Gerald Gazdar alerted me to this interesting material on a site run by Michal Zalewski under the catchy name of http://lcamtuf.coredump.cx (it appears to be a blog maintained by a Polish programmer in California). The page might need to be consulted soon, before Microsoft lawyers track down the page author and get around to drafting minatory letters to have it shut down. Basically, what this guy did was to run a search for publicly available Microsoft Word documents that are freely downloadable from the microsoft.com domain and look in them with Unix tools to see if they contained records tracking changes or displaying material that was erased from the final draft. Ten percent of them had change-tracking records and five percent had recoverable deleted material. So he took a look at what Microsoft staff had been erasing from, e.g., press releases and essays attacking Linux. It's certainly instructive, occasionally amusing, and of course, always there is the aura of the dark side, a whiff of brimstone.
But setting aside the particular glimpse we get of the content of the cesspit of the Microsoft corporate mind, there is a general point about technology and linguistic privacy to ruminate on here. As Michal notes, Microsoft has become a victim of the dangers of its own intuitive and apparently helpful word-processing technology here; as Michal puts it, "if you come up with an intuitive technology, you must next find a way to curb its use." That's a very acute insight.Posted by Geoffrey K. Pullum at April 1, 2004 11:41 AM