mbox series

[00/16] Builtin FSMonitor Part 2.5

Message ID pull.1174.git.1647033303.gitgitgadget@gmail.com (mailing list archive)
Headers show
Series Builtin FSMonitor Part 2.5 | expand

Message

Elijah Newren via GitGitGadget March 11, 2022, 9:14 p.m. UTC
I'm calling this series part 2.5. It should be inserted between parts 2 and
3. I'll send a new version of part 3 after I send part 2.5.

This series is a bunch of "fixup!" commits for part 2. It addresses feedback
on part 2 that wasn't received until after part 2 had graduated to "next".

I was originally planning to include them at the beginning of part 3, but
since there are so many, I thought it would be easier to add a fixup series
in the middle. (And since GGG has a limit of 30 commits, I couldn't add them
to either existing series without going over the limit.)

Most of the commits deal with translation markings for die/error messages
and fixing && chains in the tests. And there were a few readability
improvements.

Jeff Hostetler (16):
  t/test-lib: avoid using git on LHS of pipe
  update-index: convert advise() messages back to warning()
  compat/fsmonitor/fsm-listen-darwin: split out GCC-specific
    declarations
  t/helper/fsmonitor-client: cleanup call to parse_options()
  fsmonitor--daemon: refactor cookie handling for readability
  t/perf/p7519: use grep rather than egrep in test
  t/perf/p7519: cleanup coding style
  t7527: add parameters to start_daemon to handle args and subshell
  t7527: fix && chaining in matrix_try()
  t7527: delete unused verify_status() function
  fsmonitor-ipc: add _() to calls to die()
  compat/fsmonitor/fsm-listen-darwin: add _() to calls to error()
  compat/fsmonitor/fsm-listen-win32: add _() to calls to error()
  fsmonitor--daemon: add _() to calls to die()
  fsmonitor--daemon: add _() to calls to error()
  fsmonitor-settings: simplify initialization of settings data

 builtin/fsmonitor--daemon.c          |  63 ++++----
 builtin/update-index.c               |  12 +-
 compat/fsmonitor/fsm-darwin-gcc.h    |  92 +++++++++++
 compat/fsmonitor/fsm-listen-darwin.c |  95 +----------
 compat/fsmonitor/fsm-listen-win32.c  |   6 +-
 fsmonitor-ipc.c                      |   4 +-
 fsmonitor-settings.c                 |   3 +-
 t/helper/test-fsmonitor-client.c     |  17 +-
 t/perf/p7519-fsmonitor.sh            |  23 +--
 t/t7527-builtin-fsmonitor.sh         | 233 +++++++++++++--------------
 t/test-lib.sh                        |   3 +-
 11 files changed, 276 insertions(+), 275 deletions(-)
 create mode 100644 compat/fsmonitor/fsm-darwin-gcc.h


base-commit: 1a9241e1fee9d418e436849853f031329e792192
Published-As: https://github.com/gitgitgadget/git/releases/tag/pr-1174%2Fjeffhostetler%2Fbuiltin-fsmonitor-part2.5-v1
Fetch-It-Via: git fetch https://github.com/gitgitgadget/git pr-1174/jeffhostetler/builtin-fsmonitor-part2.5-v1
Pull-Request: https://github.com/gitgitgadget/git/pull/1174

Comments

Junio C Hamano March 13, 2022, 8:02 a.m. UTC | #1
"Jeff Hostetler via GitGitGadget" <gitgitgadget@gmail.com> writes:

> I'm calling this series part 2.5. It should be inserted between parts 2 and
> 3. I'll send a new version of part 3 after I send part 2.5.
>
> This series is a bunch of "fixup!" commits for part 2. It addresses feedback
> on part 2 that wasn't received until after part 2 had graduated to "next".

Better late than never, but I am sensing that part 2 was
under-reviewed and we pushed it too hastily?

> I was originally planning to include them at the beginning of part 3, but
> since there are so many, I thought it would be easier to add a fixup series
> in the middle.

I think that is prudent.  Let's give part 3 (which ought to be more
interesting than part 2.5) a more solid ground to build upon, and to
prepare for that, let's first avoid repeating the mistake of merging
a series in an under-reviewed state to 'next'.

Will queue (but not tonight).

Thanks.