git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: Branch switching with submodules where the submodule replaces a folder aborts unexpectedly

From
Sam Kuper <sam.kuper@uclmail.net>
Date
Jun 19, 2018, 16:03 UTC
Message-ID
<CAD-JurKx76FjVhk6QvmZ-BH=3RfmuA998KF+GRe8Kb+oB6pm8A@mail.gmail.com>
On 12 Oct 2017 at 11:48 Thomas Braun wrote:
Show 9 quoted lines
> On 9 Oct 2017 at 23:59, Stefan Beller wrote:
>> On 9 Oct 2017 at 14:29, Thomas Braun wrote:
>>> I'm currently in the progress of pulling some subprojects in a git repository of mine into their
>>> own repositories and adding these subprojects back as submodules.
>>>
>>> While doing this I enountered a potential bug as checkout complains on branch switching that a
>>> file already exists. [...]
>>>
>>> `error: The following untracked working tree files would be overwritten by checkout:`

I am currently attempting the same thing, and experiencing the same bug, using Git 2.17.1.

>> (And I presume you know about --recurse-submodules as a flag for git-checkout)

The behaviour seems to be the same regardless of whether I use the --recurse-submodules flag with git-checkout.

Show 7 quoted lines
>> This is consistent with our tests, unfortunately. [...]
>>
>>> If I'm misusing git here I'm glad for any advice.
>>
>> You are not.
>
> Glad to know that.
I was also glad to see this reassurance :)

It might have been nice if the reassurance came at an earlier stage, however: at the CLI, rather than only after searching the mailing list. A user who does not think to do the latter might well labour under the misapprehension that they have done something wrong, rather than encountered a bug. Perhaps, if the bug is not going to be fixed terribly soon, a sentence or two could be added to the error message explaining the situation and advising the user of a workaround?

Speaking of which, what is a good workaround in this case?
`git checkout --force <branch>`?
Show 5 quoted lines
>> Apart from this bug report, would you think that such filtering of
>> trees into submodules (and back again) might be an interesting feature
>> of Git or are these cases rare and special?
>
> For me not particularly. In my case it is a one time thing going from an embedded project folder to a submodule.

The option to convert an existing directory into a submodule is something that I think developers like to have available. It seems intuitive: "Oh, I see now that what this directory holds is effectively a separate project. Let me check out a new branch, and replace the directory with a submodule on that branch. Assuming it goes well, then I will afterwards merge this new branch into master."

Regardless, the bug has clearly been giving people headaches for several years (forgive me if you were already aware of these data points):

- https://ryansechrest.com/2014/03/git-error-switching-branch-replacing-directory-submodule/
- http://blog.dcycle.com/blog/105/gitsubmodulizing-and-gitflow/
- https://stackoverflow.com/q/9299063
- https://stackoverflow.com/a/48402543
- https://stackoverflow.com/q/24091246
- https://stackoverflow.com/q/29372450
- https://github.com/supercollider/supercollider/issues/2001
- https://github.com/supercollider/supercollider/issues/2221

Thank you to Thomas for reporting this issue to the mailing list, and to Stefan for the helpful reply. Thanks as always to the Git maintainers, and good luck with fixing this bug :)

Message 1 of 1 in “Re: Branch switching with submodules where the submodule replaces a folder aborts unexpectedly”
  1. Sam KuperJun 19, 2018

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.