Re: E315 ml_get invalid lnum

2016-10-17 Fir de Conversatie h_east
Hi John, 2016-10-18(Tue) 12:03:44 UTC+9 JohnBeckett: > Vim 8 deploys $VIMRUNTIME/defaults.vim which includes an autocmd > to "jump to the last known cursor position". > That can generate E315. > > Execute the following to create two files and exit. > At exit, the cursor is on the last line of

E315 ml_get invalid lnum

2016-10-17 Fir de Conversatie JohnBeckett
Vim 8 deploys $VIMRUNTIME/defaults.vim which includes an autocmd to "jump to the last known cursor position". That can generate E315. Execute the following to create two files and exit. At exit, the cursor is on the last line of each file. gvim -N -u NONE :tabe test1.tmp :put =range(1,90) :w

Re: [patch] fix #1155 (A job cannot be stopped right after being started)

2016-10-17 Fir de Conversatie Ozaki Kiichi
> Seems like a reasonable solution. So the test fails for the current > Vim? Or does it fail sometimes? Sometimes. Especially, according to travis-ci, it is highly probable to fail on Mac; about half or more. -- -- You received this message from the "vim_dev" maillist. Do not top-post! Type

Re: Patch 8.0.0041

2016-10-17 Fir de Conversatie Tommy A
On Monday, October 17, 2016 at 5:06:47 PM UTC-4, Tommy A wrote: > Thanks again, Bram! There is one (hopefully last) issue. This works > perfectly when using `` and ``, but the problem remains the same > when `` and `` are used. It looks like adding `(void)stop_arrow()` on line 1431 of edit.c

Re: Patch 8.0.0041

2016-10-17 Fir de Conversatie Tommy A
Thanks again, Bram! There is one (hopefully last) issue. This works perfectly when using `` and ``, but the problem remains the same when `` and `` are used. On Monday, October 17, 2016 at 2:47:40 PM UTC-4, Bram Moolenaar wrote: > Tommy Allen wrote: > > > This was a faster turn around than I

Re: [PATCH] syntax: Add command to control how foldlevel is computed for a line

2016-10-17 Fir de Conversatie Bram Moolenaar
Brad King wrote: > On 10/14/2016 04:04 PM, Bram Moolenaar wrote: > > Either this needs to be donein the C syntax, or we need an folding > > option to enable this behavior. > > Thanks for taking a look. It can't currently be done purely in syntax > definitions so I've revised the change to add

Re: [patch] fix #1155 (A job cannot be stopped right after being started)

2016-10-17 Fir de Conversatie Bram Moolenaar
Ozaki Kiichi wrote: > The following will fix #1155. > > https://gist.github.com/ichizok/26d571401be0685267ebf0790e964903 > > [cause of problem] > > This problem may occur on some *nix environment. > > Calling job_stop immediately after job_start, child process may receive signal > before

Re: Patch 8.0.0041

2016-10-17 Fir de Conversatie Bram Moolenaar
Tommy Allen wrote: > This was a faster turn around than I expected, thanks! Unfortunately > there's another issue that comes from this: undo items are > overwritten. If you place text on the line, then press ``, the > undo entry should be created for the text on that line (e.g. "aaa"). > If

Patch 8.0.0042

2016-10-17 Fir de Conversatie Bram Moolenaar
Patch 8.0.0042 (after 8.0.0041) Problem:When using Insert mode completion with 'completeopt' containing "noinsert" change is not saved for undo. (Tommy Allen) Solution: Call stop_arrow() before inserting for pressing Enter. Files: src/edit.c, src/testdir/test_popup.vim

[PATCH] syntax: Add command to control how foldlevel is computed for a line

2016-10-17 Fir de Conversatie Brad King
On 10/14/2016 04:04 PM, Bram Moolenaar wrote: > Either this needs to be donein the C syntax, or we need an folding > option to enable this behavior. Thanks for taking a look. It can't currently be done purely in syntax definitions so I've revised the change to add a new syntax command:

Re: [vim/vim] Fix #1168 completion problem from #875 (#1169)

2016-10-17 Fir de Conversatie h_east
2016-10-18(Tue) 0:13:48 UTC+9 Shougo: > Issue#1168 repro steps doesn't seem to asynchronous for me. Where is the > asynchronous? > > > > Because it uses complete_check() and complete_add(). (Below is Off topic) No. In this case, Those functions are not asynchronous because they are called

[patch] fix #1155 (A job cannot be stopped right after being started)

2016-10-17 Fir de Conversatie Ozaki Kiichi
Hi. The following will fix #1155. https://gist.github.com/ichizok/26d571401be0685267ebf0790e964903 [cause of problem] This problem may occur on some *nix environment. Calling job_stop immediately after job_start, child process may receive signal before reset_signals(). Therefore

Re: Patch 8.0.0041

2016-10-17 Fir de Conversatie Tommy A
Bram, This was a faster turn around than I expected, thanks! Unfortunately there's another issue that comes from this: undo items are overwritten. If you place text on the line, then press ``, the undo entry should be created for the text on that line (e.g. "aaa"). If you jump to the end of