Closed al3xandru closed 1 year ago
weird. the code is indeed the same for every scrolling move (just distance values change). some questions then:
j
pressed for a while and stopsjfi, nobody else reported. and i haven't had this issue (after v1. had something similar during dev of course, because i built the shit properly).
k
tooalright, thanks, let me know. coz this is just plain weird π οΈ
if j
keeps going down that means that somehow the j
key up event is not being sent to Scrolla, if that is of any help.
I have disabled all other apps. And tried again with TextEdit. Same behavior. I'm trying to add a short screen recording, but I'm not sure what that would clarify.
would still be nice. really wanna see.
Just seconds after posting... I noticed there is a lag in typing the j
key. I then realized that the only app I have not stopped is "kindaVim" which is set to activate using... can you guess? jk
π
I don't know what can be done or if there's anything to be done. From my pov the ticket can be closed.
gimme some time to think about it. i'll come back to you tomorrow.
ok, exploring. kV only cares about catching key presses, not key releases, so it's gonna be tricky. a question tho: how do you expect the thing to work? like if you use the jk
sequence to enter Normal Mode with kV, the j
press always has to be delayed, for the whole macOS, to wait for the next key press and check if it's your sequence. which means even if we get this to work, you'll have a delay with scrolling down with j
in Scrolla: that'll be the delay set in your kV Settings, because kV will hold the key during during that delay period. would that be ok? (not even sure it's doable. may require make big changes to kV, which i'm not really keen to π¬οΈ).
Honestly, I have no idea how this combination could be made different (or at least less unexpected). While using the jk
chord to activate kindaVim sounds great on paper, it has a bit more impact on the system. Using it in a single app (vim) is slightly better from this perspective as it doesn't impact the whole system. IMO I'd put a note on kindaVim (or a link to the site) explaining 1) the impact; 2) potential unwanted interaction with Scrolla.
While using the
jk
chord to activate kindaVim sounds great on paper, it has a bit more impact on the system. Using it in a single app (vim) is slightly better from this perspective as it doesn't impact the whole system.
agreed. i don't use it myself and it wasn't in the first 20 prod versions i think. but it's been asked by several people. could be done through Hammerspoon but i've made an integrated version. it started as typing j
directly, and removing it you typed k
fast enough. but caused issue when using j
and k
not when typing letters but when using those letters to interact with UI Elements. so i went to the only possible delay route. i would tend to think that most people who are big users of jk
know the tradeoffs.
IMO I'd put a note on kindaVim (or a link to the site) explaining 1) the impact; 2) potential unwanted interaction with Scrolla.
i'm also a big fan of avoiding complications and having texts and explanations everywhere π
οΈποΈ the thing with Scrolla is that it never receives the keyup event for j
, which is why it keeps scrolling. that definitely has to be worked out. but i think if there's a delay when using j
because kV (or any other app like Hammerspoon for that matter) has a delay for jk
, that should be quite understandable by the users. most of users of kV/Hammerspoon/Karabiner-Elements etc. know what they're getting into ποΈ
@al3xandru i'm gonna release an update on kV that will mitigate the issue by sending a key up event when the j
(or whatever first letter used) delay is passed. that will solved the problem that Scrolla never received the key up event. BUT you'll have a delay when scrolling using j
with Scrolla. i think ultimately, with A LOT OF WORK AND CHANGES, i could reduce this delay. but not sure i wanna do it. that would require massive changes in kV's event system. not sure it's worth it. so i'll release this update soon, and you try and see what you think. we can then decide to dig deeper if needed from there.
@godbout the up event idea sounds quite smart! While I went back to using shortcuts for activation so this is not directly affecting my usage, I believe it's going to be useful long term.
ποΈ
just out of curiosity, so now you don't use a sequence anymore at all? did you have to change your Vim habits? or your were just testing around?
improved with kV47: https://github.com/godbout/kindaVim.docs/releases/tag/47
@godbout activating kindaVim is not yet established so tweaking the activation shortcut is not painful. Or, differently put, having to deal with a delayed j
key is more painful than changing that activation. Honestly, it's early days for kindaVim on my side. I don't know how many overlapping users are there between kindaVim and Scrolla and out of those how many are using the chord activation, but your improvement will make their experience better.
yeah. i'm not a big fan of the delay j
, which is why i've never got into the jk
thing. personally i use escape
to enter Normal Mode, and i've remapped caps_lock
to escape
through the macOS Settings. definitely works very well for me.
Title tries to explain it all: when pressing
j
the area continues to scroll down instead of that action occurring once. That's howk
and other keys behave.