On Wed, Jan 25, 2012 at 12:05 PM, Michael Yang <[email protected]> wrote: > > twitter & other services are adopting this too > http://www.w3.org/QA/2011/05/hash_uris.html
yeah i don't think there is anything inherently bad about it; it's just a generic mechanism for sites to generate static snapshots. ... one can argue that GOOG and co. should just have JS enabled spiders, and maybe that is the best way, but the problems i see are numerous ... people write crappy JS with busy loops and gobs of crap happening (i hear my dual core laptop spin up all the time), and they tie content/links/actions/changes to lots of stupid things like mouseover + appstate + geolocation + eyecolor + ... + ... vs. strictly <a> tags. should the spider pull a page, and artificially move a mouse through every "pixel" on the screen? seems like a huge can-of-worms, to me at least. i think we just need to 301 redirect directly to the mail app, *after* enabling `#!` type uris. then, google will see that, and try to pull the _escaped_hash_ version, at which point we will perform an internal redirect to the static txt file. bang, done. should only be a couple lines fix in the app, and 2 or 3 rules in the apache config. i'm totally booked today and tomorrow (interviews, yay!), but fri and weekend i'll be able to set it up (if not already working that is). -- C Anthony
