Rust-Written Redox OS 0.8 Released With i686 Support, Audio & Multi-Display Working

[Waethorn]

Because support of closed-source drivers is one of the main features of microkernels. If a microkernel isn't supporting closed-source drivers, it is missing one of its major features. Truly, it is a big loss, because, as many have pointed out in this thread, vendors don't like to open-source their drivers.

There is, of course, another possibility, which is the one that can make me very sad. Perhaps the developers of Redox OS didn't express it correctly, and they misinterpreted what a microkernel means. I would strongly advise against such a decision, as you will be trashing the most beautiful feature of your entire OS.

You tried to turn this into a specious argument that "manufacturers love to write drivers for microkernels" despite every market statistic pointing to the contrary. The two are not equivalents. It is you that doesn't understand what a microkernel means. And microkernels aren't a magic wand to fix what you equate to a broken license.

[Lancia]

I agree with microkernels having the advantage of supporting closed sourced drivers, but the disadvantage for the open source community is that companies are no longer required to publish the source code of their drivers like with monolithic kernels and instead just give a binary blob I'm assuming?

[cj.wijtmans]

If you don't like my opinions on the subject: too bad. You're not going to shut me up with your condescending tone.

Shut up will weathorn. You make no sense whatsoever. I doubt you have a clue what you are talking about.

That devs write something commonly does not mean they love to do it. Its just a job that needs to be done. Most of the time devs hate doing something because most of the time its not the ideal way to do something but its just the way how things are currently working. I would prefer to write a microkernel driver that has ABI issues with a kernel but i am strong believer of KISS.

[Developer12]

Ok I'm not into periphery drivers - I thougth this is also managed as an USB device ?

Basically every laptop touchpad on earth is PS2. Laptop keyboards often are as well. There's no reason to over-complicate things with USB when there are often PS2 ports on the south bridge.

[Waethorn]

Shut up will weathorn. You make no sense whatsoever. I doubt you have a clue what you are talking about.

That devs write something commonly does not mean they love to do it. Its just a job that needs to be done. Most of the time devs hate doing something because most of the time its not the ideal way to do something but its just the way how things are currently working. I would prefer to write a microkernel driver that has ABI issues with a kernel but i am strong believer of KISS.

No

[Waethorn]

Basically every laptop touchpad on earth is PS2. Laptop keyboards often are as well. There's no reason to over-complicate things with USB when there are often PS2 ports on the south bridge.

Sorry, that's not correct. Though Redox targets old laptops, Windows-certified laptops haven't for a long time. This includes essentially any current Windows laptop shipping at least since Windows 10 launched 7 years ago. They commonly use one of the internal busses like I2C, sometimes internal USB. The documentation on "Windows Precision Touchpads" notes this in this first sentence.

Device Bus Connectivity (Windows Precision Touchpad) | Microsoft Learn

[Waethorn]

I agree with microkernels having the advantage of supporting closed sourced drivers, but the disadvantage for the open source community is that companies are no longer required to publish the source code of their drivers like with monolithic kernels and instead just give a binary blob I'm assuming?

There is nothing in the GPL that states that a company can't ship a binary firmware blob or closed-source driver. Nor is there any technical limitation within the Linux monolithic kernel that prevents it from installing as a closed-source module (except for the security certificate issue of the bootloader with Secure Boot). They are free to release it with their own means of distribution, whether separately, or with a complete Linux distribution. Canonical has shown that you can indeed include closed-source drivers, like those from NVIDIA, and non-GPL packages, like ZFS support, with a whole OS.

Now if a distro publisher says that "all of the packages in our distro are GPL-compatible", then they just shot themselves in the foot by imposing restictions on themselves. Debian and Fedora did this, but Canonical didn't. This isn't a question of the license of the kernel, except where a driver-writer wants to create a kernel fork and include their driver code or patches. If they create external modules, Linux still works with that, both technically and legally.

So in summary:

Does the Linux kernel work with external closed-source drivers? Yes.
Can a Linux distro include external closed-source drivers? Yes.

How is this different from a microkernel?

[Waethorn]

You can't force-feed freedom to other people, it has to be their choice.

Freedom with strings attached isn't freedom at all.

[Developer12]

Sorry, that's not correct. Though Redox targets old laptops, Windows-certified laptops haven't for a long time. This includes essentially any current Windows laptop shipping at least since Windows 10 launched 7 years ago. They commonly use one of the internal busses like I2C, sometimes internal USB. The documentation on "Windows Precision Touchpads" notes this in this first sentence.

Device Bus Connectivity (Windows Precision Touchpad) | Microsoft Learn

Have you checked the laptop models that redox runs on?

Also, that "I2C" is the SMBus, which is arguably even worse. It may even be SMBus to PS2, because touchpad manufacturers do still indeed manufacture PS2 touchpads for laptops.

Almost as jank as the MAX7311. That chip is designed (and still used) to connect ISA devices to the "I2C" SMBus, because yes, those still exist too.

[Waethorn]

Have you checked the laptop models that redox runs on?

Also, that "I2C" is the SMBus, which is arguably even worse. It may even be SMBus to PS2, because touchpad manufacturers do still indeed manufacture PS2 touchpads for laptops.

Almost as jank as the MAX7311. That chip is designed (and still used) to connect ISA devices to the "I2C" SMBus, because yes, those still exist too.

I2C is NOT SMBus. There are only I2C to USB bridges, not to PS/2. The MAX7311 is an industrial breakout controller, much like a USB dock controller. Nobody uses these for modern laptop touchpads because they're not compatible with the permitted touchpads as part of the Windows logo certification, which every major tier-1 and tier-2 OEM makes their laptops for. The evaluation kit is only designed for Windows compatibility up to Windows Vista. I would hardly call that something modern. ISA is also not accepted in the Windows Logo certification program and hasn't been for many years, since at least Windows 8.0 shipped 10 years ago.

I2C vs SMBus | Differences & Similarities (evision-webshop.de)

Application Note 6.17: Definitions & Differences Between 12C, Access.bus & SMBus - SMSC (microchip.com)