Twitter + Gnip Partnership

[Ryan Sarver]

Companies have leveraged Twitter’s open API to analyze and report on
conversations and sentiment across the network since its inception.
These products have been indispensable in helping brands, marketers
and businesses engage with their customers on Twitter. This is an area
we want to support more fully, and today we are excited to announce a
partnership with Gnip to develop and market data products specifically
for these analysis and non-display companies. Gnip will sublicense
access to our public Tweets to developers interested in analyzing
large amounts of Twitter data.

Over the past year we have spoken with many companies and
entrepreneurs throughout the ecosystem who need easier access to more
data. In particular, companies building analysis and non-display
products have asked us for greater volume and coverage. Our
partnership with Gnip is built to address this need. Gnip will focus
exclusively on creating products to meet the existing and emerging
demands of companies creating non-display products. Check out Gnip’s
blog to learn more and to see details about their initial Twitter data
products: http://blog.gnip.com/gnip-twitter-partnership/.

Many of you may wonder what this means for elevated access and
whitelisting requests. Our default levels like Spritzer, Follow and
Track will not be changing, and will remain free and available
directly from Twitter. Companies and developers are encouraged to
begin development with these free APIs, available at
http://dev.twitter.com/pages/streaming_api. This does affect companies
wishing to create products which analyze Tweets and do not display
Tweets to end-users. Moving forward, we will begin to encourage these
companies needing elevated access for analysis and non-display
products to work with Gnip to find the right data products for their
commercial needs.

We’re excited about this partnership, and the support it offers the
data analysis and non-display market. You can learn more about the
details and Gnip by visiting http://gnip.com/twitter. Please let me
know if you have any questions about how this affects you and your
products.

To contact Gnip:
web: http://gnip.com
email: in...@gnip.com
twitter: http://twitter.com/gnip

Best, Ryan

[Edward Hotchkiss]

Checkout indexing all possible Twitter data into a large dataset and
doing "api calls" from there.

Best,

--------------
Edward H. Hotchkiss
http://www.edwardhotchkiss.com/
http://www.twitter.com/edwardhotchkiss/
--------------

[Dewald Pretorius]

Ryan,

The Gnip blog post states:

[QUOTE]Twitter Decahose. This volume-based product is comprised of 10%
of the full firehose. Starting today, developers who want to access
this sample rate will access it via Gnip instead of Twitter. Twitter
will also begin to transition non-display developers with existing
Twitter Gardenhose access over to Gnip.[/QUOTE]

How does this affect the basic statuses/sample method of the Streaming
API? Are you discontinuing it? If so, when?

[Ryan Sarver]

Dewald,

The basic levels of all of the streaming APIs -- Spritzer, Follow,
Track -- will remain open, free and direct from us. Elevated levels
for non-display use will be served through Gnip.

Hope that answers the question.

Best, Ryan

> --
> Twitter developer documentation and resources: http://dev.twitter.com/doc
> API updates via Twitter: http://twitter.com/twitterapi
> Issues/Enhancements Tracker: http://code.google.com/p/twitter-api/issues/list
> Change your membership to this group: http://groups.google.com/group/twitter-development-talk
>

[Dewald Pretorius]

Ryan,

Thanks. Can I then suggest that you request Gnip to modify the
description of their Twitter Decahose feed. They refer to it as a
sample rate, which immediately creates confusion with your statuses/
sample.

[Shannon Clark]

Looking at Gnip's website they have the "contact us for pricing" links - will Twitter & Gnip be making the pricing for the various levels public?

Will companies that license the data be allowed to, in turn, sell services on top of that data - i.e. will this spark a new generation of products such as Scout Labs (now Lithium) or other analytics tools which are built by companies who have negotiated for full or partial firehose access but which are then used by clients of those companies each of whom will configure different queries and searches to monitor?

And on a more technical level will Gnip and Twitter work together to make the transition for developers who might start building/testing a tool using Twitter's free API's but then later migrate to Gnip's commercial feeds as seemless as possible? Will the API calls etc be similar (or identical but with different URL's?)

And a further query - you emphasize that this is for "non-display" services - does that mean, for example, that an analytics tool built using the new Mentions feed from Gnip cannot display the underlying Tweets that are returned by that feed? This would seem to severely limit the value and utility of such analytics to many businesses (who might want to reply to many of those messages, might want to follow people on Twitter discussing their company/brand/industry/competitors, and in almost all cases will want to view the full Tweet w/rich metadata not just a summarization of #s of tweets etc.)

And/or would a business focused Twitter client - CoTweet, Hootsuite, Tweetdeck etc be able to offer (perhaps as part of a professional version) such enhanced Mentions feeds and display them within that application?

thanks,

Shannon

(I'm not an active developer at the moment but I am consulting some business clients on a range of social media tools and as analytics and the appropriate use of them is a core part of my recommendations I'm following these developments closely and look forward to I hope new competitors in the analytics space soon)

---------------------
Real Things - http://realthings.posterous.com/
Slow Brand - http://slowbrand.com
Searching for the Moon - http://shannonclark.wordpress.com
---------------------
cell: 1.510.333.0295                 Twitter - rycaut

[M. Edward (Ed) Borasky]

Ryan, what about User Streams? I'm building something around User
Streams but it is a "non-display" analytics application. Am I at risk
for Twitter inserting another business into *my* data stream as well?
And I'm curious how some of the other Streaming consumers are going to
react to insertion of a monopoly middleman into their data source. I
briefly dealt with Gnip a while back and found their API hard to use
and their pricing exorbitant.
--
M. Edward (Ed) Borasky
http://borasky-research.net http://twitter.com/znmeb

"A mathematician is a device for turning coffee into theorems." - Paul Erdos

[Adam Green]

Ryan:

Shannon raises a lot of great points, but I'd like to hear more about
the issue of reselling data derived from a purchased stream. Right now
the TOS says that you can't resell data from the API. I've been
telling clients that eventually Twitter will decide to make money from
the API, and when that happens there would have to be a way to resell
what has been paid for. Now that you are selling access to the API,
which I strongly agree with, will you allow a free market to evolve
around that by making it possible for Twitter data retailers to grow
businesses, as well as wholesalers like Gnip? Please, say yes. I'm
hoping an Apple-style, control the distribution channel completely
mindset doesn't develop at Twitter. I'm hoping Twitter wants to help
the developer ecosystem turn into a true third party market. Letting
developers sell data or help clients sell data is essential for that.

--
Adam Green
Twitter API Consultant and Trainer
http://140dev.com
@140dev

[Dewald Pretorius]

The minimum Gnip charge is $500 per month, with a minimum of a year
contract, if you want to use Gnip in a production application.

And that's before the -- still unknown -- additional access charges
for the Twitter feeds.

You can't use Gnip in a production application if you are not an
incorporated business, so that excludes access for many developers,
even if they can afford the charges.

Maybe there's a secondary market here, for an incorporated business to
provide access for one-man developers to Gnip data for a fee. Meaning,
Reseller Inc subscribes to Gnip and gets the data feeds, and resells
them to one-man developers. I haven't checked Gnip's TOS to see if
that's expressly prohibited.

On Nov 17, 2:51 pm, "M. Edward (Ed) Borasky" <zn...@borasky-

research.net> wrote:
> Ryan, what about User Streams? I'm building something around User  
> Streams but it is a "non-display" analytics application. Am I at risk  
> for Twitter inserting another business into *my* data stream as well?  
> And I'm curious how some of the other Streaming consumers are going to  
> react to insertion of a monopoly middleman into their data source. I  
> briefly dealt with Gnip a while back and found their API hard to use  
> and their pricing exorbitant.
> --

> M. Edward (Ed) Boraskyhttp://borasky-research.nethttp://twitter.com/znmeb

[Dewald Pretorius]

By the way, if you get Twitter data from Gnip, you are not bound to
the Twitter TOS. Your business and contractual relationship is with
Gnip, not Twitter.

[Adam Green]

Dewald, I can't speak for Twitter, but I think you are missing the
path they seem to be building. As an independent developer you can
still use the streaming API at the default level of 400 keywords and
5,000 follows for free. That is plenty to get a site started or build
a proof of concept for a client at no cost. If the site gets traction
or a client likes it, then you charge for it and get the money to
scale up. The client could be the corporate purchaser of the feed, or
if you have a site that charges for access, then you'd be crazy not to
get limited liability by incorporating as an LLC or S corp. That costs
$500 to file for in most states.

I have no idea what Gnip's final prices will be. If they are
exhorbitant, Twitter will either die, or they will give wholesale
status to multiple vendors and let the market figure out the wholesale
price. I think they are smart enough to choose the later. The big
thing, the REALLY BIG thing, is that I just used the word price twice
in relation to Twitter. That means people will pay for Twitter stuff.
That means developers can get paid for Twitter stuff. Hooray! I like
getting paid. I don't mind paying others if it means I can also get
paid. As long as everything is free, nobody gets paid.

Don't you want to get paid for your work?

> --
> Twitter developer documentation and resources: http://dev.twitter.com/doc
> API updates via Twitter: http://twitter.com/twitterapi
> Issues/Enhancements Tracker: http://code.google.com/p/twitter-api/issues/list
> Change your membership to this group: http://groups.google.com/group/twitter-development-talk
>

--

[M. Edward (Ed) Borasky]

I quite frankly don't see *any* economic value in a downsampled
Firehose. Why should *anyone* pay Gnip for 10% or 50% of the Firehose
when they can negotiated *directly* with Twitter for the whole Firehose?

--
M. Edward (Ed) Borasky

http://borasky-research.net http://twitter.com/znmeb

"A mathematician is a device for turning coffee into theorems." - Paul Erdos

Quoting Dewald Pretorius <dpr...@gmail.com>:

[Dewald Pretorius]

Adam, what I wrote was not a case for or against Gnip and/or Twitter
selling the stream through Gnip. I simply quoted Gnip prices and
conditions from the Gnip pricing page, because it is relevant to this
discussion.

This time I will venture further and say that I do not see where is
the value-add for the developer in getting his Twitter stream data
from Gnip instead of directly from Twitter. Perhaps it's because the
details on Gnip's site are still very scant on the Twitter feeds, or
perhaps it's a deal where only the benefits for Twitter and Gnip were
fully considered. I'd love to hear what are the benefits for the
developer.

[Ryan Sarver]

Shannon, good questions -- answers inline below...

On Wed, Nov 17, 2010 at 12:27 PM, Shannon Clark <shanno...@gmail.com> wrote:
> Looking at Gnip's website they have the "contact us for pricing" links -
> will Twitter & Gnip be making the pricing for the various levels public?

They will be published if they aren't already and they are being
widely reported through RWW and other outlets. One of the main goal is
transparency

>
> Will companies that license the data be allowed to, in turn, sell services
> on top of that data - i.e. will this spark a new generation of products such
> as Scout Labs (now Lithium) or other analytics tools which are built by
> companies who have negotiated for full or partial firehose access but which
> are then used by clients of those companies each of whom will configure
> different queries and searches to monitor?

Companies can definitely build and sell products based on the analysis
of the data. A major market for this move is the Social Media
Monitoring (SMM) market and we expect that to grow.

>
> And on a more technical level will Gnip and Twitter work together to make
> the transition for developers who might start building/testing a tool using
> Twitter's free API's but then later migrate to Gnip's commercial feeds as
> seemless as possible? Will the API calls etc be similar (or identical but
> with different URL's?)

Gnip is offering an exact proxy of our API so that the payloads look
the same. You would just need to change the endpoint you are pointing
at and (I think) your credentials for accessing the endpoint

>
> And a further query - you emphasize that this is for "non-display" services
> - does that mean, for example, that an analytics tool built using the new
> Mentions feed from Gnip cannot display the underlying Tweets that are
> returned by that feed? This would seem to severely limit the value and
> utility of such analytics to many businesses (who might want to reply to
> many of those messages, might want to follow people on Twitter discussing
> their company/brand/industry/competitors, and in almost all cases will want
> to view the full Tweet w/rich metadata not just a summarization of #s of
> tweets etc.)

This is really about B2C vs B2B. We expect that the dashboard will
want to show tweets and we support that, but it should be for a
commercial audience that wouldn't be interested in running Twitter's
promoted products. Let me know if that doesn't make sense.

>
> And/or would a business focused Twitter client - CoTweet, Hootsuite,
> Tweetdeck etc be able to offer (perhaps as part of a professional version)
> such enhanced Mentions feeds and display them within that application?

This deal is all about elevated access. CoTweet and Hootsuite are able
to operate on the freely available, basic APIs. If however, Hootsuite
wanted to get larger volumes of data for analytics, they would want to
reach out to Gnip.

Hope that answers your questions.

Best, Ryan

[Ryan Sarver]

This deal with Gnip is all about *elevated access* you can build
whatever product you want (as long as it adheres to the Twitter API
Rules) with the basic APIs and basic levels of access.

As to the second part of your question we are setting the pricing as
to ensure that their sole position isn't exploited. With that being
said, you might find the products to be expensive, but we feel this is
premium data and we're mostly focused on consumer facing businesses
where the business model is promoted products and the data is free to
developers.

[Ryan Sarver]

Adam, it's a good question and it really comes down to what you are
trying to re-sell.

Re-syndication or re-sale of the actual tweets is strictly prohibited
and won't change on our end. We are however, ok with reselling of data
that results from analysis of the Twitter API.

So a great example is Klout. They do a lot of work to determine a
user's Klout score by analyzing the Twitter API and the content of
tweets. They *are* able to resell their score, but they would not be
able to resell the tweets that were used to determine that score.

It's nuanced, so let me know if that makes sense.

[Ryan Sarver]

That's explicitly not true. You are bound by both the Twitter API
Rules and Gnip's TOS

[Ryan Sarver]

Ed, many developers don't want or can't afford the full Firehose. The
market for Gnip is very large based on the demand that we were unable
to serve.

[Dewald Pretorius]

Ryan,

Gnip will have to extend the Twitter API Rules into their TOS,
otherwise good luck with enforcing the Twitter API Rules if the stream
consumer has a contract only with Gnip.

Your answer about elevated access answers my question about value.

For completeness, here's what ReadWriteWeb says about the prices:

Gnip will offer 50% of all the messages posted to Twitter for $360,000
per year, or 5% of all messages for $60,000 per year. [1]

http://www.readwriteweb.com/archives/twitter_to_sell_50_of_all_tweets_for_360kyear_thro.php

[Adam Green]

Ryan, I understand. I'm just happy to see you help companies put a
real value on Twitter data in any form. And I'm happy to see Twitter
find new ways to make money. You'll never hear "everything online must
be free" from me. I go way back to when people paid for software, in
a box, in stores.

I'm also willing to bet that Twitter will eventually allow a paid
market to develop in actual tweets as well as data derived from them.
When Twitter IPOs, the market will demand that. Paying a third party
to filter and rank tweets that can be displayed on a website seems
perfectly legitimate. Why should every company have to pay to do their
own API programming to display aggregated tweets, when they can pay
someone for high quality tweets as a service? It seems illogical to
me, and from the point of view of the tweet's author, the copyright
issues are identical.

[Dewald Pretorius]

As a business model, is there another company that takes content,
which its users create and enter into the company's service with no
compensation, and then turns around and sells that content to third
parties, still with no compensation to the creators of the content?

I've been trying to think of another company that does this, but I'm
striking a blank. I'm sure there must be others.

> >>> Slow Brand -http://slowbrand.com
> >>> Searching for the Moon -http://shannonclark.wordpress.com

> >>> ---------------------
> >>> cell: 1.510.333.0295                 Twitter - rycaut
>

[dshah]

Ryan,

 

What happens to users who are at 'restricted track' or 'partner track' levels for streaming API access? Also, what is the time frame for moving from twitter to Gnip and would twitter be contacting users who will no longer be able to access Twitter API and refer them thru migration process?

 

I am still not clear about usage of 'non-display' term. From your example - would current or future B2B tool vendors offering services similar to Radian6, ScoutLabs, have to go thru Gnip to get Twitter data? 

 

 

--

Thanks,

Devang.

[M. Edward (Ed) Borasky]

Quoting Ryan Sarver <rsa...@twitter.com>:
> On Wed, Nov 17, 2010 at 12:27 PM, Shannon Clark
> <shanno...@gmail.com> wrote:
>> Will companies that license the data be allowed to, in turn, sell services
>> on top of that data - i.e. will this spark a new generation of products such
>> as Scout Labs (now Lithium) or other analytics tools which are built by
>> companies who have negotiated for full or partial firehose access but which
>> are then used by clients of those companies each of whom will configure
>> different queries and searches to monitor?
>
> Companies can definitely build and sell products based on the analysis
> of the data. A major market for this move is the Social Media
> Monitoring (SMM) market and we expect that to grow.

As I've already noted, I don't see the economic / business sense in
paying a monopoly middleman for downsampled Firehose when the full
Firehose is directly available via negotiation with Twitter. IMHO, if
you've got the brains and infrastructure to create social media
monitoring business value from 10% or 50% of the Firehose, it's easy
to scale that up to 100% of the Firehose. If you don't, well, you're
one of the 95 percent of businesses that fail because *you* made a
wrong decision.

While I haven't paid much attention to the "social media monitoring"
market recently, what I've seen for much of 2010 is consolidation -
big companies like IBM buying smaller ones with *solid* business
models. What I *haven't* seen in social media monitoring / analytics
is "small nimble startups" becoming successful with "minimum viable
products".

Social media monitoring is a difficult business to be in, *especially*
at the data rates Twitter delivers and the "unnatural" aspects of
Twitter linguistics. The sales cycle for social media monitoring tools
is long and arduous, and, IMHO, Facebook, Flickr and YouTube data are
immensely richer and easier for marketers to explore and exploit than
Twitter data.

--
M. Edward (Ed) Borasky

http://borasky-research.net http://twitter.com/znmeb

[Edward Hotchkiss]

Just write your own massive dataset and filter our Twitters ads. On a
side note someone wrote about error 403 proxy. No you never need a
proxy, but use a proxy to circumvent the API sure awesome.

[John Kalucki]

Every search engine, social network, blogging platform, content aggregator, and to a certain extent, every used book store and used record store...

-John

[M. Edward (Ed) Borasky]

Quoting John Kalucki <jo...@twitter.com>:

> Every search engine, social network, blogging platform, content aggregator,
> and to a certain extent, every used book store and used record store...

Except that digital content producers can block search engines if it's
in their economic interests to do so. I'm not sure how that's working
out in "Murdoch vs. Google", but at least it's been examined. ;-)

For that matter, some "news organizations" have imposed strict rules
on how and when they may "use" Twitter.

[Tom van der Woerdt]

You can simply set your account to protected...

Tom

Sent from my iPhone

[Dewald Pretorius]

John,

I'm not sure how you draw that comparison. Google/Yahoo/Microsoft do
not sell the content of the sites that they index. Neither do
WordPress or Blogger sell the content of the blog posts. Facebook/Buzz
do not sell the content of people's status updates. They monetize
"around" the content, with ads, etc., just as Twitter does with
promoted content.

This is not a question of right or wrong. The Twitter TOS make it
clear that you can / will provide the content to third-parties with no
compensation to Twitter users.

I'm just trying to figure out who else uses the same business model.

[Matthew Terenzio]

We have every right in the world to gather this data for analysis without any permission. It's public. Redistributing it will be subject to fair use and copyright law but not gathering it and making broad analysis. That is what search engines do and so far the courts have said they have a right to cache copies on their own servers, not for public display necessarily, but in order to better analyze it. Oddly, the courts landed on the right side for once, saying that the greater good of the utility of search was a societal need and, in this case, more important than minor infringements, if any, on the site's copyrights.

On Wed, Nov 17, 2010 at 12:18 PM, Ryan Sarver <rsa...@twitter.com> wrote:

Companies have leveraged Twitter’s open API to analyze and report on
conversations and sentiment across the network since its inception.
These products have been indispensable in helping brands, marketers
and businesses engage with their customers on Twitter. This is an area
we want to support more fully, and today we are excited to announce a
partnership with Gnip to develop and market data products specifically
for these analysis and non-display companies. Gnip will sublicense
access to our public Tweets to developers interested in analyzing
large amounts of Twitter data.

Over the past year we have spoken with many companies and
entrepreneurs throughout the ecosystem who need easier access to more
data. In particular, companies building analysis and non-display
products have asked us for greater volume and coverage. Our
partnership with Gnip is built to address this need. Gnip will focus
exclusively on creating products to meet the existing and emerging
demands of companies creating non-display products. Check out Gnip’s
blog to learn more and to see details about their initial Twitter data
products: http://blog.gnip.com/gnip-twitter-partnership/.

Many of you may wonder what this means for elevated access and
whitelisting requests. Our default levels like Spritzer, Follow and
Track will not be changing, and will remain free and available
directly from Twitter. Companies and developers are encouraged to
begin development with these free APIs, available at
http://dev.twitter.com/pages/streaming_api. This does affect companies
wishing to create products which analyze Tweets and do not display
Tweets to end-users. Moving forward, we will begin to encourage these
companies needing elevated access for analysis and non-display
products to work with Gnip to find the right data products for their
commercial needs.

We’re excited about this partnership, and the support it offers the
data analysis and non-display market. You can learn more about the
details and Gnip by visiting http://gnip.com/twitter. Please let me
know if you have any questions about how this affects you and your
products.

To contact Gnip:
web: http://gnip.com
email: in...@gnip.com
twitter: http://twitter.com/gnip

Best, Ryan

[Scott J]

I would like to know the answer to this as well. What will the limits
be on the statuses/filter?

[L. Mohan Arun]

"We have every right in the world to gather this data for analysis
without
any permission. It's public."

No.

"You don't get to compile posts from a discussion forum into a
product, under the idea that such posts are "public domain."
They are not." - Unless you own the forum or have a deal with the
forum owner,
and you stated in the TOS
that all posts made in the forum can be repackaged commercially
and only you have the right to do that.

I am not saying this on my own, this is from one of the newsletters I
receive,
which covered this exact same topic, I would be happy to share the
relevant
text of the newsletter if someone is interested ...

- - -
Mohan Arun
www.mohanarun.com

[Matthew Terenzio]

I don't care what your newsletter says. I'm talking about American law.

--

[Matthew Terenzio]

Just to clarify. I never said they were Public Domain. Twitter or the user own the copyrights. Probably both. I meant it has been made public information, thereby granting some rights to those it was made public to. I wouldn't have a right to redistribute a book written by you, but I have every right to quote it in an article I write about you.
More importantly, I can read 1000 books by 1000 different people and then write a paper that says 50% of the books written contained the word 'Obama' and  and the average amount of times "Obama" was used in a book was 14.
I wouldn't be breaking any laws.
But who cares.
In the future, if you want to access the Twitter data for such usage with any sort of speed you will pay to do so. It won't even be worth the headache if you can devise an alternative.

[Edward Hotchkiss]

Well, they do have their ToS the law has so far placed in favor of usage of apps and apis regardless of ToS as long as it is "legal". Yet, due to massive litigation.

Best,

--------------

Edward H. Hotchkiss

http://www.edwardhotchkiss.com/

http://www.twitter.com/edwardhotchkiss/

--------------

[Matthew Terenzio]

Right. usage of the API is completely under Twitter control and TOS. I understand that. And yes, all of this is new and subject to litigation. Not worth the headache unless a rug was pulled out under and existing established business and agreement, which is probably only a few companies if any  and I'm sure Twitter is working with them.

[jmathai]

I agree with blogging platforms and social networks but not the rest.
Being an owner of a website does not imply that I'm a "Google user".
Nor is a musician a user of the used record store.

On Nov 17, 8:48 pm, John Kalucki <j...@twitter.com> wrote:
> Every search engine, social network, blogging platform, content aggregator,
> and to a certain extent, every used book store and used record store...
>

[Ryan Sarver]

The basic level of statuses/filter will remain unchanged

[L. Mohan Arun]

"Twitter or the user own the copyrights. Probably both. I meant it has
been made public
information, thereby granting some rights to those it was made public
to."

This is a good point, makes one think: "What exactly are the rights
someone
is implicitly giving up just by posting it in a public forum, and if
someone else is found
using such posted contents inappropriately, at a later date, what
legal recourse does
the original poster have, to have the inappropriate usage rescinded
and compensated for damages"

From an authority source:
"Merely posting a work online does not relinquish all rights. As in
other environments, merely placing property in public does not release
property rights. The Internet context, however, may indicate that some
actions with respect to the work are implicitly permitted." (as long
as it doesnt harm the poster and the forum from which the post was
taken)

This is from http://www.ipinfoblog.com/archives/intellectual-property-posting-as-implied-license.html

Facebook made the same argument.
“Anyone can opt out of appearing here by changing their Search privacy
settings.”

And someone asks "Yeah, but should they have to?"
http://techcrunch.com/2010/07/28/hacker-proves-facebooks-public-data-is-public/

I imagine this whole 'using twitter tweets only for analysis/
aggregation' is a non-issue, because you are only statistically mining
info, without any personal data attached to it. But RapLeaf removed
identifying info (name, tel, etc) from their profile databases once
complaints started coming in. To be safe, if you are mining tweets
only for analysis, I wouldnt store the userid, because userid is
'identifying info' that can be used to tie the tweet to its
originator. Most people wont bother, because the prevailing idealogy
is that if you tweeted, then you understood that the whole world knows
what you tweeted and you cant take the knowledge back.

~~~ Mohan Arun ~~~

[Adam Green]

What about users who want their tweets to be repeated? Politicians,
celebrities, product managers, and many others use Twitter as a
broadcast medium. You can argue that this is wrong, or that Twitter is
only for direct contact between one person and another, but that is
like saying paper was only intended to write letters between two
people. Twitter is a medium that will be used in many ways. It is just
in its infancy. I know one thing, Twitter will be around long after
the "power to the people" perspective of Web 2.0. Long after anyone
remembers what Web 2.0 and user generated content mean.

> --
> Twitter developer documentation and resources: http://dev.twitter.com/doc
> API updates via Twitter: http://twitter.com/twitterapi
> Issues/Enhancements Tracker: http://code.google.com/p/twitter-api/issues/list
> Change your membership to this group: http://groups.google.com/group/twitter-development-talk
>

--

[M. Edward (Ed) Borasky]

Quoting Ryan Sarver <rsa...@twitter.com>:
> Many of you may wonder what this means for elevated access and
> whitelisting requests. Our default levels like Spritzer, Follow and
> Track will not be changing, and will remain free and available
> directly from Twitter. Companies and developers are encouraged to
> begin development with these free APIs, available at
> http://dev.twitter.com/pages/streaming_api.

Is "Spritzer" still 1% of the Firehose? Since the status IDs are no
longer sequential, the previous "obvious" sampling algorithm - "status
ID mod 100 == 0" - no longer will work.

[Ryan Sarver]

Spritzer is currently at 1% of the Firehose, but as the docs say it's
subject to change without notice

[M. Edward (Ed) Borasky]

Quoting Ryan Sarver <rsa...@twitter.com>:

> Spritzer is currently at 1% of the Firehose, but as the docs say it's
> subject to change without notice

Given the Snowflake algorithm, how can a program consuming Spritzer
determine whether a Spritzer rate change has happened because

a. People are tweeting at a different rate, exclusive-or
b. Twitter has changed the proportion of Firehose being sent to Spritzer?

[Karthik K]

can some one please tell me what happens to users who are at 'restricted track' or 'partner track' levels for streaming API access?