Re: Patch 8.1.2052

On Di, 17 Sep 2019, Bram Moolenaar wrote: > Patch 8.1.2052 > Problem:Using "x" before a closed fold may delete that fold. > Solution: Do not translate 'x' do "dl". (Christian Brabandt, closes #4927) > Files:src/normal.c, src/testdir/test_fold.vim Quoting from the related issue

Re: [vim/vim] :! {cmd} quoting issues on windows 10 when using bash as shell (#4950)

In my version of git-for-windows, bash defines that environment variable as ProgramW6432. On Tuesday, September 17, 2019 at 6:10:01 PM UTC-4, CrispyDrone wrote: > > *Describe the bug* > I've exhausted all options and followed the documentation but am unable to > get :! {cmd} to work with the f

Patch 8.1.2052

Patch 8.1.2052 Problem:Using "x" before a closed fold may delete that fold. Solution: Do not translate 'x' do "dl". (Christian Brabandt, closes #4927) Files: src/normal.c, src/testdir/test_fold.vim *** ../vim-8.1.2051/src/normal.c2019-09-09 20:04:04.738340561 +0200 --- src/no

Re: [vim/vim] Add g:json_no_conceal to disable conceal for json filetype (#4948)

> there is an option named g:vim_json_warnings in this file, maybe use > g:vim_json_conceal is better, also doc need to be updated. Yes, this would be more consistent: if has('conceal') && (!exists("g:vim_json_conceal") || g:vim_json_conceal==1) Avoiding "no" is also good. Can you sug

Re: LGTM changes

Christian wrote: > we are currently using the LGTM service that runs some static analysis > on new PRs. However I just noticed that there is going to be a change > according to their blog: > https://blog.semmle.com/github-apps-lgtm-migration/ > > So according to it, we should switch to the lg

Patch 8.1.2051

Patch 8.1.2051 Problem:Double-click test is a bit flaky. Solution: Correct entry in list of flaky tests. Files: src/testdir/runtest.vim *** ../vim-8.1.2050/src/testdir/runtest.vim 2019-09-01 14:45:23.753964955 +0200 --- src/testdir/runtest.vim 2019-09-17 21:26:55.210070144 +

Patch 8.1.2050

Patch 8.1.2050 Problem:Popup window test fails in some configurations. (James McCoy) Solution: Clear the command line. Files: src/testdir/test_popupwin.vim, src/testdir/dumps/Test_popupwin_scroll_10.dump *** ../vim-8.1.2049/src/testdir/test_popupwin.vim 2019-09-15

Patch 8.1.2049

Patch 8.1.2049 (after 8.1.2048) Problem:Cannot build tiny version. Solution: Add #ifdefs. Files: src/main.c *** ../vim-8.1.2048/src/main.c 2019-09-17 20:28:34.148873666 +0200 --- src/main.c 2019-09-17 20:53:19.456620826 +0200 *** *** 1061,1071 --- 1061,1073

Patch 8.1.2048

Patch 8.1.2048 Problem:Not clear why SafeState and SafeStateAgain are not triggered. Solution: Add log statements. Files: src/getchar.c, src/main.c, src/proto/main.pro *** ../vim-8.1.2047/src/getchar.c 2019-09-16 21:58:08.792800986 +0200 --- src/getchar.c 2019-09-17 20:28

Re: Channel based plugins and scheduling

Paul Jolly wrote: > > I'm seeing some some instances of not getting callbacks via > > SafeStateAgain... but I'll have to investigate that later. > > This was actually my mistake; state() returns 'c' when in a channel > callback handler. If that is the only state, then it is safe for me to > imm

Re: [vim/vim] Add g:json_no_conceal to disable conceal for json filetype (#4948)

there is an option named g:vim_json_warnings in this file, maybe use g:vim_json_conceal is better, also doc need to be updated. -- Best regards, wsdjeg From: matsuhav (Vim Github Repository) Date: 2019-09-17 22:20 To: vim/vim

Re: Re: [vim/vim] Properly break CJK lines (#3875)

Awesome work, I also need this patch for writing chinese doc. thanks! -- Best regards, wsdjeg From: 依云 (Vim Github Repository) Date: 2019-09-17 00:57 To: vim/vim CC: Subscribed

LGTM changes

Bram, we are currently using the LGTM service that runs some static analysis on new PRs. However I just noticed that there is going to be a change according to their blog: https://blog.semmle.com/github-apps-lgtm-migration/ So according to it, we should switch to the lgtm app integration, which

Re: Channel based plugins and scheduling

> I'm seeing some some instances of not getting callbacks via > SafeStateAgain... but I'll have to investigate that later. This was actually my mistake; state() returns 'c' when in a channel callback handler. If that is the only state, then it is safe for me to immediately schedule the work. But f

Re: Channel based plugins and scheduling

> > Any ideas what might be going on here? > > Sounds like you need a nested somewhere. Have a look at > `:h autocmd-nested` That was exactly it, thanks very much. I'm seeing some some instances of not getting callbacks via SafeStateAgain... but I'll have to investigate that later. -- -- You r

Re: Channel based plugins and scheduling

On Di, 17 Sep 2019, Paul Jolly wrote: > This is great, thanks. I'm using state() and SafeStateAgain > (adding/removing the autocommand as required) > > I've got a WIP PR with the changes up at > https://github.com/govim/govim/pull/533 > > However one issue I'm seeing is that autocommands are

Re: Channel based plugins and scheduling

This is great, thanks. I'm using state() and SafeStateAgain (adding/removing the autocommand as required) I've got a WIP PR with the changes up at https://github.com/govim/govim/pull/533 However one issue I'm seeing is that autocommands are not fired when doing work in response to SafeStateAgain.