Frequently Asked Questions¶
How far back in time does SFM go when collecting a TwitterUser’s tweets?¶
The Twitter API only provides up to the most recent 3200 tweets for an account. When a new TwitterUser is added in SFM, the user_timeline script will request as many tweets as the Twitter API can provide, i.e. up to 3200.
Do I have to set up supervisord in order to use filterstreams or streamsample?¶
No, filterstreams and streamsample can also be run manually using the filterstream and streamsample management commands described in the management commands page.
The number of retweets in a TwitterItem is inconsistent with the number of retweets shown on the tweet in Twitter. Why?¶
TwitterItems are created as they appear at the time they are captured from Twitter. However, tweets on Twitter can change afterwards; they can be further retweeted, they can be deleted, etc. In fact, an advantage of using SFM is that it takes a snapshot of tweets before they change or disappear!
Currently there is no way to “update” a TwitterItem with any changes that may have occurred to the corresponding tweet. This is something we might consider if there is a use case for it.
Does the SFM web interface provide a way to view the files generated by filterstream and streamsample?¶
When I click on the link to view a raw tweet, it’s difficult to read in my browser.¶
There are a number of broswer plugins available (JSONovich, JSONView, and others) which improve the way that JSON is displayed.
Can I set this up on a Mac?¶
We haven’t been running this on a Mac, but a colleague we met at Code4Lib 2014 has done it. Check out his blog post here: http://dicarve.blogspot.com/2014/04/an-relatively-easy-way-for-installing.html