* Mark Pilgrim <[EMAIL PROTECTED]> [2005-08-29 18:20]: > On 8/26/05, Graham <[EMAIL PROTECTED]> wrote: > > So you're saying browsers should check robots.txt before > > downloading images? > > It's sad that such an inane dodge would even garner any > attention at all, much less require a response.
I’m with you on how robots.txt is to be interpreted, but to a point there is a point to the dodge. F.ex, your example of pointing an enclosure to a large file on a foreign server in order to perform a DoS against it is equally practicable by pointing an <img src=""> to it from a high traffic site. The distinction between what’s inline content and what’s not really is more arbitrary than inherent. Of course, that’s just splitting hairs, since it doesn’t actually make a difference to the interpretation. Crawlers generally don’t traverse img/@src references, and the few that do, such as Google’s and Yahoo’s image search services, respect robots.txt. Further, aggregation services do not retrieve images referenced in the content of the feeds they consume. So why should they retrieve enclosures? Regards, -- Aristotle Pagaltzis // <http://plasmasturm.org/>
