Interesting tool: it'll take any JSON and convert it to a report (including making a nice sortable table for a list). You can try it out online, too: http://ajaxstack.com/jsonreport/

"Have you ever needed to generate some fake polygons to test a geo application?  No... oh.  I have, so I created GEOIPSUM!!!" "I used polymaps for the map, rails for the app, don't go near the dateline, it's a bit crap."

"Perhaps the best feature of Google Maps API V3 is that you can now remove select map elements. Why is this API V3’s best feature, you ask? Well, because it allows you to remove map elements that are not particularly relevant to your mashup. " This looks like a very useful guide, with examples.

"Twitter RSS Links allows you to generate a (RSS|ATOM|JSON|XML) feed from your friend's tweets that contains HTTP links." Does what it says, with OAuth (so it works for private folks). I really should do a delicious-network style un-shortening UI on top of the JSON, mind you.

A pretty good introductiory article to working with the Flickr API in the raw, using JSON (hurrah), and including uploading.

The headline seems correct to me too (using the Vox API, which is Atom, is painful in a way using the snowflake APIs from Flickr and Tumblr isn't- hm, how much does (human readable) documentation matter?), but Gregorio takes the time to figure out why, and the reasons seem pretty accurate on reflection.

"JSONovich is a Firefox extension that pretty-prints and colorizes JSON content (of types application/json and text/x-json). It uses json.org's json2.js to parse and indent JSON data, unless there is a native JSON parser as in Firefox 3.1, and google-code-prettify's parser and stylesheet to colorize it."

Matt Biddulph's list of useful libraries for the iPhone, from the perspective of someone used to the flexibility of scripting languages (that'd be me too). Includes my ex-colleague Stig Brautaset's JSON library.

"building raw Ajax + JSON apps is the fastest, simplest way to get a new idea in front of other people" Well, both of my recent Flickr explorations have pretty much been made that way. CouchDB as application platform might be a nice thing to explore, along with AppJet and Reasonably Smart.

A writeup of Yahoo's YQL, with special attention paid to the fact it outputs in JSON (yay JSON!). It certainly does a good job of making what I've been calling "API joins" for a while explicit, both in terms of the fact they're like database joins, and also the fact that they're really slow.

"The YQL platform provides a single endpoint service that enables developers to query, filter and combine data across Yahoo! and beyond. YQL exposes a SQL-like SELECT syntax that that is both familiar to developers and expressive enough for getting the right data." A fairly interesting experiment, if nothing else.

"IMG·2·JSON is a simple Google App Engine python application which extracts metadata from images and returns the results as a JSON string. " Could be handy, perhaps. Impressive use of GAE too.

I really like the flickraw approach. So much so, I can actually talk about it for twenty minutes. Watch out or I'll buttonhole you...

I seem to be on a bit of a JSON crusade at the moment; hopefully I'll be able to get a few other people to sign up along with me.

I can't remember if I bothered noting this here after seeing Freebase at etech this year, but I've just seen it again, so here's the link.

Call me lazy, but I like JSON and YAML because they're an eval away from being useful data structures. XML, by contrast, always seems to be painful.

Like XSLT, but for JSON, not XML. Obviously.