<chrcav>
Not personally a wayland user but I was pretty sure xwayland should be enabled by default on sway. Did you explicitly disable xwayland in your sway config?
chrcav has quit [Quit: Lost terminal]
<mnkydeth>
I didn't disable it. I did make sure to put xwayland enable in the .config/sway/config
tilman has quit [Ping timeout: 256 seconds]
<mnkydeth>
I just tried greetd and tuigreet. But no luck there either.
tilman has joined #crux
<mnkydeth>
I may try modifying my start script a bit. I found TimB87 crux-wayland startup script. He has a little different config then what I used.
<mnkydeth>
I mostly used the oper.io example as my base for crux. switching from x11 to native-wayland.
<mnkydeth>
Not much time left for me tonight to play. So, I'll try a few things over the next couple days. Just figured I would put out a line to see what people typically do here that run wayland. config files or logins etc.
<farkuhar>
mnkydeth: if you compiled sway while xorg-xwayland was not present, then sway would have been configured with no xwayland support. Try rebuilding sway now that you have xorg-xwayland installed.
<farkuhar>
it sounds from your description "finally decided to try steam" that maybe you added the xorg ports after setting up a pure wayland desktop. If that's the case, rebuild sway and see if the pkgmk log shows that xwayland was detected.
chrcav has joined #crux
chrcav has quit [Quit: Lost terminal]
chrcav has joined #crux
<chrcav>
SiFuh: turns out I was missing CONFIG_EFI_STUB=y in my kernel config. Once I added that syslinux started working.
<jaeger>
interesting, I didn't think stub was required for syslinux
<jaeger>
Usually stub support means the kernel itself is the UEFI bootloader
<jaeger>
mnkydeth: sway needs to be recompiled after xwayland is installed, if you haven't done that already
<SiFuh>
jaeger: in the pdf I wrote for syslinux I had CONFIG_EFI_STUB=y as well
groovy4shoes has quit [Remote host closed the connection]
groovy2shoes has joined #crux
pvn has joined #crux
pvn has quit [Quit: Leaving.]
pvn has joined #crux
XXX1232 has joined #crux
<cruxbot>
[opt.git/3.7]: php: update to 8.1.13
<cruxbot>
[core.git/3.7]: tzdata: update to 2022g
frinnst has quit [Ping timeout: 265 seconds]
<cruxbot>
[core.git/3.7]: vim: update to 9.0.0979
<cruxbot>
[core.git/3.7]: xz: update to 5.2.9
<cruxbot>
[core.git/3.7]: cmake: update to 3.25.1
<cruxbot>
[opt.git/3.7]: gvim: update to 9.0.0979
<farkuhar>
stenur: for me it's the opposite, the libera irclogs page looks total bozo in a webkitgtk browser (nyxt with zoom-ratio 1.25), but fine in firefox-bin.
<SiFuh>
Thanks chrcav_. jaeger I will query and verify in #syslinux about this.
<SiFuh>
I am finding a lot of references to it, but none explaining why or if it is needed
<SiFuh>
farkuhar: I will keep you posted on this as well, if it is as chrcav_ says it might need to be specified in the documentation.
ty3r0x has quit [Ping timeout: 256 seconds]
<farkuhar>
SiFuh: thanks. I'm now seeing in the Slackware syslinux docs a few comments that I don't remember being there earlier. Especially the four bullet points after "Frankly, there are no compelling reasons not to use CSM ... On the contrary, there are reasons to avoid UEFI booting, because ..."
<SiFuh>
"In practice, the EFI stub loader is generally used with a separate boot manager program—most commonly rEFInd or gummiboot/systemd-boot, but GRUB 2 can be used this way, too."
<SiFuh>
Maybe SYSLINUX is booting the kernel directly through UEFI stub?
<SiFuh>
UEFI and stub* like a man in the middle
ty3r0x has joined #crux
<chrcav_>
That might be what's happenning. I ended up switch from GRUB2 to SYSLINUX on two systems and enabling EFI stub was the only change I needed to get running.
<SiFuh>
chrcav_: Yeah, that is cool. I reckon you found the issue
emmett1_ has joined #crux
chrcav_ is now known as chrcav
<SiFuh>
chrcav_: All my systems have had EFI STUB enabled and I even wrote that it is to be enabled in my original documentation a couple of years ago. I never really looked into what STUB was until later. I had the same idea as jaeger that it was only used from UEFI to Kernel.
<chrcav>
I mean that's what the kernel doc says about it.
<SiFuh>
Yeah, so hopefully the SYLINUX guys can clear that up. Either way, whether they confirm or not, we may need to make a note of this for future installations in the CRUX handbook
<SiFuh>
I'd prefer actual confirmation though.
<SiFuh>
I may end up examing the source instead to find the answer.
<cruxbot>
[opt.git/3.7]: glib: update to 2.74.3
<jaeger>
I generally don't have stub support in my kernels except for when I tested booting a kernel directly with it for the documentation ages ago
<SiFuh>
I am pretty sure I picked up the addition of the kernel option from Arch Linux's SYSLINUX documentation.
<SiFuh>
To be honest. SYSLINUX was such a pain in the arse to get around because there was so much trash documentation out there and limited explanations as well.
<SiFuh>
This is the reason I decided to write the cheat sheet and the basic documentation for myself. Mostly in layman's terms
SiFuh has quit [Remote host closed the connection]
SiFuh has joined #crux
pvn1 has quit [Quit: Leaving.]
guido_rokepo has quit [Quit: guido_rokepo]
emmett1_ has quit [Ping timeout: 246 seconds]
<stenur>
For EFI_STUB to make sense the kernel must be on the EFI partition, right. Otherwise you need something to chain-load. I use a static kernel on EFI to chain load another one via kexec-tools on an encrypted partition.
<SiFuh>
Yes, but this isn't what we are talking about.
<SiFuh>
Basically, we are told that EFI_STUB is to load the kernel directly without a bootloader. However, it appears that in UEFI boot with SYSLINUX we may actually need EFI_STUB enabled in the kernel or it will not boot.
<stenur>
hard to believe actually.
<SiFuh>
Yeah, and the SYSLINUX guys haven't replied either.
<SiFuh>
All the howtos and documentation I have read that mention kernel configuration options are all saying to enable it, but none explain why.