dnkl changed the topic of #foot to: Foot - fast, lightweight and minimalistic Wayland terminal emulator || 1.16.2 || https://codeberg.org/dnkl/foot || channel logs: https://libera.irclog.whitequark.org/foot
sgm has quit [Remote host closed the connection]
sgm has joined #foot
boomboxnation has quit [Ping timeout: 272 seconds]
boomboxnation has joined #foot
hspak has joined #foot
alexander has quit [Quit: ZNC 1.8.2+deb3.1 - https://znc.in]
alexander has joined #foot
hspak has quit [Quit: Ping timeout (120 seconds)]
hspak has joined #foot
<Arnavion> dnkl: For completeness, could you do the custom tarball + signature for tllist too?
hspak has quit [Quit: Ping timeout (120 seconds)]
hspak has joined #foot
__jmcantrell__ has joined #foot
jmcantrell has quit [Killed (lead.libera.chat (Nickname regained by services))]
__jmcantrell__ is now known as jmcantrell
jmcantrell_ has joined #foot
hspak has quit [Quit: Ping timeout (120 seconds)]
hspak has joined #foot
cbb has quit [Quit: WeeChat 4.2.1]
<dnkl> rjarry: the behavior you're seeing is still consistent with current master branch, *not* the MR
<dnkl> dbg: input.c:1572: pressed: space (32/0x20), seat=default, term=0x52100001b900, serial=561, mods=Shift (0x00000001), consumed=Shift+Mod5 (0x00000081), locked=Mod2 (0x00000010), repeats=1
<dnkl> that's what I'm seeing, and I get the expected output (text only), but *only on the MR, not the master branch*
<dnkl> Arnavion: done
<dnkl> rjarry: alternatively, if you really are on the right branch, the issue is that your keymap is reporting shift as enabled, but not consumed. That too would explain what you're seeing
<dnkl> 130 (in the kitty kbd) isn't num-lock only, but shift+num-lock
<dnkl> something that perhaps _should_ be changed in foot is that we currenly report locked modifiers as enabled modifiers, if there are other active, non-consumed, modifiers (e.g. alt, or ctrl)
<dnkl> kitty does the same however
<dnkl> perhaps we should change foot to only consider "pressed" keys when reporting kitty modifiers
<dnkl> that would fix another issue I'm seeing; numlock+space (that is, *holding* num-lock), for example, doesn't report as an escape sequence. It probably should
<dnkl> not sure if that would mess up e.g. sticky keys though...
<dnkl> or maybe reporting locked keys is how the protocol is supposed to work... leave it up to the application to decode
<Arnavion> Thanks
luca has quit [Remote host closed the connection]
luca has joined #foot
<dnkl> so, the kitty kbd specification says this: "The modifiers optional parameter encodes any modifiers pressed for the key event".
<rjarry> dnkl: I am positive that I am on the correct branch
<dnkl> rjarry: alright... but as long as your keymap doesn't report shift as being consumed, there's nothing foot can do
<rjarry> :(
<rjarry> am I testing this correctly ? do I need to enable kitty magic keyboard protocol before pressing shift+space ?
<dnkl> you'll obviously want to enable the kitty keyboard protocol if you want to see how foot encodes the key-press in the kitty protocol
<dnkl> but it won't change how modifiers are reported by xkb (and thus logged by foot)
<rjarry> that was my question yes
<rjarry> I have no idea why is xkb reporting this
<rjarry> maybe because I disabled caps lock ?
<dnkl> no idea either
<dnkl> I pushed another commit on the kitty-is-text branch; it changes how we look at the modifiers, from all modifiers being enabled, to only those being depressed.
<dnkl> I don't think it'll change what you're seeing rjarry , since it shouldn't affect the consumed state of shift
<dnkl> but might be worth testing anyway
<rjarry> dnkl: I will have a look later, thanks !
<dnkl> while I believe the last commit implements the spec, I'm not it's what the specification *intended*
<dnkl> alright, he updated the specification to be more clear - we should be reporting *enabled* modifiers. Not just pressed
<dnkl> he also admitted kitty wasn't following the specification 100%. So we shouldn't blindly follow kitty's implementation
<dnkl> I'll revert the last commit, and try to follow the specification more closely
<dnkl> this means you *will* get escape sequences whenever e.g caps or NumLock are enabled
<dnkl> rjarry: see comments above (i.e not much point in testing the last commit since it'll be reverted)
<dnkl> rockorager: this is probably something you'd like to take notice of too: the kitty spec was just updated to clarify locked modifiers
<dnkl> locked modifiers _should_ always cause the corresponding bit in the kitty key event to be set. When *enabled*. I.e. not just when they are pressed
<dnkl> this means having num-lock enabled will cause *all* key events to be escape sequence based, rather than text based
<dnkl> same is true for caps-lock
<dnkl> alright, just pushed another commit to kitty-is-text, with updated behavior for locked modifiers
Biolunar has quit [Quit: brb]
Biolunar has joined #foot
hspak has quit [Quit: Ping timeout (120 seconds)]
hspak has joined #foot
jmcantrell has quit [Quit: WeeChat 4.2.1]
jmcantrell_ is now known as jmcantrell
bookworm has quit []
<dnkl> PR rebased onto latest master, and PR description updated
bookworm has joined #foot
miko has quit [Remote host closed the connection]
<rjarry> dnkl: here's what I get https://0x0.st/H7zB.txt
<rjarry> so that means we need to handle this as a special case in tui apps?
<dnkl> you need to handle modifiers 🤷
<dnkl> but you always had to?
<dnkl> if you're not interrested in caps- and num-lock, just filter out those two modifiers
miko has joined #foot
<rjarry> I mean, it is expected that when numlock/capslock is pressed and/or locked, we don't get a unicode character but a CSIu sequence?
<dnkl> yes
<rjarry> ok
<rjarry> that's what I meant by "handle this as a special case"
<rjarry> we shouldn't expect shift+space to output unicode
luca_ has joined #foot
<dnkl> not just that combination. *all* key presses will generate CSIu when NumLock is active
<dnkl> or caps
luca has quit [Ping timeout: 272 seconds]
<rjarry> weird, but that's what it is
<rjarry> but I'm not sure how your patches change things
<rjarry> I get the same CSIu sequence with shift+space
<rjarry> than before your patches
<dnkl> it changes how xkb sets the consumed modifiers
<dnkl> it doesn't change it for you, because for some reason, your keymap still doesn't report shift as consumed
<dnkl> it does change things over here, for me
<rjarry> nothing to do with capslock nor numlock?
<dnkl> it also changes how caps and NumLock are reported. But that's not really related
luca_ is now known as luca
<dnkl> well, if you have NumLock enabled you'll now get a CSI regardless of what your keymap says about shift being consumed or not
<dnkl> so, related in that sense
<rjarry> here's my keymap
<dnkl> I can certainly test it, but it's not going to change how we do things in foot
<dnkl> I don't have the knowledge to debug your keymap I'm afraid
<rjarry> me neither, lol
<rjarry> I literally copy pasted an example from some SO post
<rockorager> dnkl: thanks for the notice
<rockorager> dnkl: is kitty going to change their encoding...?
<rockorager> s/encoding/implementation
<rockorager> printf "\e[>1u" && cat -v
<rockorager> in kitty, with caps on and I get plain text
<rjarry> rockorager: can it depend on the xkb layout ?
<dnkl> rockorager: don't know if kitty will change anything. According to https://github.com/kovidgoyal/kitty/issues/7183#issuecomment-1974674879, you should be getting a CSI. That's my interpretation at least
<rockorager> Hm
narodnik has quit [Quit: WeeChat 4.2.1]
narodnik has joined #foot
sentriz has quit [Ping timeout: 240 seconds]
<rockorager> TBH I'm not a fan of that
<rockorager> Seems like locked mods shouldn't be considered unless report all keys as escapes is on
narodnik has quit [Quit: WeeChat 4.2.1]
narodnik has joined #foot
romildo has joined #foot
sentriz has joined #foot
<dnkl> manio: btw, which compositor are you running?
<manio> sway
narodnik has quit [Quit: WeeChat 4.2.1]
narodnik has joined #foot
narodnik has quit [Quit: WeeChat 4.2.1]
narodnik has joined #foot
boomboxnation has quit [Ping timeout: 260 seconds]
boomboxnation has joined #foot
romildo has quit [Quit: Leaving]