<novakane>
ifreund: can confirm that unmap is fixed for me!
<ifreund>
nice :)
<novakane>
there is just the floating view below tilling view but other than that I have no other issue now
<ifreund>
ah yeah, that was a bit of an experiment but I think it's better to switch that back to something more similar to master branch
<novakane>
I think so, I'm pretty sure there will be a lot of complaints about that otherwise lol
uncomfy has joined #river
<leon-p>
ok, I found an actually funny bug
<leon-p>
I just tried to resize a floating imv view and it started to "wiggle"
<leon-p>
can reliably reproduce
<leon-p>
goes away when the view is no longer visible, either by unfocusing the tags for by hiding it behind another view
<tleydxdy[m]>
sounds like a featuree
<leon-p>
totally
<leon-p>
seems like github does not like the screencap though
<leon-p>
easiely reproducable with imv, for me at least
uncomfy has quit [Remote host closed the connection]
<ifreund>
heh, I can't reproduce
<leon-p>
I'll get you a backtrace of the crash that sometimes happens afterwards in a few minutes
<ifreund>
thanks
Szadek5 has quit [Quit: Ping timeout (120 seconds)]
Szadek5 has joined #river
<novakane>
oh I can reproduce, no crash though, it just goes to an infinite loop of transaction
<ifreund>
yeah, I just managed to reproduce as well
<ifreund>
haven't gotten any crash though
<novakane>
not for epileptic people
<leon-p>
I managed to crash it twice, but that may also be related to the screenrecording, haven't been able to crash it since then either
<ifreund>
novakane: floating views are now rendered above layout views, that commit might be a nice example to look at to better understand the scene graph stuff if you like :)
<novakane>
nice, I'll have a look!
dbuckley has joined #river
<ifreund>
leon-p: hmm, so imv is definitely buggy here but river needs to handle this special kind of buggy better
<ifreund>
in response to a configure with a new size imv literally acks the configure then commits the old buffer again
<ifreund>
*then* it resizes and commits a new buffer
<ifreund>
wtf imv
<leon-p>
realistically, wiggling still falls in the "somewhat annoying but ok"-category w.r.t. buggy / broken clients; it's the crashing I find concerning
<ifreund>
Yeah same, I haven't managed to reproduce any crash yet related to imv
<leon-p>
I'll try to reproduce again later, now I am off to a concert :)
<plumeus>
Speaking of imv, I always had an issue where closing River while its open leaves it running in the background with 100% CPU usage
<leon-p>
imv has many many many bugs...
<plumeus>
Killing the process will cause the image to be attempted to be opened on Firefox
<leon-p>
it's not a good wayland clients by any means
<leon-p>
anyway, now I am /really/ off, bye
<ifreund>
o7
<plumeus>
Bye
smach has quit []
angry_vincent has quit [Remote host closed the connection]
angry_vincent has joined #river
angry_vincent has quit [Changing host]
angry_vincent has joined #river
waleee has joined #river
angry_vincent has quit [Remote host closed the connection]
<novakane>
ifreund: so the inflight state in Output and View is a new state happening between pending and current, that's right?
<ifreund>
novakane: inflight is the state used in an ongoing transaction. Transactions can no longer be preempted like on master branch so we need to separate it from the pending state
<novakane>
so is it, pending is sent to clients, clients respond and inflight state change and then become current?
<ifreund>
no, pending is never sent directly to clients. root.applyPending() is called and if there is no inflight transaction it will update the inflight state and start one
<ifreund>
then when the transaction is completed the inflight state will be copied to current and the scene graph
Nulo has quit [Read error: Connection reset by peer]
Nulo has joined #river
<novakane>
alright, I get it, thanks, trying to catch up with the new code, just have the scene branch running but I didn't have time to really looks at the commits yet
Nulo has quit [Ping timeout: 268 seconds]
Nulo has joined #river
Guest0 has joined #river
Guest0 has quit [Client Quit]
bonic has joined #river
<bonic>
ifreund: Do you want me to help reimplement pointer constraints for the scene graph branch? I feel kinda bad that you have to reimplement it since I was the one that initially added it and I hadn't helped with it after it got merged.