Map of a Twitter status object

Twitter status objects are starting to get pretty big (especially with annotations around the corner).  Have you wondered what all the fields are these days?  Here's a useful map.  I decided to do this map in JSON as opposed to XML as we're considering not supporting XML on v2 of the API (this can be seen because we have a few endpoints that are JSON only right now) — we'll keep it on v1, of course.

Loading mentions Retweet

Comments (19)

Apr 19, 2010
Can Özmen said...
Why is the Twitter status object this verbose? Especially the embedded author info.
Apr 19, 2010
Patrick Kennedy said...
An illustrated example is meant to be verbose - think I prefer looking at the XML. The annotated idea will make it super verbose even via JSON. I think Twitter should offer different formats for basic status objects, it's good to have options, but it's understanable why things are moving towards JSON.
Apr 19, 2010
Raffi Krikorian said...
fair point.

If I have time, I'll create an equivalent example in XML.



Apr 19, 2010
5minuteargument said...
(Off-topic, but what's the point of that 'scribd' thingy? By cramming the image into a small iframe, it forces me to a) constantly scroll around, trying to view each part of the image, OR b) print the thing out, wasting paper. A plain old would have been far more accessible)
Apr 19, 2010
dscape said...
Dropping XML support is a really terrible decision on developers like myself. Please don't!?
Apr 19, 2010
Raffi Krikorian said...
don't worry!  we would never drop it on the current API - we're just evaluating whether we will bother supporting it on version 2.  version 2 will probably be fundamentally different enough that your current object parsers won't work (XML or JSON), so you probably will have to re-code either one.
Apr 20, 2010
marcellorinaldi said...
Really interesting post, thanks for sharing! :-)

Marcello Rinaldi
@marcellorinaldi
http://conduciendoaciegas.wordpress.com

Apr 20, 2010
Tripp Lilley said...
@5minuteargument: The original image is a vector image, and Scribd lets people whose browsers lack decent SVG support see it in its fullness.

The "Fullscreen" view, combined with the zoom controls, makes the utility of this pretty evident.

That said, a PNG preview, linked to the Scribd, a PDF, or a larger PNG would've also been a reasonable approach. But, hey, Scribd's what all the cool kids are doing :-P

Apr 20, 2010
coda said...
That ain't no JSON object. That there is a Ruby hash.
Apr 20, 2010
Raffi Krikorian said...
sure sure sure.  i just ran it through ruby to pretty print it.  you get the idea.
Apr 21, 2010
lnxwalt said...
I'd rather have the SVG or a PNG than Scribd. I'm not going to join another service just so I can download / print it and the Scribd version is almost unreadable.
Apr 21, 2010
Raffi Krikorian said...
Later today I'll upload a raw omnigraffle file (what I used to generate this).



Apr 21, 2010
James Anthony Campbell said...
great post that breaks down the meta data stored in every tweet
Apr 21, 2010
pattipdx said...
Thanks Raffi for letting us have a look under the hood!
Apr 21, 2010
DavidBromage said...
Why are the dates in such a bizarre format and not something sensible like ISO8601?
Apr 21, 2010
Raffi Krikorian said...
for historical reasons - its definitely something we would change if it wouldn't break backwards compatibility.  we'll probably address this in a future version of the API.
Apr 22, 2010
chrisbeach said...
Thanks for the illustration, although I can't tell you how much I hate Scribd. It breaks scrolling/zooming on the web.
Apr 27, 2010
kunalkantawala said...
Very useful illustration. Fully analytic.
May 09, 2010
vastaman said...
The callouts for the in_reply_to's and author username and screen name seem to be miss-wired.

Leave a comment...

About