There's lots of Gmail love going on, here and in the comments on the original post. But notice the first few sentences of the post: "I have in my computer eve

There's lots of Gmail love going on, here and in the comments on the original post. But notice the first few sentences of the post: "I have in my computer every e-mail message I have sent or received since 1992. Minus the obvious spam, this database comes to about half a million messages..." Gmail does not provide a local archive of mail messages, across the shifts in providers, mail clients, and addresses that occur over a couple of decades of using email. Gmail has some nice features, no doubt (I do use it myself), but it isn't a be-all, end-all email solution, particularly for those who need or would like to keep their email under local control. Gmail's spam filtering is great, but it doesn't work 100% of the time. Some spam slips through to my inbox; what's worse, I have noticed a variety of false-positives (e.g., regular emails from mailing lists that I often delete but occasionally read aren't spam), meaning that I still must give the Spam folder a once-over before I delete its contents. And, as far as I know, Gmail's spam filtering algorithm is proprietary. This means that the only way to get it is to route your email through Google, and accept the attendant trade-offs of doing so. Regardless of whether that works for you, it isn't a sustainable solution to spam to tell people to "just use Gmail." I would much prefer to see a new Internet email protocol evolve, maybe something like what's described by this commenter (http://www.cringely.com/2010/11/the-decline-and-fall-of-e-mail/comment-page-1/#comment-38525 http://www.cringely.com/2010/11/the-decline-and-fall-of-e-ma...). Basically, if we move to a new standard protocol where something like the features of PGP are mandatory, then building (and filtering mail by) a distributed web of trust will become much easier. ISPs will migrate to this protocol because it will reduce their load and provide a better experience for their customers; et voila, we will have overcome the network effects hurdle that currently makes PGP essentially useless.