Closed sharad closed 4 years ago
Let's not do that. All this does is save a tiny bit of CPU and disk-space at the cost of known as well as unknown complications.
Just as an example: one complication is that it would no longer be trivial to go to the repository of a certain package. Also until now we could assume that a local repository always corresponds to exactly one repository and as a consequence we did not have to and did not think about complications that arise due to a repository belonging to multiple packages. There likely would be issues involving time travel.
Agreed, thanks
generate repository path based on package name, which is posing multiple clone or sync from same repository if there are multiple packages present in same repository, which in turn consuming lot of CPU and disk-space
Modified method package-recipe--working-tree to generate path based on repository url, to overcome this issue.