| Message ID | pull.1691.git.git.1711049963.gitgitgadget@gmail.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | t9803-git-p4-shell-metachars.sh: update to use test_path_* functions | expand |
From: Sanchit Jindal <sanchit1053@gmail.com> replacing `test -e` with test_path_exists, `test ! -e` with test_path_is_missing `test -f` with test_path_is_file These helper functions will run the `test` command with the corresponding flags and will echo a message if the assert fails. This will provide better debugging logs for test, instead of the previous method which provided no message Signed-off-by: Sanchit Jindal <sanchit1053@gmail.com>
On Thu, Mar 21, 2024 at 3:39 PM Sanchit Jindal via GitGitGadget <gitgitgadget@gmail.com> wrote: > From: Sanchit Jindal <sanchit1053@gmail.com> > > replacing `test -e` with test_path_exists, > `test ! -e` with test_path_is_missing > `test -f` with test_path_is_file > These helper functions will run the `test` command with the > corresponding flags and will echo a message if the assert fails. > This will provide better debugging logs for test, instead of the > previous method which provided no message > > Signed-off-by: Sanchit Jindal <sanchit1053@gmail.com> When rerolling a series to address reviewer comments, you will want the fixes applied directly to the patches about which the reviewers commented. The way to do this is to use `git rebase -i` to adjust the patches as needed. In the case of this simple series, you just want to "squash" patches [1/2] and [2/2] into a single patch using the `git rebase -i` "squash" command, and adjust the commit message of the squashed patch appropriately. Finally, to resubmit it via GitGitGadget, force-push the revised series to GitGitGadget (using `git push --force <whatever> <whatever>`), and tell GitGitGadget to "/submit". Regarding the commit message, first explain the problem the patch is solving, and then explain how the patch solves it. Thus, start by explaining that `test` doesn't provide any diagnostic information when it fails, which isn't helpful to test authors. Then explain that the patch replaces `test` with the test_path_* functions which do provide useful diagnostic information. The From: and Signed-off-by: lines look good in this reroll.
On Fri, 22 Mar 2024 at 01:22, Eric Sunshine <sunshine@sunshineco.com> wrote: > > On Thu, Mar 21, 2024 at 3:39 PM Sanchit Jindal via GitGitGadget > <gitgitgadget@gmail.com> wrote: > > From: Sanchit Jindal <sanchit1053@gmail.com> > > > > replacing `test -e` with test_path_exists, > > `test ! -e` with test_path_is_missing > > `test -f` with test_path_is_file > > These helper functions will run the `test` command with the > > corresponding flags and will echo a message if the assert fails. > > This will provide better debugging logs for test, instead of the > > previous method which provided no message > > > > Signed-off-by: Sanchit Jindal <sanchit1053@gmail.com> > > When rerolling a series to address reviewer comments, you will want > the fixes applied directly to the patches about which the reviewers > commented. The way to do this is to use `git rebase -i` to adjust the > patches as needed. In the case of this simple series, you just want to > "squash" patches [1/2] and [2/2] into a single patch using the `git > rebase -i` "squash" command, and adjust the commit message of the > squashed patch appropriately. Finally, to resubmit it via > GitGitGadget, force-push the revised series to GitGitGadget (using > `git push --force <whatever> <whatever>`), and tell GitGitGadget to > "/submit". > > Regarding the commit message, first explain the problem the patch is > solving, and then explain how the patch solves it. Thus, start by > explaining that `test` doesn't provide any diagnostic information when > it fails, which isn't helpful to test authors. Then explain that the > patch replaces `test` with the test_path_* functions which do provide > useful diagnostic information. > > The From: and Signed-off-by: lines look good in this reroll. Thank You for the review I renamed the branch using github as it had a typo which made the earlier patch to be erased.I am extremely sorry about this. I have created another patch with a more descriptive commit message, and after squashing it with the earlier commit. I hope it is satisfactory I wanted to ask, Is it possible to send the patches created using `git format-patch` manually using gmail, Or the default headers applied by `send-email` required. Regards, Sanchit Jindal
Sanchit Jindal <sanchit1053@gmail.com> writes: > I have created another patch with a more descriptive commit message, > and after squashing it with the earlier commit. I hope it is satisfactory Hopefully. What GGG sent with v2 label still was a 2-patch series, whose [2/2] patch was an empty patch, though. > I wanted to ask, Is it possible to send the patches created using > `git format-patch` manually using gmail, Or the default headers > applied by `send-email` required. I do not know where you are using gmail from (I know that Android version has no way to send text-only e-mail for example), but when cutting and pasting the output from format-patch, you should make sure * Remove "From <commit object name> Mon Sep 17 00:00:00 2001" marker line that signals the beginning of each patch message. It is a mistake to leave this line in the body of your message. * Remove "Date: " header. Do not leave this line in the body of your message. * Remove "Subject: " header and move its contents to your MUA's Subject entry field. * Remove "From: " header, and arrange that your MUA puts the same "Sanchit Jindal <sanchit1053@gmail.com> on the From header. If your MUA is not cooperating, you can leave that line in the body of the message * Your cut&paste does not corrupt whitespaces, like squashing two consecutive spaces into one, removing the leading whitespaces, turning a tab into a run of spaces, folding a line at a whitespace in the middle of the message, etc. * Your MUA does not turn your text message into HTML gunk. As long as you do the above carefully, you should be OK. "git format-patch --help" has a section on MUA specific hints, which also might be helpful. Thanks.