Freedesktop SDK 21.08 Released With RISC-V Support

[phoronix]

Phoronix: Freedesktop SDK 21.08 Released With RISC-V Support

The Freedesktop SDK as a platform and SDK run-time for Flatpak apps/run-times is out with its first major stable update in one year...

https://www.phoronix.com/scan.php?pa...ktop-SDK-21.08

[gens]

Why not just rename it to "GNOMEdesktop".

[RahulSundaram]

Why not just rename it to "GNOMEdesktop".

Because the freedesktop flatpak runtime is different from the GNOME runtime and used by KDE apps for example in Flatpak?

[oiaohm]

Why not just rename it to "GNOMEdesktop".

https://docs.flatpak.org/en/latest/a...-runtimes.html

This is the link. The kde runtime and the gnome runtime of flatpak both use the core freedesktop runtime in flatpak. Lot of runtime parts gnome applications need are not in freedesktop runtime.

[gens]

It seems I did not express myself clearly enough. I apologize. What I mean is that "freedesktop" is getting to be more about things that are not "desktop", and most(all?) of it is coming from GNOME. The "spirit" of "freedesktop" is waning, so what is the point of keeping the name? Maybe call this "packagemanagerkit" or "org.software.mechanism.kit"? Freedesktop, despite sometimes lacking in the past, was one of the grounding stones that enabled linux to have a good desktop experience. But lately it seems to be more about whatever a certain linux distribution thinks is needed and less about practical desktop experience. Since the desktop is moving onto wayland, why not think up some protocols to make things like drag and drop, setting display brightness, copy/paste, notifications, window placement, etc.. possible to implement ? Coding programs to do such things is only going to make it seem forced, and is going to limit growth and flexibility of the desktop. Did we forget what X11-s design enabled everyone to do ? It enabled new mechanisms, and thus new behavior, to be implemented easily (granted, not always safely or simply). Instead of saying "this is the way you do that", why not try and give people simple mechanisms and letting them do whatever they want with them. Separate mechanisms and policies, expose everything that can not be abused, make it simple and independent of anything but the kernel as much as you can. Making things monolithic and dependent on complex, varied, stuff will hinder people who are interested, thus reducing contributions to people who have enough time to wade through all that.

[Vistaus]

Michael What's also interesting is that Apple is now exploring RISC-V: https://www.tomshardware.com/news/ap...-v-programmers

[RahulSundaram]

Freedesktop, despite sometimes lacking in the past, was one of the grounding stones that enabled linux to have a good desktop experience. But lately it seems to be more about whatever a certain linux distribution thinks is needed and less about practical desktop experience

If you look at the component list in the sdk or how it build out using Buildstream, it is very much about the shared components used by multiple desktop environments and it is not GNOME specific and explicitly not distro specific either

https://gitlab.com/freedesktop-sdk/f.../-/tree/master
https://gitlab.com/freedesktop-sdk/f...components.bst

[bridgman]

I am sure you do not understand what I am saying. And you probably never will.

Can you try revising your post ? I read your post and RahulSundaram's response and from what I could see the response seemed to address your concerns pretty well.