>there is only one rule in Linux: never break the userspace nor the ABI>Linux developers constantly break the userspace and the ABIWhy are Linux developers so allergic to respecting the idea that the software should serve the user first and foremost?
>>108531472I'm the only user. Everyone else is a guest. My code, my space.
>>108531472im gaydid you even think of that?
>>108531472Literal autism. Or what else would you describe the strong adherence to internalized rules that don't matter?
>>108531472>confusing the Linux kernel with the userspace once againthey're not sending their bestLinux (the kernel) only promises to not change the *userspace* ABI of the *kernel*. They're free (and do it all the time) to break the internal ABI of the kernel, breaking every driver that remains closed-source.what GNU, KDE, GuhNOME, etc, do is up to them, that's a dirty luser problem.
>>108531472>constantly break the userspace and the ABIShow one example.
big surprise linux software became hostile to the user around google/android corporotizqtion hoovering up and sticking shills in ever project that was worth a damn
>>108533603sysctl
>>108533697A word is not an example. How is sysctl broken?
>>108533792>goalpost-moving
>>108533826
>>108531472>>108531472Someone post the table of Windows syscall changes, showing that Microsoft randomly breaks fundamental ABIs during ordinary updates
pfffft HAHAHAHAHAHAHAhAhA
>>108533858>nooooo you can't use the stubs in ntdll.dll>because>because>because you just can't okYou sound almost as retarded as >>108533855
>>108533900So you now tell us that breaking ABI is actually a good thing, because Microshit does it randomly during updates?
>>108531472>haha, why does Linux break ABI that much>>108533900>WTF, NOO, BREAKING ABI IS FINE, STOP COMPLAINING ABOUT WINDOWS DOING ITmeds
>>108533927The ABI of the stubs is unaffected.
>>108533888>statically link your application>windows update happens>suddenly can't even open a port anymore, because the syscall changed randomly for no reasonHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAv
>>108533959You cannot "statically link" ntdll.dll, you dingus.
>>108533936>just use a compatibility layer that translates the ABI breakage How about you just don't break the ABI in the first place, you subhuman.Somehow Linux manages to keep their syscalls the same for decades and the fucking Microshit idiots can't do that?
>>108533969Of course you can simply compile syscalls into your application, you idiot.How about you write a game in assembly. Remember that this used to be a thing.
>>108533974>just don't break the ABIJust use the stubs, autist.
>>108531472If uppity faggots didn't break downstream users' shit for funsies every once in a while, linux could've actually gotten somewhere instead of wasting 90% of work on "maintaining packages" (read: unfucking other's shit). Looking at you gcc and glibc
>>108533980You. Cannot. Statically. Link. ntdll.dll.You can manually issue syscalls, but that's your problem.
>>108533969Give me ONE reason why it shouldn't be possible to write an application in Assembly and not have it randomly break between updates.You are just ok with the fact that on Windows you can't do anything in Assembly, because some fucking monkeys at Microsoft have to randomly change numbers in their busywork?Just give us a reason. Why does Microsft break its ABI.>but but you can use a compatibility layerIsn't a reason.WHY CHANGE THE ABI?
>>108531472>allergic to respecting the idea that the software should serve the userWhich one? It's legacy software created to run on mainframe computers the size of a building. With a bunch of thin client.
>>108533993I CAN WRITE AN APPLICATION IN ASSEMBLY AND DO SYSCALLS IN ASSEMBLY AND CAN EVEN CALL IT FROM C WITHOUT NTDLL.DLL(and windows would randomly break it on updates for no fucking reason)
>>108533996I'll give you two:1. To trigger your autism.2. To optimize SSDT layout.>muh assemblyCry me a river.
>>108534005No one cares except autists, and their opinions are literally invalid. Just use the stubs.
Back in the day people used to write applications in Assembly, nowadays you can't do that, because Windows breaks the ABI five times during one single release cycle.And you justify it with:>haha, because it makes you mad :))and>its optimizing, changing a code from 0x7d to 0x7c makes it faster!What's next? Will it only be possible to write applications in Javascript in the future and you will say:>yeah, but the Javascript ABI doesn't change :^), so it doesn't matter that you can't write native applications
>>108533858>>108533888It doesn't affect any programs because they're not fundamental ABIs. ABIs are for applications. Those are internal and shouldn't be used by applications.
>>108533996>Give me ONE reason why it shouldn't be possible to write an application in Assembly and not have it randomly break between updates.You use DLLs and they don't randomly break. Win32 assembly programs can run on Windows 95 to 11 without breaking anything.
