ChanServ changed the topic of #wayland to: https://wayland.freedesktop.org | Discussion about the Wayland protocol and its implementations, plus libinput | register your nick to speak
<wlb>
weston Merge request !753 closed (compositor-drm: config output to reposition, independent-CRTC clone mode not support)
maxzor__ has joined #wayland
benklop has joined #wayland
maxzor__ has quit [Ping timeout: 480 seconds]
spstarr has joined #wayland
spstarr has quit []
spstarr has joined #wayland
eroux has joined #wayland
dcz has joined #wayland
maxzor__ has joined #wayland
cvmn has joined #wayland
tzimmermann has joined #wayland
ecloud has quit [Ping timeout: 480 seconds]
MajorBiscuit has joined #wayland
kenny has quit [Quit: WeeChat 3.3]
ecloud has joined #wayland
___nick___ has joined #wayland
cvmn has quit [Ping timeout: 480 seconds]
c7s has joined #wayland
cvmn has joined #wayland
<kennylevinsen>
technomancy: re: whether xkb supports "combos", how do you think an xkb layout would produce an uppercase A or, say, an exclamation point on shift 1 (for qwerty)? :)
<kennylevinsen>
it can't *produce* combos though - a common thing to do there is just to make it produce e.g. F13-F20, which most keyboards lack anyway.
mvlad has joined #wayland
fmuellner has joined #wayland
<dottedmag>
technomancy: This is something a compositor should do. I don't know any compositors that do though.
<jadahl>
I guess it doesn't support regular expressions..
pac85 has joined #wayland
<pac85>
Hi, I sore that years ago there was a proposal for "Wayland security module", something that would allow to manage security and therefore to expose sensitive tasks in Wayland compositors. However I couldn't find any recent news on that proposal, is it still in active development?
eroux has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
cvmn has quit [Ping timeout: 480 seconds]
MajorBiscuit has quit [Ping timeout: 480 seconds]
<jadahl>
emersion: https://lore.kernel.org/all/?q=l%3Adri-devel.lists.freedesktop.org+AND+NOT+%28dfn%3A%2Fdrivers%2F*+OR+dfn%3ADocumentation%2F*+OR+dfn%3Aarch%2F*+OR+dfn%3Ainclude%2F*%29+AND+NOT+%28s%3ARe%3A+OR+s%3A%220%2F%22+OR+s%3A%2200%2F%22+OR+s%3A%22%5Bgit+pull%5D%22+OR+s%3A%22%5BBug%22+OR+s%3APATCH+OR+s%3A%22%5Bpull%5D%22%29+AND+NOT+f%3Alkp%40intel.com+AND+NOT+f%3Asyzbot seems to give somewhat ok results,
<jadahl>
but I give up trying to combine that with the one filtering on interesting files :P
<emersion>
nice
<emersion>
probably one big OR?
flacks has quit [Quit: Quitter]
<jadahl>
i tried to wrap the two things in () and put an OR between but kept getting empty results
<emersion>
hm
flacks has joined #wayland
pac85 has quit [Remote host closed the connection]
pac85 has joined #wayland
pac85 has quit [Remote host closed the connection]
<technomancy>
so it sounds like there's no wayland equivalent for the feature I want; that's fine, I should be able to run my WM inside xwayland. in fact, I was doing this up to a couple weeks ago with xwayland inside cage and it worked great.
<technomancy>
however, an apt update broke the build of cage I was using and the new one crashes with a wlroots error about refusing to run because it detected a software cursor.
<technomancy>
in order to get it back I need to either fix cage or replace it. I tried replacing it by running just a stripped down config of sway, but that doesn't work; my X WM is convinced that there's another WM running in X and refuses to launch.
<technomancy>
so I could use some help either fixing cage (wtf is a software cursor?) or fixing sway so that it acts as a cage replacement
<emersion>
what's the exact error message?
<technomancy>
"Direct scan-out disabled by software cursor" followed by "Basic output test failed for eDP-1"
<technomancy>
both those messages appear to come from wlroots rather than sway
<technomancy>
hm; well maybe the actual problem is something it's failing to log?
<technomancy>
that sucks
<emersion>
can you get a backtrace?
<technomancy>
I don't really know anything about C
<kennylevinsen>
technomancy: if it's crashing, stacktrace. If sway isn't doing what you want, the debug output.
<technomancy>
wait hang on; I thought it was crashing and getting restarted by a different script but it might just be looping internally to cage
<marex>
technomancy: maybe you can try looking around WLR_NO_HARDWARE_CURSORS=1 ?
<marex>
(it's just a guess, I had to use that with sway some time ago)
<technomancy>
marex: wow, that did the trick! thanks
<technomancy>
never would have guessed that
<marex>
:)
d42 has quit [Read error: Connection reset by peer]
d42 has joined #wayland
rgallaispou has joined #wayland
dreda has quit [Read error: Connection reset by peer]
dreda_ has joined #wayland
dreda_ is now known as dreda
rgallaispou has quit [Ping timeout: 480 seconds]
Kingsy has quit [Ping timeout: 480 seconds]
marmarek is now known as Guest140
marmarek has joined #wayland
rgallaispou has joined #wayland
Guest140 has quit [Ping timeout: 480 seconds]
johnakede has joined #wayland
johnakede has quit []
MajorBiscuit has quit [Quit: WeeChat 3.3]
pac85 has quit [Remote host closed the connection]
pac85 has joined #wayland
<pq>
technomancy, you cannot run a custom X11 WM with xwayland in a normal setup. You would need to launch xwayland manually as a rootful (non-integrated) X server to be able to run your own X11 WM. Otherwise the Wayland compositor takes the X11 WM position if it supports xwayland.
<pq>
emersion, interesting, I'll probably ask again about that lore filter once I'm ready.
<emersion>
gives me some time to try to improve it :P
tzimmermann_ has quit []
_rgallaispou has joined #wayland
<pq>
technomancy, as for key combo remapping, that's not a Wayland thing at all. It's either a XKB keymap thing, a Wayland compositor feature, or between the display server and the device driver.
<pq>
IOW pretty much anywhere except over Wayland.
<technomancy>
pq: sure; it's not necessarily an xorg thing either, except that my xorg WM happens to implement it.
<technomancy>
I know it's not part of wayland itself; I was asking more if there was a wayland-compatible replacement
<pq>
technomancy, X11 offers protocol to mangle all apps' input.
<pq>
Wayland does not.
<technomancy>
I'm not in love with xorg, but I can't operate without this feature.
rgallaispou has quit [Ping timeout: 480 seconds]
<pq>
technomancy, I'm trying to explain what "wayland" is not a keyword you'd include when searching for that feature. :-)
<technomancy>
sure; I gave up on searching because I don't know the right terms; that's why I asked in here
<technomancy>
obviously it would be ideal if they would just fix it in firefox, but that bug's so old it's gonna start going thru puberty soon
<pq>
I don't think I saw you explain what you actually want to do?
<technomancy>
ah, well basically I need to be able to use Emacs key bindings in firefox
<technomancy>
so like ctrl-n becomes down-arrow, ctrl-p is up arrow, etc
<technomancy>
firefox's extension mechanism used to support this, but they removed it, so I switched to solving the problem with my WM
<pq>
in just one app, not in all apps?
<technomancy>
I mean, theoretically it'd be great if it worked everywhere but in practice firefox is the only non-emacs program I use
ppascher has joined #wayland
<pq>
I do wonder if that's not possible with a custom xkb keymap...
<kennylevinsen>
as I was hinting at, mapping key combos to symbols is within the realm of xkb
<pq>
does firefox look at symbols though? or is it looking at keycode?
<kennylevinsen>
I mean, it's how a keyboard layout can emit "A" on shift+a
<pq>
since "arrow up" key is not quite a symbol, or is it?
<kennylevinsen>
Hmm, good point, not sure
<emersion>
i think it is
<kennylevinsen>
wev/xev can probably answer that
<kennylevinsen>
It might just be a symbol that is the same on all "layers" or what xkb calls it...
<pq>
but then, doesn't ctrl+p turn into ctrl+arrow_up if ctrl+p is translated to arrow_up?
<technomancy>
it won't prevent the "is the ctrl key pressed" from showing, but it will send an arrow-up event in a way that pretends that ctrl is not being held, somehow
<technomancy>
(the xorg wm will, I mean, no idea about wev/xev)
<pq>
yeah - quite out of my knowledge. Wasn't there some macro design brewing... possible just as an idea, not even code
pac85 has quit [Remote host closed the connection]
leon-p_ has joined #wayland
leon-p_ has quit []
<leon-p>
You can modify an XKB layout to emit the arrow keys on various key+modifier combinations. I have a layout that allows me to scroll with j and k when capslock is active
<leon-p>
technomancy: ^ I have a blog article about that, if you are interested
<marex>
does anyone here know anything about the wayland-ivi-extension ? the project seems like a horrid hack and also quite dead
<technomancy>
leon-p: I am glad to see you share the same complaint about firefox as I do =)
<technomancy>
I really don't understand how they can release software that's missing such a critical feature and consider it complete
<leon-p>
to be fair, people who don't like either the mouse or the arrow keys for scrolling are probably not the target audience, so it makes sense that this is not supported first class. Just a shame it takes such a workaround.
_rgallaispou has quit [Ping timeout: 480 seconds]
<technomancy>
I guess, but why have any distinctions on what commands can and can't be rebound?
twinkleferret5 has quit []
___nick___ has quit [Ping timeout: 480 seconds]
<dottedmag>
Spec is quite terse about what "event serial" is. Is it required to be a number of event in session, or clients will work with a more relaxed definition (e.g. only events that emit this serial may increase it)?
<dottedmag>
And more general: are clients allowed to assign any meaning to the serials, or should they treat them as opaque objects?
cvmn has joined #wayland
rasterman has quit [Quit: Gettin' stinky!]
fmuellner has quit [Remote host closed the connection]
boistordu has joined #wayland
fmuellner has joined #wayland
<emersion>
serials are arbitrary
<emersion>
clients shouldn't try to assign any meaning
<emersion>
a valid compositor could always send 42 as serial
<emersion>
well, that'd be stretching things a bit
<emersion>
but don't assume it'll always increase monotonically
<emersion>
and don't assume it's unique
mvlad has quit [Remote host closed the connection]
<dottedmag>
I'm looking at it from the compositor's side. Okay, serials are opaque.
<dottedmag>
Will send a MR to clarify it in the spec.
<dottedmag>
This poses a question: does the serial in callback for wl_server::sync have any meaning, in particular should it have any relation to the serials returned by wl_pointer::enter et al?