FWIW, The http://tweetback.holzhauer.it/ plugin uses javascript to create the tweetbacks on the page, so it is not his website server that is hammering the search API, but my browser that is doing the request instead... this distributes the requests across all of the users and not one central server (I can't speak for how the other plugins work). This will not result in wordpress installs getting blacklisted. I think this falls within the "normal" usage of the search API (tho I could be completely wrong). Matt has said they tweaked the search API servers (or something to that effect) to handle a boat-load more requests without having to rate-limit hardly anyone (I know exceptions exist, of course. I hit the limit last night with like 10 tweetgrids open.... oops).
As for the URL resolving, I think it is up to those services to provide better resolution mechanisms (some of them are currently terrible). However, it would be pretty handy to have twitter "do it for you", but I'm sure that would take a huge amount of computation cost. Anyway, this is an interesting topic, and I'd like to hear others' opinions if they have them. -Chad; (shamelessly stealing the trailing semi-colon from Matt's sig b/c I liked it so much) On Fri, Jan 9, 2009 at 11:35 AM, Nicole Simon <[email protected]> wrote: > > The author of http://tweetback.holzhauer.it/ pinged me for his tweetback > project and > mentioned that http://mashable.com/2009/01/04/twitter-blog-design/ inspired > him; > http://www.smashingmagazine.com/2009/01/09/tweetbacks-plugin-for-wordpress/ > does the same and I saw another one called chat catcher. > > With everybody revamping their blog designs and the importance of twitter - > may > I suggest somebody looks into this before it is wide spread and lots of > wordpress installs get black listed because they hammer the API, > and consequently submit a lot of questions to support? > > This is a fail whale waiting to happen. My suggestion would be to address > this issue maybe for the blog authors before they all start implementing > this or at least urging tweetback developpers to put in mechanisms > for throtteling, maybe having one article explaining a bit of the daily > limitations an api can have. > > I assume blog integration will grow in importance and also bring back up > the question if if twitter or not should decode the urls back again to > ensure a more > trackable web - it could be that the api delivers two results as message, > one with the normal tweet and one with the complete decoded tweet. > > hth > Nicole > > > > ------ > Using Twitter and having a blog? > Try Tweetback at http://tweetback.holzhauer.it/ > > > -- > Suche Beta-Tester für Experiment: > "Journalisten suchen Blogger" - http://bloxpert.de/ > > Kontakt: > http://twitter.com/NicoleSimon // http://mit140zeichen.de/ > http://crueltobekind.org // http://beissholz.de > > skype: nicole.simon / mailto:[email protected] > phone: +49 451 899 75 03 / mobile: +49 179 499 7076 > > >
