I have been using the Twitter API through the tweetpony wrapper since January 2014. I weekly requested the resource https://api.twitter.com/1.1/statuses/home_timeline.json specifying different values for the count parameter (from 4 to 20) and always got a collection of the specified size.
However, since last week (October 2014) I noticed that the size of the collections I get when requesting the same resource does not match the count parameter anymore. I have tried a different wrapper, birdy, to make sure this is not a tweetpony issue and got exactly the same results.
I know that the count parameter does not accurately fix the size of the returned collection since some of the tweets may be discarded after counted. However, for the last 9 months it has worked as I expected and suddenly it is getting some unexpected results.
Is there any relevant change in the Twitter API recently applied regarding the home timeline? I noticed that the parameter include_rts can be used with the home timeline resource even though it is not included in the specification. Is there any other parameter not included in the specification I could use to retrieve the tweets that are missed in the returned collection?
Thanks for your help.