01:01 i509vcb: Is there a reason why newer protcols which adopted RFC 2119 chose lowercase keywords rather than the newer interpretation with uppercase keywords?
01:05 emersion: the README says the words are lowercase
02:45 kchibisov: Are wl_pointer coordinates account for fractional scale when wp-fractional-scale-v1 is used? By that I mean that I'd need to mul by e.g. scale 1.5 my surface_{x,y} from wl_pointer::motion if I want to convert them to 1 buffer coordinates?
03:17 kchibisov: Though, I guess it's like that, given that surface local is the original size(prior scaling). So to get the actual coordinate it's exactly the same as with integer scaling/set_buffer_scale.
05:47 Lightsword: anyone know of any examples for a simple numeric pinpad(with a few keys like forward/back/enter/clear) hid descriptor/report format that plays well with libinput?
06:12 whot: Lightsword: libinput doesn't care much about the HID format, only evdev, so whatever the kernel accepts and parses correctly
06:13 whot: Lightsword: depending on your use-case, a uinput device is more directly controlled since you set the bits that libinput will consume
06:15 Lightsword: whot, hmm, I'm using a uhid library at the moment to send hid reports(which do show up in "libinput record --with-hidraw") but I'm messing up the format so it doesn't seem to work correctly
06:17 Lightsword: whot, the hardware I'm interfacing with just emits simple key down and key up events
06:20 Lightsword: whot, does uinput have similar key down key up event handling to hid?
07:15 whot: Lightsword: yes, you just send KEY_FOO events with value 1 or 0 for press/release
07:16 whot: Lightsword: https://python-libevdev.readthedocs.io/en/stable/examples.html#creating-a-uinput-device-and-sending-events
07:23 Lightsword: whot, ah, that does look simpler, should that play well with python's asyncio(I wrote the hardware serial protocol interface code using pyserial-asyncio)?
07:40 kennylevinsen: kchibisov: fractional scale does not change anything. The pointer is fixed_t though, so already able to move fractional logical positions.
07:41 whot: Lightsword: no idea, tbh but it doesn't rely on any event loop
07:41 kchibisov: kennylevinsen: I was more asking whether the coordinate will be scaled or not. But from what I understand it still in logical pixel(surface-local).
07:42 kchibisov: So the logic is the same as with integer scale when it comes to converting pointer events from surface-local to buffer coordinates.
07:42 whot: Lightsword: there's also https://python-evdev.readthedocs.io/en/latest/ which doesn't use libedev, and it mentions asyncio in the docs
08:55 pq: kchibisov, yes, nothing changes wrt. to input coordinates. If the client wants buffer coordinates, you get those exactly the same as before. E.g. if you use wl_viewport, then you take that into account.
11:45 naveenk2: Hi slattan:
12:03 ifreund: does anyone know of any ext-session-lock-v1 clients other than swaylock or waylock?
13:31 pq: SardemFF7, FYI, I see the Weston Gitlab "Webhook disabled" banner again, maybe from yesterday, and now the event log is empty.
13:59 SardemFF7: pq: thanks for the ping
14:00 SardemFF7: still no clue about the cause though, but I think I see an auth failure on my email client (on my phone only) about the same time a webhook seems to fail
14:00 pq: SardemFF7, should I keep pinging you when I see it the next time?
14:01 pq: err, *if I see it again :-)
14:01 SardemFF7: sure, if that doesn’t bother you :-)
14:01 pq: no, it's the banner that bothers me :-D
14:03 SardemFF7: I’ll do my best to find the cause, promise ^^'
14:03 pq: thanks!