Thanks! The tweets are indexed in real-time but the writes to DB and rendering of pages are cached, and not updated frequently. There is a net lag of about an hour or so before the updated results are visible.
On Mon, Mar 22, 2010 at 1:09 AM, M. Edward (Ed) Borasky <zzn...@gmail.com>wrote: > That is *really* nice! Is it updated in real time? > -- > M. Edward (Ed) Borasky > borasky-research.net/m-edward-ed-borasky/ > > "A mathematician is a device for turning coffee into theorems." ~ Paul > Erd?s > > > > Quoting Harshad RJ <harshad...@gmail.com>: > > To test how this works I built a streaming parser for the Spritzer feed, >> and >> it occurred to me that I could make this data available to everyone. >> >> So, here it is: >> http://tdash.org/stats/clients >> >> I dunno if the OP just wanted an approx count of the client's tweets or >> the >> actual list of tweets. Personally, I would like to have both. It will be >> great if Twitter can allow search for "source:myclient" without requiring >> a >> keyword to be specified. >> >> On Mon, Mar 15, 2010 at 10:16 PM, John Kalucki <j...@twitter.com> wrote: >> >> I'd suggest calculating the binomial proportion confidence interval >>> assuming a very large n. I think you'll be pleasantly surprised at the >>> interval given n = 2.5mm/day on the Spritzer feed... >>> >>> Well, you learn something new every day. Apparently the central limit >>> theorem apparently holds for p as small as .000002 given n of 2.5mm. So, >>> if >>> your client is generating pretty much any traffic at all, the interval >>> will >>> be pretty reasonable. >>> >>> -John Kalucki >>> http://twitter.com/jkalucki >>> Infrastructure, Twitter Inc. >>> >>> >>> >>> On Mon, Mar 15, 2010 at 9:17 AM, Harshad RJ <harshad...@gmail.com> >>> wrote: >>> >>> What I meant was that searching with "source:clientName" requires >>>> atleast >>>> one keyword to be specified. Which means that you can't get all those >>>> tweets >>>> which don't have that keyword. >>>> >>>> Moreover, searching for common english words like "a", "an", or "the" >>>> (often) doesn't return any results. >>>> >>>> The idea of sampling the streaming API is a decent alternative for >>>> already >>>> popular clients, but will yield inaccurate results for new clients, >>>> because >>>> their tweets are easily submerged in the stream. >>>> >>>> >>>> >>>> On Mon, Mar 15, 2010 at 7:13 PM, Raffi Krikorian <ra...@twitter.com >>>> >wrote: >>>> >>>> its true - search doesn't return all the tweets as it is returning "the >>>>> best tweets". unfortunately, the streaming API will not allow you to >>>>> get a >>>>> stream of all the tweets by source either. what are you trying to >>>>> achieve? >>>>> are you looking for relative volumes? if so, then just watch a sample >>>>> of >>>>> tweets and make an estimate? >>>>> >>>>> >>>>> On Sun, Mar 14, 2010 at 11:50 PM, Harshad RJ <harshad...@gmail.com >>>>> >wrote: >>>>> >>>>> Err, but this does't show *all* tweets of a client. >>>>>> >>>>>> >>>>>> On Mon, Mar 15, 2010 at 6:47 AM, Raffi Krikorian <ra...@twitter.com >>>>>> >wrote: >>>>>> >>>>>> from http://apiwiki.twitter.com/Twitter-Search-API-Method%3A-search >>>>>>> >>>>>>> http://search.twitter.com/search.atom?q=landing+source:tweetie >>>>>>> >>>>>>> >>>>>>> On Sun, Mar 14, 2010 at 3:33 PM, Christian < >>>>>>> christian.frei...@googlemail.com> wrote: >>>>>>> >>>>>>> Hi There, >>>>>>>> >>>>>>>> is it possible to reveal all Tweets placed by a specific client (my >>>>>>>> client)? >>>>>>>> >>>>>>>> Hope someone could help me >>>>>>>> >>>>>>>> THX Chris >>>>>>>> >>>>>>>> >>>>>>> >>>>>>> -- >>>> Harshad RJ >>>> http://hrj.wikidot.com >>>> >>>> >>> >>> >> >> -- >> Harshad RJ >> http://hrj.wikidot.com >> >> To unsubscribe from this group, send email to twitter-development-talk+ >> unsubscribegooglegroups.com or reply to this email with the words >> "REMOVE ME" as the subject. >> >> > -- Harshad RJ http://hrj.wikidot.com To unsubscribe from this group, send email to twitter-development-talk+unsubscribegooglegroups.com or reply to this email with the words "REMOVE ME" as the subject.