Tim had one of those I-woke-up-and-thought-of-some­thing-to-blog ex­pe­ri­ences. Which is part of the prob­lem, but what kills in­ter­net is it's growth that can't be matched by skilled sys­tem ad­min­is­tra­tors. The most easy way to se­cure a net­work is to close it down. Which makes it pret­ty use­less, but most users won't no­tice any­thing as long as you leave port 80 open to the out­side. And to keep work­sta­tions man­aged eas­i­ly, don't al­low ap­pli­ca­tions to run on them. Which makes them pret­ty use­less, but most users won't no­tice any­thing as long as they have Of­fice and In­ter­net Ex­plor­er. And that's the Web 2.0 plat­form Tim is talk­ing about: in­ter­net re­duced to a brows­er us­ing port 80. On such a plat­form there is no ssh con­nec­tion to a real ap­pli­ca­tion on a real sys­tem, or ded­i­cat­ed ap­pli­ca­tions de­signed for ded­i­cat­ed tasks us­ing ded­i­cat­ed pro­to­cols. All you get in Web 2.0 are mock-up HTML ap­pli­ca­tions bare­ly held to­geth­er by some form of JavaScript, which is a bet­ter user ex­pe­ri­ence Netscape 4 could of­fer, but still worse than any na­tive so­lu­tion. The prob­lem ob­vi­ous­ly is that na­tive so­lu­tions aren't cool. As a user of na­tive so­lu­tions you'll have to face what plat­form you're us­ing, and that con­fronts you with the fact that while you own an iPod, surf with Fire­fox and own Google stock, you're still stuck with some form of Mi­crosoft Win­dows. And you'll have to ad­mit na­tive ap­pli­ca­tions run pret­ty smooth on Win­dows, as you start up some­thing like F.E.A.R. I don't see that kind of rich user ex­pe­ri­ence any­where in Web 2.0, which con­sid­ers rich to be pre­sent­ed with an er­ror mes­sage while fill­ing out a form on a web­site with­out the lag of a roundtrip to the serv­er, hail AJAX. What will Web 3.0 be? A cross brows­er Win­dows 3.0 em­u­la­tor writ­ten in JavaScript that on a 10GHz Pen­tium V is only a bit slow­er than it orig­i­nal­ly was on a 386? But GPLed, with a fresh com­mu­ni­ty of nitwits with too much time on their hands pro­mot­ing it, backed by the non-evil pro­pri­etary ven­dors du jour, ETLA named, all ready to blow the next in­ter­net bub­ble all over again...

The point is: if you just al­low anony­mous read-only ac­cess to the data­base be­hind your site, you don't need screen scrap­ing or web­ser­vices to let oth­ers remix your con­tent. You wouldn't even need to build your own web­based front-end... But since that looks to much like Web -2.0, it's hard­ly hy­peable.