{"thread":{"id":"24908","subject":"git-add creates broken submodules","startedAt":"2010-08-30T09:09:45Z","lastAt":"2010-08-30T09:09:45Z","messageCount":1,"participants":["strk"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"149294","messageId":"20100830090945.GB13729@keybit.net","threadId":"24908","inReplyTo":null,"subject":"git-add creates broken submodules","fromName":"strk","fromEmail":"strk@keybit.net","sentAt":"2010-08-30T09:09:45Z","receivedAt":"2010-08-30T09:09:45Z","isPatch":false,"sender":{"key":"strk@keybit.net","avatar":"https://gravatar.com/avatar/9a6c11d66829bf429efe5ef4c066d50272394481cc3f8ae8116a006c81dc6cf9?d=mp&s=160"},"body":"Hi there,\nwith help of doener (#git on freenode) I found\nthe root cause of a misconfigured git repository\nbeing in the way 'git add <path>' behaves when \n<path> contains a .git repository.\n\nExample:\n$ (mkdir foo; cd foo; git init; git commit --allow-empty -m ini); \\\n  git add foo\n$ git submodule status\nNo submodule mapping found in .gitmodules for path 'foo'\n\nSuch commit ended up in our centralized repository so that all\nclones now have this 'submodule' w/out entries in .gitmodules\nand .git/config.\n\nShould 'git-add' simply consider any .git as a normal entry\ninstead ? (that's what I'd expect) \nOr, if not, should it warn about the presence of\na .git directory and refuse to continue suggesting\n'git submodule add' instead.\n\nNOTE: this happens with git version 1.6.0.2\n\n\n--strk; \n\n  ()   Free GIS & Flash consultant/developer\n  /\\   http://strk.keybit.net/services.html\n"}]}