First, is the net a "publication" or a "broadcast"? Bahlieda v. Santa, (2003), 233 D.L.R. (4th) 382; 68 O.R. (3d) 115; 20 C.C.L.T. (3d) 297; 28 C.P.R. (4th) Ont. C.A. per McMurtry, C.J.O., Goudge J.A. and Gillese J.A.

suggests that it is a broadcast with limitation periods appropriate to such. But this is a 2003 decision and I've not the time to update it.

Second, and again looking at Ontario, s. 5 starts the clock when the plaintiff becomes aware of a defamatory statement in a "newspaper" or a "broadcast". The clock runs for six weeks. I wonder if "becomes aware" may be operative regardless of the determination as to broadcast.

There is also the interesting question of whether an article posted in British Columbia to a website hosted on a server in Denmark but read in Ontario is either a "newspaper" or "broadcast" under the Act.

If the period "starts afresh" we run into some very real problems. The most basic of these is the existence of "Google cache". Essentially, Google keeps a copy of virtually everything which is published on even lightly trafficked sites so even if a person were to pull down a defamatory statement it would persist in Google cache where it might be found by search several years after the initial take down. As well, a defamatory statement might well make it into search results themselves notwithstanding the removal of the initial material. Plus, and this makes it all the more interesting, individual ISPs often keep copies of frequently searched pages on their own servers.

It is not a policy decision that is made about books. Each time someone buys a book with defamation in it, or takes it out of the library, the limitation period does not start afresh.

I could understand an argument that the Internet is more accessible than books in stores (for a while, except for the most successful) or libraries. At least courts and commentators should be clear about what they are doing.

How does a 'publisher' of defamation online stop publishing, once it is brought to his/her/its attention that the defamation should not have been published (by whatever means)? One can take a document off one's own web site, or clean it up, but information (however false) tends to spread. One cannot control mirror sites or others who copy the nastiness.

Would it make sense to have the 'permanent publication' rule apply only to more or less official sources of the defamation, and not for everywhere it turns up? Or would that protection expose the person defamed to too much risk?

Mr. Justice Hall decides that policy reasons favour following the UK/Commonwealth rule: "…I do not consider that it would be appropriate for this Court to adopt the American rule over the rule that seems to be generally accepted throughout the Commonwealth; namely, that each publication of a libel gives a fresh cause of action." The policy reason is the societal interest in not barring a person's claim while the defamatory publication is still available online. That seems reasonable to me.