Fwd: Elevated access to streaming API for CrisisCamps

[M. Edward (Ed) Borasky]

---------- Forwarded message ----------
From: John Kalucki <jo...@twitter.com>
Date: Mon, Jan 25, 2010 at 9:05 AM
Subject: Re: Elevated access to streaming API for CrisisCamps
To: "M. Edward (Ed) Borasky" <zzn...@gmail.com>
Cc: a...@twitter.com, ra...@twitter.com, a...@hook.org, cb...@meedan.net


I would suggest using the track parameter on the Streaming API. The
default access level should be sufficient to capture all #haiti
hastags at this late date.

As an example, not a practical application:
> curl -s stream.twitter.com/1/statuses/filter.xml\?track=haiti

will result in a steady stream all tweets containing Haiti or #Haiti
the the text field.

-John Kalucki

http://twitter.com/jkalucki
Infrastructure, Twitter Inc.

On Sat, Jan 23, 2010 at 10:34 PM, M. Edward (Ed) Borasky

<zzn...@gmail.com> wrote:
> I know a little more about how we are currently collecting Twitter
> data. The code I saw today appears to be getting the data from the YQL
> interface to Twitter Search. I haven't quite decoded the exact search
> queries - I don't read YQL yet.
>
> On Sat, Jan 23, 2010 at 2:59 PM, M. Edward (Ed) Borasky
> <zzn...@gmail.com> wrote:
>> I'm working with a CrisisCamp team in Portland that's building
>> applications to use Twitter data for responding to the earthquake in
>> Haiti and future disasters. We are currently getting our Twitter feeds
>> indirectly via Yahoo Query Language, and I'm guessing those paths are
>> using the Search API. So we want to migrate to the Streaming API as
>> soon as we can. I'm also guessing there are lots of searches coming at
>> your server with the "#haiti" hashtag as the search string.
>>
>> I don't have an idea yet about what access level we'll need or how
>> many users, search keywords or geotag focuses we'll be using. We may
>> be able to work with the existing "sample" and "filter" streams for
>> now, although we don't have any client code constructed yet. The
>> client libraries will most likely be in Ruby and connect to a Rails
>> database. Everything we are doing is open source whenever possible.
>>
>> The home page for our efforts is
>> http://wiki.crisiscommons.org/wiki/Portland01232010, and the home page
>> for the whole CrisisCamp project is
>> http://wiki.crisiscommons.org/wiki/Main_Page. I'm not sure we need
>> anything from you at the moment, although if we can get some working
>> open source client code for streaming, that would be nice.
>>
>> --
>> M. Edward (Ed) Borasky
>> http://borasky-research.net
>>
>> "I've always regarded nature as the clothing of God." ~Alan Hovhaness
>>
>
>
>
> --
> M. Edward (Ed) Borasky
> http://borasky-research.net
>
> "I've always regarded nature as the clothing of God." ~Alan Hovhaness
>

--
M. Edward (Ed) Borasky
http://borasky-research.net

"I've always regarded nature as the clothing of God." ~Alan Hovhaness