Setup

[mpshop]

Am new to the twitter API, and Phirehose, wondered is there a simple
setup guide to get this working and then output the stream to a
webpage with CSS to style it ? Downloaded the current version, assume
I need to use sample.php through a SSH, I have godaddy and PHP 5.2.

Saw that some people save tweets to a database, don't think thats
legal on the twitter end ? Or is it ?

I looked at 140dev's work, but the support is lacking and too slow in
response to questions on setup.

Thanks

[phil....@gmail.com]

I would agree with you... and I noticed no one bothered answering here either....

 

Little things like the samples show "username" and "password" where no where in anything I get from twitter is there a username or password other than that required to login to my account and surely you'd use something different than that for an app!

 

 

[Fenn Bailey]

Hi there,

You're right - All of the Phirehose examples are designed to get you consuming the stream and no further - ie: Nothing about what you do with the tweets once you get them.

The closest we have is the ghetto-queue example which shows how you may save them to a file for later processing.

Processing and displaying tweets on a webpage is really beyond the scope of what Phirehose is designed to help you to do, that said, it's an obvious use, so perhaps an example would be helpful.

Ultimately, displaying tweets on a webpage is no different than displaying anything else (ie: retail products, financial data, etc) so you can use any other independent examples of showing data from a database - The trick is getting the tweets into the database in the first place, which is what Phirehose _is_ for (see many examples on the mailing list, etc).

Cheers,

  Fenn.

[phil....@gmail.com]

Fenn,

 

I appreciate your work in writing and sharing the code... I've done that on several projects and I can appreciate the time it takes.

 

Processing the feed isn't the problem nor is displaying it, its getting started on consuming it... the examples seem to make an assumption of knowledge that many of us probably don't have.  Just looking at my previous post, the confusion on using the userid and password vs the keys provided to the app.  These two simple differences should be documented.

 

I also think more step by step instructions need to be included in each example.  For instance, in the user-stream example, how do you specify which users to get updates from?  Is this like the FB API where you get updates on everyone who has authorized your ap?  If so, spell that out.  Do we have to provide a list of people we want updates from (this is what the twitter API seems to indicate) then how do we pass that list? Will it handle a list of 15,000?  30,000?  60,000?

 

So much information is missing and with what Twitter docs say it leaves more questions than answers.

 

I will continue trying to get my little project working... hopefully its possible.

 

Thanks,

 

Pete

[Darren Cook]

> I also think more step by step instructions need to be included in each
> example. For instance, in the user-stream example, how do you specify
> which users to get updates from?

I had replied directly to Phil earlier, but in case others have this
question.

For user streams you follow *one user*.
I.e. the user you have the OAUTH_TOKEN for.

If you want to follow multiple users you need to:
a) set up a userstream instance for each;
b) get them to permission your app (and store the token/secret that
will give you).

(This is what sitestreams is for, but that is still a closed beta, and
hard to get in.)

Darren

--
Darren Cook, Software Researcher/Developer

http://dcook.org/work/ (About me and my work)
http://dcook.org/blogs.html (My blogs and articles)