Re: [OE-core] [meta-oe][RFC PATCH 1/2] libdecor: initial add recipe

2022-05-31 Thread Markus Volk
I have made some progress in building an image with DISTRO_FEATURE 'wayland' without X11. So far I had the problem that I had to enable x11 DISTRO_FEATURE to build many packages under wayland. So all recipes  additionally pulled  x11 support. I allowed the base xlibs to build under wayland and s

Re: [OE-core] [meta-oe][RFC PATCH 1/2] libdecor: initial add recipe

2022-05-23 Thread Markus Volk
Am 22.05.22 um 14:12 schrieb richard.pur...@linuxfoundation.org: I'd be interested to know what kind of aborts as that shouldn't have happened. I'd be interested to understand what broke as it shouldn't have done and none of our testing has shown it. Cheers, Richard All-clear for the master b

Re: [OE-core] [meta-oe][RFC PATCH 1/2] libdecor: initial add recipe

2022-05-22 Thread Markus Volk
Am 22.05.22 um 14:12 schrieb Richard Purdie: I'd be interested to know what kind of aborts as that shouldn't have happened. I'd be interested to understand what broke as it shouldn't have done and none of our testing has shown it. I'm running short of time today, but will test a clean master b

Re: [OE-core] [meta-oe][RFC PATCH 1/2] libdecor: initial add recipe

2022-05-22 Thread Richard Purdie
On Sun, 2022-05-22 at 12:22 +0200, Markus Volk wrote: > Am 22.05.22 um 11:24 schrieb richard.pur...@linuxfoundation.org: > > > What kind of issues? :/ > During the build of a systemd-based image, eudev was added. Possibly I > was to blame for this by my own changes ;) Fair enough, I'm not aware

Re: [OE-core] [meta-oe][RFC PATCH 1/2] libdecor: initial add recipe

2022-05-22 Thread Markus Volk
Am 22.05.22 um 11:24 schrieb richard.pur...@linuxfoundation.org: What kind of issues? :/ During the build of a systemd-based image, eudev was added. Possibly I was to blame for this by my own changes ;) But after updating the master branch yesterday, I also got build aborts for some python m

Re: [OE-core] [meta-oe][RFC PATCH 1/2] libdecor: initial add recipe

2022-05-22 Thread Richard Purdie
On Sun, 2022-05-22 at 10:48 +0200, Markus Volk wrote: > Those are valid objections and I wasn't happy with how invasive the > necessary changes were either. The reason I'm still working on it is > that I'd really like to have an alignment for native/target for libsdl2. > As it is, libsdl2-native

Re: [OE-core] [meta-oe][RFC PATCH 1/2] libdecor: initial add recipe

2022-05-22 Thread Markus Volk
Those are valid objections and I wasn't happy with how invasive the necessary changes were either. The reason I'm still working on it is that I'd really like to have an alignment for native/target for libsdl2. As it is, libsdl2-native unconditionally adds x11. From the perspective that we event

Re: [OE-core] [meta-oe][RFC PATCH 1/2] libdecor: initial add recipe

2022-05-21 Thread Richard Purdie
On Fri, 2022-05-20 at 14:46 +0200, Markus Volk wrote: > i have started here locally to resolve all dependencies to be able to > build pipewire-native. Besides I tried to untie the node when building > libpam-native, because otherwise I can't compile openssh-native either. > All in all, the chang

Re: [OE-core] [meta-oe][RFC PATCH 1/2] libdecor: initial add recipe

2022-05-20 Thread Markus Volk
i have started here locally to resolve all dependencies to be able to build pipewire-native. Besides I tried to untie the node when building libpam-native, because otherwise I can't compile openssh-native either. All in all, the changes were more invasive than I had hoped. Quite some recipes ne

Re: [OE-core] [meta-oe][RFC PATCH 1/2] libdecor: initial add recipe

2022-05-19 Thread Markus Volk
To my knowledge libsdl2 is the  only consumer of libdecor right now. If you want to be able to run sdl programs in windows instead of just having them pasted to the background you would want to enable it by default. I love libsdl so my answer would be ... of course it should be build with decora

Re: [OE-core] [meta-oe][RFC PATCH 1/2] libdecor: initial add recipe

2022-05-19 Thread Alexander Kanavin
If it is in core, there should be something in core that consumes and makes use of it, and preferably automated tests for it. Is this the case? Shall weston enable it by default then? Alex On Thu, 19 May 2022 at 21:43, Markus Volk wrote: > > It doesn't need to be in core, but i would call having

Re: [OE-core] [meta-oe][RFC PATCH 1/2] libdecor: initial add recipe

2022-05-19 Thread Markus Volk
It doesn't need to be in core, but i would call having window decoration a core component because otherwise things will just not work as expected for weston/gnome-shell. I could also send that recipe to meta-oe if you prefer or just store it in meta-wayland since its wayland related stuff but if

Re: [OE-core] [meta-oe][RFC PATCH 1/2] libdecor: initial add recipe

2022-05-19 Thread Alexander Kanavin
Also, does this need to be in core (as opposed to meta-oe)? Why? Alex On Thu, 19 May 2022 at 18:43, Luca Ceresoli via lists.openembedded.org wrote: > > Hi Markus, > > Il giorno Thu, 19 May 2022 08:43:11 +0200 > "Markus Volk" ha scritto: > > > libdecor is a client-side decoration library for Way

Re: [OE-core] [meta-oe][RFC PATCH 1/2] libdecor: initial add recipe

2022-05-19 Thread Luca Ceresoli via lists.openembedded.org
Hi Markus, Il giorno Thu, 19 May 2022 08:43:11 +0200 "Markus Volk" ha scritto: > libdecor is a client-side decoration library for Wayland clients. It > is used by libsdl2 for window decoration and is required to provide > decoration for shells that use client-side decoration such as > gnome-shel

[OE-core] [meta-oe][RFC PATCH 1/2] libdecor: initial add recipe

2022-05-18 Thread Markus Volk
libdecor is a client-side decoration library for Wayland clients. It is used by libsdl2 for window decoration and is required to provide decoration for shells that use client-side decoration such as gnome-shell or weston. Signed-off-by: Markus Volk --- .../libdecor/libdecor_0.1.0.bb