Improve interval check method #2

conao3 · 2020-09-13T09:10:32Z

Those commit are commit in @zk-phi's fork (repository)

Accoding his blog, this idea is like below.

So what I thought was to reduce the number of false inputs by
invoking when "one tempo delay from other keystrokes", instead of
just checking for "a short enough interval between two keystrokes".

before

 \ カタカタ /      \ バン /     \ カタカタ /
h o g e h o g e      hj       h o g e h o g e . . .
----------------------------------------------------
                  |< 0.1s|

after

 \ カタカタ /      \ バン /     \ カタカタ /
h o g e h o g e      hj       h o g e h o g e . . .
----------------------------------------------------
                 |< 0.15s|
              |> 0.1s||> 0.25s|

irigone · 2020-12-17T11:19:26Z

Why don't they accept the patch? Also, is the package unmaintained anymore?

conao3 · 2020-12-17T11:26:30Z

What do you think about this PR, @tarsius? And are there any potential maintainers?

tarsius · 2020-12-18T17:20:56Z

What do you think about this PR

Please put some more work into it before asking me to review. It appears that you are making some tweaks and then tweak the same things in the next commit or so. It is not at all obvious to me why you are doing that. Maybe there are reasons for doing that; if so, then please document them.

.gitignore

key-chord.el

tarsius · 2020-12-18T17:25:43Z

key-chord.el

@@ -183,6 +172,8 @@

 ;; ########   History   ########################################
 ;;
+;; 0.7 (2017-06-29) zk-phi
+;; 	Add option `key-chord-safety-interval'


It makes little sense to release in the middle of a patch series. Also you should add the note about this new option in the commit that adds this new option. You can add a new history entry without releasing at the same time. Just use something like Unreleased instead of 0.7 (2017-06-29) zk-phi.

Thanks, I don't make a bump version in this PR.

tarsius · 2020-12-18T17:26:27Z

key-chord.el

  "Max time delay between two press of the same key to be considered a key chord.
 This should normally be a little longer than `key-chord-two-keys-delay'.")

-(defvar key-chord-safety-interval 0.08
+(defvar key-chord-safety-interval 0.15


Please explain why you are making these changes.

Also there is a typo in the commit message.

I don't know why he makes this change but this commit is disappeared after fixup commits.

tarsius · 2020-12-18T17:27:58Z

key-chord.el

@@ -173,7 +173,7 @@
 ;; ########   History   ########################################
 ;;
 ;; 0.7 (2017-06-29) zk-phi
-;; 	Add option `key-chord-safety-interval'
+;; 	Add option `key-chord-safety-interval-backward/forward'


You are adding two options and neither one is named key-chord-safety-interval-backward/forward.

I don't make a bump version, so this History entry was the whole disapear.

key-chord.el

conao3 · 2020-12-18T19:28:28Z

Fix commit history and force pushed.

This PR was originally a separate fork by zk-phi, which he continued to develop as a fork, not to create a such PR.
It was my mistake to make it into a PR as it is. I apologize for the hassle and thanks for the good review.

tarsius · 2020-12-22T20:34:51Z

I actually wanted to try this package for a long time and now I finally have.

I also cleaned it. Please rebase this pr on top of that cleanup. Sorry for the additional work this causes.

This PR was originally a separate fork by zk-phi, which he continued to develop as a fork, not to create a such PR.
It was my mistake to make it into a PR as it is. I apologize for the hassle and thanks for the good review.

No problem. I didn't realize what was going on.

So I assume you tried this for a while. In your experience, how (much) doee it improve the user experience? I don't have enough experience using this package even without this change, so I intend to test both approaches for a while before making a decision.

conao3 · 2020-12-23T02:30:34Z

rebase done.

To be honest, I'm not an enthusiastic fan of this package, but I thought zk-phi's suggestion made sense, so I suggested this PR.
(When he wrote the blog, upstream was already inactive and he was wondering where to send the patch)

It is reasonable to assume that the "key-chord decision" will be a little slower than the surrounding input speed, and I feel that this change has reduced the number of unintended command executions.

tarsius · 2023-05-22T02:24:23Z

Finally merged.

conao3 · 2023-05-22T07:10:18Z

thanks!

BlueDrink9 · 2024-01-08T23:39:23Z

@tarsius would you consider reverting this merge?

New issues created by merge

This has introduced #6, #7 and #8. ghost-of-freedom makes what I think is a good point, albeit brusquely:
Neither author nor merger uses this package much from what we can read here (correct me if this is wrong), whereas at least 11 people that do use it have been significantly enough disrupted by this PR to come here and mention it.

No old issues solved by merge

This repo had sat for 4 years without accruing many issue reports. Two issues are referenced as being fixed by this PR, of which neither are actually solved by it. #5 had an existing correct solution found by the author (but they never closed the issue). #4 seems unrelated, presumably closed by mistake.

This reverts commit e724def. This (emacsorphanage#2) should not have been merged. Maybe as an opt-in, but the default, which worked for many people for years, should have stayed the same. Closes emacsorphanage#6, emacsorphanage#7 and emacsorphanage#8.

tarsius · 2024-01-09T14:25:45Z

I agree. Sorry about that.

tarsius · 2025-03-19T22:31:22Z

A similar change has been proposed in #12. Please help us test and review that, because otherwise my only options are to simply reject or to merge and risk a repeat of what happened in this case.

tarsius requested changes Dec 18, 2020

View reviewed changes

conao3 force-pushed the master branch from 2584710 to 0ac95de Compare December 18, 2020 19:20

conao3 force-pushed the master branch from 0ac95de to 6e9b07b Compare December 23, 2020 02:23

Add option key-chord-safety-interval

f1baa26

conao3 force-pushed the master branch from 6e9b07b to f1baa26 Compare December 23, 2020 02:34

tarsius closed this May 22, 2023

This was referenced May 22, 2023

Keyboard layout tolerance #4

Open

Timeout #5

Closed

ghost-of-freedom mentioned this pull request Aug 24, 2023

key-chord stops working after Emacs farts #8

Closed

tarsius mentioned this pull request Mar 19, 2025

Improve performance, add typing detection #12

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Improve interval check method #2

Improve interval check method #2

conao3 commented Sep 13, 2020 •

edited

Loading

irigone commented Dec 17, 2020

conao3 commented Dec 17, 2020

tarsius commented Dec 18, 2020

tarsius Dec 18, 2020

conao3 Dec 18, 2020

tarsius Dec 18, 2020

tarsius Dec 18, 2020

conao3 Dec 18, 2020

tarsius Dec 18, 2020

conao3 Dec 18, 2020

conao3 commented Dec 18, 2020 •

edited

Loading

tarsius commented Dec 22, 2020

conao3 commented Dec 23, 2020

tarsius commented May 22, 2023

conao3 commented May 22, 2023

BlueDrink9 commented Jan 8, 2024 •

edited

Loading

tarsius commented Jan 9, 2024

tarsius commented Mar 19, 2025

Improve interval check method #2

Improve interval check method #2

Conversation

conao3 commented Sep 13, 2020 • edited Loading

before

after

irigone commented Dec 17, 2020

conao3 commented Dec 17, 2020

tarsius commented Dec 18, 2020

tarsius Dec 18, 2020

Choose a reason for hiding this comment

conao3 Dec 18, 2020

Choose a reason for hiding this comment

tarsius Dec 18, 2020

Choose a reason for hiding this comment

tarsius Dec 18, 2020

Choose a reason for hiding this comment

conao3 Dec 18, 2020

Choose a reason for hiding this comment

tarsius Dec 18, 2020

Choose a reason for hiding this comment

conao3 Dec 18, 2020

Choose a reason for hiding this comment

conao3 commented Dec 18, 2020 • edited Loading

tarsius commented Dec 22, 2020

conao3 commented Dec 23, 2020

tarsius commented May 22, 2023

conao3 commented May 22, 2023

BlueDrink9 commented Jan 8, 2024 • edited Loading

New issues created by merge

No old issues solved by merge

tarsius commented Jan 9, 2024

tarsius commented Mar 19, 2025

conao3 commented Sep 13, 2020 •

edited

Loading

conao3 commented Dec 18, 2020 •

edited

Loading

BlueDrink9 commented Jan 8, 2024 •

edited

Loading