No more tags for new 9.0 patches?

[Manfred Rosenboom]

Up to patch 9.0.1677 I see a related annotated tag. But meanwhile I see the following new patches in the commit log:

* 5b0889b8b patch 9.0.1712: missing null check in object_clear()
* d3515a1e8 patch 9.0.1711: dead code in charset.c
* 4a8eb6e7a patch 9.0.1710: scrolloff options work slightly different
* e5e4709a6 patch 9.0.1709: dynamic build with python 3.12 breaks
* 8ef1fbc0c patch 9.0.1708: getcompletion() failes for user-defined commands
* badeedd91 patch 9.0.1707: Cannot wrap around in popup_filter_menu()
* 02551c9de patch 9.0.1706: typos in the xxd manpage
* b25dbb3f9 patch 9.0.1705: cursor position wrong when clicking on an unprintable char
* 0c6181fec patch 9.0.1704: Cannot use positional arguments for printf()
* 74cc13cc4 patch 9.0.1703: Vim9 Calling a method in an extended class fails
* 9c30f4389 patch 9.0.1702: Undo test is flaky
* 57a02ccf4 patch 9.0.1701: vim9 crash when class member overridden
* 32f586eec patch 9.0.1700: Cannot compile with dynamic perl < 5.38
* 6c313bbb0 patch 9.0.1699: compile warning for xdiff/xutils on MS-Windows
* 7fe108990 patch 9.0.1698: Test_map_restore_sid fails in GUI
* 73b820926 patch 9.0.1697: incsearch test not sufficient
* 8878653a4 patch 9.0.1696: sodium_mlock may still fail in CI
* f1cc4d55b patch 9.0.1695: Crash with overlong textprop above
* bacc83009 patch 9.0.1694: wrong mapping applied when replaying a char search
* 2d63e4b3c patch 9.0.1693: Ctrl-Q not handled like Ctrl-V in replace mode
* db4fd2906 patch 9.0.1692: Android not handling AI_V4MAPPED ai_flag
* 7b7b4cb6f patch 9.0.1691: wrong viewport restored for incsearch and smoothscroll
* f6cdab370 patch 9.0.1690: popup_create() not aborting on errors
* 288bf26c5 patch 9.0.1689: python 3.12 produces warnings and fails test

* ca6ac9907 patch 9.0.1688: cannot store custom data in quickfix list
* 7e0bae024 patch 9.0.1687: mapset() not properly handling script ID
* 5fee11114 patch 9.0.1686: undotree() only works for the current buffer
* 422b9dcbf patch 9.0.1685: silence Python 3.11 depreciations for gcc
* b00df7aa3 patch 9.0.1684: Update libvterm to rev 839
* 6019fed0c patch 9.0.1682: sodium encryption is not portable
* 1d7caa58e patch 9.0.1681: Build Failure with Perl 5.38
* 335c58494 patch 9.0.1680: sodium test fails in Github CI
* 84bc00e9b patch 9.0.1679: Cleanup Tests from leftover files
* ad34abee2 patch 9.0.1678: blade files are not recognized
* 4c0089d69 (tag: v9.0.1677) patch 9.0.1677: typo in syntax test input file

Don't get each patch a related annotated tag anymore?

[Christian Brabandt]

I think we only used lightweight tags? So I only added those:

```
chrisbra@debian ~/code/vim-upstream (git)-[master]- % git describe v9.0.1676
fatal: No annotated tags can describe
'16abd997c99b1a40fb0ba54889ce0cf44d7585b4'.
However, there were unannotated tags: try --tags.
chrisbra@debian ~/code/vim-upstream (git)-[master]- % git describe v9.0.1712
fatal: No annotated tags can describe
'5b0889b8bf25b0793b1949ec965c94b623900aba'.
However, there were unannotated tags: try --tags.
chrisbra@debian ~/code/vim-upstream (git)-[master]- % git cat-file -t v9.0.1676
commit
chrisbra@debian ~/code/vim-upstream (git)-[master]- % git cat-file -t v9.0.1712
commit
```


Best,
Christian
--
"Aww, if you make me cry anymore, you'll fog up my helmet."
-- "Visionaries" cartoon

[Enan Ajmain]

On Mon, 14 Aug 2023 12:18:50 +0200
Christian Brabandt <cbl...@256bit.org> wrote:
> On Mo, 14 Aug 2023, Manfred Rosenboom wrote:
> > Don't get each patch a related annotated tag anymore?
>
> I think we only used lightweight tags? So I only added those:

Not that it's super important, but I suggest strictly using annotated
tags. Cannot recall the exact reasons, but last year my team and I
faced multiple unrelated problems: all of them because of using
lightweight tags instead of annotated ones.

Here's what "git tag --help" has to say:
> Annotated tags are meant for release while lightweight tags are meant
> for private or temporary object labels. For this reason, some git
> commands for naming objects (like git describe) will ignore
> lightweight tags by default.

I don't wanna be annoying. Just wanted to leaves a word that weird
problems may pop up if using lightweight tags instead of annotated ones.

--
Enan