. good useful clone semantics, posix_spawn implemented in userspace, kernel more flexibly, microkernel fans delighted
<Ermine> if you're writing libc you're either total idiot or total genius. Like on that bell curve meme
<cloudowind> not everyone born genius , so called idiots gotta live too , i mean its ok to be an idiot i dont see anything wrong with it :/
* cloudowind says and gets ready for work
<heat> definitely, but hopefully they're not writing core system software
housemate has quit [Quit: Nothing to see here. I wasn't there. I take IRC seriously.]
housemate has joined #osdev
housemate has quit [Remote host closed the connection]
housemate has joined #osdev
housemate has quit [Remote host closed the connection]
housemate has joined #osdev
housemate has quit [Quit: Nothing to see here. I wasn't there. I take IRC seriously.]
duderonomy has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
housemate has joined #osdev
<sortie> https://sortix.org/ ← I updated my website :)
<bslsk05> ​sortix.org: The Sortix Operating System
<nikolar> sortie: what's differetn
<Ermine> git diff it
<nikolar> nah, that's work
<nikolar> i don't care that mcuh
housemate has quit [Quit: Nothing to see here. I wasn't there. I take IRC seriously.]
goliath has quit [Quit: SIGSEGV]
<sortie> It's basically a whole rewrite :)
housemate has joined #osdev
FreeFull has quit [Ping timeout: 265 seconds]
agent314 has joined #osdev
bauen1 has joined #osdev
m3a has quit [Read error: Connection reset by peer]
housemate_ has joined #osdev
housemate_ has quit [Remote host closed the connection]
housemate has quit [Ping timeout: 244 seconds]
Turn_Left has quit [Read error: Connection reset by peer]
mrpops2ko has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
mrpops2ko has joined #osdev