Incorrect commit oid's when aggregated from commit.history()

[djrenren]

If we take the example code that mirrors git log. And have it run on the aggregated commits rather than on each commit event like so:

var git = require('nodegit'),
    async = require('async');

// Open the repository in the current directory.
git.repo('.git', function(error, repository) {
  if (error) {
    throw error;
  }

  // Use the master branch.
  repository.branch('master', function(error, branch) {
    if (error) {
      throw error;
    }

    branch.history().on('end', function(err, commits){
        commits.forEach(function(commit){
                 // Print out `git log` emulation.
            async.series([
                function(callback) {
                    commit.sha(callback);
                },
                function(callback) {
                    commit.time(callback);
                },
                function(callback) {
                    commit.author(function(error, author) {
                        author.name(callback);
                    });
                },
                function(callback) {
                    commit.author(function(error, author) {
                        author.email(callback);
                    });
                },
                function(callback) {
                    commit.message(callback);
                }
            ], function printCommit(error, results) {
                console.log('SHA ' + results[0]);
                console.log(new Date(results[1] * 1000));
                console.log(results[2] + ' <' + results[3] + '>');
                console.log(results[4]);
            });
        });
    });
  });
});

Then all the commit sha's are wrong. And most if not all of them point to nonexistant commits. Compare the outputs of the two versions of the program:
(I edited out the emails)

Original in README.md:

SHA a581e81d0205f27ede6a5557431a7ad16b4d4c60
Mon Jan 03 45538 18:06:40 GMT-0500 (EST)
ademoss <------------------------>
Initial commit

SHA 38e52f3dea8aac939168b4968f7f2688b9ce5a2e
Wed Jul 04 45551 06:20:00 GMT-0400 (EDT)
Anthony De Moss <------------------------>
- initial commit

SHA 9976fc4546bca7d152453be2c0c9ca59ca9bad7a
Sat Mar 01 45552 11:03:20 GMT-0500 (EST)
John renner <------------------------>
Initial implementation

SHA 700636e2d84f29a4ca8cfcc9b1c003c3c018fc7f
Thu Mar 13 45552 10:10:00 GMT-0400 (EDT)
John renner <------------------------>
Fixed silly naming conventions

Modified aggregating version:

SHA 0000000046bca7d1f01d00d4607f0000d0c92101
Mon Jan 03 45538 18:06:40 GMT-0500 (EST)
ademoss <------------------------>
Initial commit

SHA 0000000046bca7d1f01d00d4607f0000d0c92101
Wed Jul 04 45551 06:20:00 GMT-0400 (EDT)
Anthony De Moss <------------------------>
- initial commit

SHA 0000000046bca7d1f01d00d4607f0000d0c92101
Sat Mar 01 45552 11:03:20 GMT-0500 (EST)
John renner <------------------------>
Initial implementation

SHA 48fcde000000000070cd21010000000020fcde00
Thu Mar 13 45552 10:10:00 GMT-0400 (EDT)
John renner <------------------------>
Fixed silly naming conventions

The original matches the git log perfectly. The same thing happens if I try to manually aggregate them instead of using the aggregate provided in history().on('end').

I'm guessing this issue is due to some variable being reused improperly when creating the commits/oid's.

[nkallen]

This may be an unfair request, but consider using the wip branch, where I believe this issue is fixed. The API has changed significantly however, so this may be troublesome for you.

[djrenren]

ah okay, good to know it's getting fixed. Is there any sort of ETA / progress update towards a next release?

[nkallen]

Fixed in https://github.com/nodegit/nodegit/tree/wip