Follow-up refactor patches, mixed with unrelated patches in the middle,
is normal in vim-dev when the original patch is not fully tested
across all builds via CI or reviewed by others.
Unless core maintainers push "vim-patch:" directly to master/main branch
without running the full test-suite,
there is little reason to merge incomplete ports that will fail
on Nvim's CI or code review.
Relevant changes in patches like v8.2.0514 are either ported
or (will) become N/A.
Ignoring incompatible implementation,
TerminalOpen and TermOpen events are not 1-1.
"TerminalWinOpen" was accepted as N/A in
commit c7ee6af777 .
I planned to not do this to have more test cases
after manipulating the hunks header to filter out more hunks
but Justin is eager to just mark these N/A
to bump the Vim major.minor version.
Time to move on from v8.1.x.
Target v8.1.2195 .
Nvim did not port Vim's ":terminal" opts.
Incompatible implementations.
Ex-command was ported from C to Lua.
Vim needs them partly because of splitting the current window.
I keep forgetting `++close` option so I either run ":qall!"
or kill the parent process (ie. terminal emulator).
Unsatisfied users should create their Ex-command that runs jobstart().
Vim's TerminalWinOpen seems to be required because of Vim
buffer-job-popupwin implementation.
Based on the patch, I'm puzzled why fzf needs this on Vim.
Nvim's TermOpen, TermEnter, and detection mechanisms to know
if buffer is on a (active,visible) window should suffice to not port it.
If there was a feature request or issue without a merged fix,
then I can't find it.
https://github.com/junegunn/fzf/pull/2000
Vim core did not leverage it to override/customize bell/beep.
It wasn't used for custom sounds for system/user (autocmd) events.
It should be in-scope for GUI, unlike terminal, even as a plugin
by leveraging some internal option similar to `set guioptions+=!`.
No progress as of Vim 9.2 so I quit.
vim-patch.sh fails to detect n/a patches
because of ifdef FEAT_ guards and reserved Vim9script error codes.
Ignore all conditional directives for Vim's "FEAT_" guards.
https://cppreference.com/c/preprocessor/conditional
Following remain relevant:
- src/testdir/Makefile (default setup in Unix-like OS)
- src/testdir/Make_all.mak (define old/new style Vim tests)
Rest are n/a because Nvim runs Vim's tests on GHA Windows runner.
* build(vim-patch): n/a func in version.c
init_longVersion() sets datetime for ":version".
Nvim removed it for more deterministic builds.
* build(vim-patch): n/a header from amiga
* build(vim-patch): n/a vim9 err msg for declaring vars
* build(vim-patch): nvim support.txt replaces *os-support*
* build(vim-patch): vim xdg is n/a
* build(vim-patch): n/a 1-line err msg
Vim error message can be 1-line (globals.h), not multi-line (errors.h).
Target v8.2.1032
Target v8.2.0543
This can catch stray '#define' but they're most likely 'TTFLAG' for
Vim9 types. N/A.
TTFLAG is used in 'static_types[]'.
See v9.0.0623.
Problem:
Cannot assign Lua functions/closures to "func" ('completefunc',
'tagfun', …) or "expr" ('foldexpr', 'indentexpr', …) options.
Solution:
- Store "func"/"expr" options as `Callback` instead of string.
- Delete oceans of copy-pasted code.
- BREAKING: LuaRef returned via RPC/Vimscript is now represented as
`"<Lua N: file:line>"` (like what `:map` shows) instead of `nil`.
- Note: `man.vim` still uses `v:lua` string, bc it's a vimscript ftplugin.
Helped-by: Lewis Russell <lewis6991@gmail.com>
1. Vim9 error messages, ex commands, are N/A.
2. ":scriptversion" is N/A:
3. Most `#.*define.*ex_ni$` are N/A
because `FEAT_` compile guards are N/A.
----
Target patch: v8.2.2238
Actual patch list:
```
vim-patch:8.2.4240: error for using flatten() in Vim9 script is unclear
vim-patch:9.0.0170: various minor code formatting issues
vim-patch:9.1.0356: MS-Windows: --remote may change working directory
vim-patch:9.2.0251: Link error when building without channel feature
vim-patch:9.2.0714: Coverity warns for NULL deref
vim-patch:9.2.0818: tests: client-server test fails without X11 server
vim-patch:9.2.0868: GTK: Window Manager hint prevents giving focus to dialog
```
----
Ignore diff blanklines via regex ("-I"), not "--ignore-blank-lines".
Almost useless.
I tried with/without "--word-diff=porcelain", "--word-diff-regex",
and diff algorithms like `--histogram`.
Nvim diverged from Vim's ":version".
- no date
- different compiled features list
- different config file and directory
- no compiler flags
* build(vim-patch): n/a doc for balloon feature (8.2.3917)
Vim patches for Vim help files must use Vim tags as diff hunk header
to reliably detect if a hunk is N/A or not.
Git does not know which files are Vim help files
so that it can use custom "diff.<filetype>.xfuncname" regexp
to create the header.
Update Neovim's gitattributes to detect Vim help files
and then use it on vim-patch.sh to scan Vim patches.
Add non-function tags from ":h vim_diff"
to detect N/A Vim runtime patches.
Problem: scripts/genappimage.sh runs `make` with no extra flags and
emits a fixed nvim-linux-<arch>.appimage filename, so packaging a
variant build (e.g. with the bundled PUC-Rio Lua interpreter)
requires editing the Makefile or post-renaming the AppImage, which
breaks the AppImage zsync URL.
Solution: forward DEPS_CMAKE_FLAGS and CMAKE_EXTRA_FLAGS to `make`,
accept OUTPUT so the AppImage (and its zsync Filename:) follows the
caller's chosen name, and mark CPACK_PACKAGE_FILE_NAME as CACHE
STRING on Linux so the resulting tarball/deb follow as well. Default
behavior is unchanged when none of these env vars are set.
"vim-patch.sh -L" includes gpg signature for "log.showSignature=true"
in user's .gitconfig.
vim-patch.sh expects no signature for 1-line output per patch.
Problem:
Options parsing is still painful for dict-style options.
Solution:
schema-maxxing => better `opt:get()` (will be the basis for `vim.o()`),
unified (and more-detailed) err msgs.
- Drop bespoke structure-builder in `_core/options.lua`.
- Define `schema` for all non-primitive options (except 'guicursor' and
statusline-style options); generate reified keysets `OptKeyDict`).
- Generate 'fillchars' => `fcs_tab`, 'listchars' => `lcs_tab`.
- `nvim_set_option_value`:
- Return the improved structures. Also from `vim.opt.x:get()`.
- Eliminate api <=> lua roundtrip, centralize option structure
handling.
- Improve/unify errors.
- Bump ERR_BUFLEN 80 → 256 so the "one of" list isn't truncated.
- Eliminate old 'diffopt' order-dependence (`iwhiteall` before `iwhite`)
Error samples:
Typed-key path (opt_strings_check → diffopt/mousescroll/breakindentopt):
E474: Unknown item 'foo'
E474: 'context' requires a number
E474: 'ver' number is out of range
E474: 'algorithm' must be one of: myers, minimal, patience, histogram
E474: 'filler' does not take a value
Related:
- #31084
- #34661
- #31820
- #14739
- #20107
- fix#18875
- :get() returns `{ sbr = true, shift = '3' }` (reified-keyset) instead of `{'sbr', 'shift:3'}`
- Setting via table now works too. `object_as_optval_for` `is_map` now recognizes struct options.
- fix#30296
- instead of `E474: Invalid argument`, errors now look like:
```
E474: Invalid value 'x', expected one of: single, double: ambiwidth=x
E474: Unknown item 'foo': diffopt=foo
E474: 'context' requires a number: diffopt=context:x
```
simplify `win_float_parse_option` from #26799.
Vim's code and test for channel,job API are already N/A.
Blacklisting runtime/doc/channel.txt ensures most patches for it
will be detected as N/A.
Vim's API that is too incompatible with Neovim must be handled
outside vim-patch.sh .
Jan Edmund Lazo
Justin M. Keyes
phanium
Barrett Ruth