Cloning repo appends extra path segments. about devhome HOT 2 CLOSED

My idea for conflict resolution is

1. Separate by provider
2. Separate by account name
3. Separate by repo name.

Or, another idea is to not append to clone path and change the repo directory name.

But that could make the name of the directory too long.

Examples.

I have a Github and ADO provider and I want to download a repo called MyRepo from both. Clone Path is C:\users\dhoehna\Desktop\MyRepos. ADO provider happens to start cloning first. Github provider runs into a conflict.

Github will instead download to [Clone Path]\[ProviderName]\[RepoName] in this example C:\users\dhoehna\Desktop\MyRepos\MyRepo

Then, after path resolution if a repository exists then -N where N is the number of folders is appended to the folder name.

Lets say the repo existed. Then, the path would be C:\users\dahoehna\Desktop\MyRepos\MyRepos-1

Discussion is encouraged.

Basically, how do we avoid naming conflicts so a repo is cloned no matter what?

from devhome.

@dhoehna There is no need to prevent collision a priori. Collisions can happen for many reasons, many of them not in dev home's control. If we find a collision (i.e. the cloning fails) for whatever reason, we should just show an error. We should not insert artificial separation of repos based on provider or account, please make it so repos get cloned to the expected location of FolderPathIEnteredInTheUI\NameOfTheRepo.

from devhome.