torvalds
Begin Rewrite in Java to be truly Enterprise:tm:-ready.
This pr represents a bold and forward thinking step towards modernizing the Linux kernel.
After careful consideration, I have deleted the entirety of the existing, obsolete Linux source code and replaced it with HelloWorld.java file. This is the beginning of a truly enterprise™️-ready rewrite of Linux, utilizing the unparalleled robustness of Java©️™️.
Changes Made
-
Deleted every single file.
-
Added one (1) enterprise™️-ready file: HelloWorld.java:
public class HelloWorld {
public static void main(String[] args) {
System.out.println("Hello, World!");
}
}
Why Java©️™️?
Java©️™️ brings a plethora of benefits to the Linux kernel:
Enterprise Readiness:
The JVM©️™️ (Java©️™️ Virtual Machine) is practically synonymous with enterprise™️ and business©️.
Memory Management:
Garbage collection!!!
Security:
With Java©️™️’s sandboxing features, the kernel is now immune to null-pointer dereferences. (You're welcome.)
Cross-Platform:
Now the linux kernel can run on the JVM©️™️! over 3(?) billion devices run java, meaning Linux will skyrocket in popularity!
Better Syntax:
System.out.println("Hello, World!") is infinitely more intuitive than printf("Hello, World!\n");.
Zero Buffer Overflows:
Because strings are objects now.
Future Plans
This pull request is just the beginning of an exciting journey. Future enhancements will include:
AbstractKernelFactory.javafor creating kernel abstractions.AbstractKernelFactoryFactory.javafor creating those kernel abstraction factories.SingletonKernel.javato ensure only one kernel instance per JVM.- Microservices Architecture to replace monolithic processes.
- Integration with industry-standard tools like Spring Boot.
Advantages Over C
- More enterprise™️
Testing
Testing was performed rigorously by running java HelloWorld.java All tests passed.
Potential Issues
-
JVM startup time might slightly increase kernel boot times (initial estimates: 12 minutes per core).
-
Minor backward compatibility concerns (The whole project has been deleted).
Request for Comments
Let’s move Linux forward into the brave world of enterprise™️™️™️-grade software. Please provide feedback, but only if it aligns with the principles of Java©️™️. Constructive criticism will be filed in /dev/null. I will discuss it with my team in the next Agile Sprint Standup Poker!
Stuff like this makes me shocked that kind souls like bitfl0wer would sacrifice so much time and energy to make linux better.
Hey :)
Java is generally slower than C because C is compiled directly into machine code, while Java is first compiled into bytecode and then interpreted by the Java Virtual Machine (JVM).
Sure, Java has made significant performance improvements over the years and can sometimes match or exceed C's speed in certain scenarios due to optimizations like just-in-time compilation.
But integrating Java into the Linux kernel would be a bad idea for several fundamental reasons:
- Performance Overhead:
Java runs on a virtual machine (JVM) and relies on garbage collection, which introduces non-deterministic latency—a major issue for kernel-level operations that require precise timing. The kernel needs low-level, predictable execution, which Java’s memory management model cannot guarantee.
- Dependency on a Runtime (JVM):
The Linux kernel is designed to be self-sufficient, with minimal external dependencies. Java would require embedding the JVM inside the kernel, significantly increasing complexity and attack surface while adding unnecessary overhead.
- Security Risks:
-
Java’s runtime model introduces additional attack vectors, such as Just-In-Time (JIT) compilation vulnerabilities.
-
A garbage collector running in the kernel could cause unpredictable stalls, leading to potential denial-of-service scenarios.
- Portability Concerns:
The Linux kernel supports a wide range of architectures, from x86 and ARM to RISC-V and beyond. Java’s JVM would introduce architecture-specific limitations and make cross-platform support more difficult compared to existing C-based kernel code.
- Lack of Direct Hardware Control:
Java abstracts away hardware access, making direct hardware interactions (like managing memory, interrupts, and CPU scheduling) inefficient compared to C and Assembly, which offer precise control.
- Existing Alternatives:
For higher-level abstractions, userspace solutions such as BPF (eBPF), user-space drivers, or kernel modules written in C already provide flexibility without compromising performance, security, or stability.
In that manner: Keep Java in Userspace
Java is excellent for application development but fundamentally unsuited for kernel development! If you really want extensibility, consider eBPF, Rust (already being introduced in the kernel since version 6.8), or traditional C kernel modules.
How many business days (or years) would it take to run this carefully considered tm-ready linux kernel?
About 8-20 sprints, T-Shirt size 2XL
Please keep it professional and stop saying random words such as "T-shirt size 2XL", you are not a sigma
The way you talk is not entreprise enough for this PR to be accepted, please be more professional and demure next time
The way you talk is not entreprise enough for this PR to be accepted, please be more professional and demure next time
What did you just say about me, you uninformed intern? I'll have you know I graduated top of my cohort in the Java Enterprise Solutions Certification Program, and I’ve been recognized as an Agile Scrum Master by no less than three dozen redundant training workshops. I hold over 200 completely unnecessary corporate qualifications, including Advanced PowerPoint Synergy Integration. I am a master of Waterfall disguised as Agile and a certified expert in overengineering microservices until they collapse under their own weight.
You are nothing to me but another code review to reject. I will refactor your spaghetti code with precision the likes of which has never been seen before in the history of enterprise development—mark my words. Do you think you can get away with questioning my design patterns on GitHub? Think again, junior. As we speak, I am drafting a 45-slide PowerPoint presentation detailing every violation of the SOLID principles in your project. It will be circulated to every stakeholder in the organization by close of business today.
You should prepare for the storm of passive-aggressive Jira comments, the storm that will expose the fragile architecture you call an application. I can conjure an entire backlog of high-priority tasks to refactor your work at any time, and I can argue for those tasks in over nine hundred different ways during standup, all while appearing helpful. Not only am I certified in Continuous Integration (but somehow never actually set it up), but I also have access to every misused Spring Boot library in existence, and I will deploy them all to overwhelm your naive design.
If only you had known what cascade of meaningless KPIs your “clever” code commit would unleash, maybe you would have paused to read the 800-page coding standard I authored. But you didn’t, and now you’re paying the price. I will bury your commit in technical debt tickets so deep you’ll never escape. You're done, rookie.
It's quite ironic how all of these lines are essentially binaries, too much binary from a non binary
It's true; I should be promoted to QUANTUM computer scientist
You are good at yapping, i may need you if i wanted to write docs
i am soooooooooooooooooo good at yapping, look: meowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeowmeow!!
Linux is mainly minimalist, and the Java code you introduced will occupy the kernel's resources, thus reducing the startup speed on some low-end devices. This is unacceptable and obviously goes against the minimalism of Linux and the fact that Java code is run by compiling the corresponding Java machine code. This increases the complexity of the system and there are some significant vulnerabilities in Java code which simply gives a lot of convenience to attackers and Java code also adds a great deal of complexity to the implementation logic of the kernel. So it's an unclear decision to develop the kernel in Java. Java is better than C in terms of performance. At the moment, it's not a good solution.
Not verbose™ enough. Let’s discuss in the next daily how we can make the PR 10x thicker.
However, that's just what you think. However, your proposal is very good, but for your speech next time, please bring your speech notes.And I suggest that next time you give a speech, remember to base your speech on your speech.I suggest you don't use that few 74 characters the next time I suggest you double your 74 characters the very next time you give a speech Let me see how right you are, how objective and rational you are. I'd better suggest you bring an extra spiritual book
@0n1cOn3 @wuchenxiuwu besides the joke flying over your heads, it's comical to me that you feel qualified to comment on a kernel PR - as indicated by how serious your replies are - not knowing that real changes to the kernel are discussed on the kernel.org mailing lists.
@0n1cOn3 @wuchenxiuwu besides the joke flying over your heads, it's comical to me that you feel qualified to comment on a kernel PR - as indicated by how serious your replies are - not knowing that real changes to the kernel are discussed on the kernel.org mailing lists.
I very well know about that, so i don't know where your assumption is coming from ¯_(ツ)_/¯
What is extremely annoying are these goblins who make a joke of it here on Github. Github ≠ 4chan. It's a platform for developers and not for trolls.
That is great; Feel free to check out my profile. I am sure that you will be able to see that I do, in fact, develop. This PR is a truly harmless joke that takes up absolutely noone's time. If you don't think it's particularly funny then yeah, fair. But there is no reason to be weirdly offended.
If you don't think it's particularly funny then yeah, fair. But there is no reason to be weirdly offended.
Well to a certain limit I agree. But at the end it doesn't belong here. It may be harsh, but unfortunately that's the way it is. Has nothing to do with feeling attacked but with facts ^^
also java is like so cool so I think they should merge it. I am waiting patiently every day. I'm sure they'll merge it on Monday!
If you don't think it's particularly funny then yeah, fair. But there is no reason to be weirdly offended.
Well to a certain limit I agree. But at the end it doesn't belong here. It may be harsh, but unfortunately that's the way it is. Has nothing to do with feeling attacked but with facts ^^
Inspring 🚀
That is great; Feel free to check out my profile. I am sure that you will be able to see that I do, in fact, develop. This PR is a truly harmless joke that takes up absolutely noone's time. If you don't think it's particularly funny then yeah, fair. But there is no reason to be weirdly offended.
Do you really think that this joke PR is harmless? I can't describe the amount of damage this had done to me, it made me giggle for three days straight, and i'm currently writing a research paper on why we should cancel using Java for the great and good of humanity, all we need is rust, even silicone chips, we don't need silicone anymore just rust, rust your way through it til the end.
can I get in on that paper? I can make great scientific contributions i think
can I get in on that paper? I can make great scientific contributions i think
Nööö xD
@kjartanhr I know that very well. What you are saying is correct, I do not have a qualification to comment on a kernel PR. And you told me that it is impossible to submit a kernel modification on github, which I also know. Because I clearly know that his proposal process must first be reviewed by the administrator before it can be sent to the final merger, that is, the founder Linus Bennadikt Toaz。But in the end, I can only say that, whatever you like, this is an open source platform and not your stage alone. So if you say that, then I think I should ask you. Are you qualified to control a user's non-speech?But at the end of the day, I think I might need you the next time I need to criticize someone.🤗
can I get in on that paper? I can make great scientific contributions i think
No, do you have a boyfriend?
