I used to visit Microsoft Research site regularly because they had some really interesting papers and projects there.  Since blogging, I haven't been back (blogging sucked up all my spare time), but thanks to some .NET bloggers aggregated at weblogs.asp.net, I found these two great papers.

Is It Just My Imagination? by Suzanne Ross

Suzanne describes a way to use inkblots, meaningless smears of ink psychologists use, as visual hints to evoke passwords.  In a way, this technique is somewhere between what you know and who you are.  Cool.  This is how it works.  Show users a series of computer generated inkblots and collect their responses.  From each response, take the first letter of the first word and last letter of the last word.

sad angel

Neat idea.  I spent a great deal of time thinking about visual passwords and this paper's psychological angle was like a fresh breath of air for me.  Unless I read it wrong, Suzanne seems to be recommending ten inkblots for both password generation and verification to get 20 character password.  I think that is too much for verification phase.  I would, instead, show fewer but random squence of those ten inkblots.

While I can see many possible problems with inkblot scheme, one-track mind for example, but the core idea is interesting and worth further studies.

Thanks to Frans Bouma

Distributed Computing Economics by Jim Gray

Yup.  That is Jim "Transaction" Gray, whose thick book sits prominently on my bookshelf.  There are many books I regret buying and his is not one of them.  In this paper, Jim talks about the cost of computing and how it affects where location of computing.  Good read.

Thanks to Randy Holloway.