04:38 YaLTeR[m]: how normal is it for two monitors from the same manufacturer to have the exact same model, make, and serial? I thought you could rely on model+make+serial being more or less unique to store things like window positions across monitors
04:50 soreau: easier to use the crtc/connector name
05:10 YaLTeR[m]: then it won't work across connector name changes. Like if you have two GPUs, their connector names can come up in parallel and race, and you have different ones every boot. Or displayport docks bump connector name with every monitor
07:17 wlb: weston Issue #959 opened by () Segfault in simple-egl example https://gitlab.freedesktop.org/wayland/weston/-/issues/959
07:50 wlb: weston Issue #959 closed \o/ (Segfault in simple-egl example https://gitlab.freedesktop.org/wayland/weston/-/issues/959)
07:57 bbhtt: did wayland-protocol 1.38 bump the wayland scanner version required?
07:57 bbhtt: getting an error trying to build `linux-dmabuf/linux-dmabuf-v1.xml:113: element event: validity error : No declaration for attribute deprecated-since of element event`
08:11 bbhtt: ok, seems it affects the tests only
08:36 kennylevinsen: YalTeR[m]: note, I have occasionally seen that as a mishap when the monitors do actually have unique serials but EDID handling messed up and gave the same result for both monitors. Double check the EDID to make sure it’s actually foo.
08:43 YaLTeR[m]: huh
09:16 zamundaaa[m]: YaLTeR: two monitors having the exact same EDID is a thing, you can't fully rely on it
09:18 zamundaaa[m]: KWin has a system where it'll use the edid serial + manufacturer + model + manufacturing date when it's unique, a hash of the entire edid when it's not, mst path or connector name when that's not unique either
09:19 zamundaaa[m]: And yes, connector names are not something anyone should ever use for persistent data, but we don't always have a better option
09:21 kennylevinsen: I’d be a little bit surprised if the serial is the same but the rest of the EDID differs to give a different hash
09:22 kennylevinsen: Ah I have seen zero-serial eDP panels so I guess that’s one way for that to happen
09:27 kennylevinsen: zamundaaa[m]: how does that work if, say, you start with one monitor, then plug in one that conflicts?
09:30 zamundaaa[m]: kennylevinsen: yep, zero serial is what that's for
09:31 zamundaaa[m]: About plugging in a second display with the exact same EDID... it'll just load the display configuration according to changed rules
09:32 zamundaaa[m]: As it's pretty much exclusively about external displays, that scenario is luckily pretty uncommon
09:33 kennylevinsen: Yeah that seems fair
10:13 YaLTeR[m]: zamundaaa @zamundaaa:kde.org: i see, that's interesting
11:30 wlb: wayland Issue #503 opened by () Wayland bug's https://gitlab.freedesktop.org/wayland/wayland/-/issues/503
11:34 wlb: wayland Issue #503 closed \o/ (Wayland bug's https://gitlab.freedesktop.org/wayland/wayland/-/issues/503)
12:19 emersion: bl4ckb0ne: would you be interested in replacing Drew DeVault for drm-lease protocol maintainership?
12:19 bl4ckb0ne: sure
12:23 wlb: wayland-protocols Merge request !351 opened by () drm-lease: nominate Simon Zeni as maintainer https://gitlab.freedesktop.org/wayland/wayland-protocols/-/merge_requests/351
12:23 emersion: there :)
12:24 bl4ckb0ne: yay
12:27 fililip: are the time fields in the core protocol meant to be 32bit or 64bit? it says uint everywhere
12:27 emersion: https://wayland.freedesktop.org/docs/html/ch04.html#sect-Protocol-Wire-Format
12:29 fililip: is that intentional? 32bit millisecond time can only be measured for around 49 days
12:30 emersion: time is usually represented with multiple uints
12:50 fililip: unless for something like wl_pointer.motion time doesn't have to constantly increase and is just relative
12:50 fililip: that's the part I don't get
12:51 emersion: then it just wraps around, and clients need to handle this
12:51 fililip: oh, ok, thanks for clarifying!
13:02 wlb: wayland-protocols Merge request !352 opened by () ext-image-copy-capture: fix the error in the protocol description https://gitlab.freedesktop.org/wayland/wayland-protocols/-/merge_requests/352