* git-rebase --abort eats files @ 2010-06-26 12:53 Madhu 2010-06-26 18:09 ` Johannes Sixt 2013-08-31 15:26 ` git-rebase --continue eats commits Madhu 0 siblings, 2 replies; 6+ messages in thread From: Madhu @ 2010-06-26 12:53 UTC (permalink / raw) To: git Don't know if this has been resolved-by-debate here before, But adding a file via git-add in the middle of an interactive rebase and aborting the rebase deletes the hitherto untracked file. It should not. rm -rfv /tmp/t1 ; mkdir -pv /tmp/t1 cd /tmp/t1 && git-init-db cd /tmp/t1 && touch file1 file2 file3 cd /tmp/t1 && git-add file1 cd /tmp/t1 && git-commit -m "Explet1" file1 cd /tmp/t1 && git-add file2 cd /tmp/t1 && git-commit -m "Explet2" file2 cd /tmp/t1 && git-rebase --abort cd /tmp/t1 && EDITOR="sed -i -e 's/^pick/edit/'" git rebase -i 'HEAD^' cd /tmp/t1 && git-add file3 cd /tmp/t1 && git-status cd /tmp/t1 && git-rebase --abort test -e /tmp/t1/file3 || echo bwaaah Maybe something like this to fix it? diff --git a/git-rebase--interactive.sh b/git-rebase--interactive.sh index 436b7f5..2702536 100755 --- a/git-rebase--interactive.sh +++ b/git-rebase--interactive.sh @@ -749,6 +749,7 @@ first and then run 'git rebase --continue' again." git symbolic-ref HEAD $HEADNAME ;; esac && + git-reset && output git reset --hard $HEAD && rm -rf "$DOTEST" exit -- Madhu ^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: git-rebase --abort eats files 2010-06-26 12:53 git-rebase --abort eats files Madhu @ 2010-06-26 18:09 ` Johannes Sixt 2010-06-28 9:05 ` Ramkumar Ramachandra 2013-08-31 15:26 ` git-rebase --continue eats commits Madhu 1 sibling, 1 reply; 6+ messages in thread From: Johannes Sixt @ 2010-06-26 18:09 UTC (permalink / raw) To: Madhu; +Cc: git On Samstag, 26. Juni 2010, Madhu wrote: > Don't know if this has been resolved-by-debate here before, But adding > a file via git-add in the middle of an interactive rebase and aborting > the rebase deletes the hitherto untracked file. It should not. > > Maybe something like this to fix it? > > +++ b/git-rebase--interactive.sh > @@ -749,6 +749,7 @@ first and then run 'git rebase --continue' again." > git symbolic-ref HEAD $HEADNAME > ;; > esac && > + git-reset && > output git reset --hard $HEAD && > rm -rf "$DOTEST" > exit No, it can't be that simple. If rebase stopped due to a conflict on a commit that added new files, then your version of rebase --abort will leave these new files behind as untracked. -- Hannes ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: git-rebase --abort eats files 2010-06-26 18:09 ` Johannes Sixt @ 2010-06-28 9:05 ` Ramkumar Ramachandra 2010-06-29 1:23 ` Madhu 0 siblings, 1 reply; 6+ messages in thread From: Ramkumar Ramachandra @ 2010-06-28 9:05 UTC (permalink / raw) To: Johannes Sixt; +Cc: Madhu, git Hi Madhu, Johannes Sixt wrote: > On Samstag, 26. Juni 2010, Madhu wrote: > > +++ b/git-rebase--interactive.sh > > @@ -749,6 +749,7 @@ first and then run 'git rebase --continue' again." > > git symbolic-ref HEAD $HEADNAME > > ;; > > esac && > > + git-reset && > > output git reset --hard $HEAD && > > rm -rf "$DOTEST" > > exit > > No, it can't be that simple. If rebase stopped due to a conflict on a commit > that added new files, then your version of rebase --abort will leave these > new files behind as untracked. Right. The interactive rebase has to be able to differentiate between files that you added to resolve a conflict and files that you added to retain at the end of the rebase -- and the interactive rebase has no information about this. Hence, this problem can't be fixed without explicitly finding out the intent of the user. In my opinion, you should simply stash your changes before aborting the rebase instead of adding files and figuring out some complex way of expressing intent. -- Ram ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: git-rebase --abort eats files 2010-06-28 9:05 ` Ramkumar Ramachandra @ 2010-06-29 1:23 ` Madhu 2010-06-29 6:47 ` Pete Harlan 0 siblings, 1 reply; 6+ messages in thread From: Madhu @ 2010-06-29 1:23 UTC (permalink / raw) To: Ramkumar Ramachandra; +Cc: j6t, git |Date: Mon, 28 Jun 2010 11:05:17 +0200 |From: Ramkumar Ramachandra <artagnon@gmail.com> |Cc: Madhu <enometh@meer.net>, git@vger.kernel.org |Content-Type: text/plain; charset=us-ascii |Content-Disposition: inline | |> No, it can't be that simple. If rebase stopped due to a conflict |> on a commit that added new files, then your version of rebase |> --abort will leave these new files behind as untracked. | |Right. The interactive rebase has to be able to differentiate |between files that you added to resolve a conflict and files that |you added to retain at the end of the rebase -- and the interactive |rebase has no information about this. Hence, this problem can't be |fixed without explicitly finding out the intent of the user. Wrong. Rebase has to be able to differentaiate between two cases 1. when there is a conflict, and the user is prompted to fix it, and then continue with a git-add, git-commit, and git-rebase --continue and 2. when the user is given a commit, which he is asked to git-commit --amend, and then git-rebase --continue Rebase is already aware of when each situation occurs. |In my opinion, you should simply stash your changes before aborting |the rebase instead of adding files and figuring out some complex |way of expressing intent. This does not make sense. -- Madhu ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: git-rebase --abort eats files 2010-06-29 1:23 ` Madhu @ 2010-06-29 6:47 ` Pete Harlan 0 siblings, 0 replies; 6+ messages in thread From: Pete Harlan @ 2010-06-29 6:47 UTC (permalink / raw) To: Madhu; +Cc: Ramkumar Ramachandra, j6t, git On 06/28/2010 06:23 PM, Madhu wrote: > |Date: Mon, 28 Jun 2010 11:05:17 +0200 > |From: Ramkumar Ramachandra <artagnon@gmail.com> > |Cc: Madhu <enometh@meer.net>, git@vger.kernel.org > | > |> No, it can't be that simple. If rebase stopped due to a conflict > |> on a commit that added new files, then your version of rebase > |> --abort will leave these new files behind as untracked. > | > |Right. The interactive rebase has to be able to differentiate > |between files that you added to resolve a conflict and files that > |you added to retain at the end of the rebase -- and the interactive > |rebase has no information about this. Hence, this problem can't be > |fixed without explicitly finding out the intent of the user. > > Wrong. Rebase has to be able to differentaiate between two cases > > 1. when there is a conflict, and the user is prompted to fix it, and > then continue with a git-add, git-commit, and git-rebase --continue > > and > > 2. when the user is given a commit, which he is asked to git-commit > --amend, and then git-rebase --continue > > Rebase is already aware of when each situation occurs. If I read your original question correctly, the problem is: if you have an untracked file in your directory before you start a rebase, and then you add that file during the rebase, and then abort the rebase, Git will delete your file from the working directory instead of just returning it to its untracked state. So aborting a rebase doesn't simply roll back time: it can be destructive in a way the user may not expect. The followups to the original question seem to me to be clouding the issue with the question of what to do with any new material added during a rebase, not just files that were originally present but untracked. Maybe it's not easy to solve the original problem, or maybe it's not worth doing; maybe it's worth documenting. (And documenting how to recover the files, since they're in the object database for a while.) I'd guess that to solve it rebase would have to do a "git status" when it started, to see which files it should leave behind in the working directory. I'm not a Git hacker so that's just speculation. But I think it's worth discussing this behavior on its own, separately from the question about other material added during a rebase. [As to what to do with other material added during a rebase, I'd like it nuked. When I abort a rebase it's because I've gummed things up and want to start over.] --Pete > > |In my opinion, you should simply stash your changes before aborting > |the rebase instead of adding files and figuring out some complex > |way of expressing intent. > > This does not make sense. > > -- > Madhu ^ permalink raw reply [flat|nested] 6+ messages in thread
* git-rebase --continue eats commits 2010-06-26 12:53 git-rebase --abort eats files Madhu 2010-06-26 18:09 ` Johannes Sixt @ 2013-08-31 15:26 ` Madhu 1 sibling, 0 replies; 6+ messages in thread From: Madhu @ 2013-08-31 15:26 UTC (permalink / raw) To: git Don't know if this has been resolved-by-debate here before, But if `git-rebase' finds a hitherto untracked file in the worktree, which it wants to create, it then aborts asking you to remove the file. So if you remove it and ask git to continue with `git-rebase --continue', it then deletes the commit that was being applied from the branch. ---Madhu (setenv "TDIR" "/dev/shm/foo/") mkdir -pv $TDIR && cd $TDIR && git-init (cd $TDIR && echo a > a && git add a && git commit -m "a") (cd $TDIR && echo b > b && git add b && git commit -m "b") (cd $TDIR && echo c > c && git add c && git commit -m "c") (cd $TDIR && EDITOR="sed -i -e 's/^pick/edit/'" git rebase -i 'HEAD^^') (cd $TDIR && echo fubar > c) (cd $TDIR && git-rebase --continue) Rebasing (2/2) error: The following untracked working tree files would be overwritten by merge: c Please move or remove them before you can merge. Aborting Could not apply ddd6f51... (cd $TDIR && rm -fv c) (cd $TDIR && git-rebase --continue) Rebasing (2/2) Successfully rebased and updated refs/heads/master. (cd $TDIR && git log) ;; commit `c' is gone ^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2013-08-31 15:56 UTC | newest] Thread overview: 6+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2010-06-26 12:53 git-rebase --abort eats files Madhu 2010-06-26 18:09 ` Johannes Sixt 2010-06-28 9:05 ` Ramkumar Ramachandra 2010-06-29 1:23 ` Madhu 2010-06-29 6:47 ` Pete Harlan 2013-08-31 15:26 ` git-rebase --continue eats commits Madhu
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).