When I built the 140dev framework 3 years ago, my goal was to provide a set of training wheels. Coders who were new to the API could get a complete working system in miniature. My emphasis was on simplicity, certainly not production level functionality. Since the Twitter IPO and the $200M acquisition of Topsy, collecting data from Twitter has taken on a greater sense of urgency. I’ve been getting so many emails for a common set of improvements that I decided to start from scratch and build a new version of this code. I’ve learned a lot about tweet collection in the last 3 years. My goal will be to include as many of the production level techniques as I can in this new version.
The code will still be free, and I’ll also be writing a new book to provide more context for maintaining and enhancing the system. I’ll leave the current version available for download for a long time, but I’ll have a beta test version of the new code available fairly fast. More details on downloads of the new code in later posts.
The first task in creating a new version of the streaming code is replacing the tweet display code. This module shows me to be guilty of trying to do too much and ending up with a worse result. The fundamental flaw was in including Ajax code to live refresh the count of new tweets. This is cute, but adds a configuration step that is too confusing. Ajax imposes a set of cross server limits that I shouldn’t have brought into this code.
My new version of the tweet display code will use the report system from the engagement programming book. It can become a universal tweet browser for all my projects. I’ll also include a user browser to show each user’s profile. The starting code is here. I’ll post a new download link as I build out the beta version of this new code.
I’d like to get feedback on this project as I propose and make changes. Please let me know what you think in comments or tweets.
.@140dev, the Twitter streaming API framework really needs this enhancement:
