Interesting. I did something like this a few years back, sample output: http://uguu-archive.appspot.com/misaka/replay.html http://uguu-archive.appspot.com/mis

Interesting. I did something like this a few years back, sample output: http://uguu-archive.appspot.com/misaka/replay.html http://uguu-archive.appspot.com/misaka/replay.html http://ioccc.org/2013/misaka/spoiler.html http://ioccc.org/2013/misaka/spoiler.html (mirror, in case if I ran out of quota) I would be curious in some implementation details for this project, particularly how the intermediate deltas are stored.