{"thread":{"id":"2823","subject":"git commit improvement request","startedAt":"2005-12-13T19:49:51Z","lastAt":"2005-12-13T19:49:51Z","messageCount":1,"participants":["Jon Nelson"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"13575","messageId":"Pine.LNX.4.63.0512131347000.6812@gheavc.wnzcbav.cig","threadId":"2823","inReplyTo":null,"subject":"git commit improvement request","fromName":"Jon Nelson","fromEmail":"jnelson-git@jamponi.net","sentAt":"2005-12-13T19:49:51Z","receivedAt":"2005-12-13T19:49:51Z","isPatch":false,"sender":{"key":"jnelson-git@jamponi.net","avatar":null},"body":"\nThere is a small improvement I would like to see it 'git commit'.\nI am using 0.99.9i at this moment.\n\nIf I perform a 'git commit' on a series of files, and then decide not to \nperform the commit, I can back out of the operation with a blank commit \nmessage. Additionally, it would be nice for git commit to recognize a \nnon-zero exit code theoretically indicating 'aborted commit'. In any \ncase, the problem is that, even though git /states/ that it has aborted \nthe commit due to the empty commit message, git diff /appears/ to show \nthat the commit took place. a 'git reset' takes care of things for me \nbut it was rather confusing at first and it seems counter-intuitive. I'm \nhoping that the behavior here can be changed before 1.0 is released, \nwhich I am in eager anticipation for.  Thanks!\n\n\n--\nJon Nelson <jnelson-git@jamponi.net>\n"}]}