Sorry, but it is time to make a familiar point all over again.
The other day, I noted that -- if you want insights into the mindsets of editors wrestling with the tricky, hot-button religion angles in the Charlie Hebdo massacre -- it is very important to study the early versions of stories in an elite publication (think The New York Times, in this case) and then contrast them with the versions that ran later.
This is hard to do because of the evolving WWW-era practice of actually removing earlier versions of the story from the online record. This raises all kinds of questions (including for media critics), such as: Did the earlier versions count? Is it accurate to say that a publication like the Times published something if the material no longer "exists" on the record? If a digital tree is removed from a digital forest, how do you discuss whether or not it existed in the first place?
Screen shots help, but it's impossible to screen shot everything. I suspect that stories are now changing so fast that those online time-machine search programs cannot catch everything. There are, of course, critics out there making their own copies of the earlier stories. Thus, via Mediaite.com, we have this gripping passage from an early Times report, quoting survivors of the massacre: